Investigation of T cell-mediated immune surveillance against tumor-specific
antigens in genetically engineered mouse models of cancer
by
Michel Justin Porter Du Page
B.A. Molecular and Cell Biology
University of California, Berkeley, 2000
SUBMITTED TO THE DEPARTMENT OF BIOLOGY IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY IN BIOLOGY
AT THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
FEBRUARY 2011
© 2011 Michel J.P. Du Page. All rights reserved.
The author hereby grants to MIT permission to reproduce and to distribute publicly
paper and electronic copies of this thesis document in whole or in part in any medium
now known or hereafter created.
Signature of Author:______________________________________________________
Department of Biology
September 30, 2010
Certified by:____________________________________________________________
Tyler Jacks
Professor of Biology
Thesis Supervisor
Accepted by:___________________________________________________________
Robert T. Sauer
Luria Professor of Biology
Chairman, Committee for Graduate Students
Investigation of T cell-mediated immune surveillance against tumor-specific
antigens in genetically engineered mouse models of cancer
by
Michel Justin Porter Du Page
Submitted to the Department of Biology on January 31, 2011
in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in Biology
ABSTRACT
The association of tumor cells and lymphocytes has led to the hypothesis that our
immune system actively inhibits the formation and progression of cancer, a phenomenon called
tumor immune surveillance. T cells specific to mutant proteins have been identified in cancer
patients and the recent success of cancer immunotherapies provides evidence that the immune
system can fight this disease. Yet the frequent occurrence of malignant disease despite T cell
recognition presents a significant medical problem. Only after we determine how tumors bypass
the immune system can immunotherapeutic approaches be improved.
To understand how tumors subvert immune responses, tumor transplantation or
transgenic mice expressing tumor-associated antigens have been used to model cancer. To
assess the role of anti-tumor T cells in models that more accurately reflect the human disease, I
developed new systems to introduce exogenous antigens, to mimic neoantigens, into
genetically engineered mouse models of lung cancer and sarcomas.
Utilizing the mouse model of lung cancer, I show that endogenous T cells respond to
and infiltrate lung tumors, delaying malignant progression. Despite continued antigen
expression, T cell infiltration does not persist and tumors ultimately escape immune attack.
Transplantation of cell lines derived from lung tumors that express these antigens or
prophylactic vaccination against autochthonous tumors, however, results in rapid tumor
eradication or selection of tumors that lose antigen expression. These results support clinical
data that suggest a role for the immune system in cancer suppression rather than prevention.
Tumor immune surveillance and immunoediting have largely been defined using
carcinogen-driven models of sarcomagenesis. Using a genetically engineered model of
sarcomagenesis, I show that immunoediting requires potent T cell antigens and that
lymphocytes drive the evolution of less immunogenic tumors by selecting for antigen loss.
Finally, immunotherapies have historically been ineffective in treating cancer patients. I
show that vaccination against specific antigens expressed in mouse lung cancers leads to
sustained anti-tumor T cell responses that eradicate recently initiated tumors. Vaccination also
stimulates anti-tumor T cell responses in an antigen-independent fashion by enhancing the
expansion and activity of T cells that recognize antigens only expressed in tumors.
Thesis supervisor: Tyler Jacks
Title: Professor of Biology
2
Michel DuPage
Curriculum Vitae
Koch Institute for Integrative Cancer Research, MIT
Dr. Tyler Jacks Laboratory
77 Massachusetts Ave, E17-517
Cambridge, MA 02139
dupage@mit.edu; (617) 452-3821
EDUCATION
Massachusetts Institute of Technology.
Ph.D. candidate in the Department of Biology (2003-present)
University of California, Berkeley.
Bachelor of Arts in Molecular and Cell Biology with Honors in the emphasis of Genetics
and a Minor in Anthropology (May 2000)
RESEARCH EXPERIENCE
Doctoral Research (2004-present)
Investigation of T cell-mediated immune surveillance against tumor antigens in
genetically engineered mouse models of cancer
Dr. Tyler Jacks Laboratory, David H. Koch Professor of Biology at MIT, Investigator
of the Howard Hughes Medical Institute, member of the National Academy of
Sciences, Director of the Koch Institute for Integrative Cancer Research, MIT.
Research Associate (2000-2003)
Antibody Technology
Department of Immunology, Genentech, Inc.
Honors Independent Undergraduate Research (1999-2000)
Sex Determination in the Germline of Drosophila melanogaster
Dr. Thomas W. Cline Laboratory, Professor of Genetics at the University of
California, Berkeley and member of the National Academy of Sciences.
HONORS & AWARDS
Margaret A. Cunningham Immune Mechanisms in Cancer Fellowship (2006-2007)
Dupont MIT Alliance fellow (2003-2004)
Outstanding Scholar of the Department of Molecular and Cell Biology (2000)
Honors Undergraduate Research in Molecular and Cell Biology (1999-2000)
Phi Beta Kappa Society (2000-present)
Golden Key National Honor Society (1997-2000)
Robert C. Byrd Honors Scholarship recipient (1996-2000)
TEACHING & MENTORSHIP EXPERIENCE
Mentored graduate student rotation projects in Jacks laboratory (2010)
Mentored MIT Undergraduate Research Opportunities Program (2005-present)
Teaching assistant, Immunology, MIT (H. Ploegh, J. Chen, L. Steiner, 2006)
Teaching assistant, Introductory Biology, MIT (E. Lander, R. Weinberg, 2004)
3
PRESENTATIONS
DuPage M, Cheung AF, Mazumdar C, Winslow MM, Bronson R, Crowley D, Chen J, Jacks
T. Endogenous T cell responses to lung adenocarcinoma neoantigens delay malignant
tumor progression. AACR Annual Meeting, Washington DC (2010) Poster.
DuPage M. Endogenous T cell responses to lung adenocarcinoma neoantigens delay
malignant tumor progression. Koch Institute Focus Seminar Series, Cambridge, MA
(2010) Oral presentation.
DuPage M. Endogenous T cell responses to lung adenocarcinoma neoantigens delay
malignant tumor progression. Broad Institute Cancer Program and Initiative Meeting,
Cambridge, MA (2009) Oral presentation.
DuPage M. Endogenous T cell responses to lung adenocarcinoma neoantigens delay
malignant tumor progression. Koch Institute Retreat, Waterville Valley, NH (2008) Oral
presentation.
DuPage M, Cheung AF, Jacks T. T cell mediated immune surveillance of novel tumor
antigens in a mouse model of lung adenocarcinoma. Smith Family New Investigator
Awards Program, Boston (2007) Poster.
DuPage M. T cell mediated immune surveillance to novel tumor antigens in a mouse
model of lung adenocarcinoma. Colrain Meeting, Colrain, MA (2007) Oral presentation.
DuPage M, Cheung AF, Jacks T. T cell mediated immune surveillance of novel tumor
antigens in a mouse model of lung adenocarcinoma. Cancer Immunotherapy Symposium,
New York City (2006) Poster.
DuPage M, Cheung AF, Jacks T. T cell mediated immune surveillance to novel tumor
antigens in a mouse model of lung adenocarcinoma. Mouse Models of Human Cancer
Consortium Meeting, “Tumor inflammation and immunity,” Washington DC (2006) Poster.
PUBLICATIONS
DuPage M, Mazumdar C, Jacks T. Tumor neoantigen expression dictates immunoediting
of sarcomas. (manuscript in preparation)
DuPage M, Cheung AF, Mazumdar C, Winslow MM, Bronson R, Crowley D, Chen J, Jacks
T. Endogenous T cell responses to lung adenocarcinoma neoantigens delay malignant
tumor progression. Cancer Cell (2011) 19: 72-85.
DuPage M, Dooley AL, Jacks T. Conditional mouse lung cancer models using adenoviral
or lentiviral delivery of Cre recombinase. Nat Protoc. (2009) 4: 1064-72.
Cheung AF, DuPage M, Dong HK, Chen J, Jacks T. Regulated expression of a tumorassociated antigen reveals multiple levels of T-cell tolerance in a mouse model of lung
cancer. Cancer Res. (2008) 68:9459-68.
4
ACKNOWLEDGEMENTS
During the course of my doctoral training I’ve had the fortune of excellent scientific
mentorship, critical training and technical support, and most importantly, close friendships. All of
these relationships have made this experience exciting, rewarding and fun. I would like to call
special attention to several individuals that were particularly important during my graduate
student tenure.
I want to begin by thanking my most influential scientific mentor, Tyler Jacks. Tyler is a
tremendous leader and by simply trying to emulate his high standards, his lab enjoys great
success. Perhaps the greatest lesson I’ve learned from Tyler is that there is no formula, no “best
way” to do science, each of us must find our own calling. This is what leads to the creativity and
excitement that brings about true breakthroughs. Tyler always “joked” (I think), “have you made
any breakthroughs? Are you pushing back the frontiers?” By entrusting me with the freedom
and support to do my best, his confidence in me became my own.
Many faculty members have been instrumental in my development as a scientist.
Richard Hynes and Jianzhu Chen, both members of my thesis committee, took me into their
labs for rotations my first year. I made friendships with them and their labs that have fostered
important scientific interactions over the years. I also want to thank Herman Eisen, Robert
Weinberg, and Richard Goldsby for going out of their way to provide advice and support. Arlene
Sharpe, also on my defense committee, has not only been a benevolent collaborator with us
over the years, but has also provided important guidance to me on my thesis research. Roderick
Bronson was instrumental in training me (in my limited capacity) everything I know about
pathology. I’ve enjoyed his company, his insight and his technical expertise.
The Jacks lab has been an incredible environment to grow as a scientist. Many
members of the lab have not only provided terrific advice and technical assistance over the
years but have also been great friends. I want to especially thank Ann Cheung. She was a real
pioneer in our lab, taking on a unique interest in our lab at the time – the role of the adaptive
immune system in cancer. I wonder whether I would have had the confidence to start this
challenging new direction for the lab if it weren’t for her interest and enthusiasm. I want to thank
Nik Joshi and Leah Schmidt for recently joining in our endeavors, I hope that Ann and I have
laid the foundation for fruitful discoveries in the realm of cancer immunology in the Jacks lab. I
want to thank Claire Mazumdar, an undergraduate at MIT whom I had the fortune of supervising
and now beginning her first year as a graduate student in biology at Stanford. Her hard work
was essential to many aspects of my research, but she also provided me with an opportunity to
learn how to become a better mentor. I benefited from the support of two incredibly generous
technicians in our lab, Margaret Magendantz and Kim Mercer, thank you! Besides making sure
that I could print this thesis, Judy Teixeira has been incredibly helpful to me throughout my time
in the lab and we are all lucky to have her in our group. A special thanks also to Keara Lane,
Alison Dooley, Monte Winslow, Peter Sandy, David Kirsch, and Carla Kim who were not only
great scientific colleagues but are also good friends.
I have benefited from the help of many skilled and generous members of the MIT
community. Members of the Koch Institute’s core facilities, in particular Denise Crowley, Glenn
Paradis, Eliza Vasile, Carol McKinley, and Scott Malstrom all made significant contributions to
my research. Kris Hewes and Becky King of the Department of Comparative Medicine have
been extremely helpful with mouse veterinary support. I would also like to thank Betsey Walsh
5
for taking care of me and all the biology graduate students. If it weren’t for her forging signatures
on registration day, many of us would be out a substantial amount of money!
I have made many close friends here at MIT that provided scientific support, but more
importantly, outlets from the science that were essential to my success. Nate Young joined the
Jacks lab with me as a graduate student and together we ripped this lab up. We put our own
stamp on this era of the Jacks lab and I think we had a lot of fun doing it, both scientifically and
also not so scientifically. I could always count on Nate to make me feel OK because no one can
look at a full glass and see it half empty like this guy. He and Amanda have been close friends
throughout graduate school and I hope will continue to be lifelong friends once we’re all
relocated in California. Unfortunately, many other close friends will be staying in MA (but we will
find a way to keep up!). Anne Deconinck and her family have been dear friends to my wife and
myself. Besides being irreplaceable as our lab manager, Anne has been a proponent of my
scientific success from the beginning and she is a close friend. David Feldser, although never a
proponent of my research (because: “I don’t understand it”), has been one my favorite people to
spend time with in and outside of the lab. He and Heidi are true friends that I hope we will share
many more adventures with in the future. David McFadden has been my athletic outlet through
the years. We’ve enjoyed many evenings of basketball and mornings cycling, all the while
discussing science and sports. Members of our recently formed band, “Ames and Main,” Phil
Santiago, Steve Katz, Nate Reticker-Flynn, and Jon Roberts have provided a musical outlet that
has been one of the most exciting additions to my life in the last year. I know that we will be
remembered for many years to come for our performance of Sweet Caroline at the Jacks lab
reunion of 2010. To all of these individuals, thank you for making this experience so much fun!
I would like to thank my mother and father for their enduring support of my educational
aspirations throughout my life. My mother’s interest in biology greatly influenced my own interest
in becoming a biological scientist. I believe her attention to detail and critical thinking have
greatly influenced who I am as a person and a scientist. My father’s work ethic and drive to
succeed are traits that I have aspired to emulate and without which, success in science is not
possible.
Most importantly, I attribute the success of nearly all of my endeavors to the unwavering
support of my wife, Amy Grace. She is an inspiration to me everyday. Her intelligence, positive
spirit, compassion, and love drives me in all of my pursuits. Thanks honey! We did it!
6
TABLE OF CONTENTS
ABSTRACT ...................................................................................................................................2
CURRICULUM VITAE ...................................................................................................................3
ACKNOWLEDGEMENTS..............................................................................................................5
TABLE OF CONTENTS ................................................................................................................7
CHAPTER 1: INTRODUCTION.....................................................................................................9
I. HISTORICAL PERSPECTIVE .........................................................................................11
II. IMMUNE PROMOTION OR SUPPRESSION OF CANCER ...........................................21
III. T CELL RESPONSES TO CANCER ...............................................................................25
A. DETECTION.........................................................................................................25
Tumor antigens..................................................................................................27
T cell priming .....................................................................................................31
T cell trafficking to tumors..................................................................................37
B. REGULATION ......................................................................................................40
Direct killing of tumor cells .................................................................................40
Cytokine-mediated cancer regulation ................................................................41
NKG2D receptor mediated anti-tumor T cell activity .........................................43
C. EVASION .............................................................................................................43
i. Cancer immunoediting ....................................................................................44
Immunoediting by passive selection...........................................................45
Immunoediting by active mechanisms of immune suppression .................47
ii. Immune tolerance of cancer ..........................................................................53
IV. TUMOR IMMUNE SURVEILLANCE IN CANCER PATIENTS ........................................55
Cancer in genetically or pharmacologically immunosuppressed patients ................56
Tumor infiltrating lymphocytes and patient prognoses .............................................58
Cancer immunotherapy ............................................................................................59
V. MODELING IMMUNE-TUMOR INTERACTIONS IN MICE .............................................61
Cancer by transplantation.........................................................................................65
Cancer by chemical induction...................................................................................66
Cancer by germline genetically engineered mice .....................................................68
Cancer by conditional genetically engineered mice .................................................69
Investigating T cell responses to mouse models of cancer ......................................70
Mouse cancer models alter anti-tumor T cell responses ..........................................72
Goals for the next generation of mouse models of cancer .......................................73
REFERENCES .......................................................................................................................76
CHAPTER 2: Endogenous T cell responses to antigens expressed in lung
adenocarcinomas delay malignant tumor progression ..............................................................103
ABSTRACT...........................................................................................................................104
INTRODUCTION ..................................................................................................................105
RESULTS .............................................................................................................................107
DISCUSSION .......................................................................................................................127
7
MATERIALS AND METHODS ..............................................................................................132
ACKNOWLEDGEMENTS .....................................................................................................136
SUPPLEMENTARY FIGURES (Figures S1 – S11) ..............................................................137
REFERENCES .....................................................................................................................150
CHAPTER 3: Tumor antigen expression is required for immunoediting against sarcomas ......156
ABSTRACT...........................................................................................................................157
INTRODUCTION ..................................................................................................................158
RESULTS .............................................................................................................................160
DISCUSSION .......................................................................................................................170
MATERIALS AND METHODS ..............................................................................................174
ACKNOWLEDGEMENTS .....................................................................................................176
SUPPLEMENTARY DATA (Supplementary Figures 1 – 4) ..................................................177
REFERENCES .....................................................................................................................182
CHAPTER 4: Therapeutic vaccinations with influenza enhance anti-tumor T cell responses
against mouse lung cancer via antigen dependent and independent mechanisms ..................185
ABSTRACT...........................................................................................................................186
INTRODUCTION ..................................................................................................................187
RESULTS .............................................................................................................................189
DISCUSSION .......................................................................................................................199
MATERIALS AND METHODS ..............................................................................................204
ACKNOWLEDGEMENTS .....................................................................................................206
REFERENCES .....................................................................................................................207
CHAPTER 5: DISCUSSION AND FUTURE DIRECTIONS.......................................................210
I. Tumor antigens in immunotherapy: TAA versus TSA....................................................211
II. Different models of cancer have altered T cell responses.............................................215
III. Tumor origin affects the T cell response and a tumor’s immunogenicity.......................217
IV. Cancer vaccination ........................................................................................................219
V. Immune tolerance or immunoediting? ...........................................................................220
VI. Timing in adaptive immune responses: a unifying theme?............................................221
REFERENCES .....................................................................................................................224
APPENDIX 1: Conditional mouse lung cancer models using adenoviral or lentiviral
delivery of Cre recombinase......................................................................................................227
ABSTRACT...........................................................................................................................228
INTRODUCTION ..................................................................................................................229
EXPERIMENTAL DESIGN ...................................................................................................234
MATERIALS .........................................................................................................................237
PROCEDURE .......................................................................................................................240
ANTICIPATED RESULTS ....................................................................................................245
FIGURES ..............................................................................................................................249
SUPPLEMENTARY FIGURES .............................................................................................253
ACKNOWLEDGEMENTS .....................................................................................................256
REFERENCES .....................................................................................................................257
8
CHAPTER 1
INTRODUCTION
9
Cancer is a disease that results from the uncontrolled proliferation of cells that ultimately
invade and disrupt the function of vital organs, causing death. Cancer is a complex disease that
can originate from many different tissues and progress through many different stages during its
development. Despite their diversity, cancers are unified by a shared etiology. Cancer is a
genetic disease that arises as a result of the culmination of multiple disruptions in intrinsic
cellular pathways, most importantly, mutations in the DNA encoding proto-oncogenes and tumor
suppressors. However, after over three decades of research since the initial discovery that
mutations in cellular genes underlie this disease, it is clear that cancer is much more
complicated than the culmination of disrupted genes in hyperproliferating cells; cancer is also a
product of the surrounding environment in which it originates and grows. Normal cells and
factors surrounding tumors are clearly implicated in shaping the disease, but the cells of the
immune system stand at the forefront as critical regulators of cancer. Indeed, the immune
system is viewed by many to be a “hallmark” of cancer (Condeelis and Pollard, 2006; Dunn et
al., 2006; Hanahan and Weinberg, 2000; Mantovani, 2009; Zitvogel et al., 2006), as it seems
that immune cells can act in opposing fashions to either promote or inhibit the initiation and
progression of this disease (de Visser et al., 2006; Grivennikov et al., 2010; Swann et al., 2008).
This thesis will focus on the ways in which T cells, primarily CD8+ cytotoxic T
lymphocytes (CTLs), may act to prevent or inhibit cancer progression. Utilizing new
experimental models in mice to investigate CTL responses to cancer, this thesis provides
insights into the natural role of CTLs reacting against cancer. This thesis also reveals that CTL
responses to cancers may be altered by the nature of the antigenic elements that target T cells
to the cancer cells as well as by the origin of the disease. Lessons learned from this work may
have important implications for how the immune system may be used to treat patients with
cancer.
10
I. HISTORTICAL PERSPECTIVE
The role of the immune system in cancer has been an active area of basic and clinical
research since the turn of the 20th century; however, a clear understanding of how the immune
system interacts with and potentially regulates cancer has only recently come into focus. The
convergence of our understanding of the biology of cancer and the immune system combined
with the acquisition of new tools to study their complexities has allowed greater insight into
immune-tumor interactions. However, a century’s worth of research was required to reach our
present understanding of tumor immunology, the study of immune responses to cancer. A field
steeped in this much history requires a brief historical perspective in order to appreciate not only
how we reached our current understanding of immune-tumor interactions, but also to
acknowledge the many important discoveries made from research in this field. Key discoveries
in the history of cancer and tumor immunology are highlighted in Figure 1.
The potential for important interactions between a developing tumor and the surrounding
immune system has been appreciated for over a century. Upon some of the first histological
examinations of cancer during the 19th century, Rudolf Virchow, the leader in cellular pathology
at the time, reported that most cancers were infiltrated by leukocytes (Balkwill and Mantovani,
2001). It wasn’t long before a functional connection between inflammatory cells and cancer
could be shown. The work of William Coley, and the generation of Coley’s toxins derived from
bacterial components, demonstrated that the induction of immune responses, at least when
provoked by pathogens, could suppress and even eradicate some tumors, particularly sarcomas
(Dranoff, 2004). However, it wasn’t until 1975 that the active agent in Coley’s toxin, tumor
necrosis factor (TNF), was purified (Carswell et al., 1975), providing a fitting example of the
11
12
1893
1909
1915
1930
Ames Test:
carcinogens
5
are mutagens
1952
1974
SV40 virus
causes tumors
Mitchison shows
lymphocytes mediate
3
tumor rejection
1976
Kohler & Milstein
develop
technique to
7
produce mAbs
TNF discovered
as active agent
6
in Coley’s toxins
Discovery
of NK cells
1967
3
2
1
First human
tumor antigen
14
discovered
First endogenous
GEM cancer models
15
targeting p53
RD Schreiber
coins “tumor
immunoediting”
FDA approves
HPV vaccine –
prevents
cervical cancer
FDA approves first
therapeutic cancer
21
vaccine (Sipuleucel-T)
CTLA-4 antibodies
increase survival of
metastatic melanoma
22
patients
ACT therapy
effective in ~50% of
metastatic
19
melanoma patients
ACT therapy with
lymphocytes transduced
20
with recombinant TCRs
Rous, P., A Sarcoma of the Fowl Transmissible by an Agent Separable from the Tumor Cells. J Exp Med 13 (4), 397 (1911).
Gross, Ludwik, Intradermal Immunization of C3H Mice against a Sarcoma That Originated in an Animal of the Same Line. Cancer Research 3 (5), 326 (1943).
Mitchison, N. A., Studies on the immunological response to foreign tumor transplants in the mouse. I. The role of lymph node cells in conferring immunity by adoptive transfer. J Exp
Med 102 (2), 157 (1955).
Rosenberg
begins using
TILs + IL-2 to
16
treat melanoma
Generation of
tetramers to
18
track T cells
First peptide
vaccines in
cancer patients
Dvorak: “Tumors:
wounds that do
12
not heal”
First transgenic
GEM cancer
11
models
1979 1981 1987 1991 1996 2001 2006 2010
Jenkins & Schwartz
discover T cell
13
costimulation
FDA approves
HBV vaccine –
prevents HCC
p53
discovered
as a tumor
10
antigen
Doherty & Zinkernagel
discover mechanism of
8
T cell recognition
Burnet & Thomas
coin “tumor
immunosurveillance”
HeLa cells first
cultured tumor
cell line
1943
Gross: mice can
be immunized to
2
their own tumor
Snell discovers
MHC locus in mice
Dibenzanthracene
isolated from coal
tar as active agent
Rous passages
tumor via cell
1
free filtrates
Coley’s toxins
cause tumor
regression
1863
Ehrlich posits
immune system
regulates cancer
Virchow notes intimate
association of immune
cells in cancer
Coal tar first
used to induce
tumors
Prehn & Main:
MCA-induced
tumors harbor TSA
that mediate
4
transplant rejection
Varmus & Bishop
discover Src is
9
normal cellular gene
!
Figure 1: A timeline of major events in the history of tumor immunology.
!
CTLA-4 blockade
promotes tumor
17
rejection in mice
4
Prehn, R. T. and Main, J. M., Immunity to methylcholanthrene-induced sarcomas. J Natl Cancer Inst 18 (6), 769 (1957).
Ames, B. N., Durston, W. E., Yamasaki, E., and Lee, F. D., Carcinogens are mutagens: a simple test system combining liver
homogenates for activation and bacteria for detection. Proc Natl Acad Sci U S A 70 (8), 2281 (1973).
6
Carswell, E. A. et al., An endotoxin-induced serum factor that causes necrosis of tumors. Proc Natl Acad Sci U S A 72 (9), 3666
(1975).
7
Kohler, G. and Milstein, C., Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256 (5517), 495
(1975).
8
Zinkernagel, R. M. and Doherty, P. C., Immunological surveillance against altered self components by sensitised T lymphocytes in
lymphocytic choriomeningitis. Nature 251 (5475), 547 (1974); Zinkernagel, R. M. and Doherty, P. C., Restriction of in vitro T cellmediated cytotoxicity in lymphocytic choriomeningitis within a syngeneic or semiallogeneic system. Nature 248 (450), 701 (1974).
9
Stehelin, D., Varmus, H. E., Bishop, J. M., and Vogt, P. K., DNA related to the transforming gene(s) of avian sarcoma viruses is
present in normal avian DNA. Nature 260 (5547), 170 (1976).
10
DeLeo, A. B. et al., Detection of a transformation-related antigen in chemically induced sarcomas and other transformed cells of
the mouse. Proc Natl Acad Sci U S A 76 (5), 2420 (1979).
11
Brinster, R. L. et al., Transgenic mice harboring SV40 T-antigen genes develop characteristic brain tumors. Cell 37 (2), 367
(1984); Stewart, T. A., Pattengale, P. K., and Leder, P., Spontaneous mammary adenocarcinomas in transgenic mice that carry
and express MTV/myc fusion genes. Cell 38 (3), 627 (1984).
12
Dvorak, H. F., Tumors: wounds that do not heal. Similarities between tumor stroma generation and wound healing. N Engl J Med
315 (26), 1650 (1986).
13
Jenkins, M. K. and Schwartz, R. H., Antigen presentation by chemically modified splenocytes induces antigen-specific T cell
unresponsiveness in vitro and in vivo. J Exp Med 165 (2), 302 (1987).
14
van der Bruggen, P. et al., A gene encoding an antigen recognized by cytolytic T lymphocytes on a human melanoma. Science
254 (5038), 1643 (1991).
15
Donehower, L. A. et al., Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 356
(6366), 215 (1992); Jacks, T. et al., Tumor spectrum analysis in p53-mutant mice. Curr Biol 4 (1), 1 (1994).
16
Rosenberg, S. A. et al., Treatment of patients with metastatic melanoma with autologous tumor-infiltrating lymphocytes and
interleukin 2. J Natl Cancer Inst 86 (15), 1159 (1994).
17
Leach, D. R., Krummel, M. F., and Allison, J. P., Enhancement of antitumor immunity by CTLA-4 blockade. Science 271 (5256),
1734 (1996).
18
Altman, J. D. et al., Phenotypic analysis of antigen-specific T lymphocytes. Science 274 (5284), 94 (1996).
19
Dudley, M. E. et al., Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science
298 (5594), 850 (2002); Rosenberg, S. A. and Dudley, M. E., Cancer regression in patients with metastatic melanoma after the
transfer of autologous antitumor lymphocytes. Proc Natl Acad Sci U S A 101 Suppl 2, 14639 (2004).
20
Morgan, R. A. et al., Cancer regression in patients after transfer of genetically engineered lymphocytes. Science 314 (5796), 126
(2006).
21
Kantoff, Philip W. et al., Sipuleucel-T Immunotherapy for Castration-Resistant Prostate Cancer. New England Journal of Medicine
363 (5), 411 (2010).
22
Hodi, F. S. et al., Improved Survival with Ipilimumab in Patients with Metastatic Melanoma. N Engl J Med (2010).
5
13
tremendous lag time (80 years!) between key observations in tumor immunology and a true
biological understanding of the phenomenon.
At about the same time – the turn of the 20th century – investigators had begun to
propagate spontaneous tumors in animals by serially transplanting the tumors from one animal
to the next. Through the course of this procedure, it was discovered that first, tumors could not
be passaged into different species, but more interestingly, some tumors would spontaneously
regress after transplantation into animals of the same species. Subsequent experiments
showed that animals could be immunized against a transplanted tumor if they were first
inoculated with a sub-lethal challenge of the same tumor (Schreiber, 2003; Scott, 1991). In
1909, perhaps as a result of these findings, or the work of William Coley, Paul Ehrlich first
introduced the idea that cancer may be regulated by the immune system. He speculated that
the immune system served to protect the host from the progression of neoplasms into malignant
cancer, and that an attenuation of the immune response was responsible for tumor outgrowth
(Dunn et al., 2004a; Witkowski, 1990). Ehrlich was truly ahead of his time.
Peyton Rous soon published his landmark paper describing the transmission of an avian
sarcoma from one chicken to another “by means of a cell free filtrate” (Rous, 1911). Although it
took some time before it was understood that viruses within the cell-free filtrate were
responsible for transmitting the cancer (Rous, 1965), it took much longer for the development of
vaccines to prevent the infection of such viruses as hepatitis B virus (HBV) or human
papillomavirus (HPV) that can cause hepatocellular carcinomas or cervical cancers,
respectively.
Experimentation with transplantation of spontaneous tumors continued for several
decades in the early 20th century, but typically with inconclusive and irreproducible results. The
discovery that normal, untransformed tissues grafted from one animal to another were rejected
in a similar fashion to transplanted tumors temporarily sidelined the progress of tumor
14
immunology because of a fundamental lack of understanding about the immunobiology of
allograft rejection. However, these discoveries bolstered interest in the genetics of allograft
rejection, leading such mouse geneticists as Clarence Cook Little and George Snell to begin
generating inbred mouse strains at the Jackson Laboratory. Investigations into the mechanisms
of graft rejection ultimately led to the discovery of the major histocompatibility complex as the
key determinant in allograft rejection.
By the 1940s, the development of inbred mice and chemically induced carcinogenesis
provided new opportunities for experimentation by tumor immunologists. Ludwik Gross was now
able to demonstrate for the first time that mice could be specifically immunized against cancer
by using the carcinogen methylcholanthrene (MCA) to induce tumors and transplant them within
mice of the same inbred strain (Gross, 1943). Thus, tumor transplant rejection could no longer
be explained by “residual heterozygosity,” i.e. inherent differences between different strains of
mice. Although work by E.J. Foley, confirmed the results of Gross with MCA-induced sarcomas;
parallel experiments utilizing mammary tumors that arose spontaneously revealed that
immunization against spontaneous tumors was not possible (Foley, 1953a, b). Therefore, a
fundamental difference was observed between spontaneously arising tumors and carcinogeninduced tumors – only the latter tumors could be successfully immunized against. This is in fact
a critical issue that will be revisited throughout this thesis: do carcinogen-induced animal models
of cancer accurately reflect the disease in humans, specifically with respect to immune
responses?
Nevertheless, the work of R.T. Prehn and J.M. Main in the late 1950s, solidified the
importance of Gross’ earlier work by reproducing his results with tumor-specific immunization
but also demonstrating that normal tissue grafts did not protect mice from subsequently
transplanted tumors, and prior challenge with these tumors did not lead to the rejection of
subsequently grafted normal tissues, conclusively ruling out residual heterozygosity as the
15
means of rejection (Prehn and Main, 1957). Taking the system one step further, George Klein
showed that individual mice could be immunized against their own autochthonous tumors by
inducing sarcomas with MCA, surgically removing them, and then subsequently retransplanting
them back into the original host (Klein et al., 1960). Indeed, the foundation for tumor-specific
immunity had been laid. Yet, because most spontaneous tumors and some MCA-induced
tumors were not immunogenic, even these early investigators responded cautiously to their
influential work, stating that “the altered antigenicity [of the MCA-induced tumors] was not an
intrinsic part of the carcinogenic process but rather an incidental concomitant change” (Scott,
1991). The final piece of the puzzle came together with the work of Avrion Mitchison who
showed that lymphocytes were the critical mediators of the immune response that was
responsible for tumor rejection (Clayton, 2006; Mitchison, 1953, 1955).
These experiments formed the foundation for the theory that the immune system
naturally protects hosts from cancer through lymphocyte recognition of tumor-specific antigens.
Thus it is not surprising that Macfarlane Burnett and Lewis Thomas re-instigated this concept in
the early 1960s, coining the phenomenon tumor immune surveillance. Although both agreed
that the immune system served as an extrinsic tumor suppressor of cancer; Burnett believed
that cancer was recognized by the immune system due to the generation of distinct neoantigens
in tumors, whereas Thomas believed the response was due to the neoplastic process itself –
the role of the immune system was to monitor and maintain tissue homeostasis (Dunn et al.,
2002; Smyth et al., 2001). Despite evidence that the immune system could provide protection
from cancer, proof was lacking that tumor immune surveillance occurred naturally.
The theory of tumor immune surveillance was soon shattered in the 1970s by the work
of Osias Stutman who began using T cell deficient athymic nude (nu/nu) mice to induce
sarcoma formation. By comparing the susceptibility to MCA-induced tumor formation of nu/nu
and nu/+ littermates, Stutman found no difference in susceptibility to MCA-induced or
16
spontaneous cancers (Stutman, 1974). This led researchers to conclude that the tumor immune
surveillance theory was an artifact of transplantation and not representative of naturally arising
cancers. However, such conclusions may have been premature. First, the strain of mice used in
these studies is thought to be abnormally susceptible to MCA-induced tumor formation,
potentially causing tumors to arise “at a rate that overwhelms host immunological defense
mechanisms” (Dunn et al., 2002). Second, MCA was injected into newborn mice. This could be
of critical importance because we now know that particularly in mice, the adaptive immune
system does not completely develop until after birth. Therefore, if transformation began before
the complete development of T cells, it is possible that tumors would not be rejected due to the
establishment of tolerance during the first weeks after birth (Adkins et al., 2004; Billingham et
al., 1953; Medawar, 1961). Finally, and perhaps most importantly, nude mice are not completely
immune deficient. They possess some mature T cells and normal numbers of antibodyproducing B cells and natural killer (NK) cells, an innate population of lymphocytes with
cytotoxic activity similar to CTLs. Interestingly enough, NK cells were discovered as a direct
consequence of the investigation of immune responses to cancer (Kiessling and Klein, 1973;
Kiessling et al., 1976). Although originally thought to be an artifact of in vitro experiments, NK
cells were discovered and named because of their “natural” (antigen-independent) ability to kill
tumor cell lines in culture.
Support for tumor immune surveillance was largely stalled until new technologies
allowed for more refined investigations of immune-tumor interactions. A key advance was the
development by Kohler and Milstein of techniques to produce and maintain monoclonal
antibodies, allowing the identification and characterization of the different immune cell subsets
based on the expression of cell surface antigens. Perhaps more important to the field of tumor
immunology was the use of monoclonal antibodies injected into mice as a means for depleting
specific immune cell subsets or blocking key receptor-ligand interactions in mouse cancer
17
models. Depletion of a critical immune-regulatory cell population (Tregs), using the monoclonal
antibody PC-61 that targets CD25, demonstrated a key role for these cells in suppressing antitumor immune responses (Shimizu et al., 1999). The use of monoclonal antibodies to block
interactions with an inhibitory receptor on T cells (CTLA-4) in vivo, proved to enhance anti-tumor
immunity in mice (Leach et al., 1996). Recently antibodies against CTLA-4 have shown efficacy
in Phase III clinical trials of patients with metastatic melanoma (Hodi et al., 2010).
The advent of molecular cloning techniques led to the isolation and recombinant
expression of several genes encoding cytokines, of which interleukin-2 (IL-2) was the first
(Taniguchi et al., 1983). Shortly thereafter, Steven Rosenberg and colleagues began testing
recombinant IL-2 in clinical trials to treat cancer with the hope that it would stimulate existing
anti-tumor T cell responses (Lotze et al., 1986; Rosenberg et al., 1994a). The next trials
administered so-called “lymphokine-activated killer cells” (LAKs) in conjunction with IL-2 therapy
(Rosenberg et al., 1987; Rosenberg et al., 1985). LAKs were lymphocytes harvested from
patients’ blood and activated in vitro with IL-2. The final step for Rosenberg’s team was the first
incarnation of the modern form of adoptive cell transfer therapy (ACT). In this therapeutic
regimen, tumor-infiltrating lymphocytes (TILs), rather than LAKs, are harvested directly from a
patients tumors, expanded in vitro, and used as the anti-cancer immune effector cells (Dudley et
al., 2002; Rosenberg and Dudley, 2004; Rosenberg et al., 1986; Rosenberg et al., 1994b).
Molecular cloning also allowed the generation of cDNA libraries that provided a path to
the discovery of the first antigen expressed in tumors that could be recognized by T cells (van
der Bruggen et al., 1991). The use of cDNA libraries to screen for novel tumor antigens (called
SEREX for serological identification of antigens by recombinant expression cloning) has proven
to be the most rapid method available for discovering shared tumor antigens (Old and Chen,
1998).
18
With an understanding of how T cells recognize antigens, by complexing their TCR with
peptide-loaded MHC molecules, came the recombinant engineering of so-called “tetramers,”
multivalent MHC molecules that could be loaded with any peptide antigen of interest (Altman et
al., 1996). The use of tetramers has dramatically accelerated our capacity to identify and
characterize T cells responding to cancers in patients and animal models (Lee et al., 1999;
Romero et al., 1998; Savage et al., 2008; Yee et al., 1999). Finally, molecular cloning has
provided the means to artificially endow tumor-specificity to a cancer patient’s lymphocytes by
manually introducing recombinant TCRs that recognize tumor antigens into their T cells for use
in ACT-based therapy (Cohen et al., 2005; Morgan et al., 2006; Pule et al., 2008).
Finally, the technology of targeted gene disruption in mice allowed the generation of
better defined models of immunodeficiency to study tumor immune surveillance. Seminal
studies using mice that lack critical components of immunity – interferon-γ or its receptor,
perforin, or recombinase-activating gene deficient mice – were shown to have increased
susceptibility to carcinogen-induced or spontaneous tumor formation compared to wild-type
mice (Kaplan et al., 1998; Shankaran et al., 2001; van den Broek et al., 1996). The recent work
of Robert Schreiber and colleagues has expanded the notion of tumor immune surveillance to
encompass a role for the immune system in shaping tumor development. They propose that the
immune system plays a role in the evolution of cancer by selecting for tumors with decreased
immunogenicity, a concept termed cancer immunoediting (Dunn et al., 2002; Dunn et al., 2006;
Dunn et al., 2004b). This concept is supported by experiments showing that tumors that arise in
immune-deficient mice are susceptible to rejection upon subsequent transplantation into
immune competent mice (Dunn et al., 2005; Shankaran et al., 2001). It is proposed that tumors
of reduced immunogenicity arise either through the outgrowth of tumors that lack (or have lost)
antigens recognizable by the immune system or through the outgrowth of tumors that gain the
capacity to suppress an anti-tumor immune response (Pardoll, 2003).
19
Despite our increased understanding of the immune system and cancer, the availability
of new tools to study immune responses to cancer, and the technological leap to using
genetically engineered mouse models of immunodeficiency to reveal a role for immune
surveillance of cancer; most investigations continue with highly artificial tumor models. In these
models, carcinogen-induced tumors or long-term cultured tumor cells are transplanted into mice
to assess immune-tumor interactions. As discussed before, carcinogen-induced cancers may
harbor abnormally large numbers of mutations which may lead to artificially robust immune
responses (Scott, 1991), while the act of transplantation itself presents a plethora of
complications that likely make it a poor model of cancers that arise naturally in the human
population (Frese and Tuveson, 2007; Garbe et al., 2006; Willimsky and Blankenstein, 2000).
Tumor cells administered via injection are at a great disadvantage immunologically because
such routes can initiate proinflammatory responses or traffic tumor cells directly to lymphoid
organs making tumors more susceptible to rejection (Khong and Restifo, 2002). More
importantly, transplant models are inadequate because they do not allow researchers to study
the dynamics of the immune-tumor interface as normal cells become transformed, evolve, and
obtain the genetic changes necessary to initiate, promote and progress toward malignancy.
Finally, due to the rapid growth kinetics of transplanted cells, researchers are forced to study an
acute immune response to the tumor, whereas immune-tumor interactions in cancer patients
are more likely to be a chronic affair. Therefore, it is imperative that we begin to use tractable
and reproducible models of autochthonous disease that more accurately recapitulate the
stepwise progression of human cancer, both histopathologically and genetically, to study tumor
immune surveillance (Frese and Tuveson, 2007; Khong and Restifo, 2002; Tuveson and Jacks,
2002; Van Dyke and Jacks, 2002). The major goal of this thesis research was to employ the use
of such models to investigate the role of immune-tumor interactions in regulating the
20
development and progression of cancer (see Chapters 2 and 3). In the next section I will briefly
discuss evidence that the immune system can paradoxically promote cancer in some settings.
II. IMMUNE PROMOTION OR SUPPRESSION OF CANCER
Before continuing with a description of how T cells respond to and regulate cancer, it is
important to point out that the immune system has also been shown to promote cancer in
several settings. Notably, cells of the innate immune system have typically been associated with
aiding cancer progression, but adaptive immune cells such as B cells and CD4+ T cells have
also been implicated. CD8+ T cells have not been implicated in tumor promotion.
When Virchow first reported the presence of immune cells in multiple forms of cancer, he
actually speculated that the inflammatory cells associated with cancer might promote the
disease. Much more recently, it has been proposed that the tumor microenvironment, often
harboring a large composition of inflammatory cells, shares many properties with wounds. In
1986, Harold Dvorak first posed the notion that cancers are what he described as “wounds that
do not heal” (Dvorak, 1986). Recent studies comparing gene expression data from wounded
tissues and cancers in the same tissues have corroborated this hypothesis (Chang et al., 2004;
Riss et al., 2006).
Inflammatory cells recruited to sites of injury are clearly important for providing protection
from pathogens. However, inflammatory cells also play an important role in wound healing,
promoting tissue regeneration, cell migration, and new blood vessel formation through the
secretion of growth factors, chemokines and pro-angiogenic cues, all factors that could
conceivably promote cancer progression (see reviews: (Coussens and Werb, 2002; Grivennikov
et al., 2010; Mantovani et al., 2008)). Even the production of reactive oxygen species by innate
21
inflammatory cells, as a means to remove pathogenic invaders, could serve as a mutagenic
agent promoting the initiation or progression of cancers (Weitzman and Gordon, 1990). In fact,
we know that chronic inflammatory syndromes are associated with several forms of cancer in
humans, such as hepatitis and hepatocellular carcinomas, pancreatitis and pancreatic
carcinomas, inflammatory bowel disease and colorectal carcinomas, and bronchitis and lung
carcinomas (Coussens and Werb, 2002). Although these are only correlative associations, not
causal links between inflammation and cancer induction, several lines of evidence point to a
positive role for the immune system in cancer formation and progression.
Screens for cDNAs that could modify the activity of p53 provided one of the first direct
links between inflammation and cancer. The identification of macrophage migration inhibitory
factor (MIF), a cytokine produced by immune cells, as an inhibitor of the tumor suppressor p53,
provided a means for tumor induction by inflammatory cells (Hudson et al., 1999). Elegant work
in mouse models of cancer have provided direct evidence that certain cells of the immune
system can promote cancer formation and progression (Qian and Pollard, 2010). Several
investigations have demonstrated that inflammatory cells recruited to the tumor
microenvironment induce the activation of NF-κB signaling in cancer cells, a critical pathway
driving the progression of many forms of cancer (Greten et al., 2004; Meylan et al., 2009;
Pikarsky et al., 2004). The activation of STAT3 signaling in tumor cells as a consequence of
inflammation has also been shown to not only promote multiple aspects of tumorigenesis, but
also potentially inhibit aspects of anti-tumor immunity (Yu et al., 2009). Tumor-associated
macrophages, neutrophils, and mast cells have all been shown to promote tumor angiogenesis
in mouse models of breast, skin, and pancreatic cancers, primarily through the activity of matrix
metalloproteinases that can release vascular endothelial growth factor (VEGF) from the
extracellular matrix (Coussens et al., 1999; Coussens et al., 2000; Lin et al., 2006; Nozawa et
al., 2006). Metastasis is promoted in models of mammary tumorigenesis through the
22
establishment of paracrine chemotactic signals between macrophages and tumor cells
(DeNardo et al., 2009; Lin et al., 2001; Wyckoff et al., 2004). Notably, adaptive immune cells
such as B cells or CD4+ T cells have been implicated in initiating the recruitment of these innate
inflammatory cells that ultimately carry out the tumor-promoting activities of the immune system
(Daniel et al., 2003; de Visser et al., 2005; DeNardo et al., 2009). Therefore, cells of the immune
system can play critical roles in promoting cancer in some settings. However, depending on the
type of immune cells recruited to the tumor, and the phenotype of these cells, the immune
response can function in opposing ways to either promote or inhibit cancer initiation and
progression (Figure 2).
The remainder of this thesis will now focus exclusively on how T cell responses to
cancer can act to negatively regulate the progression of this disease. I will begin by providing
the mechanisms by which T cells can recognize, respond to, and regulate cancer.
23
Effector
Immune
cells
Effect on tumor
Effector
Immune
cells
Effect on tumor
Tumor cell
death
DC
Promote DC maturation and
antigen presentation to T cells
ROS
Neutrophil
(N)
DNA damage, promote
mutation
Tumor
antigens
DC
Engulfed by DCs presented to
T cells
MIF
M, N
Block p53, promote loss of
tumor suppression
Stress
ligands
NK, CTL
Alert/activate anti-tumor
immune response
Antibodies
B
(CD4 T)
Recruitment of pro-tumor
inflammatory cells
Pep/MHC
CTL
Target T cells to tumors
MMPs
M, N, MC
Release growth factors,
promote angiogenesis
IFN-!
CTL
CD4 T
Increase pep/MHC on tumors
Inhibit angiogenesis
TNF
M, N
Activate NF-!B in tumors
FAS/FASL
Perforin/Gz
NK, CTL
Mediate tumor cell killing
EGF/CSF1
Mac/
Tumor
Promote metastasis via
paracrine signals
ROS
MIF
antibodies
!"
!" !"
CTLs
pep/MHC
IFN-!
ECM
stress
ligands
MMPs
tumor
antigens
CD4+ Ts
cell death
mast cells
EGF
CSF1
NK cells
dendritic cells
ANTI-TUMOR
neutrophils
macrophages
PRO-TUMOR
Figure 2. Immune cells and their effectors can act to suppress or promote cancer.
Critical effectors and cells of the immune system that provide anti-tumor activities are depicted
in red (left side) and immune effectors and cells that act in a pro-tumor fashion are depicted in
green (right side). Several examples of anti-tumor and pro-tumor effects are listed above.
DC, dendritic cell; NK, natural killer cell; CTL, cytotoxic T lymphocyte; M or Mac, macrophage;
N, neutrophil; MC, mast cell; Gz, granzyme; ROS, reactive oxygen species; MIF, macrophage
inhibitory factor; EGF, epidermal growth factor; CSF1, colony-stimulating factor 1; ECM,
extracellular matrix; MMPs, matrix metalloproteinases.
24
III. T CELL RESPONSES TO CANCER
To understand how T cell interactions with cancer can shape this disease we will explore
three phases or parameters that describe how these interactions take place. These interactions
begin with the detection of cancer by T cells, are followed by the regulation of cancer by T cells,
and finally end with the cessation of the T cell response against cancer during tumor evasion
(Figure 3).
A. Detection
At the outset, in order for T cells to respond to cancers they require a means to “see” the
tumor. Cytotoxic T lymphocytes (CTLs) have evolved to recognize and kill host cells harboring
cellular pathogens. Impressively, CTLs have nearly a limitless capacity to recognize foreign
invaders because of an acute capacity to recognize foreign or “non-self” proteins (called
antigens) produced by the pathogen. Importantly, unlike antibodies produced from B cells that
can recognize nearly any foreign chemical structure, T cells are restricted to proteinaceous
antigens. This is because T cells utilize a surface bound receptor, the T cell receptor (TCR), to
recognize protein fragments (peptides) produced within the target cell but presented on its
surface in conjunction with the major histocompatibility proteins (MHC). These interactions
between the TCR and peptide/MHC complexes provide a means for CTLs to sample and screen
the inner contents of host cells for the presence of foreign proteins within each cell’s complex
proteome. In this way, CTLs can promote the clearance of cellular infections by seeking out and
destroying all pathogen-infected cells within the host. This presents a problem for understanding
how T cells recognize cancer cells, as these are “self” cells and only in a minority of cases do
cancer cells produce antigens from foreign pathogens. Therefore, alternative classes of
antigens must be produced by the tumors that allow for engagement by T cells. We will begin
25
Selection
- Antigen loss
- MHC presentation
TGF-"
Tumor Ags
B7-H1
DC
Suppression
soluble
MIC
- Modulatory receptors
- Cytokines
- Regulatory cells
- Migration
DAMPs
Invasion
Metastasis
CD8
CD40/
CD40L
MHCI
Kill:
Tolerance
FAS/FASL
Perforin/Granzymes
NKG2D
CD4
- Inhibitory receptors
- Chronic Ag exposure
IFN-!
B7/CD28
LAG3
MHCII
Chemokines
MDSC
CTL
TREG
DETECTION
REGULATION
PD1
Tolerant CTL
EVASION
Figure 3. The three phases of T cell responses to cancer.
The detection phase (in blue, left) involves the acquisition of tumor antigens at the tumor site by
dendritic cells (DC) that migrate to the draining lymph node to present antigens to CD8 and CD4
T cells that with additional signals (green arrows) can activate T cells which migrate back to the
tumor site and engage tumors cells directly via TCR-pep/MHCI interactions. The regulation
phase (in red, middle) is mediated by CTL recognition and direct killing of tumor cells or indirect
mechanisms involving cytokine secretion. The evasion phase (in green, right) can occur through
a variety of processes involving selection, suppression, or T cell tolerance (details are listed in
the green boxes).
26
the discussion of how T cells detect cancer by introducing the classes of antigens, called tumor
antigens, which drive the T cell response against cancer.
Tumor antigens
Tumor antigens recognized by T cells have been discovered in patients with cancer and
in experimental models of cancer (Novellino et al., 2005; Schreiber, 2003). These antigens can
be broadly subdivided into two major classes: tumor-specific antigens (TSAs) and tumorassociated antigens (TAAs) (Table 1). The significance of the class of antigen, whether it’s
specific to the tumor or not, in determining the reactivity and function of T cells responding to
cancers is the subject of great debate (Finn, 2008b; Parmiani et al., 2007; Schietinger et al.,
2008). TSAs are postulated to be better targets for immunotherapy because these antigens
have been shown to be more immunogenic than TAAs and T cell responses against them do
not pose the threat of inducing autoimmunity due to their exclusive expression in tumors.
However, TSAs are usually specific to each individual’s cancer and thus necessitate
personalized therapies, whereas TAAs have been identified in multiple types of cancer and
present the potential for broad applicability to many patients and cancers.
TABLE 1: Classification of tumor antigens
Antigen Type:
Class:
Mutation
Protein fusion
Tumor-specific antigens
Viral
Modified glycosylation
Cancer-testis
Tumor-associated antigens Oncofetal
Differentiation
Overexpressed
Cryptic transcription
Non-classical antigens
Protein splicing
Examples:
Ras, CDK4, β-catenin
BCR-ABL
HPV E6 and E7, EBNA-1
Cosmc (mouse)
MAGE-1, NY-ESO-1
CEA
Gp100, MART-1, Tyrosinase
Her2/Neu, CycB, hTERT, p53
Gp75
Gp100, FGF-5
27
Tumor-specific antigens
Antigens that are expressed uniquely in cancer cells represent enticing targets for
cancer immunotherapy because of the potential for T cells to clearly differentiate cancer cells
from normal cells. Tumor-specific antigens (TSAs) are the products of mutated proteins, fusion
proteins, and viral proteins. The role of TSAs in the rejection of chemically-induced and UVinduced mouse models of cancer has been appreciated since the 1950s (Schreiber, 2003). In
seminal experiments in which a multitude of distinct carcinogen-induced tumors were screened
for the capacity to provide protection from subsequent transplants after prophylactic vaccination,
it was demonstrated that protection only occurred after vaccination with the same tumor
(Mondal et al., 1970; Prehn, 1968). This provided the first evidence that the tumor antigens
capable of eliminating cancers may be unique to each individual cancer. It wasn’t until several
decades later that the first TSA arising from a mutation in a UV-induced mouse tumor was
identified (Monach et al., 1995). In this case, the neoantigen required for tumor rejection was
derived from a single point mutation in the L9 ribosomal protein. Although this mutation has not
been identified to play a role in driving the development of cancer, mutations in the critical tumor
suppressor p53 from methylcholanthrene-induced mouse sarcomas have also been shown to
generate peptides that can be recognized by T cells (Noguchi et al., 1994). More importantly,
several mutations have been identified in human cancers that are both oncogenic and antigenic,
such as mutations in Ras (Fossum et al., 1994a; Fossum et al., 1994b; Fossum et al., 1995),
CDK4 (Wolfel et al., 1995), CDC27 (Wang et al., 1999), and β-catenin (Robbins et al., 1996).
These may be ideal targets for immunotherapy because the neoantigens are in fact essential for
tumorigenesis, making it unlikely that tumors would lose expression of these proteins during the
course of targeted immunotherapy (Cavallo et al., 2007). The recent sequencing of multiple
cancer genomes has revealed that hundreds of point mutations exist in any given tumor,
increasing the likelihood that one or more of these mutant proteins can be recognized by T cells
28
(Sjoblom et al., 2006). In fact, recent computational analyses, based on the cancer genomes
from breast and colorectal cancer patients, predict that individual tumors harbor from 7-10 novel
antigens capable of being presented to T cells on the most ubiquitous MHC class I allele (Segal
et al., 2008).
Chromosomal rearrangements and translocations can generate new antigens through
the fusion of normally distinct polypeptide sequences and are common in some cancers,
especially hematopoietic malignancies. Indeed, T cells recognizing the BCR-ABL or TEL-AML1
fusion proteins have been found in patients with chronic myelogenous leukemia or acute
lymphoblastic leukemia, respectively (Novellino et al., 2005).
Finally, viral proteins produced by DNA viruses that stably integrate into the host cell’s
genome and are associated with certain malignancies can serve as TSAs. Viral proteins from
human papillomavirus (E6 and E7) expressed in cervical cancers, Epstein-Barr virus (EBNA-1)
expressed in Burkitt's lymphoma and nasopharyngeal-carcinomas, and Hepatitis B virus
expressed in hepatocellular carcinomas have all been found to be targeted by T cells in patients
with these cancers (de Vos van Steenwijk et al., 2010; Finn, 2008a; Fogg et al., 2009;
Schreiber, 2003). Although it would seem that viral antigens would instigate the most robust
anti-tumor T cell responses, surprisingly, viral-peptide-specific T cells are found to be largely
non-functional in cancer patients (de Vos van Steenwijk et al., 2010; Fogg et al., 2009). This
may be the result of viral mechanisms of immune escape during the history of the infection,
however, the inactivity of anti-tumor T cells is a characteristic of natural T cell responses to
nearly all cancers, regardless of the class of tumor antigens targeted.
Tumor-associated antigens
Tumor-associated antigens (TAAs) represent antigens that can be recognized by T cells
in patients with cancer, but are not exclusively expressed in tumors. The overwhelming data
29
from carcinogen-induced mouse tumors indicated that the antigens that cause tumor rejection
were almost always unique to the individual tumors (Schreiber, 2003). Therefore, it came as a
surprise when the first human tumor antigen identified in a patient with melanoma was the
melanoma antigen (MAGE-1), a TAA (van der Bruggen et al., 1991). This class of antigens is
grouped into four categories based primarily on their expression patterns in normal versus
tumor tissues.
MAGE-1 is now known as a cancer-testis antigen, a class of antigens that are normally
only expressed in the male germline or the placenta but that are often re-activated and
expressed in various tumors (Novellino et al., 2005). Cancer-testis antigens may have
similarities to TSAs because they are expressed only in germ cells – tissues that do not express
MHC molecules and are therefore not recognized by T cells. However, these antigens have also
been found to be expressed in the thymus, making T cells potentially reactive to these antigens
susceptible to mechanisms of central and peripheral tolerance, a caveat of all TAAs (Schietinger
et al., 2008). Similar to cancer-testis antigens, oncofetal antigens, such as carcinoembryonic
antigen (CEA), are normally expressed only in embryonic tissues but are also expressed in
many cancers. Differentiation antigens represent antigens that are expressed in a specific
lineage of cells that may also give rise to a tumor. These have mostly been characterized in
melanoma, and melanocyte-specific genes, such as Gp100, MART-1, and Tyrosinase,
exemplify this class of TAAs. Overexpressed antigens are ubiquitous antigens that are found to
have elevated expression levels in tumors. Often these antigens are directly associated with the
neoplastic state of the transformed cells (Finn, 2008b). Her2/Neu, p53, cyclin B1, and hTERT
are examples of these overexpressed antigens.
30
Non-conventional tumor antigens
In addition to TSAs and TAAs, a class of non-conventional, alternative tumor antigens
resulting from abnormal transcripts or post-translational modifications have been recently
defined. T cells have been found that recognize novel open reading frames initiated from nonclassical start codons in tumors (Schwab et al., 2003; Wang et al., 1996), and T cell clones have
been isolated from cancer patients that recognize post-translationally spliced proteins (Hanada
et al., 2004; Vigneron et al., 2004; Warren et al., 2006). These types of tumor antigens may
represent a new type of target for anti-tumor T cells, though it remains to be seen whether these
antigens are specific to cancer. In contrast to these examples, a unique TSA was recently
described in a mouse fibrosarcoma that resulted from a mutation disrupting the function of a
glycosyltransferase chaperone that altered the glycosylation of a protein and as a result,
generated a neoantigen (Schietinger et al., 2006). In addition, loss of function mutations in
multiple components of the machinery for generating peptides for presentation by MHC
molecules, such as TAP and ERAAP, have been shown to lead to the presentation of novel
immunogenic antigens that can be recognized by T cells in experimental tumor models
(Hammer et al., 2007; van Hall et al., 2006). Therefore, it is clear that there are many routes to
generating many types of tumor antigens in cancer.
T cell priming
Although the presence of tumor antigens is essential for T cell responses to cancer,
potentially more relevant is the context in which the antigens are first encountered by T cells.
The manner in which a naïve T cell first encounters its cognate antigen can directly influence a
diversity of T cell fates including death, proliferation, effector function, anergy, trafficking, and
memory development. The importance of proper priming for generating T cell responses against
31
tumor antigens is exemplified by the fact that in the absence of active vaccination against tumor
antigens expressed in tumors, many patients lack T cells specific to these antigens.
Exactly where and how T cells first encounter tumor antigens is an actively debated
topic. Some of the earliest experiments using multiple types of transplantable mouse tumors
indicated that T cells were primed directly by tumor cells after dissemination into lymph nodes
and as a result, T cells remain “ignorant” of tumors (unprimed) until the latter stages of disease
when tumor cells eventually metastasize to lymphoid organs (Ochsenbein et al., 1999;
Ochsenbein et al., 2001). These conclusions were supported by early studies using transgenic
mouse cancer models (RIP-Tag2) in which T cell responses to pancreatic β-cell tumors
expressing a model TAA (LCMV-GP) were absent without a viral vaccination against the TAA to
initiate a T cell response (Speiser et al., 1997). However, in subsequent, and more sensitive
studies, it was determined that with increasing tumor burden, T cells were stimulated by the
TAA in pancreatic tumors. More importantly, in this model, as well as other transplantable tumor
models, priming did not occur by the direct recognition of tumor cells by T cells, but by bone
marrow-derived antigen-presenting cells that had acquired the TAAs from the tumors and crosspresented them to T cells in the tumor draining lymph nodes (Nguyen et al., 2002; Spiotto et al.,
2002). Although tumor cells may directly prime naïve T cells in certain contexts, the prevailing
view is that bone marrow-derived dendritic cells are in fact responsible for the initial presentation
of tumor antigens to T cells in lymph nodes (Dranoff, 2004). As such, dendritic cells (DCs)
represent a critical link between tumors and T cells. DCs may ultimately be the master
regulators of the anti-tumor T cell response because they can dramatically alter the potency of
the T cell response depending on additional signals they supply to T cells during their initial
encounter with antigen.
32
Dendritic cells prime T cell responses
Just as T cells have evolved to respond to and eliminate foreign pathogens, DCs have
evolved to prime T cell responses specifically in response to foreign intruders. Therefore, it is
questionable whether DCs can induce productive T cell responses against tumors, and how
they can do so, when their purpose is to respond to infections. Indeed, this remains to be a
leading theory as to why anti-tumor T cell responses are so often ineffective; DCs do not
adequately prime T cells to respond to tumors or DCs may even induce the differentiation of
non-functional T cells specific for tumor antigens (Vicari et al., 2002).
A clear difference between antigens acquired from a tumor and antigens acquired from
the site of an infection is the presence of pathogen-associated materials and inflammation at
sites of infections. To recognize foreign pathogens, DCs and other innate immune cells have
evolved mechanisms to recognize various building blocks commonly used by many types of
pathogens, but absent from the host organism, called pathogen-associated molecular patterns
(PAMPs). PAMPs are recognized by a multitude of pattern recognition receptors (PRRs) on
DCs and the ligation of PRRs and PAMPs on DCs can provide signals that instruct DCs to
engulf antigens, migrate to lymph nodes, and fully prime T cells. The major class of PRRs are
the Toll-like receptors (TLRs) that recognize a range of pathogen materials such as sugars
(LPS) and certain nucleic acid structures such as unmethylated CpG DNA and double-stranded
RNA (Rakoff-Nahoum and Medzhitov, 2009). The absence of such molecules in the
microenvironment of tumors may limit inflammation and not allow for the proper maturation of
DCs presenting tumor antigens. It has long been appreciated that antigen presentation by DCs
in the absence of inflammation does not induce T cell responses (Jenkins and Schwartz, 1987).
The recognition of PAMPs by DCs induces the upregulation of co-stimulatory molecules (B7-1
and B7-2) on the surface of DCs and these provide a critical “second signal” to T cells (via
CD28) that synergizes with TCR signals to fully activate T cells. The importance of PAMPs in
33
inducing anti-tumor immune responses is supported by experiments that have shown enhanced
immunity to tumors by the administration of TLR ligands or by selective blockade of pathways
found to negatively regulate TLR signaling in DCs (Shen et al., 2004; Verdeil et al., 2008;
Whitmore et al., 2004).
However, it is clear that T cells respond to many cancers and therefore, must not require
PAMPs (Albert et al., 1998b; Preynat-Seauve et al., 2006). A breakthrough in our understanding
of how this occurs came with the discovery of alternative, non-pathogenic ligands that could
activate innate immune cells. These are collectively called danger-associated molecular
patterns (DAMPs) and represent a means to alert the immune system of damage, stress, or cell
death, traits common to many cancers, in the absence of infection (Gallucci et al., 1999;
Gallucci and Matzinger, 2001; Petrilli et al., 2007). DAMPs have now been shown to contribute
to DC activation and the induction of anti-tumor T cell responses against cancer in much the
same way PAMPs can during infections, often utilizing the same receptors used to recognize
microbial structures. During natural or chemotherapy-induced death of cancer cells, the
translocation of calreticulin and phosphatidylserine (PS) to the surface of tumor cells can signal
DCs to engulf the dying cells via the coordinated activities of scavenger receptors, such as
CD36, CD91, and the PS receptor, and integrin αvβ5 (Albert et al., 1998a; Basu et al., 2001;
Obeid et al., 2007). Uric acid released from dying cells can serve as a potent adjuvant, leading
to the upregulation of costimulatory molecules on DCs (Shi et al., 2003). Heat shock proteins
released or surface-exposed during tumor cell death have also been shown to induce the
maturation of DCs (Asea et al., 2002; Chen et al., 2009; Feng et al., 2002; Millar et al., 2003),
and DC recognition of high-mobility-group box 1 (HMGB1) released from dying tumor cells via
TLR-4 can promote antigen processing and presentation (Apetoh et al., 2007). Finally, ATP
shed during tumor cell death can induce an inflammatory response in the tumor
microenvironment by activating the NLRP3 inflammasome in DCs (Ghiringhelli et al., 2009).
34
Therefore, there are a host of signals generated in the tumor microenvironment that can
promote the functional maturation of DCs that acquire tumor antigens. Additionally,
chemotherapeutic drugs may provide an enticing means for enhancing therapeutic anti-tumor T
cell responses by magnifying these cues through increased tumor cell stress and death
(Zitvogel et al., 2008).
A role for CD4 T cell help
Although DCs are essential for inducing CD8 T cell responses, CD4 (helper) T cells can
play an important role in augmenting CD8 responses to pathogens and potentially cancers
(Rocha and Tanchot, 2004; Toes et al., 1999). Like CD8 T cells, CD4 T cells utilize the TCR to
recognize peptide antigens presented on major histocompatibility molecules. However, CD4 T
cells engage a different class of MHC, MHC class II, that presents a distinct pool of peptides
due to different characteristics of the MHC class II peptide-binding pocket. In addition, unlike
MHC class I which is expressed on nearly all nucleated cells of the host, MHC class II
expression is restricted to a select group of immune cells called professional antigen-presenting
cells – DCs, B cells, and macrophages. Due to the unique role of CD4 T cells in interacting
specifically with professional antigen-presenting cells, they are believed to be the key
orchestrators of adaptive immune responses, guiding and regulating the cellular (CTL) and
humoral (B cell) effector arms of the adaptive immune response.
CD4 T cells have been found to be critical to the quality of CD8 T cell responses for
quite some time, as a lack of CD4 T cells curtails or disrupts CD8 T cell responses in several
settings (Cardin et al., 1996; Keene and Forman, 1982). How CD4 T cells help CD8 responses
has since been narrowed primarily to the initial encounter of cross-presented antigens to CD8 T
cells by DCs (Bennett et al., 1997). Upon encounter of its cognate antigen on a DC, a CD4 T
cell can relay back to the DC a positive signal that boosts the activity of the DC to prime CD8 T
35
cell responses, a phenomenon called “licensing” (Lanzavecchia, 1998). This communication is
mediated by the expression of CD40L at the surface of CD4 T cells that interacts with CD40 on
DCs (Bennett et al., 1998; Ridge et al., 1998; Schoenberger et al., 1998). In response to self
antigens expressed in tissues, it was demonstrated that while CD4 T cell licensing of DCs via
the CD40/CD40L axis was required for CD8 proliferative responses, a second signal produced
by DCs, the cytokine IL-12, was required for the acquisition of CD8 effector function (Hernandez
et al., 2002). Although DC licensing by CD4 T cells has been shown in multiple situations, it is
clear that in many scenarios, CD8 T cells do not require CD4 T cell help for efficient priming
(Lanzavecchia, 1998). Instead, CD4 T cells have more recently been shown to play a critical
role in setting up the proper differentiation of CD8 T cells into memory cells (Janssen et al.,
2003; Shedlock and Shen, 2003; Sun and Bevan, 2003). Finally, CD4 T cells can even play a
role in helping CD8 T cells gain access (or home) to sites of infection (Nakanishi et al., 2009).
Thus, in particular contexts, CD4 T cells can promote seemingly all facets of the CD8 T cell
response.
Indeed, CD4 T facilitation of CD8 responses against tumors has also been described in
experimental models of cancer. Early experiments hinted that CD4 T cells were not efficiently
primed by transplanted tumors expressing their cognate antigen, leading to the development of
non-functioning, “anergic,” T cell responses (Staveley-O'Carroll et al., 1998). However, in the
same experimental model, it was shown that in vitro activation of CD4 T cells into TH1 type
effectors prior to their transfer could induce the rejection of tumors in a CD8 T cell-dependent
manner (Nishimura et al., 1999b). Likewise, in a TRAMP autochthonous model of prostate
cancer, naïve T cells were not adequately primed by the endogenous tumors. However, by
inducing massive tumor apoptosis with androgen ablation, CD4 T cells responded to tumor
antigens with a massive proliferative response and acquired effector functions (Drake et al.,
2005). Additional experiments in this model showed that effective anti-tumor immunity is
36
mediated by CD8 T cells, but that CD4 T cells were required to fully stimulate CD8 T cell
priming by activating DCs via the CD40/CD40L axis (Shafer-Weaver et al., 2009). In response
to endogenous pancreatic tumors in a RIP-Tag2 model, coordinate CD4 T cell responses were
demonstrated to help prime naïve CD8 T cells to develop effector function and proliferate in
response to antigen encounter (Lyman et al., 2004), as well as facilitate the accumulation of
greater numbers of activated CD8 T cells in the tumors (Wong et al., 2008).
T cell trafficking to tumors
In the final phase of detection, T cells activated in the lymph nodes that drain the tumor
site must home back to and recognize the tumors. This presents yet another obstacle because
as compared to sites of infection, the tumor microenvironment lacks foreign pathogens that can
initiate inflammatory and chemotactic cues that normally direct the recruitment of immune cells.
However, there is mounting evidence that cancers also may provide signals that recruit cells of
the immune system.
The activity of many oncogenes has been found to induce the expression of
inflammatory molecules. Oncogenic K-ras mutations are among the most common genetic
lesions in cancer. Experiments in human cell lines and autochthonous mouse models of lung
cancer have shown that signaling by oncogenic K-ras directly activates the expression of the
orthologous chemokines IL-8 and ENA-78 in humans or CXCL1 (KC), CXCL2 (MIP-2), and LIX
(CXCL5) in mice (Ji et al., 2006; Sparmann and Bar-Sagi, 2004). Consequently, in vivo
experiments showed that the expression of these cytokines in K-ras-expressing tumors led to
the recruitment of immune cells (Sparmann and Bar-Sagi, 2004). Expression of the Src
oncoprotein has been shown to activate the NF-κB pathway and lead to increased expression of
the proinflammatory cytokine, IL-6 (Iliopoulos et al., 2009). In a similar fashion, the expression of
the RET/PTC3 fusion protein, commonly found in thyroid cancers, could activate NF-κB as well,
37
but its expression was associated with the upregulation of the chemokine MCP1 and the
cytokine GM-CSF (Russell et al., 2004). Tumor cell expression of CCL21, a chemokine normally
associated with immune cell homing to lymphoid organs, is a common feature of several
cancers and can lead to the recruitment of immune cells in tumors. However, expression of
CCL21 has been associated with poor prognosis in patients with ovarian cancer and poor
outcomes in mouse models of melanoma due to the preferential recruitment of immune
regulatory cell populations (Curiel et al., 2004; Shields et al., 2010). From these studies it is
clear that even the earliest steps of tumorigenesis – cellular transformation by oncogenes – may
initiate a proinflammatory network of molecules that alerts the immune system and brings about
the recruitment of immune cells to tumors.
As tumors progress to more invasive stages of disease, it is speculated that disruptions
to the normal architecture of organs (penetration of the basement membrane), tissue damage,
and tumor cell death can all trigger the release of proinflammatory signals that beckon immune
cells to cancers (Pardoll, 2003). Molecular cues may also alert immune cells to the stresses
tumor cells undergo during their progression. The process of tumorigenesis itself has been
shown to cause the upregulation of stress ligands, like the Rae-1 family and H60, in murine
tumors due to the activation of DNA damage response (DDR) pathways in cancer cells (Gasser
et al., 2005). Indeed, genomic instability and active DDR signaling are common features of most
cancers (Jackson and Bartek, 2009). The homologous stress ligands MIC-A/B in human cells
have also been found to be highly expressed on various forms of human cancers (Groh et al.,
1999). By engaging the NKG2D receptor on NK cells and T cells, these stress ligands can serve
as instigators of anti-tumor immune responses by providing a complementary means (to TCR
engagement) for immune cell detection of cancer cells in the tumor microenvironment (Raulet
and Guerra, 2009).
38
The capacity of CTLs to efficiently extravasate from blood vessels may also affect their
ability to home to tumors. Extravasation from the blood vessels into tumors may be regulated by
endothelin receptors. Low expression of endothelin B receptor (ETBR) on endothelial cells in
ovarian cancers was associated with higher amounts of infiltrating T cells and ET BR antagonists
improved CTL homing to tumors in transplantable mouse models of ovarian cancer
(Buckanovich et al., 2008). The functionality of anti-tumor CTLs may also affect their ability to
extravasate from the blood. TNF-α production from CD8 T cells was shown to be critical for their
recruitment to mouse pancreatic tumors by mediating TNFR1-dependent upregulation of the
adhesion molecule VCAM-1 on pancreatic endothelial cells, allowing for CTL extravasation into
the tumor bed (Calzascia et al., 2007).
T cell recognition of the peptide/MHC molecules on tumor cells to which they were
originally primed is the final requirement for tumor detection. Two-photon microscopy studies
that followed CTL migration into transplanted mouse tumors, in real time, found that the
recognition of cognate antigen in tumors was the critical determinant for CTL recruitment in
tumors, arguing against the importance of chemokine gradients for tumor infiltration by CTLs
(Boissonnas et al., 2007; Mrass et al., 2006). In addition, the functionality of the CTLs may be
critical for tumor cell detection. IFN-γ is known to upregulate the expression of MHC molecules
on the surface of cells, and the capacity for T cells to secrete IFN-γ in the tumor
microenvironment may be necessary for directly detecting the peptide/MHC molecules derived
from tumor antigens on the surface of tumor cells (Riond et al., 2009). The importance of IFN-γ
to CTL detection is supported by the fact that tumors expressing a dominant negative IFN-γ
receptor are resistant to rejection after transplantation (Dighe et al., 1994).
39
B. Regulation
After T cells are primed in the tumor-draining lymph nodes, they migrate into the tumor
and interact directly with tumor cells via TCR engagement of peptide/MHC molecules presenting
tumor antigens. Here CTLs can kill tumor cells, providing the major mechanism for T celldependent regulation of cancer. The ways in which T cells may kill tumor cells upon
engagement is no different than responses during the engagement and killing of pathogeninfected cells. T cell cytotoxicity is mediated by either of two pathways, the perforin/granzyme
pathway or the FasL/Fas pathway (Kagi et al., 1994b; Lowin et al., 1994b). Both pathways
ultimately converge through the activation of caspases that induce apoptosis in targeted cells.
Perforin and granzymes are contained within lytic granules in CTLs and they are released into a
synaptic cleft that forms between T cells and target cells as a consequence of TCR
engagement. Perforin polymerizes to form pores in target cells, causing membrane
destabilization as well as supporting the entry of serine proteases called granzymes. The
granzymes, particularly granzyme B, cleave several substrates, including multiple caspases,
initiating a pro-apoptotic cascade in target cells. T cells from mice deficient for perforin by gene
targeting have greatly diminished cytotoxic activity in vitro and in vivo (Kagi et al., 1994a; Lowin
et al., 1994a). CTLs also induce apoptosis of target cells by engaging the Fas receptor (CD95 or
Fas/APO-1) with Fas ligand (FasL). FasL on T cells can activate Fas on target cells, leading to
the recruitment and activation of caspase-8 through the Fas-associated death domain of Fas,
initiating apoptosis. The importance of these two pathways in mediating tumor cell death and
cancer regulation has been highlighted in various experimental cancer models.
Direct killing of tumor cells
Indeed, the first report of perforin knockout mice included experiments that showed a
decreased capacity to control the growth of transplanted fibrosarcomas (Kagi et al., 1994a).
40
Subsequent work found similar results with a multitude of cell lines from multiple origins and
also showed that tumor induction with the carcinogen methylcholanthrene (MCA) was enhanced
in perforin-deficient mice (though apparently in a CTL-independent fashion) (van den Broek et
al., 1996). Perforin-deficient mice also have an increased incidence of spontaneous
disseminated lymphomas in p53 wild-type as well as p53 heterozygous backgrounds and tumor
protection appears to be CD8 T cell-dependent (Smyth et al., 2001). The importance of
granzymes in mediating tumor cell killing was demonstrated when the overexpression of a
serpin (SPI-6, an inhibitor of granzymes) in a T cell lymphoma protected the transplanted
tumors from CTL-mediated killing (Medema et al., 2001). In a similar experimental fashion, the
overexpression of cFLIP, an inhibitor of Fas-mediated cell death, in tumor cells lines allowed the
outgrowth of tumors that were normally rejected by CTLs (Medema et al., 1999). Two-photon
imaging of CTL-mediated tumor regression has been recently documented in vivo (Boissonnas
et al., 2007; Breart et al., 2008; Mrass et al., 2006). In these studies, investigators could
visualize that direct contacts between individual CTLs and tumor cells immediately preceded
tumor cell apoptosis (Breart et al., 2008; Mrass et al., 2006). Therefore, CTLs can regulate
cancer by inducing apoptosis in tumor cells after making direct cell-cell contacts.
Cytokine-mediated cancer regulation
Complementing these direct pathways of inducing target cell death, cytokine release by
T cells upon target cell recognition is also a critical effector function of T cells. Cytokines
released by T cells may orchestrate the responses of other immune cells, induce the expression
of adhesion molecules on target cells, alter the expression or activity of proteins responsible for
antigen processing and presentation on target cells, or even affect the proliferative capacity of
target cells. The production and secretion of two key effector cytokines, IFN-γ and TNF-α, have
been shown to be important in the regulation of cancer, though the mechanisms of action are
41
not as clear. IFN-γ receptor-deficient mice (IFNGR1-/-) develop MCA-induced sarcomas with an
increased penetrance and decreased latency compared to wild-type mice and develop
spontaneous tumors with decreased latency in a p53-deficient background (Kaplan et al., 1998).
IFNGR1 expression was found to be critical in the tumor cells themselves and restoration of the
receptor in tumor cell lines made the tumors susceptible to rejection by adaptive immune cells,
presumably T cells, in an IFN-γ-dependent manner (Kaplan et al., 1998). Spontaneous
development of lymphomas was also increased in IFN-γ-deficient mice, although the critical
cells lacking IFN-γ that contribute to the increased cancer predisposition (tumor cells or T cells)
could not be determined (Street et al., 2002). Although TNF-α-deficient mice do not have an
increased incidence of lymphomagenesis (Street et al., 2002), and TNF-α deficiency in mice
has been shown to actually reduce the experimental induction of skin cancer (Moore et al.,
1999), the incidence of MCA-induced sarcomas was enhanced in the absence of TNF-α (Swann
et al., 2008). More importantly, studies in which TNF-α deficiency was targeted specifically to
the anti-tumor CTLs showed that TNF-α production from CTLs is essential for the regulation of
endogenous pancreatic tumors (Calzascia et al., 2007). In a transplantable fibrosarcoma model,
secretion of IFN-γ and TNF-α from anti-tumor CTLs that were activated and transferred into
tumor-bearing mice was necessary to cause the complete regression of established tumors
(Zhang et al., 2008a). In addition, production of IFN-γ and TNF-α by CD4 T cells was shown to
regulate tumor proliferation by inhibiting tumor angiogenesis in a transplantable sarcoma model
as well as an endogenous pancreatic cancer model (RIP-Tag2) (Muller-Hermelink et al., 2008;
Qin and Blankenstein, 2000). Therefore, effector cytokines also provide a critical means of
regulating cancer.
42
NKG2D receptor mediated anti-tumor T cell activity
In addition to its role in detecting tumors, the NKG2D receptor can also act to further
stimulate the activity of the anti-tumor T cell response. Enforced expression of NKG2D ligands
on tumor cells led to enhanced CTL-mediated cytotoxicity in vitro and tumor rejection in vivo
(Diefenbach et al., 2001). Treatment of mice with antibodies that block NKG2D interactions with
its ligands enhanced the penetrance of MCA-induced sarcomagenesis in a manner that was
dependent on adaptive immune cells, implying that NKG2D activity is critical in controlling the
initiation of cancer (Smyth et al., 2005). However, NKG2D-deficient mice had a reduced
capacity to control the progression of cancer in prostate (TRAMP) and lymphoma (Eµ-myc)
models, indicating that the activity of NKG2D on T cells may also influence their capacity to
regulate cancer progression (Guerra et al., 2008).
C. Evasion
The fact that cancer is a leading cause of morbidity worldwide indicates that the immune
response must ultimately be ineffective in preventing the incidence or progression of this
disease. Therefore, if a T cell response against a tumor is induced, ultimately mechanisms must
arise that allow the tumor to evade this response. Two competing, but not mutually exclusive,
theories to account for the evasion of T cell responses by tumors have been postulated. The
first, called cancer immunoediting, hypothesizes that just as tumors evolve to bypass intrinsic
cellular pathways that inhibit tumorigenesis, tumors also evolve to bypass extrinsic mechanisms
of tumor suppression, such as anti-tumor immune responses, during the course of their
development. Therefore, tumors are “shaped” by anti-tumor immune responses and this can be
exemplified by identifying changes in tumor cells that make them less susceptible to immune
43
recognition and destruction. The second hypothesis, called immune tolerance, asserts that T
cell responses to cancer are likely to be fundamentally different than responses to acute
pathogenic infections which the immune system has evolved to resist (Restifo et al., 2002). As a
consequence, immune responses against cancer are incomplete and ultimately result in the
immune tolerance of the cancer. Similarities in immune phenotypes in response to cancer and
chronic infections have been described (Pardoll, 2002). It is possible that immune tolerance to
cancer results from the induction of immune regulatory pathways that have evolved to prevent
autoimmune disease in the setting of chronic immune responses toward self tissues or
persistent disease.
i. Cancer Immunoediting
The term immunoediting, although described indirectly for many years, was only recently
coined by Robert Schreiber and colleagues in 2001 (Shankaran et al., 2001). Evidence of
immunoediting in human and mouse cancers can be measured by the occurrence of disruptions
in nearly all of the critical parameters previously described in this section. Seemingly every facet
of a productive immune response to cancer, from the expression of T cell-targeted tumor
antigens, to the proper priming, homing and effector function of anti-tumor T cells, have been
found to be subverted in one way or another by various cancers. Although there is a wealth of
evidence for cancer immunoediting in response to multiple immune cell populations, I will only
focus on immunoediting in response to anti-tumor T cell responses here. For the purposes of
organization, I have classified examples of immunoediting into two forms: those that are
recessive, arising passively by immunoselection and those that act in a dominant fashion,
actively suppressing T cell responses.
44
Immunoediting by passive selection
The clearest example of cancer immunoediting is the selective outgrowth of tumors that
have lost the expression of antigen(s) targeted by T cells, thereby preventing T cell recognition
of the tumor (Khong and Restifo, 2002). This has been described extensively in mouse cancer
models and also in the context of active immunotherapy against specific antigens in cancer
patients. The first descriptions of antigen loss as a mechanism by which tumors could be
selected that were resistant to rejection came from the laboratories of Thierry Boon and Hans
Schreiber (Urban et al., 1982; Uyttenhove et al., 1983). Using different types of transplantable
tumors, P815 mastocytoma or UV-induced fibrosarcoma, both groups showed that tumors
capable of growing out after transplantation into syngeneic mice had selectively lost the
antigenic determinants required for CTL killing and showed that the “antigen-loss variants” had
increased tumorigenicity upon secondary transplantation. More recent studies have indicated
that reductions in target antigen expression in tumors in combination with decreased T cell
activity may be sufficient to allow transplanted renal cell tumors to escape (Zhou et al., 2004).
Thus, combinations of immunoediting mechanisms may be involved in tumor evasion. In
melanoma cancer patients receiving immunotherapy, specific reductions in the expression of
the targeted antigens have been noted, providing evidence that immune selection of antigenloss variants also occurs in cancer patients (Jager et al., 1996; Riker et al., 1999).
Thus some caution must be lent to the hope that T cell-based immunotherapy can
ultimately cure cancer. However, it is possible that by targeting multiple tumor antigens
simultaneously, the likelihood of tumor escape via antigen-loss may be dramatically reduced
(Schreiber, 2003). Alternatively, recent work has shown that tumor antigen-loss variants can still
be destroyed indirectly by anti-tumor T cell responses (Spiotto et al., 2004; Zhang et al., 2007;
Zhang et al., 2008b). T cell-mediated destruction of stromal cells that cross-present the tumor
antigens in the tumor microenvironment could also lead to tumor eradication (Spiotto et al.,
45
2004). In Chapters 2 and 3 of this thesis, I will discuss the role of antigen loss as a mechanism
of tumor evasion and provide data to support the contention that the type of cancer, and the
model used – autochthonous or transplantable, can greatly influence whether antigen-loss
occurs in the tumors.
Similar to the selective loss of antigens targeting T cell responses to tumors, selective
loss of the machinery required to present the peptide antigens at the surface of tumor cells
would also impede T cell recognition of cancer cells. Early experiments using the transplantation
of UV-induced tumors, however, failed to demonstrate a loss of surface expression of the MHC
proteins on progressively growing tumors (Ward et al., 1990). Furthermore, mice deficient in
antigen processing by targeted mutation of TAP-I (transporter associated with antigen
processing) or LMP-2 (a component of the 20S proteasome) were not found to be more
susceptible to spontaneous tumor development in the context of p53 deficiency (Johnsen et al.,
2001). Despite these examples in mice, there are numerous pieces of data from human cancers
that suggest a selective pressure is in place that drives the outgrowth of cancers with defects in
antigen presentation (Algarra et al., 2004; Chang and Ferrone, 2007; Khong and Restifo, 2002;
Pardoll, 2003; Restifo et al., 1993).
Reduced surface expression of MHC class I alleles is a common feature of multiple
cancers of epithelial origin (Hicklin et al., 1998; Maeurer et al., 1996b; Restifo et al., 1993;
Sanda et al., 1995). Furthermore, reductions in MHCI on the surface of tumors can often be
correlated with the progression of the disease, with metastatic tumors having the lowest levels
of MHCI expression (Cromme et al., 1994b; Maeurer et al., 1996a; Vitale et al., 1998). The
ability of many cancer cell lines to re-express MHCI upon manipulation with cytokines such as
IFN-γ in vitro indicates that MHCI down modulation may frequently be epigenetic in nature
(Khong and Restifo, 2002; Pardoll, 2003; Raffaghello et al., 2005). However, there are also
numerous examples of mutations that permanently inactivate antigen processing or
46
presentation in human cancers, such as disruption of β2 microglobulin a common component
required for the stability of all MHCI alleles (Chang et al., 2006; Chen et al., 1996a; Chen et al.,
1996b; Cromme et al., 1994a; Hicklin et al., 1998). A study of human neuroblastoma tumor
samples examined multiple components of the antigen-processing and presentation machinery
using mRNA expression and a panel of antibodies to multiple different components. They
reported that defects in different components accounted for disrupted antigen presentation in
each tumor (Raffaghello et al., 2005). Evidence that the modifications in antigen presentation
identified in human cancers are in fact mechanisms providing for tumor evasion of T cell
responses can only be correlative, but it is intriguing that multiple studies have found that
disruptions in antigen processing and presentation in many different forms of cancer are
associated with poorer prognoses for patients (Anichini et al., 2006; Leys et al., 2007; Ogino et
al., 2006).
Immunoediting by active mechanisms of immune suppression
By definition, all edited tumors arise as a consequence of selective forces imposed by
the immune system. However, in this form of immunoediting, cancers are selected that have
acquired a new capacity to thwart the anti-tumor T cell response rather than become inert or
invisible to responding T cells. A tremendous wealth of research has been aimed at identifying
mechanisms by which tumors evade T cell responses by inducing immune suppression, far too
many to examine all of them here (for a more complete review see: (Rabinovich et al., 2007)).
Therefore I will limit this section to only a few mechanisms of tumor-induced immune
suppression that highlight the diversity of ways tumors can evolve to evade T cell responses.
47
Altered expression of immune modulatory ligands
Costimulatory receptors expressed on T cells are likely critical for the efficient execution
of effector function upon encounter of tumor cells. However, T cells also express inhibitory
receptors that act to moderate T cell responses and prevent the occurrence of autoimmune
reactions. A key mechanism of tumor evasion of T cell responses is the co-opted expression of
ligands for these inhibitory receptors (Zou and Chen, 2008). The first example of tumor evasion
by over-expressing a ligand recognized by T cells was paradoxically the MIC-A/B ligands.
Indeed, these are stimulatory ligands for the NKG2D receptor, but in patients with various forms
of cancer that over-expressed MIC-A/B, these ligands were found to be shed from the tumors
(Groh et al., 2002). Soluble MIC-A/B was discovered to bind NKG2D and cause its
internalization, thus reducing NKG2D expression on the surface of anti-tumor T cells and
presumably decreasing the CTLs capacity to kill tumor cells. Further work from Spies and
colleagues has provided evidence that soluble MIC-A/B may also serve to recruit a unique
population of CD4+ T cells that have immunosuppressive activities (Groh et al., 2006).
The best characterized over-expressed T cell inhibitory ligand on cancers it B7-H1, or
programmed cell death ligand 1 (PD-L1). PD-L1 expression on tumor cells can engage PD-1
expressed on T cells to transmit inhibitory signals that disrupt the anti-tumor CTL effector
function (Blank et al., 2004). In the original work implicating B7-H1 in tumor evasion, it was
demonstrated that B7-H1 over-expression on tumor cells could direct the apoptosis of
responding CTLs (Dong et al., 2002). Subsequent work has shown that B7-H1 may also serve
as an autocrine signal in tumor cells to promote cell survival and resistance to T cell-mediated
cytotoxicity (Azuma et al., 2008). In any case, common oncogenic mutations have been shown
to directly feed into the upregulation of B7-H1 on tumors (Marzec et al., 2008; Parsa et al.,
2007), and the over-expression of B7-H1 is associated with poor prognosis in patients with renal
48
cell carcinoma as well as many other forms of cancer (Thompson et al., 2006; Zou and Chen,
2008).
The inhibitory receptor B and T lymphocyte attenuator (BTLA) expressed on T cells can
also be engaged by tumor cell expression of its ligands. Although there is some disagreement
about the ligands for BTLA (Sedy et al., 2005; Watanabe et al., 2003), both predicted ligands
B7x and HVEM are found to be upregulated on various cancers, and BTLA was found highly
expressed on melanoma-antigen-specific CTLs in melanoma patients (Derre et al., 2010; Zang
et al., 2003). In addition, B7x expression on prostate cancers correlates with a poor prognosis
(Zang et al., 2007), and like the MIC-A/B ligands, B7x has been found at high concentrations in
the serum of renal cell carcinoma patients (Thompson et al., 2008).
Secretion of immunosuppressive molecules
A characteristic feature of cancer is the mass production and secretion of transforming
growth factor-β (TGFβ) (Flavell et al., 2010; Siegel and Massague, 2003). TGFβ was originally
thought to impede cancer progression by binding TGFβ receptors on tumor cells and signaling
growth inhibitory signals to tumor cells via activation of the SMAD transcription factors (Siegel
and Massague, 2003). This hypothesis was confirmed by the frequent detection of inactivating
mutations in TGFβ receptors in human colon carcinomas or deletions of SMAD4/DPC4 in
pancreatic carcinomas (Hahn et al., 1996; Markowitz et al., 1995). However, TGFβ signals are
now recognized to have pleiotropic effects in cancer progression, potentially acting to suppress
tumorigenesis early, and then acting to promote the process later through inhibition of antitumor immune responses (Siegel and Massague, 2003; Yang and Moses, 2008). TGFβ
production by tumor cells has been shown to inhibit CTL effector function by inducing SMAD
transcription factors to bind and inhibit the expression of multiple genes involved in cytolytic
activity in CTLs such as granzyme A, granzyme B, Fas ligand (FasL), and IFN-γ (Thomas and
49
Massague, 2005). Furthermore, CTLs recovered after peptide vaccination of melanoma patients
had reduced capacity to secrete IFN-γ and TNF-α or upregulate the expression of the T-bet
transcription factor (a marker of effector function) by the addition of TGFβ in vitro (Ahmadzadeh
and Rosenberg, 2005). Acting as a growth inhibitor, tumor-derived TGFβ can also reduce the
proliferative capacity of anti-tumor CTLs in mouse models (Gorelik and Flavell, 2001). Most
importantly, inhibition of TGFβ signaling specifically in tumor-reactive CTLs can lead to tumor
protection or eradication in transplantable mouse cancer models (Gorelik and Flavell, 2001;
Zhang et al., 2005a).
In addition to TGFβ, tumor production and/or secretion of multiple other proteins or
metabolites are associated with tumor-induced immune suppression. Tumors produce vascular
endothelial growth factor (VEGF) and prostaglandin E2 (PGE2), generated in excess by
heightened expression of the enzyme cyclooxygenase-2 (COX-2), both of which can inhibit APC
maturation (Khong and Restifo, 2002; Zou, 2005). Tumor production of the enzyme indoleamine
2,3-dioxygenase (IDO) can lead to the depletion of tryptophan in the tumor microenvironment,
reducing the proliferative capacity of anti-tumor CTLs or potentially inducing apoptosis
(Uyttenhove et al., 2003). Interestingly, blockade of IDO in transplantable mouse cancer models
using competitive inhibitors or RNA interference could lead to tumor regression (Uyttenhove et
al., 2003; Zheng et al., 2006). Thus there are many examples from mouse models and cancer
patients that indicate tumor-derived molecules are produced as a means of tumor evasion.
Recruitment of immunoregulatory cells
Tumor-induced recruitment of immunoregulatory cells is now considered to be the critical
mechanism underlying tumor evasion in multiple settings (Nagaraj and Gabrilovich, 2008;
Rabinovich et al., 2007; Zou, 2006). T regulatory cells (Tregs), defined by the surface expression
of CD4 and CD25 and the lineage-specific transcription factor Foxp3, may be the most critical
50
immunoregulatory cell population recruited to the tumor microenvironment to suppress antitumor CTL responses from their initial priming in the DLN to their effector function upon
encounter with tumor cells (Zou, 2006). Even before Tregs were implicated in human cancer, the
mere discovery of Tregs led to experiments that showed that the specific depletion of these cells
could promote the rejection of transplanted tumors (Sakaguchi et al., 1995; Shimizu et al.,
1999). Since then, numerous studies have found that Tregs are specifically recruited to the tumor
microenvironment in patients with various forms of cancer and that their presence is associated
with poor outcomes (Curiel et al., 2004; DeLong et al., 2005; Liyanage et al., 2002; Viguier et
al., 2004; Woo et al., 2001). Recruitment is likely a direct consequence of chemokine
production, such as CCL21 or CCL22, in tumor cells that preferentially attract these
immunosuppressive cells (Curiel et al., 2004; Shields et al., 2010). Tregs from cancer patients
have been demonstrated to potently inhibit the proliferation of tumor-infiltrating lymphocytes
(TILs) in an antigen-specific manner (Bonertz et al., 2009; Wang et al., 2004; Woo et al., 2002).
In mouse models of cancer, Tregs have also been shown to block the priming phase of CD8 T
cell responses (Ercolini et al., 2005). In addition to blocking T cell activation, Tregs can block the
effector function of TILs (Chen et al., 2005). In one particular study, Tregs were shown to prevent
the degranulation of lytic vesicles, thus blocking the final stage of T cell-mediated cytotoxicity
(Mempel et al., 2006). Additional mechanisms by which Tregs can mediate the suppression of
anti-tumor CTLs include TGFβ secretion, direct cell-cell contacts, and competition for growth
signals such as IL-2 (Zou, 2006).
Myeloid-derived suppressor cells (MDSCs), defined as Gr-1+CD11b+ cells in mice,
represent a diverse population of antigen-presenting cells (APCs) that like Tregs, are T cellsuppressive in nature and often found in association with tumors (Nagaraj and Gabrilovich,
2008; Zou, 2005). Although MDSCs are APCs, they seem to inhibit anti-tumor CTL responses in
an indirect manner. MDSCs have been shown to interfere with the initial priming of T cells in
51
tumor-draining lymph nodes (DLNs) by the secretion of TGFβ which can directly inhibit the
proliferation of CD8 T cells or indirectly inhibit the CD8 response by promoting the proliferation
of Tregs (Ghiringhelli et al., 2005; Yang and Moses, 2008). MDSC production of nitric oxide (NO),
which can disrupt IL-2 receptor signaling, and IFN-γ production can also suppress T cell
proliferation in tumor DLNs (Kusmartsev et al., 2000; Mazzoni et al., 2002; Movahedi et al.,
2008). MDSCs in the tumor bed can suppress the cytolytic activity of T cells by producing TGFβ
(Terabe et al., 2003) or by disrupting the capacity of the TCR-CD8 complex to interact with
peptide/MHC on tumor cells (Nagaraj et al., 2007). Interestingly, by producing reactive oxygen
species and peroxynitrite that induce the nitration of tyrosines in the TCR complex, MDSCs
were found to impede TCR engagement of peptide/MHC (Nagaraj et al., 2007). Finally, it is
important to note that MDSCs have been shown to also directly enhance tumor progression by
promoting tumor vascularization and metastasis (Yang et al., 2004; Yang et al., 2008). As of yet,
it is unclear what signals drive the accumulation of MDSCs in tumors and tumor DLNs, but
experiments ex vivo with tumor cell lines at least suggest that tumor-derived factors can be
secreted that induce the suppressive phenotype of MDSCs (Ghiringhelli et al., 2005).
Inhibition of immune cell migration in tumors
Limited T cell access to tumors has long been considered a mechanism by which tumors
escape CTL responses. Experimental systems in mice indicate that promoting T cell migration
into tumors artificially by the enforced expression of the TNF family member LIGHT, can
dramatically improve anti-tumor CTL responses to transplanted tumors (Yu et al., 2004).
Although potentially useful for immunotherapeutic strategies, such experiments do not indicate
that tumors actively engage in mechanisms to block the migration of immune cells to evade T
cell responses. However, as mentioned previously while describing T cell detection of tumors,
the expression of endothelin B receptor (ETBR) on endothelial cells in human ovarian cancers
52
was found to correlate with reduced T cell infiltration into tumors and reduced patient survival
(Buckanovich et al., 2008). It is speculated that ETBR may be upregulated in the tumor
vasculature as a direct consequence of over-expression of the ligand, endothelin-1 (ET-1)
(Buckanovich et al., 2008). Indeed, endothelins have been found to be overexpressed in
multiple cancers (Nelson et al., 2003). Alternatively, tumor expression of liver X receptor (LXR)
ligands can prevent the migration of dendritic cells (DCs) out of the tumor site by causing the
down-modulation of the chemokine receptor CCR7 required for lymph node migration
(Villablanca et al., 2010). Thus, tumors evaded the immune system by preventing the migration
of DCs to lymph nodes where they could initiate an anti-tumor T cell response.
ii. Immune tolerance of cancer
The idea that tumor evasion of T cell responses occurs as a result of the selective
evolution of tumors that can hide from or suppress T cell responses is not universally accepted
(Restifo et al., 2002). Indeed a large body of literature supports the idea that T cell responses to
cancer are incomplete or wither away as a result of natural immune regulatory networks that are
in place to prevent inappropriate immune responses that are harmful to the host. Thus T cell
responses against cancer may become inactivated by mechanisms that normally prevent
autoimmune disease. In this light, it may not be surprising that T cell responses against cancer
have many features in common with T cell responses against pathogens during chronic
infections (Kim and Ahmed, 2010; Pardoll, 2002). In both cases, a persistent and prolonged
immune response may trigger the induction of tolerance (or exhaustion) as a means to prevent
what may be “perceived” as autoimmune reactions. This may be why successful
immunotherapy in cancer patients is so often associated with autoimmune side effects (Caspi,
2008). Here I will briefly discuss some evidence to support the contention that autoregulation of
anti-tumor immune responses provides an alternative explanation for tumor evasion.
53
Importantly, autoregulatory mechanisms may complement, rather than exclude, tumor-intrinsic
mechanisms of immune evasion.
Tolerance phenotypes are characteristic of anti-tumor T cells
In response to multiple forms of cancer in humans as well as autochthonous mouse
cancers, tumor-specific T cells commonly exhibit features associated with tolerance or
exhaustion (Anderson et al., 2007; Bai et al., 2008; Cheung et al., 2008; de Vos van Steenwijk
et al., 2010; Inokuma et al., 2007; Kuss et al., 2003; Lyman et al., 2004; Roberts et al., 2009;
Savage et al., 2008; Willimsky and Blankenstein, 2005). These characteristics most often
include a decreased capacity to produce effector cytokines, such as IFN-γ, TNF-α, or IL-2, and
a reduced ability to kill target cells in vivo or in vitro. Such characteristics are identical to those
of viral-specific T cells during chronic infections (Wherry and Ahmed, 2004; Wherry et al., 2003).
Decreased functionality is often associated with high surface expression of the PD-1 inhibitory
receptor on T cells, considered to be a bona fide marker of exhausted T cells (Bai et al., 2008;
Barber et al., 2006; Day et al., 2006; Wherry et al., 2007). Importantly, PD-1, along with CTLA-4
and LAG-3, which are also inhibitory receptors upregulated on tumor-specific and chronic viralspecific T cells, appear to play functional, albeit distinct roles, in maintaining the tolerant state of
the T cells. This is supported by experiments in which antibody blockade of receptor-ligand
interactions greatly restored CTL activity in the context of chronic viral infections or cancer
(Barber et al., 2006; Freeman et al., 2006; Grosso et al., 2009; Grosso et al., 2007; Kaufmann
and Walker, 2009; Liakou et al., 2008). Nevertheless, these inhibitory pathways did not evolve
to suppress potentially beneficial T cell responses against viruses and cancers; they most likely
evolved as safety mechanisms to prevent autoimmune reactions. This is exemplified by the fact
that mice with targeted disruptions in these receptors (CTLA-4 or PD-1), or in some cases their
ligands (PD-L1/2), all have various forms of autoimmune disease (Keir et al., 2006; Nishimura et
54
al., 1999a; Waterhouse et al., 1995). These autoimmune reactions are triggered by the
breakdown of pathways mediated by these receptors that establish or maintain the peripheral
tolerance of T cells, i.e. after development in the thymus (Fife et al., 2006; Martin-Orozco et al.,
2006; Perez et al., 1997). Therefore, many of the inhibitory pathways engaged on most tumorreactive T cells are the same pathways that are critical for maintaining tolerance to host tissues.
Chronic antigen exposure drives T cell tolerance
It is essential to the host that the immune system properly regulates the induction of
immunity or tolerance; therefore, specific signals must be responsible for instructing T cells to
switch to a tolerant state in response to cancer. One potential mechanism to explain tolerance
induction in both cancer and chronic infections is prolonged exposure to antigen (Kim and
Ahmed, 2010). Persistent antigen exposure has long been known to induce tolerance or
exhaustion in reactive T cells (Redmond and Sherman, 2005). Recent work utilizing two
different models of chronic viral infection in mice showed that persistent exposure to antigen
was sufficient to drive T cell exhaustion (Bucks et al., 2009; Mueller and Ahmed, 2009).
Therefore, T cell responses to cancer may be influenced by immune regulatory pathways that
have evolved to terminate T cell responses to antigens that persist as a means to prevent
autoimmune disease.
IV. TUMOR IMMUNE SURVEILLANCE IN CANCER PATIENTS
Evidence presented in the previous section provides overwhelmingly support for the
contention that T cells can recognize and respond to many forms of human cancer. The fact that
human cancers often have reduced expression (or mutations) in proteins related to antigen
55
processing and presentation provides some coincidental evidence that human cancers may be
under the selective pressure of T cell responses. However, the role of anti-tumor T cells in
regulating cancers in these patients has not been fully addressed. Several lines of investigation
point to a clear role of tumor immune surveillance in the human population by: (1) examining the
incidence of cancer in genetically or pharmacologically immunosuppressed individuals, (2)
looking at the degree of T cell reactivity (and T cell phenotypes) in individual cancers and
correlating the T cell response to patient outcome, and (3) interpreting the results of multiple
clinical trials utilizing various forms of immunotherapy to treat cancer.
Cancer in genetically or pharmacologically immunosuppressed patients
Just as the importance of immune surveillance against cancer was revealed in various
immune-compromised mice, individuals with inherited immune deficiencies or patients receiving
immunosuppressive drugs to tolerate organ transplants can provide evidence to support tumor
immune surveillance in humans. However, analyzing sporadic tumor formation in people with
immunodeficiencies presents several important complications. First, it is important to consider
that most animal studies providing evidence for tumor immune surveillance used high doses of
carcinogens to rapidly provoke a very high incidence of cancer, a phenomenon that occurs
rarely in the human population (see review: (Dunn et al., 2006)). In addition, the few animal
studies that did examine spontaneous tumor development were often performed in a tumorprone background such as p53 deficiency, a characteristic uncommon to the majority of human
cancers (Dunn et al., 2006). Therefore, the evidence from mouse models of increased
spontaneous tumor development with immune deficiency is sparse. Second, in the few cases
where enhanced tumorigenesis was found in immune-compromised mice, it was necessary to
age the animals for extended periods of time to detect spontaneous cancers (Shankaran et al.,
2001; Street et al., 2002). Indeed, spontaneous cancers typically arise late in life, making similar
56
observations in the human population difficult because genetically comparable, severely
immune-compromised individuals do not typically live past their fourth decade (Pardoll, 2003).
Third, immune-compromised individuals are inherently more susceptible to various infections.
This leads to a substantially increased incidence of cancers that are pathogen-induced,
potentially obscuring the incidence of spontaneous cancers that could arise with a greater
latency in these individuals (Schreiber, 2003). Nevertheless, immune-compromised individuals,
either genetically or pharmacologically, are much more susceptible than the general population
to develop certain types of lymphomas, sarcomas, or epithelial cancers of the stomach or cervix
(Penn, 1988). However, as is the case with individuals with acquired immune-deficiency
syndromes (AIDS), it is clear that all of these cancers are pathogen-associated – herpesvirus-8
with Kaposi’s sarcoma, Epstein-Barr virus with non-Hodgkin’s lymphoma, human papillomavirus
with cervical cancer, hepatitis B virus with hepatocellular carcinoma, and Helicobacter pylori
with gastric cancer (Boshoff and Weiss, 2002). The most common cancers of the general
population – epithelial cancers of the lung, colon, prostate, or breast – are not typically
increased in these individuals. However, there is some evidence from long-term analyses of
immune-suppressed patients receiving organ transplants that they also have an increased risk
of developing epithelial cancers of nonviral etiology (Dunn et al., 2004b; Finn, 2008a).
A handful of studies of individuals with more minor immune defects, or polymorphisms in
immunoregulatory genes, also point to a role for the immune system in regulating human
cancers. One study showed that childhood hematological malignancies of nonviral origin were
increased 50% in children who inherited two mutant versions of the perforin gene which
severely reduced their T cell’s cytotoxic activity (Chia et al., 2009). In addition, women with a
particular polymorphism in FasL, the other effector of CTL cytotoxicity, had a threefold increase
in the risk of cervical cancer (Sun et al., 2005). A multitude of other investigations have linked
polymorphisms in immunoregulatory genes, such as cytokines, CTLA-4, or PD-1, with cancer
57
incidence (Cozar et al., 2007; Dehaghani et al., 2009; Nikolova et al., 2007; Welsh et al., 2009;
Zhang et al., 2005b). Interestingly, polymorphisms in CTLA-4 associated with autoimmune
disease were also associated with decreased incidence of basal or squamous cell carcinomas
of the skin (Welsh et al., 2009).
Tumor infiltrating lymphocytes and patient prognoses
A critical role for tumor immune surveillance is also evident by correlating the degree of
tumor-infiltrating lymphocytes (TILs) in various forms of cancer and patient outcome. Multiple
studies have reported that immune infiltrates in tumors correlate with lower grade or nonmetastatic disease and/or improved survival (Dunn et al., 2002; Kohrt et al., 2005; Pages et al.,
2005; Piersma et al., 2007). One particularly exhaustive study examined immunohistochemically
stained biopsies from patients with colon cancer and determined that the degree of infiltrating T
cells, and particularly CTLs, was a better indicator of the patient’s survival than predicted by
standard histopathological criteria, such as the size and invasiveness of the tumors (Galon et
al., 2006). Flow cytometry-based analyses indicate that the presence of large numbers of tumorspecific CTLs in the tumor, blood or tumor-draining lymph nodes of cancer patients are
predictive of better outcomes (Goodyear et al., 2005; Molldrem et al., 2000; Pages et al., 2005;
Wahlin et al., 2007). Finally, microarray-based experiments of human tumor tissues and lymph
nodes from breast or hepatocellular carcinoma patients have indicated that the expression of
genes associated with cytotoxic anti-tumor immune responses are predictive of better patient
prognoses (Budhu et al., 2006; Finak et al., 2008). In Chapter 2, I will present data to show that
the presence of increased numbers of lymphocytes in lung adenocarcinomas (by gene
expression analysis) is also predictive of prolonged patient survival.
58
Cancer immunotherapy
The successful treatment of cancer with immune-based therapies, particularly those that
harness CTLs, indicates that the immune system is capable of regulating cancer, at least under
the right conditions. Rather than detail the various forms of immunotherapy that have been
tested in cancer patients, my goal in this section will be to draw on the commonalities of
successful cancer immunotherapies as a means to better understand natural immune
responses to cancer in humans.
By far the most clinically tested form of cancer immunotherapy is the cancer vaccine
(Goldman and DeFrancesco, 2009). Cancer vaccines utilizing peptides, tumor cells, dendritic
cells, and various immune adjuvants have been used to try to actively boost T cell responses
against patients’ tumors. However, a sobering review by Steven Rosenberg and colleagues
pointed out that vaccine-based clinical trials have been largely unsuccessful (Rosenberg et al.,
2004). In fact, at the time of his review, no therapeutic cancer vaccine had ever been approved
for use in treating cancer. The failure of cancer vaccines is important because it tells us that it is
in fact very difficult to stimulate T cell responses against cancers in patients with the disease.
This may indicate that the immune environment in cancer patients is not conducive to activating
T cell responses. It is also reminiscent of failed attempts to therapeutically vaccinate patients
with chronic infections, such as HIV (Kim and Ahmed, 2010; Smith, 2001). These failures may
be the result of T cell tolerance induced by chronic antigenic stimulation; especially since the
targeted antigens are often tumor-associated antigens, rather than tumor-specific antigens
(Goldman and DeFrancesco, 2009).
The first FDA-approved cancer vaccine, Sipuleucel-T, which has proven to be
efficacious in the treatment of advanced prostate cancer, utilizes a patient’s own antigenpresenting cells (APCs) that are manipulated ex vivo and then transferred back into the patients
to stimulate anti-tumor T cell responses (Kantoff et al., 2010). It is interesting that rather than try
59
to actively manipulate the APCs in cancer patients (as so many failed cancer vaccines have
tried to do in the past), this technique removes the APCs from the patients and activates,
matures and loads them with tumor antigens in culture. Therefore, a critical parameter that may
contribute to the success of this therapy is the manipulation of immune cells outside of the host,
potentially avoiding a multitude of immunoregulatory mechanisms that act to maintain immune
tolerance against the cancer in the host. The importance of manipulating immune cells ex vivo is
echoed by the prior success of adoptive cell transfer therapy (ACT), perhaps the most effective
immunotherapy against cancer (Dudley et al., 2002; Rosenberg and Dudley, 2004). In this
approach, T cells infiltrating a patient’s tumor are harvested and then activated and expanded in
culture before being transferred back into the patient (Rosenberg et al., 2008). Thus, while
vaccines that attempt to provoke the anti-tumor immune response in cancer patients are
ineffective, techniques that stimulate the immune cells outside of the cancer patient can be
effective.
The only example of a cancer immunotherapy that has worked to boost the endogenous
anti-tumor immune response within cancer patients is Ipilimumab, a monoclonal antibody that
targets CTLA-4, to treat metastatic melanoma (Hodi et al., 2010). However, this immunotherapy
is fundamentally different than vaccine therapies. Rather than provoking the induction of new
anti-tumor T cell responses, anti-CTLA-4 antibodies modulate the existing T cells by blocking
signals from this key inhibitory receptor. Thus, Ipilimumab acts as an immunomodulatory drug
that can unleash preexisting anti-tumor T cell responses in cancer patients. The success of
Ipilimumab reinforces the notion that immune regulatory networks in cancer patients are the
major obstacles for successful immunotherapy, as targeting them directly in patients or
subverting them by manipulation of immune cells ex vivo provide the best clinical results.
Perhaps the most important lesson from these cancer immunotherapy trials is that T cell
responses against cancer are naturally engaged during the course of tumor progression. We
60
know this is true because Ipilimumab and ACT would not work if T cell responses weren’t
already established in the cancer patients. The success of these particular forms of therapy also
lends support to the theory that tumors escape immune surveillance because T cell tolerance is
induced in response to cancer. Chapter 4 of this thesis is dedicated to investigating the potential
for therapeutic vaccinations against human cancers utilizing our lung adenocarcinoma mouse
model.
V. MODELING IMMUNE-TUMOR INTERACTIONS IN MICE
Nearly all of the hallmarks of cancer can be, and were, defined through experimentation
in vitro with cells grown in culture. Self-sufficiency in growth signals, insensitivity to antigrowth
signals, evasion of apoptosis, limitless replicative potential, were all demonstrated with
transformed cells in culture, whereas sustained angiogenesis and tissue invasion and
metastasis were largely inferred directly from clinical observations. Although in vitro assays for
invasion and metastasis have contributed to our understanding, the underlying mechanisms
governing tumor vascularization and metastasis required the use of animal models. Similarly,
the role of T cells in regulating cancer can be inferred from clinical data. However, investigating
the role of the immune system in human cancer absolutely necessitates the use of sophisticated
animal models to faithfully recapitulate the interactions of immune cells with cancers in humans.
Therefore, this section is devoted to introducing the models of cancer used to study immunetumor interactions in mice, as well as to discussing the strengths and weaknesses of these
models and some attainable goals for the next generation of cancer models. See Figure 4 and
Table 2 for a comparison of some of the key features of the different mouse cancer models.
61
3**7$+".A@+"#+#,%"B*
!"#$%&'()*+"#,-)".*
/0123-*
45623-*
67023-*
/**9C)D,(+@2,"&'(#,%"B*
893*
:;*
<8/3=7>3*
-)")$+#)*
()@@*@,").*
&)?,")*+"#,-)".*
9**H)$D@,")2H48B*
GF*
4>*
E4F*
%&$
7-*
!"#$
JA%"#+")%'.*
#'D%$.*KL*#,D)*
F"(%-)")*
I*
7-*
!"#$
7'D%$*3"#,-)"*
<**9%"&,#,%"+@2H48B*
D'#+#,%"*
5J5*
4>*
?@%M)&*
4>*
5J5*
!"#$
9$)4R7*
!"#$
9$)24R*
7R4*
!&)$
7'D%$*3"#,-)"*P*F"(%-)")*
73*
!"#$
<%M27$+".+(#,Q+#%$*
J7F>*
J7F>*
!
JA+#,%#)DA%$+@@O2(%"#$%@@)&*
#'D%$*,"&'(#,%"*KL*3-*
=7+D*'(*
=<%M*'(*
=5)"#,29$)N3-*
7'D%$*3"#,-)"*
=5)"#,2S@AN*"*+3-*'(*
=5)"#,29$)N!&)+3-*
JA+#,%#)DA%$+@@O2(%"#$%@@)&*
#'D%$*,"&'(#,%"*
=7+D*'(*
=<%M*
R)-'@+#)&*+"#,-)"*
)MA$)..,%"*
Figure 4. Mouse models of cancer to study antigen-specific T cell responses.
(A) Transplantable models allow the introduction of model antigens in culture prior to
transplantation.
62
(B) Chemically-induced models of cancer do not allow the study of antigen-specific T cell
responses in the primary setting, but endogenous tumor antigens may be identified after culture
and investigated upon transplantation.
(C) Germline genetically engineered mice (GEM) develop tumors spontaneously due to targeted
genetic knock-out (KO) of endogenous loci or transgenic (Tg) expression of oncogenes from
tissue-specific promoters (TSP). Antigen-specific responses can be studied by crossing tumorprone mice with mice that express model antigens transgenically under the control of TSP.
(D) Conditional GEM provide spatial and temporal control of tumor induction utilizing tissuespecific expression of Cre-ER or doxycycline-inducible transactivators (TA) in combination with
timed administration of tamoxifen (Tam), doxycycline (Dox) or viruses (such as lenti-viruses)
that express Cre ± tumor antigens. Model antigen expression can be regulated concurrently with
the cancer-promoting genetic events (top, black arrow) or subsequently using combinations of
these technologies (bottom, red arrows). Brackets denote whether antigen expression is
controlled using a combination of LSL+CreERT+Tam or TRE+TA+Dox technologies.
Ag, tumor antigen; EP, endogenous promoter; TSP, tissue-specific promoter; LSL, LoxP-STOPLoxP cassette inhibits gene expression until it is removed by Cre recombinase; floxed, LoxP
sites flank exons of a gene such that Cre-mediated recombination removes these exons and
disrupts the gene; TRE, Tet(Dox)-regulated promoter elements; TA, a doxycycline-inducible
transactivator; Cre, Cre recombinase; Flp, Flp recombinase; triangles represent LoxP (or
potentially FRT) sites, boxes represent cDNAs or exons.
63
Table 2: A comparison of the different mouse models of cancer
Features:
Transplanted
Chemicallyinduced
GermlineGEM
ConditionalGEM
Cancers modeled:
All
Limited
All
All
Examples:
B16 melanoma
EL4 lymphoma
MC57 fibrosarcoma
Lewis Lung carc.
TRAMPC prostate
MCA sarcoma
UV fibrosarcoma
DMBA+TPA skin carc.
RIP-Tag2 pancreatic
β-cell hyperplasia
Erbb2/Neu mammary
TRAMP (Pro-Tag2)
prostate
LSL-SV40 many type:
sarc., lymph., carc.
Kras
lung
LSL-G12D
f/f
Kras
;PTEN
ovarian
LSL-G12D
f/f
Kras
;p53
sarcoma
LSL-G12D
Kras
pancreatic
Time to progression:
(survival time)
Genetics alterations
mimic human cancers?
Histopathology mimics
human cancers?
LSL-G12D
0-4 weeks
2-4 months
2-6 months
2-12 months
(after transplant)
(after induction)
(mouse age)
(after induction)
Unknown
Unknown/yes
Yes
Yes
(some models)
Limited cases
Yes
(limited tumor types)
Yes
Yes
Yes
Yes
No
Yes
Tumor initiated by
transformation of
normal cells?
No
Yes
Timing of tumor
initiation controlled?
Yes
(variable latency &
penetrance)
(empirically defined)
Location of tumor
formation controlled?
Yes
Yes
(orthotopic)
(limited tumor types)
Yes (transgenic);
No (tumor
Partially
suppressor KO)
Yes
(limited technology)
Tumors in natural
microenvironment?
Yes
Yes
(maybe orthotopic)
(carcinogens may
affect environment)
(oncogenic events not
restricted to tumor)
Yes
Multifocal disease?
No
?
Yes
Yes
Track tumor antigen
specific T cell
responses?
Yes
No
Yes
Yes
Restrict tumor antigen
expression to tumors?
Yes
NA
No
Possible
Possible
No
No
Possible
Regulated tumor
antigen expression?
No
GEM, genetically engineered mice; carc., carcinoma; RIP-Tag2, rat insulin promoter driven
SV40 Tag; TRAMP, rat probasin promoter driven SV40 Tag; ?, unknown; NA, not applicable.
64
Cancer by transplantation
Transplantable models of cancer have historically been the models of choice for the
investigation of T cell responses against cancer. The cancers used for transplantation
experiments are typically cell lines that were generated from a tumor that arose spontaneously
in a mouse. Cell lines derived from a multitude of different types of cancer (B16 melanoma, EL4
lymphoma, MC57 sarcoma, or 4T1 breast-, Lewis lung- and TRAMP-C prostate-carcinomas)
are routinely transplanted subcutaneously as models of the human cancers. While these tumor
models provide a rapid and reproducible system to assess T cell responses to cancer, they
have several critical limitations (Becher and Holland, 2006; Frese and Tuveson, 2007),
especially with regards to evaluating the activity of anti-tumor T cell responses (Scott, 1991;
Williams, 1982).
Transplantable models require the introduction of large numbers of cells via injection.
The injection itself can incite the immune system by inducing pro-inflammatory reactions at the
site of implantation (Khong and Restifo, 2002). Transplanting cells from culture into ectopic
environments in mice inevitably leads to tremendous tumor cell death. However, large numbers
of dead or dying tumor cells can be engulfed by phagocytic cells that initiate anti-tumor T cell
responses, a phenomenon that does not occur at the initiation of human cancer. Transplanted
tumors do not develop the normal architectural elements of endogenously arising tumors, and
lack key cellular and extracellular components of their human counterparts. Even orthotopic
transplants, in which tumor cells are put back into the organ from which they originally arose, fail
to recapitulate the process of human tumorigenesis. Unlike human cancer which begins with the
transformation of normal cells in situ followed by the progression of genetically heterogeneous
descendent cells that acquire additional mutations, transplanted tumors deliver large numbers of
fully progressed, homogeneous tumor cell clones that simply grow and rarely progress
phenotypically. The growth of transplanted tumors is significantly faster than most human
65
cancers, with doubling rates of < 2 days (Sausville and Burger, 2006). This rapid growth rate is
an important consideration for immunotherapeutic investigations because it necessitates that
therapies work quickly, and in an acute setting unlike clinical trials in cancer patients. Another
consideration when using these models is that these are cell lines derived from a single tumor
that has been passaged in vitro and in vivo by multiple investigators, for extended periods of
time. The role of selective pressures and genetic drift on the immunogenicity as well as other
features of these cells during passage are impossible to predict, but are likely significant
(Becher and Holland, 2006; Scott, 1991). Thus, it may be dangerous to put too much stock into
the results of experiments using these “venerable” cell lines (Scott, 1991). Cautions were raised
as many as 30 years ago that transplantable models of cancer may not be appropriate models
of the human disease (Alexander, 1977; Williams, 1982). Indeed the limitations of these models
have recently been demonstrated in experiments that revealed differences in the
neovascularization and therapeutic response of transplanted compared to autochthonous
cancers in mice (Olive et al., 2009; Sikder et al., 2003). Such findings may explain the failure of
clinical trials in which transplantable cancer models were used to predict the success of
therapies in cancer patients (Becher and Holland, 2006). In Chapter 2 of this thesis, I will
provide additional evidence that T cell responses against transplanted and autochthonous
models of the same mouse cancer are inherently different, and that T cell responses to
autochthonous tumors more closely mimic the situation in human cancer.
Cancer by chemical induction
The theory of tumor immune surveillance is predicated upon the use of chemically (or
carcinogen)-induced mouse models of cancer. Methylcholanthrene (MCA)-induced cancer
models were used to demonstrate that various immune-deficient mice have an increased
susceptibility to cancer and that tumors harbor tumor-specific, transplant-rejection antigens.
66
Carcinogen-induced cancers provide vast improvements over transplantable models of cancer.
They allow investigators to roughly control the location and timing of tumor onset, and tumors
arise from the transformation of normal cells in their natural setting, surrounded by a
physiologically normal architecture. Importantly, carcinogens are known to be an important
driver of cancer in humans and synonymous mutations are often induced in human and mouse
tumors.
Nevertheless, there are only a limited number of models available for chemical cancer
induction, often requiring particular genetic backgrounds and only providing a few types of
cancer to study. Although the administration of a carcinogen allows for some control over tumor
onset, tumor induction is often incomplete with low doses of carcinogens. There can also be
considerable variability in the progression of cancer, likely as a consequence of differences in
the underlying genetics of each independently arising tumor. However, there are greater
concerns with regards to the immunological aspects of carcinogen-induced cancer models. The
application of carcinogens may induce local inflammation preceding or during transformation by
the upregulation of stress ligands (NKG2D ligands) on premalignant and stromal cells (Khong
and Restifo, 2002). To increase the penetrance and reduce the latency of tumorigenesis, high
doses of carcinogens are often used. This can lead to abnormally rapid tumor progression and
the potential to generate artificially large numbers of carcinogen-induced neoantigens that can
serve as transplant-rejection antigens (Khong and Restifo, 2002). The generation of
neoantigens as a byproduct of massive carcinogenesis, rather than tumorigenesis, has been a
concern since the 1950s (Foley, 1953a, b; Prehn and Main, 1957). Early comparisons of
spontaneously arising tumors and MCA-induced tumors led investigators to the realization that
most spontaneous tumors did not contain rejection antigens, thus antigenicity was not an
intrinsic property of all cancers (Prehn and Main, 1957). Indeed, many consider the
immunogenicity of MCA-induced tumors to be unrepresentative of human cancers (Hewitt et al.,
67
1976). Inherent in these concerns, however, is the necessity to transplant these tumors in order
to assess immune-tumor interactions, primarily because tumor induction with carcinogens is so
variable and once tumors become palpable, their growth is so rapid.
Cancer by germline genetically engineered mice
Germline genetically engineered mice (GEM) encompass two types of GEM that are
typically separated but, because of their many shared properties in the context of cancer
immunology, will be grouped together here. Germline GEM include transgenic GEM and
endogenous GEM since they both harbor constitutive (not conditional) germline phenotypes
(Frese and Tuveson, 2007). Transgenic GEM utilize pronuclear injection of cDNAs into zygotes
to randomly insert oncogenes, either mammalian or viral (i.e. SV40 Tag), into the mouse
genome. Endogenous GEM disrupt the endogenous proto-oncogenes or tumor suppressors in
the mouse genome to mimic the natural disruptions of these genes in human cancers.
Transgenic GEM can be engineered and selected for tissue-specific expression of oncogenes,
allowing the development of tumors in selected organs. While endogenous GEM benefit from
the fact that the actual loci in mice are disrupted, they cannot restrict the mutations to specific
cells in the animal due to the manipulation of the endogenous genes. In any case, both systems
can be considered germline in nature because the expression of oncogenes or loss of tumor
suppressors cannot be controlled temporally or restricted specifically to tumors.
By causing the transformation of normal cells in situ with defined genetic events,
germline GEM can generate nearly all forms of human cancer, while recapitulating many of the
pathophysiological characteristics observed during the progression of the human disease. In
addition to expanding the spectrum of tumors that can be studied, germline GEM also provide
more tractable and reproducible models of cancer compared to carcinogen-induced models.
However, due to the constitutive nature of germline GEM, it is not possible to control tumor
68
initiation, tumor multiplicity, or restrict genetic disruptions exclusively to tumors. Although these
are features in common with inherited forms of cancer, they may not accurately reflect the
physiologic setting of sporadic forms of human cancer in which a somatic mutation in a single
cell initiates tumor development.
Cancer by conditional genetically engineered mice
Conditional GEM are similar to the germline GEM in that the genetic events driving the
cancers as well as the progression of the disease closely recapitulates features of the human
disease. However, conditional GEM provide considerably more control over tumor development
because the genetic disruptions can be controlled in a spatiotemporal fashion. In this way, the
activation of oncogenes or deletion of tumor suppressors is completely dependent upon specific
manipulations made by the investigator in adult animals. Control of the timing and location of the
genetic disruptions, and thus tumor formation, can be achieved using variations of two basic
strategies. The first strategy utilizes the combination of ligand-regulated transcriptional
activators (TAs) and TA-regulated promoters (TREs), such as the bacterial tetracycline operon,
to regulate the transcription of transgenic loci. The second strategy uses site-specific
recombinases (SSRs), such as Cre/LoxP, to control the expression of endogenous genetic loci.
Tissue-specific expression of TAs and timed administration of doxycycline – a tetracycline
analog that activates the TA to induce the expression of TRE-regulated genes – can allow the
controlled expression of oncogenes. The use of LoxP recombination sites, engineered so as to
delete tumor suppressor genes or allow the expression of oncogenes upon Cre-mediated
recombination, can be controlled by tissue-specific expression of a ligand-regulatable Cre, such
as the tamoxifen-regulated CreERT (Cre-estrogen receptor fusion protein), or alternatively, by
sporadic infection of cells in a given tissue with viral vectors that express Cre. Finally, the
combination of both systems can provide a means for spatiotemporal control of endogenous
69
Cre/LoxP-conditional oncogenes or tumor suppressors by way of a tissue-specific TA and TREregulated Cre recombinases.
The use of these conditional GEM provides investigators with exquisite control over the
timing of tumor initiation in specific tissues, making it possible to follow the complete evolution of
tumor progression that is not possible with any other system. However, these models are limited
by the development of reagents to specifically control gene expression in targeted organs, i.e.
mice with tissue-specific TAs or CreERT alleles (Gidekel Friedlander et al., 2009). However,
tissue-specific expression with transgenic and even endogenously targeted TAs or SSRs can be
difficult due to ectopic expression in multiple tissues. Furthermore, the ligand-regulated TAs and
SSRs are often leaky, leading to sporadic oncogene activation or tumor suppressor loss. Finally,
even if these reagents are not leaky and are specific for cells of a given tissue, administration of
the regulating ligands can lead to the widespread activity of TAs or SSRs throughout the entire
targeted organ, making it difficult to induce the genetic events sporadically in a given tissue.
While the delivery of viral vectors that express Cre to specific organs can provide temporal
control and the induction of the genetic changes in a sporadic fashion, it is currently limited to
only a few cancer models (Dinulescu et al., 2005; Jackson et al., 2001; Kirsch et al., 2007;
Willimsky and Blankenstein, 2005). In addition, it is difficult to target viral infection exclusively to
the tumor-initiating cells, although the use of avian sarcoma leukosis virus and targeted
expression of its cell entry receptor may provide a means for cell-type specific infectivity (Frese
and Tuveson, 2007).
Investigating T cell responses to mouse models of cancer
To understand the role of T cell responses against cancer, investigators can analyze the
phenotypes of bulk CD8 T cell populations or assess CTL anti-tumor activity by depleting the
entire CTL population with anti-CD8 antibodies, but ideally, investigators would narrow their
70
analysis to the tumor antigen-specific T cell populations responding to the cancers. Such
analyses are impossible to do with carcinogen-induced cancers as they develop in their
endogenous setting because the tumor antigens are unknown when tumors arise. However, by
generating cell lines from carcinogen-induced tumors, tumor-specific antigens can be identified
that allow the investigation of antigen-specific T cell responses upon transplantation. In addition,
by generating cell lines, all transplantable models, whether spontaneous or carcinogen-induced,
benefit from the ease of introducing model tumor antigens into the tumor cell lines.
Transplantable models offer one of the few systems to obtain truly tumor-specific expression of
model tumor antigens, thus avoiding the complexities of T cell responses to antigens expressed
in tumors and other tissues of the mouse.
To study T cell responses to cancers arising in germline GEM, the approach has been to
cross the GEM to mice that express model tumor antigens in the organ that gives rise to the
tumors, usually employing the same tissue-specific promoters that drive the expression of the
oncogenes, such as the rat insulin promoter (RIP) in pancreatic β-islet cells or the rat probasin
promoter in prostate epithelial cells (Drake et al., 2005; Lyman et al., 2004; Speiser et al., 1997).
Alternatively, the oncogene itself can serve as the tumor antigen, such as with SV40 Tag or
Her2/Neu transgenic mice (Anderson et al., 2007; Ercolini et al., 2005; Muller-Hermelink et al.,
2008; Willimsky and Blankenstein, 2005). However, expression of model tumor antigens
throughout the targeted organs can limit investigations because the antigens are also expressed
in normal cells of the targeted tissue, and well before tumor formation. Studies have indicated
that the recognition of antigens in the context of tumors or normal tissues can induce distinct
phenotypes in responding T cells (Getnet et al., 2009). T cells responding to the expression of
model antigens in normal tissues are often deleted or rendered tolerant to these antigens (Adler
et al., 1998; Morgan et al., 1998; Ohashi et al., 1991). Typically model antigens in these
systems are also expressed in the thymus, leading to the deletion of endogenous antigen-
71
specific T cells and forcing investigators to study the effect of transferred TCR-transgenic T cells
on tumor development. While this may provide a powerful system for preclinical
immunotherapeutic trials, it likely does not reproduce the interactions that occur in the natural
setting of cancer progression.
Very few laboratories have utilized conditional GEM as models to study T cell responses
to tumors, potentially because of the relatively recent generation of these models. Examples to
date have used Cre/LoxP technology to prevent the expression of model antigens until Cre is
activated or introduced into cancer initiating cells, leading to the simultaneous expression of
both oncogenes and model antigens (Cheung et al., 2008; Huijbers et al., 2006; Soudja et al.,
2010; Willimsky and Blankenstein, 2005; Willimsky et al., 2008). Although these systems are
advantageous because they utilize the most sophisticated mouse models of cancer and can
potentially provide a means for tumor-specific model antigen expression, in practice, it has been
difficult to adequately control the expression of the model antigens in mice, leading to leaky
expression in the thymus or other cells prior to tumor formation that may alter the phenotype of
responding anti-tumor T cells (Cheung et al., 2008; Willimsky and Blankenstein, 2005). In
Chapter 2, I will describe a novel method that utilizes conditional GEM in combination with
lentiviral tumor-induction to introduce exogenous model tumor antigens into tumors, allowing the
study of endogenous T cell responses to lung tumors expressing model neoantigens.
Mouse cancer models alter anti-tumor T cell responses
A major divide exists between the phenotype of T cells responding to tumors that are
transplanted versus those that arise endogenously. While T cell responses to transplanted
tumors have been demonstrated to be effective, often leading to tumor regression or even
complete tumor eradication, attempts to replicate these results using autochthonous cancer
models have been unsuccessful. Evidence of cancer immunoediting, via antigen loss and other
72
mechanisms, is commonly found only in the context of tumor transplantation (Johnsen et al.,
2001; Shankaran et al., 2001; Spiotto et al., 2004; Urban et al., 1982). Almost universally, T cell
tolerance, albeit in many different forms, is induced in response to autochthonous tumors and
importantly, this is also what has been observed in cancer patients (Drake, 2010). This
dichotomy may indicate that transplanted tumors induce more potent T cell responses that force
the selective outgrowth of tumors that can evade the immune response. Autochthonous tumors,
in contrast, may be ineffective at inducing potent T cell responses, making T cell tolerance the
major mechanism of tumor evasion in these models. Significant evidence to support this
hypothesis is provided in Chapter 2 of this thesis. In any case, these observations make it clear
that careful consideration should be paid to the mouse cancer models used for investigations if
we hope to learn valuable information about how to better treat the human disease.
Goals for the next generation of mouse models of cancer
Multiplicity, nature, and expression of tumor antigens
Clearly the next generation of mouse models used to study T cell responses to cancer in
the mouse should employ conditional GEM because these provide the most relevant
representations of the human disease and allow for significant control of tumor development. A
major hurdle for the next generation of cancer models will be to better recapitulate the T cell
responses found in human cancers by modulating the tumor antigens that direct the T cell
response. There are several key aspects to address. First, it is not clear whether the multiplicity
of tumor antigens targeted by T cells alters the effectiveness of T cell responses. New methods
that allow the introduction of multiple independent T cell antigens will allow investigators to
determine the importance of tumor escape via the loss of multiple antigens, epitope spreading in
promoting effective anti-tumor T cell responses, and CD4 T cell help in promoting CD8 T cell
responses. Second, the nature of the tumor antigens targeted by T cells may be relevant to the
73
effectiveness of T cell responses. Though it has been argued that tumor-specific antigens
(TSAs) would be better immunotherapeutic targets than tumor-associated antigens (TAAs)
(Schietinger et al., 2008), T cells are found to be similarly tolerant in autochthonous models of
cancer expressing TSAs or TAAs (see Chapter 2 and (Bai et al., 2008; Cheung et al., 2008;
Willimsky and Blankenstein, 2005)). However, it is unclear whether true TSA expression has
been achieved in these autochthonous models. In addition to the relevance of TSA versus TAA,
the relevance of T cell responses targeting TSAs that are selected for by the tumor because
they promote tumor progression versus TSAs that are bystander events during tumor
progression is not known. The development of models that link tumor antigens to the “driving”
(or oncogenic) event will be useful for investigating whether T cell responses to oncogenic
antigens, whose expression in tumors cannot easily be downregulated to evade an immune
response, instigate alternative mechanisms of tumor escape. Third, it is important to develop
systems that allow the independent control of tumor antigen expression and tumor initiation. For
instance, models that separate tumor initiation and the expression of tumor neoantigens can be
used to determine whether T cell responses are different when the induction of neoantigen
expression occurs in established tumors, where most mutated proteins likely arise in human
cancers. In addition, the use of TRE promoters to control antigen expression could help
elucidate the role of chronic antigen exposure in driving T cell tolerance by enabling
investigators to reversibly turn on, and then turn off, the expression of the model tumor antigens
and assess T cell function.
T cell responses in different forms of cancer
Finally, discoveries made in a particular model of cancer have often been accepted as
broadly applicable to all immune-tumor interactions. However, there is a tremendous diversity of
anti-tumor T cell responses that can be observed in different models of the disease. It is likely
74
that the immune response and tumor response can vary greatly depending on the originating
tissue and the genetic pathology of different forms of cancer. In support of this concept,
Chapters 2 and 3 of this thesis will describe two very different interactions between tumors and
T cells in two different models of autochthonous cancer. Therefore, an important goal of the next
generation of mouse cancer models will be to extend our analyses to other genetically
engineered mouse models of cancer to discern whether T cell responses and tumor
development change depending on the context of the disease.
75
REFERENCES
Adkins, B., Leclerc, C., and Marshall-Clarke, S. (2004). Neonatal adaptive immunity comes of
age. Nat Rev Immunol 4, 553-564.
Adler, A.J., Marsh, D.W., Yochum, G.S., Guzzo, J.L., Nigam, A., Nelson, W.G., and Pardoll,
D.M. (1998). CD4+ T cell tolerance to parenchymal self-antigens requires presentation by bone
marrow-derived antigen-presenting cells. J Exp Med 187, 1555-1564.
Ahmadzadeh, M., and Rosenberg, S.A. (2005). TGF-beta 1 attenuates the acquisition and
expression of effector function by tumor antigen-specific human memory CD8 T cells. J
Immunol 174, 5215-5223.
Albert, M.L., Pearce, S.F., Francisco, L.M., Sauter, B., Roy, P., Silverstein, R.L., and Bhardwaj,
N. (1998a). Immature dendritic cells phagocytose apoptotic cells via alphavbeta5 and CD36,
and cross-present antigens to cytotoxic T lymphocytes. J Exp Med 188, 1359-1368.
Albert, M.L., Sauter, B., and Bhardwaj, N. (1998b). Dendritic cells acquire antigen from
apoptotic cells and induce class I-restricted CTLs. Nature 392, 86-89.
Alexander, P. (1977). Back to the drawing board--the need for more realistic model systems for
immunotherapy. Cancer 40, 467-470.
Algarra, I., Garcia-Lora, A., Cabrera, T., Ruiz-Cabello, F., and Garrido, F. (2004). The selection
of tumor variants with altered expression of classical and nonclassical MHC class I molecules:
implications for tumor immune escape. Cancer Immunol Immunother 53, 904-910.
Altman, J.D., Moss, P.A., Goulder, P.J., Barouch, D.H., McHeyzer-Williams, M.G., Bell, J.I.,
McMichael, A.J., and Davis, M.M. (1996). Phenotypic analysis of antigen-specific T
lymphocytes. Science 274, 94-96.
Anderson, M.J., Shafer-Weaver, K., Greenberg, N.M., and Hurwitz, A.A. (2007). Tolerization of
tumor-specific T cells despite efficient initial priming in a primary murine model of prostate
cancer. J Immunol 178, 1268-1276.
Anichini, A., Mortarini, R., Nonaka, D., Molla, A., Vegetti, C., Montaldi, E., Wang, X., and
Ferrone, S. (2006). Association of antigen-processing machinery and HLA antigen phenotype of
melanoma cells with survival in American Joint Committee on Cancer stage III and IV
melanoma patients. Cancer Res 66, 6405-6411.
Apetoh, L., Ghiringhelli, F., Tesniere, A., Obeid, M., Ortiz, C., Criollo, A., Mignot, G., Maiuri,
M.C., Ullrich, E., Saulnier, P., et al. (2007). Toll-like receptor 4-dependent contribution of the
immune system to anticancer chemotherapy and radiotherapy. Nat Med 13, 1050-1059.
Asea, A., Rehli, M., Kabingu, E., Boch, J.A., Bare, O., Auron, P.E., Stevenson, M.A., and
Calderwood, S.K. (2002). Novel signal transduction pathway utilized by extracellular HSP70:
role of toll-like receptor (TLR) 2 and TLR4. J Biol Chem 277, 15028-15034.
76
Azuma, T., Yao, S., Zhu, G., Flies, A.S., Flies, S.J., and Chen, L. (2008). B7-H1 is a ubiquitous
antiapoptotic receptor on cancer cells. Blood 111, 3635-3643.
Bai, A., Higham, E., Eisen, H.N., Wittrup, K.D., and Chen, J. (2008). Rapid tolerization of virusactivated tumor-specific CD8+ T cells in prostate tumors of TRAMP mice. Proc Natl Acad Sci U
S A 105, 13003-13008.
Balkwill, F., and Mantovani, A. (2001). Inflammation and cancer: back to Virchow? Lancet 357,
539-545.
Barber, D.L., Wherry, E.J., Masopust, D., Zhu, B., Allison, J.P., Sharpe, A.H., Freeman, G.J.,
and Ahmed, R. (2006). Restoring function in exhausted CD8 T cells during chronic viral
infection. Nature 439, 682-687.
Basu, S., Binder, R.J., Ramalingam, T., and Srivastava, P.K. (2001). CD91 is a common
receptor for heat shock proteins gp96, hsp90, hsp70, and calreticulin. Immunity 14, 303-313.
Becher, O.J., and Holland, E.C. (2006). Genetically engineered models have advantages over
xenografts for preclinical studies. Cancer Res 66, 3355-3358, discussion 3358-3359.
Bennett, S.R., Carbone, F.R., Karamalis, F., Flavell, R.A., Miller, J.F., and Heath, W.R. (1998).
Help for cytotoxic-T-cell responses is mediated by CD40 signalling. Nature 393, 478-480.
Bennett, S.R., Carbone, F.R., Karamalis, F., Miller, J.F., and Heath, W.R. (1997). Induction of a
CD8+ cytotoxic T lymphocyte response by cross-priming requires cognate CD4+ T cell help. J
Exp Med 186, 65-70.
Billingham, R.E., Brent, L., and Medawar, P.B. (1953). Actively acquired tolerance of foreign
cells. Nature 172, 603-606.
Blank, C., Brown, I., Peterson, A.C., Spiotto, M., Iwai, Y., Honjo, T., and Gajewski, T.F. (2004).
PD-L1/B7H-1 inhibits the effector phase of tumor rejection by T cell receptor (TCR) transgenic
CD8+ T cells. Cancer Res 64, 1140-1145.
Boissonnas, A., Fetler, L., Zeelenberg, I.S., Hugues, S., and Amigorena, S. (2007). In vivo
imaging of cytotoxic T cell infiltration and elimination of a solid tumor. J Exp Med 204, 345-356.
Bonertz, A., Weitz, J., Pietsch, D.H., Rahbari, N.N., Schlude, C., Ge, Y., Juenger, S., Vlodavsky,
I., Khazaie, K., Jaeger, D., et al. (2009). Antigen-specific Tregs control T cell responses against
a limited repertoire of tumor antigens in patients with colorectal carcinoma. J Clin Invest 119,
3311-3321.
Boshoff, C., and Weiss, R. (2002). AIDS-related malignancies. Nat Rev Cancer 2, 373-382.
Breart, B., Lemaitre, F., Celli, S., and Bousso, P. (2008). Two-photon imaging of intratumoral
CD8+ T cell cytotoxic activity during adoptive T cell therapy in mice. J Clin Invest 118, 13901397.
77
Buckanovich, R.J., Facciabene, A., Kim, S., Benencia, F., Sasaroli, D., Balint, K., Katsaros, D.,
O'Brien-Jenkins, A., Gimotty, P.A., and Coukos, G. (2008). Endothelin B receptor mediates the
endothelial barrier to T cell homing to tumors and disables immune therapy. Nat Med 14, 28-36.
Bucks, C.M., Norton, J.A., Boesteanu, A.C., Mueller, Y.M., and Katsikis, P.D. (2009). Chronic
antigen stimulation alone is sufficient to drive CD8+ T cell exhaustion. J Immunol 182, 66976708.
Budhu, A., Forgues, M., Ye, Q.H., Jia, H.L., He, P., Zanetti, K.A., Kammula, U.S., Chen, Y., Qin,
L.X., Tang, Z.Y., and Wang, X.W. (2006). Prediction of venous metastases, recurrence, and
prognosis in hepatocellular carcinoma based on a unique immune response signature of the
liver microenvironment. Cancer Cell 10, 99-111.
Calzascia, T., Pellegrini, M., Hall, H., Sabbagh, L., Ono, N., Elford, A.R., Mak, T.W., and
Ohashi, P.S. (2007). TNF-alpha is critical for antitumor but not antiviral T cell immunity in mice.
J Clin Invest 117, 3833-3845.
Cardin, R.D., Brooks, J.W., Sarawar, S.R., and Doherty, P.C. (1996). Progressive loss of CD8+
T cell-mediated control of a gamma-herpesvirus in the absence of CD4+ T cells. J Exp Med
184, 863-871.
Carswell, E.A., Old, L.J., Kassel, R.L., Green, S., Fiore, N., and Williamson, B. (1975). An
endotoxin-induced serum factor that causes necrosis of tumors. Proc Natl Acad Sci U S A 72,
3666-3670.
Caspi, R.R. (2008). Immunotherapy of autoimmunity and cancer: the penalty for success. Nat
Rev Immunol 8, 970-976.
Cavallo, F., Calogero, R.A., and Forni, G. (2007). Are oncoantigens suitable targets for antitumour therapy? Nat Rev Cancer 7, 707-713.
Chang, C.C., and Ferrone, S. (2007). Immune selective pressure and HLA class I antigen
defects in malignant lesions. Cancer Immunol Immunother 56, 227-236.
Chang, C.C., Ogino, T., Mullins, D.W., Oliver, J.L., Yamshchikov, G.V., Bandoh, N., Slingluff,
C.L., Jr., and Ferrone, S. (2006). Defective human leukocyte antigen class I-associated antigen
presentation caused by a novel beta2-microglobulin loss-of-function in melanoma cells. J Biol
Chem 281, 18763-18773.
Chang, H.Y., Sneddon, J.B., Alizadeh, A.A., Sood, R., West, R.B., Montgomery, K., Chi, J.T.,
van de Rijn, M., Botstein, D., and Brown, P.O. (2004). Gene expression signature of fibroblast
serum response predicts human cancer progression: similarities between tumors and wounds.
PLoS Biol 2, E7.
Chen, H.L., Gabrilovich, D., Tampe, R., Girgis, K.R., Nadaf, S., and Carbone, D.P. (1996a). A
functionally defective allele of TAP1 results in loss of MHC class I antigen presentation in a
human lung cancer. Nat Genet 13, 210-213.
78
Chen, H.L., Gabrilovich, D., Virmani, A., Ratnani, I., Girgis, K.R., Nadaf-Rahrov, S., FernandezVina, M., and Carbone, D.P. (1996b). Structural and functional analysis of beta2 microglobulin
abnormalities in human lung and breast cancer. Int J Cancer 67, 756-763.
Chen, M.L., Pittet, M.J., Gorelik, L., Flavell, R.A., Weissleder, R., von Boehmer, H., and
Khazaie, K. (2005). Regulatory T cells suppress tumor-specific CD8 T cell cytotoxicity through
TGF-beta signals in vivo. Proc Natl Acad Sci U S A 102, 419-424.
Chen, T., Guo, J., Han, C., Yang, M., and Cao, X. (2009). Heat shock protein 70, released from
heat-stressed tumor cells, initiates antitumor immunity by inducing tumor cell chemokine
production and activating dendritic cells via TLR4 pathway. J Immunol 182, 1449-1459.
Cheung, A.F., Dupage, M.J., Dong, H.K., Chen, J., and Jacks, T. (2008). Regulated expression
of a tumor-associated antigen reveals multiple levels of T-cell tolerance in a mouse model of
lung cancer. Cancer Res 68, 9459-9468.
Chia, J., Yeo, K.P., Whisstock, J.C., Dunstone, M.A., Trapani, J.A., and Voskoboinik, I. (2009).
Temperature sensitivity of human perforin mutants unmasks subtotal loss of cytotoxicity,
delayed FHL, and a predisposition to cancer. Proc Natl Acad Sci U S A 106, 9809-9814.
Clayton, J. (2006). Lymphocytes: not useless after all. J Exp Med 203, 487.
Cohen, C.J., Zheng, Z., Bray, R., Zhao, Y., Sherman, L.A., Rosenberg, S.A., and Morgan, R.A.
(2005). Recognition of fresh human tumor by human peripheral blood lymphocytes transduced
with a bicistronic retroviral vector encoding a murine anti-p53 TCR. J Immunol 175, 5799-5808.
Condeelis, J., and Pollard, J.W. (2006). Macrophages: obligate partners for tumor cell migration,
invasion, and metastasis. Cell 124, 263-266.
Coussens, L.M., Raymond, W.W., Bergers, G., Laig-Webster, M., Behrendtsen, O., Werb, Z.,
Caughey, G.H., and Hanahan, D. (1999). Inflammatory mast cells up-regulate angiogenesis
during squamous epithelial carcinogenesis. Genes Dev 13, 1382-1397.
Coussens, L.M., Tinkle, C.L., Hanahan, D., and Werb, Z. (2000). MMP-9 supplied by bone
marrow-derived cells contributes to skin carcinogenesis. Cell 103, 481-490.
Coussens, L.M., and Werb, Z. (2002). Inflammation and cancer. Nature 420, 860-867.
Cozar, J.M., Romero, J.M., Aptsiauri, N., Vazquez, F., Vilchez, J.R., Tallada, M., Garrido, F.,
and Ruiz-Cabello, F. (2007). High incidence of CTLA-4 AA (CT60) polymorphism in renal cell
cancer. Hum Immunol 68, 698-704.
Cromme, F.V., Airey, J., Heemels, M.T., Ploegh, H.L., Keating, P.J., Stern, P.L., Meijer, C.J.,
and Walboomers, J.M. (1994a). Loss of transporter protein, encoded by the TAP-1 gene, is
highly correlated with loss of HLA expression in cervical carcinomas. J Exp Med 179, 335-340.
Cromme, F.V., van Bommel, P.F., Walboomers, J.M., Gallee, M.P., Stern, P.L., Kenemans, P.,
Helmerhorst, T.J., Stukart, M.J., and Meijer, C.J. (1994b). Differences in MHC and TAP-1
expression in cervical cancer lymph node metastases as compared with the primary tumours. Br
J Cancer 69, 1176-1181.
79
Curiel, T.J., Coukos, G., Zou, L., Alvarez, X., Cheng, P., Mottram, P., Evdemon-Hogan, M.,
Conejo-Garcia, J.R., Zhang, L., Burow, M., et al. (2004). Specific recruitment of regulatory T
cells in ovarian carcinoma fosters immune privilege and predicts reduced survival. Nat Med 10,
942-949.
Daniel, D., Meyer-Morse, N., Bergsland, E.K., Dehne, K., Coussens, L.M., and Hanahan, D.
(2003). Immune enhancement of skin carcinogenesis by CD4+ T cells. J Exp Med 197, 10171028.
Day, C.L., Kaufmann, D.E., Kiepiela, P., Brown, J.A., Moodley, E.S., Reddy, S., Mackey, E.W.,
Miller, J.D., Leslie, A.J., DePierres, C., et al. (2006). PD-1 expression on HIV-specific T cells is
associated with T-cell exhaustion and disease progression. Nature 443, 350-354.
de Visser, K.E., Eichten, A., and Coussens, L.M. (2006). Paradoxical roles of the immune
system during cancer development. Nat Rev Cancer 6, 24-37.
de Visser, K.E., Korets, L.V., and Coussens, L.M. (2005). De novo carcinogenesis promoted by
chronic inflammation is B lymphocyte dependent. Cancer Cell 7, 411-423.
de Vos van Steenwijk, P.J., Heusinkveld, M., Ramwadhdoebe, T.H., Lowik, M.J., van der Hulst,
J.M., Goedemans, R., Piersma, S.J., Kenter, G.G., and van der Burg, S.H. (2010). An
unexpectedly large polyclonal repertoire of HPV-specific T cells is poised for action in patients
with cervical cancer. Cancer Res 70, 2707-2717.
Dehaghani, A.S., Kashef, M.A., Ghaemenia, M., Sarraf, Z., Khaghanzadeh, N., Fattahi, M.J.,
and Ghaderi, A. (2009). PDCD1, CTLA-4 and p53 gene polymorphism and susceptibility to
gestational trophoblastic diseases. J Reprod Med 54, 25-31.
DeLong, P., Carroll, R.G., Henry, A.C., Tanaka, T., Ahmad, S., Leibowitz, M.S., Sterman, D.H.,
June, C.H., Albelda, S.M., and Vonderheide, R.H. (2005). Regulatory T cells and cytokines in
malignant pleural effusions secondary to mesothelioma and carcinoma. Cancer Biol Ther 4,
342-346.
DeNardo, D.G., Barreto, J.B., Andreu, P., Vasquez, L., Tawfik, D., Kolhatkar, N., and Coussens,
L.M. (2009). CD4(+) T cells regulate pulmonary metastasis of mammary carcinomas by
enhancing protumor properties of macrophages. Cancer Cell 16, 91-102.
Derre, L., Rivals, J.P., Jandus, C., Pastor, S., Rimoldi, D., Romero, P., Michielin, O., Olive, D.,
and Speiser, D.E. (2010). BTLA mediates inhibition of human tumor-specific CD8+ T cells that
can be partially reversed by vaccination. J Clin Invest 120, 157-167.
Diefenbach, A., Jensen, E.R., Jamieson, A.M., and Raulet, D.H. (2001). Rae1 and H60 ligands
of the NKG2D receptor stimulate tumour immunity. Nature 413, 165-171.
Dighe, A.S., Richards, E., Old, L.J., and Schreiber, R.D. (1994). Enhanced in vivo growth and
resistance to rejection of tumor cells expressing dominant negative IFN gamma receptors.
Immunity 1, 447-456.
80
Dinulescu, D.M., Ince, T.A., Quade, B.J., Shafer, S.A., Crowley, D., and Jacks, T. (2005). Role
of K-ras and Pten in the development of mouse models of endometriosis and endometrioid
ovarian cancer. Nat Med 11, 63-70.
Dong, H., Strome, S.E., Salomao, D.R., Tamura, H., Hirano, F., Flies, D.B., Roche, P.C., Lu, J.,
Zhu, G., Tamada, K., et al. (2002). Tumor-associated B7-H1 promotes T-cell apoptosis: a
potential mechanism of immune evasion. Nat Med 8, 793-800.
Drake, C.G. (2010). Prostate cancer as a model for tumour immunotherapy. Nat Rev Immunol
10, 580-593.
Drake, C.G., Doody, A.D., Mihalyo, M.A., Huang, C.T., Kelleher, E., Ravi, S., Hipkiss, E.L.,
Flies, D.B., Kennedy, E.P., Long, M., et al. (2005). Androgen ablation mitigates tolerance to a
prostate/prostate cancer-restricted antigen. Cancer Cell 7, 239-249.
Dranoff, G. (2004). Cytokines in cancer pathogenesis and cancer therapy. Nat Rev Cancer 4,
11-22.
Dudley, M.E., Wunderlich, J.R., Robbins, P.F., Yang, J.C., Hwu, P., Schwartzentruber, D.J.,
Topalian, S.L., Sherry, R., Restifo, N.P., Hubicki, A.M., et al. (2002). Cancer regression and
autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science 298,
850-854.
Dunn, G.P., Bruce, A.T., Ikeda, H., Old, L.J., and Schreiber, R.D. (2002). Cancer
immunoediting: from immunosurveillance to tumor escape. Nat Immunol 3, 991-998.
Dunn, G.P., Bruce, A.T., Sheehan, K.C., Shankaran, V., Uppaluri, R., Bui, J.D., Diamond, M.S.,
Koebel, C.M., Arthur, C., White, J.M., and Schreiber, R.D. (2005). A critical function for type I
interferons in cancer immunoediting. Nat Immunol 6, 722-729.
Dunn, G.P., Koebel, C.M., and Schreiber, R.D. (2006). Interferons, immunity and cancer
immunoediting. Nat Rev Immunol 6, 836-848.
Dunn, G.P., Old, L.J., and Schreiber, R.D. (2004a). The immunobiology of cancer
immunosurveillance and immunoediting. Immunity 21, 137-148.
Dunn, G.P., Old, L.J., and Schreiber, R.D. (2004b). The three Es of cancer immunoediting.
Annu Rev Immunol 22, 329-360.
Dvorak, H.F. (1986). Tumors: wounds that do not heal. Similarities between tumor stroma
generation and wound healing. N Engl J Med 315, 1650-1659.
Ercolini, A.M., Ladle, B.H., Manning, E.A., Pfannenstiel, L.W., Armstrong, T.D., Machiels, J.P.,
Bieler, J.G., Emens, L.A., Reilly, R.T., and Jaffee, E.M. (2005). Recruitment of latent pools of
high-avidity CD8(+) T cells to the antitumor immune response. J Exp Med 201, 1591-1602.
Feng, H., Zeng, Y., Graner, M.W., and Katsanis, E. (2002). Stressed apoptotic tumor cells
stimulate dendritic cells and induce specific cytotoxic T cells. Blood 100, 4108-4115.
81
Fife, B.T., Guleria, I., Gubbels Bupp, M., Eagar, T.N., Tang, Q., Bour-Jordan, H., Yagita, H.,
Azuma, M., Sayegh, M.H., and Bluestone, J.A. (2006). Insulin-induced remission in new-onset
NOD mice is maintained by the PD-1-PD-L1 pathway. J Exp Med 203, 2737-2747.
Finak, G., Bertos, N., Pepin, F., Sadekova, S., Souleimanova, M., Zhao, H., Chen, H.,
Omeroglu, G., Meterissian, S., Omeroglu, A., et al. (2008). Stromal gene expression predicts
clinical outcome in breast cancer. Nat Med 14, 518-527.
Finn, O.J. (2008a). Cancer immunology. N Engl J Med 358, 2704-2715.
Finn, O.J. (2008b). Immunological weapons acquired early in life win battles with cancer late in
life. J Immunol 181, 1589-1592.
Flavell, R.A., Sanjabi, S., Wrzesinski, S.H., and Licona-Limon, P. (2010). The polarization of
immune cells in the tumour environment by TGFbeta. Nat Rev Immunol 10, 554-567.
Fogg, M.H., Wirth, L.J., Posner, M., and Wang, F. (2009). Decreased EBNA-1-specific CD8+ T
cells in patients with Epstein-Barr virus-associated nasopharyngeal carcinoma. Proc Natl Acad
Sci U S A 106, 3318-3323.
Foley, E.J. (1953a). Antigenic properties of methylcholanthrene-induced tumors in mice of the
strain of origin. Cancer Res 13, 835-837.
Foley, E.J. (1953b). Attempts to induce immunity against mammary adenocarcinoma in inbred
mice. Cancer Res 13, 578-580.
Fossum, B., Breivik, J., Meling, G.I., Gedde-Dahl, T., 3rd, Hansen, T., Knutsen, I., Rognum,
T.O., Thorsby, E., and Gaudernack, G. (1994a). A K-ras 13Gly-->Asp mutation is recognized by
HLA-DQ7 restricted T cells in a patient with colorectal cancer. Modifying effect of DQ7 on
established cancers harbouring this mutation? Int J Cancer 58, 506-511.
Fossum, B., Gedde-Dahl, T., 3rd, Breivik, J., Eriksen, J.A., Spurkland, A., Thorsby, E., and
Gaudernack, G. (1994b). p21-ras-peptide-specific T-cell responses in a patient with colorectal
cancer. CD4+ and CD8+ T cells recognize a peptide corresponding to a common mutation
(13Gly-->Asp). Int J Cancer 56, 40-45.
Fossum, B., Olsen, A.C., Thorsby, E., and Gaudernack, G. (1995). CD8+ T cells from a patient
with colon carcinoma, specific for a mutant p21-Ras-derived peptide (Gly13-->Asp), are
cytotoxic towards a carcinoma cell line harbouring the same mutation. Cancer Immunol
Immunother 40, 165-172.
Freeman, G.J., Wherry, E.J., Ahmed, R., and Sharpe, A.H. (2006). Reinvigorating exhausted
HIV-specific T cells via PD-1-PD-1 ligand blockade. J Exp Med 203, 2223-2227.
Frese, K.K., and Tuveson, D.A. (2007). Maximizing mouse cancer models. Nat Rev Cancer 7,
645-658.
Gallucci, S., Lolkema, M., and Matzinger, P. (1999). Natural adjuvants: endogenous activators
of dendritic cells. Nat Med 5, 1249-1255.
82
Gallucci, S., and Matzinger, P. (2001). Danger signals: SOS to the immune system. Curr Opin
Immunol 13, 114-119.
Galon, J., Costes, A., Sanchez-Cabo, F., Kirilovsky, A., Mlecnik, B., Lagorce-Pages, C.,
Tosolini, M., Camus, M., Berger, A., Wind, P., et al. (2006). Type, density, and location of
immune cells within human colorectal tumors predict clinical outcome. Science 313, 1960-1964.
Garbe, A.I., Vermeer, B., Gamrekelashvili, J., von Wasielewski, R., Greten, F.R., Westendorf,
A.M., Buer, J., Schmid, R.M., Manns, M.P., Korangy, F., and Greten, T.F. (2006). Genetically
induced pancreatic adenocarcinoma is highly immunogenic and causes spontaneous tumorspecific immune responses. Cancer Res 66, 508-516.
Gasser, S., Orsulic, S., Brown, E.J., and Raulet, D.H. (2005). The DNA damage pathway
regulates innate immune system ligands of the NKG2D receptor. Nature 436, 1186-1190.
Getnet, D., Maris, C.H., Hipkiss, E.L., Grosso, J.F., Harris, T.J., Yen, H.R., Bruno, T.C., Wada,
S., Adler, A., Georgantas, R.W., et al. (2009). Tumor recognition and self-recognition induce
distinct transcriptional profiles in antigen-specific CD4 T cells. J Immunol 182, 4675-4685.
Ghiringhelli, F., Apetoh, L., Tesniere, A., Aymeric, L., Ma, Y., Ortiz, C., Vermaelen, K.,
Panaretakis, T., Mignot, G., Ullrich, E., et al. (2009). Activation of the NLRP3 inflammasome in
dendritic cells induces IL-1beta-dependent adaptive immunity against tumors. Nat Med 15,
1170-1178.
Ghiringhelli, F., Puig, P.E., Roux, S., Parcellier, A., Schmitt, E., Solary, E., Kroemer, G., Martin,
F., Chauffert, B., and Zitvogel, L. (2005). Tumor cells convert immature myeloid dendritic cells
into TGF-beta-secreting cells inducing CD4+CD25+ regulatory T cell proliferation. J Exp Med
202, 919-929.
Gidekel Friedlander, S.Y., Chu, G.C., Snyder, E.L., Girnius, N., Dibelius, G., Crowley, D., Vasile,
E., DePinho, R.A., and Jacks, T. (2009). Context-dependent transformation of adult pancreatic
cells by oncogenic K-Ras. Cancer Cell 16, 379-389.
Goldman, B., and DeFrancesco, L. (2009). The cancer vaccine roller coaster. Nat Biotechnol
27, 129-139.
Goodyear, O., Piper, K., Khan, N., Starczynski, J., Mahendra, P., Pratt, G., and Moss, P. (2005).
CD8+ T cells specific for cancer germline gene antigens are found in many patients with
multiple myeloma, and their frequency correlates with disease burden. Blood 106, 4217-4224.
Gorelik, L., and Flavell, R.A. (2001). Immune-mediated eradication of tumors through the
blockade of transforming growth factor-beta signaling in T cells. Nat Med 7, 1118-1122.
Greten, F.R., Eckmann, L., Greten, T.F., Park, J.M., Li, Z.W., Egan, L.J., Kagnoff, M.F., and
Karin, M. (2004). IKKbeta links inflammation and tumorigenesis in a mouse model of colitisassociated cancer. Cell 118, 285-296.
Grivennikov, S.I., Greten, F.R., and Karin, M. (2010). Immunity, inflammation, and cancer. Cell
140, 883-899.
83
Groh, V., Rhinehart, R., Secrist, H., Bauer, S., Grabstein, K.H., and Spies, T. (1999). Broad
tumor-associated expression and recognition by tumor-derived gamma delta T cells of MICA
and MICB. Proc Natl Acad Sci U S A 96, 6879-6884.
Groh, V., Smythe, K., Dai, Z., and Spies, T. (2006). Fas-ligand-mediated paracrine T cell
regulation by the receptor NKG2D in tumor immunity. Nat Immunol 7, 755-762.
Groh, V., Wu, J., Yee, C., and Spies, T. (2002). Tumour-derived soluble MIC ligands impair
expression of NKG2D and T-cell activation. Nature 419, 734-738.
Gross, L. (1943). Intradermal Immunization of C3H Mice against a Sarcoma That Originated in
an Animal of the Same Line. Cancer Research 3, 326-333.
Grosso, J.F., Goldberg, M.V., Getnet, D., Bruno, T.C., Yen, H.R., Pyle, K.J., Hipkiss, E., Vignali,
D.A., Pardoll, D.M., and Drake, C.G. (2009). Functionally distinct LAG-3 and PD-1 subsets on
activated and chronically stimulated CD8 T cells. J Immunol 182, 6659-6669.
Grosso, J.F., Kelleher, C.C., Harris, T.J., Maris, C.H., Hipkiss, E.L., De Marzo, A., Anders, R.,
Netto, G., Getnet, D., Bruno, T.C., et al. (2007). LAG-3 regulates CD8+ T cell accumulation and
effector function in murine self- and tumor-tolerance systems. J Clin Invest 117, 3383-3392.
Guerra, N., Tan, Y.X., Joncker, N.T., Choy, A., Gallardo, F., Xiong, N., Knoblaugh, S., Cado, D.,
Greenberg, N.M., and Raulet, D.H. (2008). NKG2D-deficient mice are defective in tumor
surveillance in models of spontaneous malignancy. Immunity 28, 571-580.
Hahn, S.A., Schutte, M., Hoque, A.T., Moskaluk, C.A., da Costa, L.T., Rozenblum, E.,
Weinstein, C.L., Fischer, A., Yeo, C.J., Hruban, R.H., and Kern, S.E. (1996). DPC4, a candidate
tumor suppressor gene at human chromosome 18q21.1. Science 271, 350-353.
Hammer, G.E., Gonzalez, F., James, E., Nolla, H., and Shastri, N. (2007). In the absence of
aminopeptidase ERAAP, MHC class I molecules present many unstable and highly
immunogenic peptides. Nat Immunol 8, 101-108.
Hanada, K., Yewdell, J.W., and Yang, J.C. (2004). Immune recognition of a human renal cancer
antigen through post-translational protein splicing. Nature 427, 252-256.
Hanahan, D., and Weinberg, R.A. (2000). The hallmarks of cancer. Cell 100, 57-70.
Hernandez, J., Aung, S., Marquardt, K., and Sherman, L.A. (2002). Uncoupling of Proliferative
Potential and Gain of Effector Function by CD8+ T Cells Responding to Self-Antigens. J. Exp.
Med. 196, 323-333.
Hewitt, H.B., Blake, E.R., and Walder, A.S. (1976). A critique of the evidence for active host
defence against cancer, based on personal studies of 27 murine tumours of spontaneous origin.
Br J Cancer 33, 241-259.
Hicklin, D.J., Wang, Z., Arienti, F., Rivoltini, L., Parmiani, G., and Ferrone, S. (1998). beta2Microglobulin mutations, HLA class I antigen loss, and tumor progression in melanoma. J Clin
Invest 101, 2720-2729.
84
Hodi, F.S., O'Day, S.J., McDermott, D.F., Weber, R.W., Sosman, J.A., Haanen, J.B., Gonzalez,
R., Robert, C., Schadendorf, D., Hassel, J.C., et al. (2010). Improved Survival with Ipilimumab in
Patients with Metastatic Melanoma. N Engl J Med.
Hudson, J.D., Shoaibi, M.A., Maestro, R., Carnero, A., Hannon, G.J., and Beach, D.H. (1999). A
proinflammatory cytokine inhibits p53 tumor suppressor activity. J Exp Med 190, 1375-1382.
Huijbers, I.J., Krimpenfort, P., Chomez, P., van der Valk, M.A., Song, J.Y., Inderberg-Suso,
E.M., Schmitt-Verhulst, A.M., Berns, A., and Van den Eynde, B.J. (2006). An inducible mouse
model of melanoma expressing a defined tumor antigen. Cancer Res 66, 3278-3286.
Iliopoulos, D., Hirsch, H.A., and Struhl, K. (2009). An epigenetic switch involving NF-kappaB,
Lin28, Let-7 MicroRNA, and IL6 links inflammation to cell transformation. Cell 139, 693-706.
Inokuma, M., dela Rosa, C., Schmitt, C., Haaland, P., Siebert, J., Petry, D., Tang, M., Suni,
M.A., Ghanekar, S.A., Gladding, D., et al. (2007). Functional T cell responses to tumor antigens
in breast cancer patients have a distinct phenotype and cytokine signature. J Immunol 179,
2627-2633.
Jackson, E.L., Willis, N., Mercer, K., Bronson, R.T., Crowley, D., Montoya, R., Jacks, T., and
Tuveson, D.A. (2001). Analysis of lung tumor initiation and progression using conditional
expression of oncogenic K-ras. Genes Dev 15, 3243-3248.
Jackson, S.P., and Bartek, J. (2009). The DNA-damage response in human biology and
disease. Nature 461, 1071-1078.
Jager, E., Ringhoffer, M., Karbach, J., Arand, M., Oesch, F., and Knuth, A. (1996). Inverse
relationship of melanocyte differentiation antigen expression in melanoma tissues and CD8+
cytotoxic-T-cell responses: evidence for immunoselection of antigen-loss variants in vivo. Int J
Cancer 66, 470-476.
Janssen, E.M., Lemmens, E.E., Wolfe, T., Christen, U., von Herrath, M.G., and Schoenberger,
S.P. (2003). CD4+ T cells are required for secondary expansion and memory in CD8+ T
lymphocytes. Nature 421, 852-856.
Jenkins, M.K., and Schwartz, R.H. (1987). Antigen presentation by chemically modified
splenocytes induces antigen-specific T cell unresponsiveness in vitro and in vivo. J Exp Med
165, 302-319.
Ji, H., Houghton, A.M., Mariani, T.J., Perera, S., Kim, C.B., Padera, R., Tonon, G., McNamara,
K., Marconcini, L.A., Hezel, A., et al. (2006). K-ras activation generates an inflammatory
response in lung tumors. Oncogene 25, 2105-2112.
Johnsen, A.K., France, J., Nagy, N., Askew, D., Abdul-Karim, F.W., Gerson, S.L., Sy, M.S., and
Harding, C.V. (2001). Systemic deficits in transporter for antigen presentation (TAP)-1 or
proteasome subunit LMP2 have little or no effect on tumor incidence. Int J Cancer 91, 366-372.
Kagi, D., Ledermann, B., Burki, K., Seiler, P., Odermatt, B., Olsen, K.J., Podack, E.R.,
Zinkernagel, R.M., and Hengartner, H. (1994a). Cytotoxicity mediated by T cells and natural
killer cells is greatly impaired in perforin-deficient mice. Nature 369, 31-37.
85
Kagi, D., Vignaux, F., Ledermann, B., Burki, K., Depraetere, V., Nagata, S., Hengartner, H., and
Golstein, P. (1994b). Fas and perforin pathways as major mechanisms of T cell-mediated
cytotoxicity. Science 265, 528-530.
Kantoff, P.W., Higano, C.S., Shore, N.D., Berger, E.R., Small, E.J., Penson, D.F., Redfern,
C.H., Ferrari, A.C., Dreicer, R., Sims, R.B., et al. (2010). Sipuleucel-T Immunotherapy for
Castration-Resistant Prostate Cancer. New England Journal of Medicine 363, 411-422.
Kaplan, D.H., Shankaran, V., Dighe, A.S., Stockert, E., Aguet, M., Old, L.J., and Schreiber, R.D.
(1998). Demonstration of an interferon gamma-dependent tumor surveillance system in
immunocompetent mice. Proc Natl Acad Sci U S A 95, 7556-7561.
Kaufmann, D.E., and Walker, B.D. (2009). PD-1 and CTLA-4 inhibitory cosignaling pathways in
HIV infection and the potential for therapeutic intervention. J Immunol 182, 5891-5897.
Keene, J.A., and Forman, J. (1982). Helper activity is required for the in vivo generation of
cytotoxic T lymphocytes. J Exp Med 155, 768-782.
Keir, M.E., Liang, S.C., Guleria, I., Latchman, Y.E., Qipo, A., Albacker, L.A., Koulmanda, M.,
Freeman, G.J., Sayegh, M.H., and Sharpe, A.H. (2006). Tissue expression of PD-L1 mediates
peripheral T cell tolerance. J Exp Med 203, 883-895.
Khong, H.T., and Restifo, N.P. (2002). Natural selection of tumor variants in the generation of
"tumor escape" phenotypes. Nat Immunol 3, 999-1005.
Kiessling, R., and Klein, E. (1973). Cytotoxic potential of mouse spleen cells on H-2 antibodytreated target cells. J Exp Med 137, 527-532.
Kiessling, R., Petranyi, G., Karre, K., Jondal, M., Tracey, D., and Wigzell, H. (1976). Killer cells:
a functional comparison between natural, immune T-cell and antibody-dependent in vitro
systems. J Exp Med 143, 772-780.
Kim, P.S., and Ahmed, R. (2010). Features of responding T cells in cancer and chronic
infection. Curr Opin Immunol 22, 223-230.
Kirsch, D.G., Dinulescu, D.M., Miller, J.B., Grimm, J., Santiago, P.M., Young, N.P., Nielsen,
G.P., Quade, B.J., Chaber, C.J., Schultz, C.P., et al. (2007). A spatially and temporally
restricted mouse model of soft tissue sarcoma. Nat Med 13, 992-997.
Klein, G., Sjogren, H.O., Klein, E., and Hellstrom, K.E. (1960). Demonstration of resistance
against methylcholanthrene-induced sarcomas in the primary autochthonous host. Cancer Res
20, 1561-1572.
Kohrt, H.E., Nouri, N., Nowels, K., Johnson, D., Holmes, S., and Lee, P.P. (2005). Profile of
immune cells in axillary lymph nodes predicts disease-free survival in breast cancer. PLoS Med
2, e284.
Kusmartsev, S.A., Li, Y., and Chen, S.H. (2000). Gr-1+ myeloid cells derived from tumorbearing mice inhibit primary T cell activation induced through CD3/CD28 costimulation. J
Immunol 165, 779-785.
86
Kuss, I., Donnenberg, A.D., Gooding, W., and Whiteside, T.L. (2003). Effector CD8+CD45ROCD27-T cells have signalling defects in patients with squamous cell carcinoma of the head and
neck. Br J Cancer 88, 223-230.
Lanzavecchia, A. (1998). Immunology. Licence to kill. Nature 393, 413-414.
Leach, D.R., Krummel, M.F., and Allison, J.P. (1996). Enhancement of antitumor immunity by
CTLA-4 blockade. Science 271, 1734-1736.
Lee, P.P., Yee, C., Savage, P.A., Fong, L., Brockstedt, D., Weber, J.S., Johnson, D., Swetter,
S., Thompson, J., Greenberg, P.D., et al. (1999). Characterization of circulating T cells specific
for tumor-associated antigens in melanoma patients. Nat Med 5, 677-685.
Leys, C.M., Nomura, S., LaFleur, B.J., Ferrone, S., Kaminishi, M., Montgomery, E., and
Goldenring, J.R. (2007). Expression and prognostic significance of prothymosin-alpha and
ERp57 in human gastric cancer. Surgery 141, 41-50.
Liakou, C.I., Kamat, A., Tang, D.N., Chen, H., Sun, J., Troncoso, P., Logothetis, C., and
Sharma, P. (2008). CTLA-4 blockade increases IFNgamma-producing CD4+ICOShi cells to
shift the ratio of effector to regulatory T cells in cancer patients. Proc Natl Acad Sci U S A 105,
14987-14992.
Lin, E.Y., Li, J.F., Gnatovskiy, L., Deng, Y., Zhu, L., Grzesik, D.A., Qian, H., Xue, X.N., and
Pollard, J.W. (2006). Macrophages regulate the angiogenic switch in a mouse model of breast
cancer. Cancer Res 66, 11238-11246.
Lin, E.Y., Nguyen, A.V., Russell, R.G., and Pollard, J.W. (2001). Colony-stimulating factor 1
promotes progression of mammary tumors to malignancy. J Exp Med 193, 727-740.
Liyanage, U.K., Moore, T.T., Joo, H.G., Tanaka, Y., Herrmann, V., Doherty, G., Drebin, J.A.,
Strasberg, S.M., Eberlein, T.J., Goedegebuure, P.S., and Linehan, D.C. (2002). Prevalence of
regulatory T cells is increased in peripheral blood and tumor microenvironment of patients with
pancreas or breast adenocarcinoma. J Immunol 169, 2756-2761.
Lotze, M.T., Chang, A.E., Seipp, C.A., Simpson, C., Vetto, J.T., and Rosenberg, S.A. (1986).
High-dose recombinant interleukin 2 in the treatment of patients with disseminated cancer.
Responses, treatment-related morbidity, and histologic findings. JAMA 256, 3117-3124.
Lowin, B., Beermann, F., Schmidt, A., and Tschopp, J. (1994a). A null mutation in the perforin
gene impairs cytolytic T lymphocyte- and natural killer cell-mediated cytotoxicity. Proc Natl Acad
Sci U S A 91, 11571-11575.
Lowin, B., Hahne, M., Mattmann, C., and Tschopp, J. (1994b). Cytolytic T-cell cytotoxicity is
mediated through perforin and Fas lytic pathways. Nature 370, 650-652.
Lyman, M.A., Aung, S., Biggs, J.A., and Sherman, L.A. (2004). A spontaneously arising
pancreatic tumor does not promote the differentiation of naive CD8+ T lymphocytes into effector
CTL. J Immunol 172, 6558-6567.
87
Maeurer, M.J., Gollin, S.M., Martin, D., Swaney, W., Bryant, J., Castelli, C., Robbins, P.,
Parmiani, G., Storkus, W.J., and Lotze, M.T. (1996a). Tumor escape from immune recognition:
lethal recurrent melanoma in a patient associated with downregulation of the peptide transporter
protein TAP-1 and loss of expression of the immunodominant MART-1/Melan-A antigen. J Clin
Invest 98, 1633-1641.
Maeurer, M.J., Gollin, S.M., Storkus, W.J., Swaney, W., Karbach, J., Martin, D., Castelli, C.,
Salter, R., Knuth, A., and Lotze, M.T. (1996b). Tumor escape from immune recognition: loss of
HLA-A2 melanoma cell surface expression is associated with a complex rearrangement of the
short arm of chromosome 6. Clin Cancer Res 2, 641-652.
Mantovani, A. (2009). Cancer: Inflaming metastasis. Nature 457, 36-37.
Mantovani, A., Allavena, P., Sica, A., and Balkwill, F. (2008). Cancer-related inflammation.
Nature 454, 436-444.
Markowitz, S., Wang, J., Myeroff, L., Parsons, R., Sun, L., Lutterbaugh, J., Fan, R.S.,
Zborowska, E., Kinzler, K.W., Vogelstein, B., and et al. (1995). Inactivation of the type II TGFbeta receptor in colon cancer cells with microsatellite instability. Science 268, 1336-1338.
Martin-Orozco, N., Wang, Y.H., Yagita, H., and Dong, C. (2006). Cutting Edge: Programmed
death (PD) ligand-1/PD-1 interaction is required for CD8+ T cell tolerance to tissue antigens. J
Immunol 177, 8291-8295.
Marzec, M., Zhang, Q., Goradia, A., Raghunath, P.N., Liu, X., Paessler, M., Wang, H.Y.,
Wysocka, M., Cheng, M., Ruggeri, B.A., and Wasik, M.A. (2008). Oncogenic kinase NPM/ALK
induces through STAT3 expression of immunosuppressive protein CD274 (PD-L1, B7-H1). Proc
Natl Acad Sci U S A 105, 20852-20857.
Mazzoni, A., Bronte, V., Visintin, A., Spitzer, J.H., Apolloni, E., Serafini, P., Zanovello, P., and
Segal, D.M. (2002). Myeloid suppressor lines inhibit T cell responses by an NO-dependent
mechanism. J Immunol 168, 689-695.
Medawar, P.B. (1961). Immunological tolerance. Nature 189, 14-17.
Medema, J.P., de Jong, J., Peltenburg, L.T., Verdegaal, E.M., Gorter, A., Bres, S.A., Franken,
K.L., Hahne, M., Albar, J.P., Melief, C.J., and Offringa, R. (2001). Blockade of the granzyme
B/perforin pathway through overexpression of the serine protease inhibitor PI-9/SPI-6
constitutes a mechanism for immune escape by tumors. Proc Natl Acad Sci U S A 98, 1151511520.
Medema, J.P., de Jong, J., van Hall, T., Melief, C.J., and Offringa, R. (1999). Immune escape of
tumors in vivo by expression of cellular FLICE-inhibitory protein. J Exp Med 190, 1033-1038.
Mempel, T.R., Pittet, M.J., Khazaie, K., Weninger, W., Weissleder, R., von Boehmer, H., and
von Andrian, U.H. (2006). Regulatory T cells reversibly suppress cytotoxic T cell function
independent of effector differentiation. Immunity 25, 129-141.
88
Meylan, E., Dooley, A.L., Feldser, D.M., Shen, L., Turk, E., Ouyang, C., and Jacks, T. (2009).
Requirement for NF-kappaB signalling in a mouse model of lung adenocarcinoma. Nature 462,
104-107.
Millar, D.G., Garza, K.M., Odermatt, B., Elford, A.R., Ono, N., Li, Z., and Ohashi, P.S. (2003).
Hsp70 promotes antigen-presenting cell function and converts T-cell tolerance to autoimmunity
in vivo. Nat Med 9, 1469-1476.
Mitchison, N.A. (1953). Passive transfer of transplantation immunity. Nature 171, 267-268.
Mitchison, N.A. (1955). Studies on the immunological response to foreign tumor transplants in
the mouse. I. The role of lymph node cells in conferring immunity by adoptive transfer. J Exp
Med 102, 157-177.
Molldrem, J.J., Lee, P.P., Wang, C., Felio, K., Kantarjian, H.M., Champlin, R.E., and Davis,
M.M. (2000). Evidence that specific T lymphocytes may participate in the elimination of chronic
myelogenous leukemia. Nat Med 6, 1018-1023.
Monach, P.A., Meredith, S.C., Siegel, C.T., and Schreiber, H. (1995). A unique tumor antigen
produced by a single amino acid substitution. Immunity 2, 45-59.
Mondal, S., Iype, P.T., Griesbach, L.M., and Heidelberger, C. (1970). Antigenicity of cells
derived from mouse prostate cells after malignant transformation in vitro by carcinogenic
hydrocarbons. Cancer Res 30, 1593-1597.
Moore, R.J., Owens, D.M., Stamp, G., Arnott, C., Burke, F., East, N., Holdsworth, H., Turner, L.,
Rollins, B., Pasparakis, M., et al. (1999). Mice deficient in tumor necrosis factor-alpha are
resistant to skin carcinogenesis. Nat Med 5, 828-831.
Morgan, D.J., Kreuwel, H.T., Fleck, S., Levitsky, H.I., Pardoll, D.M., and Sherman, L.A. (1998).
Activation of low avidity CTL specific for a self epitope results in tumor rejection but not
autoimmunity. J Immunol 160, 643-651.
Morgan, R.A., Dudley, M.E., Wunderlich, J.R., Hughes, M.S., Yang, J.C., Sherry, R.M., Royal,
R.E., Topalian, S.L., Kammula, U.S., Restifo, N.P., et al. (2006). Cancer regression in patients
after transfer of genetically engineered lymphocytes. Science 314, 126-129.
Movahedi, K., Guilliams, M., Van den Bossche, J., Van den Bergh, R., Gysemans, C., Beschin,
A., De Baetselier, P., and Van Ginderachter, J.A. (2008). Identification of discrete tumor-induced
myeloid-derived suppressor cell subpopulations with distinct T cell-suppressive activity. Blood
111, 4233-4244.
Mrass, P., Takano, H., Ng, L.G., Daxini, S., Lasaro, M.O., Iparraguirre, A., Cavanagh, L.L., von
Andrian, U.H., Ertl, H.C., Haydon, P.G., and Weninger, W. (2006). Random migration precedes
stable target cell interactions of tumor-infiltrating T cells. J Exp Med 203, 2749-2761.
Mueller, S.N., and Ahmed, R. (2009). High antigen levels are the cause of T cell exhaustion
during chronic viral infection. Proc Natl Acad Sci U S A 106, 8623-8628.
89
Muller-Hermelink, N., Braumuller, H., Pichler, B., Wieder, T., Mailhammer, R., Schaak, K.,
Ghoreschi, K., Yazdi, A., Haubner, R., Sander, C.A., et al. (2008). TNFR1 signaling and IFNgamma signaling determine whether T cells induce tumor dormancy or promote multistage
carcinogenesis. Cancer Cell 13, 507-518.
Nagaraj, S., and Gabrilovich, D.I. (2008). Tumor escape mechanism governed by myeloidderived suppressor cells. Cancer Res 68, 2561-2563.
Nagaraj, S., Gupta, K., Pisarev, V., Kinarsky, L., Sherman, S., Kang, L., Herber, D.L., Schneck,
J., and Gabrilovich, D.I. (2007). Altered recognition of antigen is a mechanism of CD8+ T cell
tolerance in cancer. Nat Med 13, 828-835.
Nakanishi, Y., Lu, B., Gerard, C., and Iwasaki, A. (2009). CD8(+) T lymphocyte mobilization to
virus-infected tissue requires CD4(+) T-cell help. Nature 462, 510-513.
Nelson, J., Bagnato, A., Battistini, B., and Nisen, P. (2003). The endothelin axis: emerging role
in cancer. Nat Rev Cancer 3, 110-116.
Nguyen, L.T., Elford, A.R., Murakami, K., Garza, K.M., Schoenberger, S.P., Odermatt, B.,
Speiser, D.E., and Ohashi, P.S. (2002). Tumor growth enhances cross-presentation leading to
limited T cell activation without tolerance. J Exp Med 195, 423-435.
Nikolova, P.N., Pawelec, G.P., Mihailova, S.M., Ivanova, M.I., Myhailova, A.P., Baltadjieva,
D.N., Marinova, D.I., Ivanova, S.S., and Naumova, E.J. (2007). Association of cytokine gene
polymorphisms with malignant melanoma in Caucasian population. Cancer Immunol
Immunother 56, 371-379.
Nishimura, H., Nose, M., Hiai, H., Minato, N., and Honjo, T. (1999a). Development of lupus-like
autoimmune diseases by disruption of the PD-1 gene encoding an ITIM motif-carrying
immunoreceptor. Immunity 11, 141-151.
Nishimura, T., Iwakabe, K., Sekimoto, M., Ohmi, Y., Yahata, T., Nakui, M., Sato, T., Habu, S.,
Tashiro, H., Sato, M., and Ohta, A. (1999b). Distinct role of antigen-specific T helper type 1
(Th1) and Th2 cells in tumor eradication in vivo. J Exp Med 190, 617-627.
Noguchi, Y., Chen, Y.T., and Old, L.J. (1994). A mouse mutant p53 product recognized by
CD4+ and CD8+ T cells. Proc Natl Acad Sci U S A 91, 3171-3175.
Novellino, L., Castelli, C., and Parmiani, G. (2005). A listing of human tumor antigens
recognized by T cells: March 2004 update. Cancer Immunol Immunother 54, 187-207.
Nozawa, H., Chiu, C., and Hanahan, D. (2006). Infiltrating neutrophils mediate the initial
angiogenic switch in a mouse model of multistage carcinogenesis. Proc Natl Acad Sci U S A
103, 12493-12498.
Obeid, M., Tesniere, A., Ghiringhelli, F., Fimia, G.M., Apetoh, L., Perfettini, J.L., Castedo, M.,
Mignot, G., Panaretakis, T., Casares, N., et al. (2007). Calreticulin exposure dictates the
immunogenicity of cancer cell death. Nat Med 13, 54-61.
90
Ochsenbein, A.F., Klenerman, P., Karrer, U., Ludewig, B., Pericin, M., Hengartner, H., and
Zinkernagel, R.M. (1999). Immune surveillance against a solid tumor fails because of
immunological ignorance. Proc Natl Acad Sci U S A 96, 2233-2238.
Ochsenbein, A.F., Sierro, S., Odermatt, B., Pericin, M., Karrer, U., Hermans, J., Hemmi, S.,
Hengartner, H., and Zinkernagel, R.M. (2001). Roles of tumour localization, second signals and
cross priming in cytotoxic T-cell induction. Nature 411, 1058-1064.
Ogino, T., Shigyo, H., Ishii, H., Katayama, A., Miyokawa, N., Harabuchi, Y., and Ferrone, S.
(2006). HLA class I antigen down-regulation in primary laryngeal squamous cell carcinoma
lesions as a poor prognostic marker. Cancer Res 66, 9281-9289.
Ohashi, P.S., Oehen, S., Buerki, K., Pircher, H., Ohashi, C.T., Odermatt, B., Malissen, B.,
Zinkernagel, R.M., and Hengartner, H. (1991). Ablation of "tolerance" and induction of diabetes
by virus infection in viral antigen transgenic mice. Cell 65, 305-317.
Old, L.J., and Chen, Y.T. (1998). New paths in human cancer serology. J Exp Med 187, 11631167.
Olive, K.P., Jacobetz, M.A., Davidson, C.J., Gopinathan, A., McIntyre, D., Honess, D., Madhu,
B., Goldgraben, M.A., Caldwell, M.E., Allard, D., et al. (2009). Inhibition of Hedgehog signaling
enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science 324, 14571461.
Pages, F., Berger, A., Camus, M., Sanchez-Cabo, F., Costes, A., Molidor, R., Mlecnik, B.,
Kirilovsky, A., Nilsson, M., Damotte, D., et al. (2005). Effector memory T cells, early metastasis,
and survival in colorectal cancer. N Engl J Med 353, 2654-2666.
Pardoll, D. (2003). Does the immune system see tumors as foreign or self? Annu Rev Immunol
21, 807-839.
Pardoll, D.M. (2002). Spinning molecular immunology into successful immunotherapy. Nat Rev
Immunol 2, 227-238.
Parmiani, G., De Filippo, A., Novellino, L., and Castelli, C. (2007). Unique human tumor
antigens: immunobiology and use in clinical trials. J Immunol 178, 1975-1979.
Parsa, A.T., Waldron, J.S., Panner, A., Crane, C.A., Parney, I.F., Barry, J.J., Cachola, K.E.,
Murray, J.C., Tihan, T., Jensen, M.C., et al. (2007). Loss of tumor suppressor PTEN function
increases B7-H1 expression and immunoresistance in glioma. Nat Med 13, 84-88.
Penn, I. (1988). Tumors of the immunocompromised patient. Annu Rev Med 39, 63-73.
Perez, V.L., Van Parijs, L., Biuckians, A., Zheng, X.X., Strom, T.B., and Abbas, A.K. (1997).
Induction of peripheral T cell tolerance in vivo requires CTLA-4 engagement. Immunity 6, 411417.
Petrilli, V., Dostert, C., Muruve, D.A., and Tschopp, J. (2007). The inflammasome: a danger
sensing complex triggering innate immunity. Curr Opin Immunol 19, 615-622.
91
Piersma, S.J., Jordanova, E.S., van Poelgeest, M.I., Kwappenberg, K.M., van der Hulst, J.M.,
Drijfhout, J.W., Melief, C.J., Kenter, G.G., Fleuren, G.J., Offringa, R., and van der Burg, S.H.
(2007). High number of intraepithelial CD8+ tumor-infiltrating lymphocytes is associated with the
absence of lymph node metastases in patients with large early-stage cervical cancer. Cancer
Res 67, 354-361.
Pikarsky, E., Porat, R.M., Stein, I., Abramovitch, R., Amit, S., Kasem, S., Gutkovich-Pyest, E.,
Urieli-Shoval, S., Galun, E., and Ben-Neriah, Y. (2004). NF-kappaB functions as a tumour
promoter in inflammation-associated cancer. Nature 431, 461-466.
Prehn, R.T. (1968). Tumor-specific antigens of putatively nonviral tumors. Cancer Res 28, 13261330.
Prehn, R.T., and Main, J.M. (1957). Immunity to methylcholanthrene-induced sarcomas. J Natl
Cancer Inst 18, 769-778.
Preynat-Seauve, O., Schuler, P., Contassot, E., Beermann, F., Huard, B., and French, L.E.
(2006). Tumor-infiltrating dendritic cells are potent antigen-presenting cells able to activate T
cells and mediate tumor rejection. J Immunol 176, 61-67.
Pule, M.A., Savoldo, B., Myers, G.D., Rossig, C., Russell, H.V., Dotti, G., Huls, M.H., Liu, E.,
Gee, A.P., Mei, Z., et al. (2008). Virus-specific T cells engineered to coexpress tumor-specific
receptors: persistence and antitumor activity in individuals with neuroblastoma. Nat Med 14,
1264-1270.
Qian, B.Z., and Pollard, J.W. (2010). Macrophage diversity enhances tumor progression and
metastasis. Cell 141, 39-51.
Qin, Z., and Blankenstein, T. (2000). CD4+ T cell--mediated tumor rejection involves inhibition
of angiogenesis that is dependent on IFN gamma receptor expression by nonhematopoietic
cells. Immunity 12, 677-686.
Rabinovich, G.A., Gabrilovich, D., and Sotomayor, E.M. (2007). Immunosuppressive strategies
that are mediated by tumor cells. Annu Rev Immunol 25, 267-296.
Raffaghello, L., Prigione, I., Bocca, P., Morandi, F., Camoriano, M., Gambini, C., Wang, X.,
Ferrone, S., and Pistoia, V. (2005). Multiple defects of the antigen-processing machinery
components in human neuroblastoma: immunotherapeutic implications. Oncogene 24, 46344644.
Rakoff-Nahoum, S., and Medzhitov, R. (2009). Toll-like receptors and cancer. Nat Rev Cancer
9, 57-63.
Raulet, D.H., and Guerra, N. (2009). Oncogenic stress sensed by the immune system: role of
natural killer cell receptors. Nat Rev Immunol 9, 568-580.
Redmond, W.L., and Sherman, L.A. (2005). Peripheral tolerance of CD8 T lymphocytes.
Immunity 22, 275-284.
92
Restifo, N.P., Antony, P.A., Finkelstein, S.E., Leitner, W.W., Surman, D.P., Theoret, M.R., and
Touloukian, C.E. (2002). Assumptions of the tumor 'escape' hypothesis. Semin Cancer Biol 12,
81-86.
Restifo, N.P., Esquivel, F., Kawakami, Y., Yewdell, J.W., Mule, J.J., Rosenberg, S.A., and
Bennink, J.R. (1993). Identification of human cancers deficient in antigen processing. J Exp Med
177, 265-272.
Ridge, J.P., Di Rosa, F., and Matzinger, P. (1998). A conditioned dendritic cell can be a
temporal bridge between a CD4+ T-helper and a T-killer cell. Nature 393, 474-478.
Riker, A., Cormier, J., Panelli, M., Kammula, U., Wang, E., Abati, A., Fetsch, P., Lee, K.H.,
Steinberg, S., Rosenberg, S., and Marincola, F. (1999). Immune selection after antigen-specific
immunotherapy of melanoma. Surgery 126, 112-120.
Riond, J., Rodriguez, S., Nicolau, M.L., al Saati, T., and Gairin, J.E. (2009). In vivo major
histocompatibility complex class I (MHCI) expression on MHCIlow tumor cells is regulated by
gammadelta T and NK cells during the early steps of tumor growth. Cancer Immun 9, 10.
Riss, J., Khanna, C., Koo, S., Chandramouli, G.V., Yang, H.H., Hu, Y., Kleiner, D.E.,
Rosenwald, A., Schaefer, C.F., Ben-Sasson, S.A., et al. (2006). Cancers as wounds that do not
heal: differences and similarities between renal regeneration/repair and renal cell carcinoma.
Cancer Res 66, 7216-7224.
Robbins, P.F., El-Gamil, M., Li, Y.F., Kawakami, Y., Loftus, D., Appella, E., and Rosenberg,
S.A. (1996). A mutated beta-catenin gene encodes a melanoma-specific antigen recognized by
tumor infiltrating lymphocytes. J Exp Med 183, 1185-1192.
Roberts, W.K., Deluca, I.J., Thomas, A., Fak, J., Williams, T., Buckley, N., Dousmanis, A.G.,
Posner, J.B., and Darnell, R.B. (2009). Patients with lung cancer and paraneoplastic Hu
syndrome harbor HuD-specific type 2 CD8+ T cells. J Clin Invest 119, 2042-2051.
Rocha, B., and Tanchot, C. (2004). Towards a cellular definition of CD8+ T-cell memory: the
role of CD4+ T-cell help in CD8+ T-cell responses. Curr Opin Immunol 16, 259-263.
Romero, P., Dunbar, P.R., Valmori, D., Pittet, M., Ogg, G.S., Rimoldi, D., Chen, J.L., Lienard,
D., Cerottini, J.C., and Cerundolo, V. (1998). Ex vivo staining of metastatic lymph nodes by
class I major histocompatibility complex tetramers reveals high numbers of antigen-experienced
tumor-specific cytolytic T lymphocytes. J Exp Med 188, 1641-1650.
Rosenberg, S.A., and Dudley, M.E. (2004). Cancer regression in patients with metastatic
melanoma after the transfer of autologous antitumor lymphocytes. Proc Natl Acad Sci U S A
101 Suppl 2, 14639-14645.
Rosenberg, S.A., Lotze, M.T., Muul, L.M., Chang, A.E., Avis, F.P., Leitman, S., Linehan, W.M.,
Robertson, C.N., Lee, R.E., Rubin, J.T., and et al. (1987). A progress report on the treatment of
157 patients with advanced cancer using lymphokine-activated killer cells and interleukin-2 or
high-dose interleukin-2 alone. N Engl J Med 316, 889-897.
93
Rosenberg, S.A., Lotze, M.T., Muul, L.M., Leitman, S., Chang, A.E., Ettinghausen, S.E., Matory,
Y.L., Skibber, J.M., Shiloni, E., Vetto, J.T., and et al. (1985). Observations on the systemic
administration of autologous lymphokine-activated killer cells and recombinant interleukin-2 to
patients with metastatic cancer. N Engl J Med 313, 1485-1492.
Rosenberg, S.A., Restifo, N.P., Yang, J.C., Morgan, R.A., and Dudley, M.E. (2008). Adoptive
cell transfer: a clinical path to effective cancer immunotherapy. Nat Rev Cancer 8, 299-308.
Rosenberg, S.A., Spiess, P., and Lafreniere, R. (1986). A new approach to the adoptive
immunotherapy of cancer with tumor-infiltrating lymphocytes. Science 233, 1318-1321.
Rosenberg, S.A., Yang, J.C., and Restifo, N.P. (2004). Cancer immunotherapy: moving beyond
current vaccines. Nat Med 10, 909-915.
Rosenberg, S.A., Yang, J.C., Topalian, S.L., Schwartzentruber, D.J., Weber, J.S., Parkinson,
D.R., Seipp, C.A., Einhorn, J.H., and White, D.E. (1994a). Treatment of 283 consecutive
patients with metastatic melanoma or renal cell cancer using high-dose bolus interleukin 2.
JAMA 271, 907-913.
Rosenberg, S.A., Yannelli, J.R., Yang, J.C., Topalian, S.L., Schwartzentruber, D.J., Weber, J.S.,
Parkinson, D.R., Seipp, C.A., Einhorn, J.H., and White, D.E. (1994b). Treatment of patients with
metastatic melanoma with autologous tumor-infiltrating lymphocytes and interleukin 2. J Natl
Cancer Inst 86, 1159-1166.
Rous, P. (1911). A Sarcoma of the Fowl Transmissible by an Agent Separable from the Tumor
Cells. J Exp Med 13, 397-411.
Rous, P. (1965). Viruses and tumour causation. An appraisal of present knowledge. Nature 207,
457-463.
Russell, J.P., Engiles, J.B., and Rothstein, J.L. (2004). Proinflammatory mediators and genetic
background in oncogene mediated tumor progression. J Immunol 172, 4059-4067.
Sakaguchi, S., Sakaguchi, N., Asano, M., Itoh, M., and Toda, M. (1995). Immunologic selftolerance maintained by activated T cells expressing IL-2 receptor alpha-chains (CD25).
Breakdown of a single mechanism of self-tolerance causes various autoimmune diseases. J
Immunol 155, 1151-1164.
Sanda, M.G., Restifo, N.P., Walsh, J.C., Kawakami, Y., Nelson, W.G., Pardoll, D.M., and
Simons, J.W. (1995). Molecular characterization of defective antigen processing in human
prostate cancer. J Natl Cancer Inst 87, 280-285.
Sausville, E.A., and Burger, A.M. (2006). Contributions of human tumor xenografts to anticancer
drug development. Cancer Res 66, 3351-3354, discussion 3354.
Savage, P.A., Vosseller, K., Kang, C., Larimore, K., Riedel, E., Wojnoonski, K., Jungbluth, A.A.,
and Allison, J.P. (2008). Recognition of a ubiquitous self antigen by prostate cancer-infiltrating
CD8+ T lymphocytes. Science 319, 215-220.
94
Schietinger, A., Philip, M., and Schreiber, H. (2008). Specificity in cancer immunotherapy.
Semin Immunol 20, 276-285.
Schietinger, A., Philip, M., Yoshida, B.A., Azadi, P., Liu, H., Meredith, S.C., and Schreiber, H.
(2006). A mutant chaperone converts a wild-type protein into a tumor-specific antigen. Science
314, 304-308.
Schoenberger, S.P., Toes, R.E., van der Voort, E.I., Offringa, R., and Melief, C.J. (1998). T-cell
help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions. Nature 393, 480483.
Schreiber, H. (2003). Tumor Immunology. In Fundamental Immunology, W.E. Paul, ed.
(Philadelphia: Lippincott Williams & Wilkins), pp. 1557-1592.
Schwab, S.R., Li, K.C., Kang, C., and Shastri, N. (2003). Constitutive display of cryptic
translation products by MHC class I molecules. Science 301, 1367-1371.
Scott, O.C. (1991). Tumor transplantation and tumor immunity: a personal view. Cancer Res 51,
757-763.
Sedy, J.R., Gavrieli, M., Potter, K.G., Hurchla, M.A., Lindsley, R.C., Hildner, K., Scheu, S.,
Pfeffer, K., Ware, C.F., Murphy, T.L., and Murphy, K.M. (2005). B and T lymphocyte attenuator
regulates T cell activation through interaction with herpesvirus entry mediator. Nat Immunol 6,
90-98.
Segal, N.H., Parsons, D.W., Peggs, K.S., Velculescu, V., Kinzler, K.W., Vogelstein, B., and
Allison, J.P. (2008). Epitope landscape in breast and colorectal cancer. Cancer Res 68, 889892.
Shafer-Weaver, K.A., Watkins, S.K., Anderson, M.J., Draper, L.J., Malyguine, A., Alvord, W.G.,
Greenberg, N.M., and Hurwitz, A.A. (2009). Immunity to murine prostatic tumors: continuous
provision of T-cell help prevents CD8 T-cell tolerance and activates tumor-infiltrating dendritic
cells. Cancer Res 69, 6256-6264.
Shankaran, V., Ikeda, H., Bruce, A.T., White, J.M., Swanson, P.E., Old, L.J., and Schreiber,
R.D. (2001). IFNgamma and lymphocytes prevent primary tumour development and shape
tumour immunogenicity. Nature 410, 1107-1111.
Shedlock, D.J., and Shen, H. (2003). Requirement for CD4 T cell help in generating functional
CD8 T cell memory. Science 300, 337-339.
Shen, L., Evel-Kabler, K., Strube, R., and Chen, S.Y. (2004). Silencing of SOCS1 enhances
antigen presentation by dendritic cells and antigen-specific anti-tumor immunity. Nat Biotechnol
22, 1546-1553.
Shi, Y., Evans, J.E., and Rock, K.L. (2003). Molecular identification of a danger signal that alerts
the immune system to dying cells. Nature 425, 516-521.
95
Shields, J.D., Kourtis, I.C., Tomei, A.A., Roberts, J.M., and Swartz, M.A. (2010). Induction of
lymphoidlike stroma and immune escape by tumors that express the chemokine CCL21.
Science 328, 749-752.
Shimizu, J., Yamazaki, S., and Sakaguchi, S. (1999). Induction of tumor immunity by removing
CD25+CD4+ T cells: a common basis between tumor immunity and autoimmunity. J Immunol
163, 5211-5218.
Siegel, P.M., and Massague, J. (2003). Cytostatic and apoptotic actions of TGF-beta in
homeostasis and cancer. Nat Rev Cancer 3, 807-821.
Sikder, H., Huso, D.L., Zhang, H., Wang, B., Ryu, B., Hwang, S.T., Powell, J.D., and Alani, R.M.
(2003). Disruption of Id1 reveals major differences in angiogenesis between transplanted and
autochthonous tumors. Cancer Cell 4, 291-299.
Sjoblom, T., Jones, S., Wood, L.D., Parsons, D.W., Lin, J., Barber, T.D., Mandelker, D., Leary,
R.J., Ptak, J., Silliman, N., et al. (2006). The consensus coding sequences of human breast and
colorectal cancers. Science 314, 268-274.
Smith, K.A. (2001). To cure chronic HIV infection, a new therapeutic strategy is needed. Curr
Opin Immunol 13, 617-624.
Smyth, M.J., Godfrey, D.I., and Trapani, J.A. (2001). A fresh look at tumor immunosurveillance
and immunotherapy. Nat Immunol 2, 293-299.
Smyth, M.J., Swann, J., Cretney, E., Zerafa, N., Yokoyama, W.M., and Hayakawa, Y. (2005).
NKG2D function protects the host from tumor initiation. J Exp Med 202, 583-588.
Soudja, S.M., Wehbe, M., Mas, A., Chasson, L., de Tenbossche, C.P., Huijbers, I., Van den
Eynde, B., and Schmitt-Verhulst, A.M. (2010). Tumor-initiated inflammation overrides protective
adaptive immunity in an induced melanoma model in mice. Cancer Res 70, 3515-3525.
Sparmann, A., and Bar-Sagi, D. (2004). Ras-induced interleukin-8 expression plays a critical
role in tumor growth and angiogenesis. Cancer Cell 6, 447-458.
Speiser, D.E., Miranda, R., Zakarian, A., Bachmann, M.F., McKall-Faienza, K., Odermatt, B.,
Hanahan, D., Zinkernagel, R.M., and Ohashi, P.S. (1997). Self antigens expressed by solid
tumors Do not efficiently stimulate naive or activated T cells: implications for immunotherapy. J
Exp Med 186, 645-653.
Spiotto, M.T., Rowley, D.A., and Schreiber, H. (2004). Bystander elimination of antigen loss
variants in established tumors. Nat Med 10, 294-298.
Spiotto, M.T., Yu, P., Rowley, D.A., Nishimura, M.I., Meredith, S.C., Gajewski, T.F., Fu, Y.X.,
and Schreiber, H. (2002). Increasing tumor antigen expression overcomes "ignorance" to solid
tumors via crosspresentation by bone marrow-derived stromal cells. Immunity 17, 737-747.
Staveley-O'Carroll, K., Sotomayor, E., Montgomery, J., Borrello, I., Hwang, L., Fein, S., Pardoll,
D., and Levitsky, H. (1998). Induction of antigen-specific T cell anergy: An early event in the
course of tumor progression. Proc Natl Acad Sci U S A 95, 1178-1183.
96
Street, S.E., Trapani, J.A., MacGregor, D., and Smyth, M.J. (2002). Suppression of lymphoma
and epithelial malignancies effected by interferon gamma. J Exp Med 196, 129-134.
Stutman, O. (1974). Tumor development after 3-methylcholanthrene in immunologically
deficient athymic-nude mice. Science 183, 534-536.
Sun, J.C., and Bevan, M.J. (2003). Defective CD8 T cell memory following acute infection
without CD4 T cell help. Science 300, 339-342.
Sun, T., Zhou, Y., Li, H., Han, X., Shi, Y., Wang, L., Miao, X., Tan, W., Zhao, D., Zhang, X., et
al. (2005). FASL -844C polymorphism is associated with increased activation-induced T cell
death and risk of cervical cancer. J Exp Med 202, 967-974.
Swann, J.B., Vesely, M.D., Silva, A., Sharkey, J., Akira, S., Schreiber, R.D., and Smyth, M.J.
(2008). Demonstration of inflammation-induced cancer and cancer immunoediting during
primary tumorigenesis. Proc Natl Acad Sci U S A 105, 652-656.
Taniguchi, T., Matsui, H., Fujita, T., Takaoka, C., Kashima, N., Yoshimoto, R., and Hamuro, J.
(1983). Structure and expression of a cloned cDNA for human interleukin-2. Nature 302, 305310.
Terabe, M., Matsui, S., Park, J.M., Mamura, M., Noben-Trauth, N., Donaldson, D.D., Chen, W.,
Wahl, S.M., Ledbetter, S., Pratt, B., et al. (2003). Transforming growth factor-beta production
and myeloid cells are an effector mechanism through which CD1d-restricted T cells block
cytotoxic T lymphocyte-mediated tumor immunosurveillance: abrogation prevents tumor
recurrence. J Exp Med 198, 1741-1752.
Thomas, D.A., and Massague, J. (2005). TGF-beta directly targets cytotoxic T cell functions
during tumor evasion of immune surveillance. Cancer Cell 8, 369-380.
Thompson, R.H., Kuntz, S.M., Leibovich, B.C., Dong, H., Lohse, C.M., Webster, W.S.,
Sengupta, S., Frank, I., Parker, A.S., Zincke, H., et al. (2006). Tumor B7-H1 is associated with
poor prognosis in renal cell carcinoma patients with long-term follow-up. Cancer Res 66, 33813385.
Thompson, R.H., Zang, X., Lohse, C.M., Leibovich, B.C., Slovin, S.F., Reuter, V.E., Cheville,
J.C., Blute, M.L., Russo, P., Kwon, E.D., and Allison, J.P. (2008). Serum-soluble B7x is elevated
in renal cell carcinoma patients and is associated with advanced stage. Cancer Res 68, 60546058.
Toes, R.E., Ossendorp, F., Offringa, R., and Melief, C.J. (1999). CD4 T cells and their role in
antitumor immune responses. J Exp Med 189, 753-756.
Tuveson, D.A., and Jacks, T. (2002). Technologically advanced cancer modeling in mice. Curr
Opin Genet Dev 12, 105-110.
Urban, J.L., Holland, J.M., Kripke, M.L., and Schreiber, H. (1982). Immunoselection of tumor cell
variants by mice suppressed with ultraviolet radiation. J Exp Med 156, 1025-1041.
97
Uyttenhove, C., Maryanski, J., and Boon, T. (1983). Escape of mouse mastocytoma P815 after
nearly complete rejection is due to antigen-loss variants rather than immunosuppression. J Exp
Med 157, 1040-1052.
Uyttenhove, C., Pilotte, L., Theate, I., Stroobant, V., Colau, D., Parmentier, N., Boon, T., and
Van den Eynde, B.J. (2003). Evidence for a tumoral immune resistance mechanism based on
tryptophan degradation by indoleamine 2,3-dioxygenase. Nat Med 9, 1269-1274.
van den Broek, M.E., Kagi, D., Ossendorp, F., Toes, R., Vamvakas, S., Lutz, W.K., Melief, C.J.,
Zinkernagel, R.M., and Hengartner, H. (1996). Decreased tumor surveillance in perforindeficient mice. J Exp Med 184, 1781-1790.
van der Bruggen, P., Traversari, C., Chomez, P., Lurquin, C., De Plaen, E., Van den Eynde, B.,
Knuth, A., and Boon, T. (1991). A gene encoding an antigen recognized by cytolytic T
lymphocytes on a human melanoma. Science 254, 1643-1647.
Van Dyke, T., and Jacks, T. (2002). Cancer modeling in the modern era: progress and
challenges. Cell 108, 135-144.
van Hall, T., Wolpert, E.Z., van Veelen, P., Laban, S., van der Veer, M., Roseboom, M., Bres,
S., Grufman, P., de Ru, A., Meiring, H., et al. (2006). Selective cytotoxic T-lymphocyte targeting
of tumor immune escape variants. Nat Med 12, 417-424.
Verdeil, G., Marquardt, K., Surh, C.D., and Sherman, L.A. (2008). Adjuvants targeting innate
and adaptive immunity synergize to enhance tumor immunotherapy. Proc Natl Acad Sci U S A
105, 16683-16688.
Vicari, A.P., Caux, C., and Trinchieri, G. (2002). Tumour escape from immune surveillance
through dendritic cell inactivation. Semin Cancer Biol 12, 33-42.
Vigneron, N., Stroobant, V., Chapiro, J., Ooms, A., Degiovanni, G., Morel, S., van der Bruggen,
P., Boon, T., and Van den Eynde, B.J. (2004). An antigenic peptide produced by peptide
splicing in the proteasome. Science 304, 587-590.
Viguier, M., Lemaitre, F., Verola, O., Cho, M.S., Gorochov, G., Dubertret, L., Bachelez, H.,
Kourilsky, P., and Ferradini, L. (2004). Foxp3 expressing CD4+CD25(high) regulatory T cells
are overrepresented in human metastatic melanoma lymph nodes and inhibit the function of
infiltrating T cells. J Immunol 173, 1444-1453.
Villablanca, E.J., Raccosta, L., Zhou, D., Fontana, R., Maggioni, D., Negro, A., Sanvito, F.,
Ponzoni, M., Valentinis, B., Bregni, M., et al. (2010). Tumor-mediated liver X receptor-alpha
activation inhibits CC chemokine receptor-7 expression on dendritic cells and dampens
antitumor responses. Nat Med 16, 98-105.
Vitale, M., Rezzani, R., Rodella, L., Zauli, G., Grigolato, P., Cadei, M., Hicklin, D.J., and
Ferrone, S. (1998). HLA class I antigen and transporter associated with antigen processing
(TAP1 and TAP2) down-regulation in high-grade primary breast carcinoma lesions. Cancer Res
58, 737-742.
98
Wahlin, B.E., Sander, B., Christensson, B., and Kimby, E. (2007). CD8+ T-cell content in
diagnostic lymph nodes measured by flow cytometry is a predictor of survival in follicular
lymphoma. Clin Cancer Res 13, 388-397.
Wang, H.Y., Lee, D.A., Peng, G., Guo, Z., Li, Y., Kiniwa, Y., Shevach, E.M., and Wang, R.F.
(2004). Tumor-specific human CD4+ regulatory T cells and their ligands: implications for
immunotherapy. Immunity 20, 107-118.
Wang, R.F., Parkhurst, M.R., Kawakami, Y., Robbins, P.F., and Rosenberg, S.A. (1996).
Utilization of an alternative open reading frame of a normal gene in generating a novel human
cancer antigen. J Exp Med 183, 1131-1140.
Wang, R.F., Wang, X., Atwood, A.C., Topalian, S.L., and Rosenberg, S.A. (1999). Cloning
genes encoding MHC class II-restricted antigens: mutated CDC27 as a tumor antigen. Science
284, 1351-1354.
Ward, P.L., Koeppen, H.K., Hurteau, T., Rowley, D.A., and Schreiber, H. (1990). Major
histocompatibility complex class I and unique antigen expression by murine tumors that
escaped from CD8+ T-cell-dependent surveillance. Cancer Res 50, 3851-3858.
Warren, E.H., Vigneron, N.J., Gavin, M.A., Coulie, P.G., Stroobant, V., Dalet, A., Tykodi, S.S.,
Xuereb, S.M., Mito, J.K., Riddell, S.R., and Van den Eynde, B.J. (2006). An antigen produced
by splicing of noncontiguous peptides in the reverse order. Science 313, 1444-1447.
Watanabe, N., Gavrieli, M., Sedy, J.R., Yang, J., Fallarino, F., Loftin, S.K., Hurchla, M.A.,
Zimmerman, N., Sim, J., Zang, X., et al. (2003). BTLA is a lymphocyte inhibitory receptor with
similarities to CTLA-4 and PD-1. Nat Immunol 4, 670-679.
Waterhouse, P., Penninger, J.M., Timms, E., Wakeham, A., Shahinian, A., Lee, K.P.,
Thompson, C.B., Griesser, H., and Mak, T.W. (1995). Lymphoproliferative disorders with early
lethality in mice deficient in Ctla-4. Science 270, 985-988.
Weitzman, S.A., and Gordon, L.I. (1990). Inflammation and cancer: role of phagocyte-generated
oxidants in carcinogenesis. Blood 76, 655-663.
Welsh, M.M., Applebaum, K.M., Spencer, S.K., Perry, A.E., Karagas, M.R., and Nelson, H.H.
(2009). CTLA4 variants, UV-induced tolerance, and risk of non-melanoma skin cancer. Cancer
Res 69, 6158-6163.
Wherry, E.J., and Ahmed, R. (2004). Memory CD8 T-cell differentiation during viral infection. J
Virol 78, 5535-5545.
Wherry, E.J., Blattman, J.N., Murali-Krishna, K., van der Most, R., and Ahmed, R. (2003). Viral
persistence alters CD8 T-cell immunodominance and tissue distribution and results in distinct
stages of functional impairment. J Virol 77, 4911-4927.
Wherry, E.J., Ha, S.J., Kaech, S.M., Haining, W.N., Sarkar, S., Kalia, V., Subramaniam, S.,
Blattman, J.N., Barber, D.L., and Ahmed, R. (2007). Molecular signature of CD8+ T cell
exhaustion during chronic viral infection. Immunity 27, 670-684.
99
Whitmore, M.M., DeVeer, M.J., Edling, A., Oates, R.K., Simons, B., Lindner, D., and Williams,
B.R. (2004). Synergistic activation of innate immunity by double-stranded RNA and CpG DNA
promotes enhanced antitumor activity. Cancer Res 64, 5850-5860.
Williams, C.J. (1982). Immunotherapy reassessed. Br Med J (Clin Res Ed) 284, 920-921.
Willimsky, G., and Blankenstein, T. (2000). Interleukin-7/B7.1-encoding adenoviruses induce
rejection of transplanted but not nontransplanted tumors. Cancer Res 60, 685-692.
Willimsky, G., and Blankenstein, T. (2005). Sporadic immunogenic tumours avoid destruction by
inducing T-cell tolerance. Nature 437, 141-146.
Willimsky, G., Czeh, M., Loddenkemper, C., Gellermann, J., Schmidt, K., Wust, P., Stein, H.,
and Blankenstein, T. (2008). Immunogenicity of premalignant lesions is the primary cause of
general cytotoxic T lymphocyte unresponsiveness. J Exp Med 205, 1687-1700.
Witkowski, J.A. (1990). The inherited character of cancer--an historical survey. Cancer Cells 2,
228-257.
Wolfel, T., Hauer, M., Schneider, J., Serrano, M., Wolfel, C., Klehmann-Hieb, E., De Plaen, E.,
Hankeln, T., Meyer zum Buschenfelde, K.H., and Beach, D. (1995). A p16INK4a-insensitive
CDK4 mutant targeted by cytolytic T lymphocytes in a human melanoma. Science 269, 12811284.
Wong, S.B., Bos, R., and Sherman, L.A. (2008). Tumor-specific CD4+ T cells render the tumor
environment permissive for infiltration by low-avidity CD8+ T cells. J Immunol 180, 3122-3131.
Woo, E.Y., Chu, C.S., Goletz, T.J., Schlienger, K., Yeh, H., Coukos, G., Rubin, S.C., Kaiser,
L.R., and June, C.H. (2001). Regulatory CD4(+)CD25(+) T cells in tumors from patients with
early-stage non-small cell lung cancer and late-stage ovarian cancer. Cancer Res 61, 47664772.
Woo, E.Y., Yeh, H., Chu, C.S., Schlienger, K., Carroll, R.G., Riley, J.L., Kaiser, L.R., and June,
C.H. (2002). Cutting edge: Regulatory T cells from lung cancer patients directly inhibit
autologous T cell proliferation. J Immunol 168, 4272-4276.
Wyckoff, J., Wang, W., Lin, E.Y., Wang, Y., Pixley, F., Stanley, E.R., Graf, T., Pollard, J.W.,
Segall, J., and Condeelis, J. (2004). A paracrine loop between tumor cells and macrophages is
required for tumor cell migration in mammary tumors. Cancer Res 64, 7022-7029.
Yang, L., DeBusk, L.M., Fukuda, K., Fingleton, B., Green-Jarvis, B., Shyr, Y., Matrisian, L.M.,
Carbone, D.P., and Lin, P.C. (2004). Expansion of myeloid immune suppressor Gr+CD11b+
cells in tumor-bearing host directly promotes tumor angiogenesis. Cancer Cell 6, 409-421.
Yang, L., Huang, J., Ren, X., Gorska, A.E., Chytil, A., Aakre, M., Carbone, D.P., Matrisian, L.M.,
Richmond, A., Lin, P.C., and Moses, H.L. (2008). Abrogation of TGF beta signaling in mammary
carcinomas recruits Gr-1+CD11b+ myeloid cells that promote metastasis. Cancer Cell 13, 2335.
100
Yang, L., and Moses, H.L. (2008). Transforming growth factor beta: tumor suppressor or
promoter? Are host immune cells the answer? Cancer Res 68, 9107-9111.
Yee, C., Savage, P.A., Lee, P.P., Davis, M.M., and Greenberg, P.D. (1999). Isolation of high
avidity melanoma-reactive CTL from heterogeneous populations using peptide-MHC tetramers.
J Immunol 162, 2227-2234.
Yu, H., Pardoll, D., and Jove, R. (2009). STATs in cancer inflammation and immunity: a leading
role for STAT3. Nat Rev Cancer 9, 798-809.
Yu, P., Lee, Y., Liu, W., Chin, R.K., Wang, J., Wang, Y., Schietinger, A., Philip, M., Schreiber,
H., and Fu, Y.X. (2004). Priming of naive T cells inside tumors leads to eradication of
established tumors. Nat Immunol 5, 141-149.
Zang, X., Loke, P., Kim, J., Murphy, K., Waitz, R., and Allison, J.P. (2003). B7x: a widely
expressed B7 family member that inhibits T cell activation. Proc Natl Acad Sci U S A 100,
10388-10392.
Zang, X., Thompson, R.H., Al-Ahmadie, H.A., Serio, A.M., Reuter, V.E., Eastham, J.A.,
Scardino, P.T., Sharma, P., and Allison, J.P. (2007). B7-H3 and B7x are highly expressed in
human prostate cancer and associated with disease spread and poor outcome. Proc Natl Acad
Sci U S A 104, 19458-19463.
Zhang, B., Bowerman, N.A., Salama, J.K., Schmidt, H., Spiotto, M.T., Schietinger, A., Yu, P.,
Fu, Y.X., Weichselbaum, R.R., Rowley, D.A., et al. (2007). Induced sensitization of tumor
stroma leads to eradication of established cancer by T cells. J Exp Med 204, 49-55.
Zhang, B., Karrison, T., Rowley, D.A., and Schreiber, H. (2008a). IFN-gamma- and TNFdependent bystander eradication of antigen-loss variants in established mouse cancers. J Clin
Invest 118, 1398-1404.
Zhang, B., Zhang, Y., Bowerman, N.A., Schietinger, A., Fu, Y.X., Kranz, D.M., Rowley, D.A.,
and Schreiber, H. (2008b). Equilibrium between host and cancer caused by effector T cells
killing tumor stroma. Cancer Res 68, 1563-1571.
Zhang, Q., Yang, X., Pins, M., Javonovic, B., Kuzel, T., Kim, S.J., Parijs, L.V., Greenberg, N.M.,
Liu, V., Guo, Y., and Lee, C. (2005a). Adoptive transfer of tumor-reactive transforming growth
factor-beta-insensitive CD8+ T cells: eradication of autologous mouse prostate cancer. Cancer
Res 65, 1761-1769.
Zhang, X., Miao, X., Sun, T., Tan, W., Qu, S., Xiong, P., Zhou, Y., and Lin, D. (2005b).
Functional polymorphisms in cell death pathway genes FAS and FASL contribute to risk of lung
cancer. J Med Genet 42, 479-484.
Zheng, X., Koropatnick, J., Li, M., Zhang, X., Ling, F., Ren, X., Hao, X., Sun, H., Vladau, C.,
Franek, J.A., et al. (2006). Reinstalling antitumor immunity by inhibiting tumor-derived
immunosuppressive molecule IDO through RNA interference. J Immunol 177, 5639-5646.
101
Zhou, G., Lu, Z., McCadden, J.D., Levitsky, H.I., and Marson, A.L. (2004). Reciprocal changes
in tumor antigenicity and antigen-specific T cell function during tumor progression. J Exp Med
200, 1581-1592.
Zitvogel, L., Apetoh, L., Ghiringhelli, F., Andre, F., Tesniere, A., and Kroemer, G. (2008). The
anticancer immune response: indispensable for therapeutic success? J Clin Invest 118, 19912001.
Zitvogel, L., Tesniere, A., and Kroemer, G. (2006). Cancer despite immunosurveillance:
immunoselection and immunosubversion. Nat Rev Immunol 6, 715-727.
Zou, W. (2005). Immunosuppressive networks in the tumour environment and their therapeutic
relevance. Nat Rev Cancer 5, 263-274.
Zou, W. (2006). Regulatory T cells, tumour immunity and immunotherapy. Nat Rev Immunol 6,
295-307.
Zou, W., and Chen, L. (2008). Inhibitory B7-family molecules in the tumour microenvironment.
Nat Rev Immunol 8, 467-477.
102
CHAPTER 2
Endogenous T cell responses to antigens expressed in lung adenocarcinomas
delay malignant tumor progression
Michel DuPage1, Ann Cheung1, Claire Mazumdar1, Monte M. Winslow1, Roderick
Bronson2, Leah M. Schmidt1, Denise Crowley1, Jianzhu Chen1, and Tyler Jacks1,3
1
Koch Institute for Integrative Cancer Research and Department of Biology, Massachusetts
Institute of Technology, Cambridge, MA 02139, USA.
2
Department of Biomedical Sciences, Tufts Cummings School of Veterinary Medicine, North
Grafton, MA 01536, USA.
3
Howard Hughes Medical Institute, Massachusetts Institute of Technology, Cambridge, MA
02139, USA.
The author generated the reagents and performed all experiments described with some
assistance from A. Cheung. C. Mazumdar, an undergraduate student supervised by the author,
performed IHC staining and quantification. M.M. Winslow analyzed human adenocarcinoma
array data. R. Bronson assisted with tumor grading analysis. L.M. Schmidt generated TRELucOS. D. Crowley prepared tissue sections for all stains. J. Chen provided TCR transgenic
mice. Experiments were performed in the laboratory of Tyler Jacks.
103
ABSTRACT
Neoantigens derived from somatic mutations in tumors may provide a critical link between the
adaptive immune system and cancer. However, it has been difficult to properly recapitulate
immune responses against tumor-specific neoantigens in models of the human disease. Here
we describe a novel system to introduce exogenous antigens into genetically engineered mouse
lung cancers to mimic tumor neoantigens. We show that endogenous T cells respond to and
infiltrate tumors, resulting in a significant delay in malignant progression. Despite continued
antigen expression, T cell infiltration does not persist and tumors ultimately escape immune
attack. In contrast, transplantation of cell lines derived from lung tumors that express the
antigens or prophylactic vaccination against autochthonous tumors, results in rapid tumor
eradication or selection of tumors that lose antigen expression. These results provide insight
into the dynamic nature of the immune response to naturally arising tumors and support clinical
data that suggest a role for the immune system in cancer suppression rather than prevention.
This model provides a valuable system for the optimization of immune therapies that can
eradicate human cancers.
104
INTRODUCTION
The potential for functionally important interactions between a developing tumor and the
immune system has been appreciated for over a century (Balkwill and Mantovani, 2001). The
association of tumor cells and lymphocytes has led researchers to postulate that the immune
system actively inhibits the formation and progression of transformed cells and ultimately
“shapes” nascent tumors by forcing the selective evolution of tumor cells that can evade the
immune response, a phenomenon called tumor immunoediting (Dunn et al., 2002; Khong and
Restifo, 2002; Schreiber, 2003). Significant numbers of T cells specific to mutated tumor
proteins have been identified in cancer patients (Finn, 2008; Lee et al., 1999; Novellino et al.,
2005; Parmiani et al., 2007; Rubio et al., 2003) and several studies have correlated immune
infiltration in tumors with improved prognosis (Buckanovich et al., 2008; Budhu et al., 2006;
Finak et al., 2008; Galon et al., 2006; Molldrem et al., 2000; Pages et al., 2005; Piersma et al.,
2007). However, the persistence of malignant disease despite immune recognition presents an
important medical and therapeutic question: how do tumors escape immune surveillance?
Several mouse models have been used to gain insights into the mechanisms by which
tumors may subvert immune responses, but each of these has critical limitations.
Transplantation of primary or cultured tumor cells is commonly used, but these models are
limited because they ectopically introduce large numbers of fully-developed tumor cells that
grow rapidly, and transplantation can initiate proinflammatory responses or traffic tumor cells
directly to lymphoid organs (Frese and Tuveson, 2007; Khong and Restifo, 2002; Ochsenbein et
al., 1999; Ochsenbein et al., 2001; Spiotto et al., 2002). Carcinogen-induced mouse models of
cancer have focused predominantly on sarcomas rather than cancers of epithelial origin and
these tumors are expected to harbor large numbers of mutations that may lead to artificially
robust immune responses (Dunn et al., 2002; Khong and Restifo, 2002). Transgenic mouse
models of cancer that develop tumors spontaneously and express model tumor antigens
105
throughout the targeted organs fail to fully recapitulate immune responses against the human
disease because antigen expression in normal tissues likely alters the T cell response, and
thymic deletion of antigen-specific T cells typically prevents the study of endogenous T cell
responses to tumors (Bai et al., 2008; Drake et al., 2005; Ercolini et al., 2005; Forster et al.,
1995; Getnet et al., 2009; Lyman et al., 2004; Muller-Hermelink et al., 2008; Nguyen et al.,
2002; Pellegrini et al., 2009; Savage et al., 2008; Smith et al., 1997; Speiser et al., 1997).
Furthermore, as transgenic models rely on sporadic tumor-initiating events that can vary widely
from tumor to tumor and mouse to mouse (Frese and Tuveson, 2007), it is difficult to follow the
dynamics of the T cell response over time.
In contrast, genetically engineered mouse models of many human cancers accurately
recapitulate both the genetic and histopathologic progression of the human disease from its
earliest lesions to metastasis and can provide spatiotemporal control of tumor onset (Frese and
Tuveson, 2007). However, few studies have employed these models to elucidate the interplay
between tumors and the immune system (Cheung et al., 2008; Clark et al., 2007; Huijbers et al.,
2006; Willimsky and Blankenstein, 2005). This is especially true in the context of lung cancer,
which is responsible for more than one million deaths each year worldwide.
In an effort to investigate the effect of T cell responses against tumor-specific
neoantigens on tumor progression, we developed a system to induce potentially immunogenic
autochthonous lung adenocarcinomas in a genetically engineered mouse model of the disease.
This model allowed us to carefully compare tumors that originate in situ that lack or express
model antigens in tumor cells, as well as to closely follow the immune response to tumors over
the entire course of tumor progression. These studies reveal a dynamic relationship between
the tumor and the immune system that ultimately influences tumor development. These data
have clear implications for understanding the role of the immune system in controlling human
lung cancer development.
106
RESULTS
Autochthonous lung tumors expressing tumor antigens are infiltrated by lymphocytes
early during tumor development
To generate autochthonous lung tumors in a spatiotemporally controlled fashion, we
utilized a model of human lung adenocarcinoma driven by the conditional expression of
oncogenic K-rasG12D in combination with the loss of p53 (Figure S1) (DuPage et al., 2009;
Jackson et al., 2005; Jackson et al., 2001). Tumor formation was initiated in K-rasLSL-G12D/+;p53fl/fl
mice by inhalation of lentiviral vectors expressing Cre recombinase with or without additional
antigens. To induce tumors that lack expression of model antigens, we used a lentiviral vector
expressing Cre alone (Lenti-x, Figure 1A). These tumors exhibited little to no lymphocyte
infiltration throughout their development as assessed by hematoxylin and eosin (H&E) staining
and immunohistochemistry (Figure 1B-1E). To induce tumors that express model tumor
antigens, we engineered lentiviral vectors that, in addition to Cre, express the T cell antigens
SIYRYYGL (SIY) and two antigens from ovalbumin – SIINFEKL (SIN, OVA257-264) and OVA323-339
– fused to the C-terminus of luciferase (Lenti-LucOS, Figure 1A). Fusing the antigens to
luciferase enabled us to monitor antigen expression in tumors. As shown in Figure 1, unlike
tumors induced with Lenti-x, Lenti-LucOS-induced tumors had large numbers of infiltrating
lymphocytes including both T and B cells at 8 weeks after tumor initiation (Figure 1B-1E). We
did not observe any differences in innate immune cell infiltrates between tumor types (Figure
S2). Based on the differences in immune cell infiltration, we designated Lenti-x-induced tumors
as non-immunogenic and Lenti-LucOS-induced tumors as immunogenic (Figure 1A and Figure
S1).
107
/?"
4
344
E44
D44
B44
C444
C344
B
RRR
-
!()*+
4
344
E44
D44
B44
C444
!"#
B
RRR
CD
R
!
!"#$&!()*+
!"#$%&'
@>. +<7&!=>
2334
@""FG5H8G$5$(98?5%#%$%I$%8#
CD
RRR
/01
+<7&!=> ψ )AA= $%&
/?"
@>. +<7&!=>
!"#$%&!()*+56<99(#8:"#%);
+<7&!=> ψ )AA= !"#
!"#$%&'5678#&%99(#8:"#%);
J5)"KKGL59935$(98?
#
"
J5)"KKGL59935$(98?
!"#$%&'
!"#$%&!()*+
0(AI:"5'()*+,M5%&'()*+,
$
,-.
4
34
E4
D4
B4
C44
A"?)"#$58N5$(98?G5O%$P
KQ9HP8)Q$%)5%#N%K$?I$"G
108
B
CD
3E
@""FG5H8G$5$(98?5%#%$%I$%8#
!"#$%&!()*+
!"#$%&'
/01
2334
Figure 1. Induction of lung tumors using lentiviral vectors.
(A) Design of Lenti-x and Lenti-LucOS constructs (SIN-LTR: self-inactivating long terminal
repeat, ψ: HIV packaging signal, cPPT: central polypurine tract, PGK: phosphoglycerate kinase
promoter, WRE: woodchuck post-transcriptional regulatory element, UbC: ubiquitin C promoter).
(B) Lung tumors induced with Lenti-x or Lenti-LucOS were stained with H&E, anti-CD3, or antiB220, eight weeks after tumor initiation. Scale = 50µm (inset 10µm).
(C and D) Quantification of immune infiltrates by IHC for CD3 and B220. P-values are ~10-8, and
~10-3 for CD3 and ~10-15 and 0.02 for B220 at 8 and 16 weeks, respectively. n= 2-5 mice, 16100 tumors, per group, mean ± SEM.
(E) Percent of lung tumors/ mouse containing infiltrating lymphocytes at 8, 16, and 24 weeks
after tumor initiation. n= 9-22 mice, 49-1594 tumors, per group, mean ± SEM.
109
T cell responses against tumor antigens are generated but not maintained
To examine the T cell response to LucOS expression in tumors over time, we assessed
the degree of T cell infiltration at 16 and 24 weeks after tumor initiation. In contrast to the robust
lymphocytic infiltration of immunogenic tumors at 8 weeks after tumor initiation, there was a
dramatic reduction in tumor-infiltrating lymphocytes at the later time-points, indicating that the
immune response was not sustained or tumors had escaped the immune response (Figure 1C1E). To determine whether the immune infiltrates in Lenti-LucOS tumors contained T cells
specific for the antigens engineered into the tumors, we performed flow cytometry with MHCI/Kb
DimerX reagents loaded with SIY or SIN peptides. To improve the sensitivity of the DimerX
reagent, we utilized both PE and APC labeled dimers to co-stain CD8+ T cells. At several time
points after tumor initiation with Lenti-LucOS, but not with Lenti-x, we detected CD8+ T cells
reactive to SIY/Kb and SIN/Kb in the lungs and the mediastinal lymph nodes draining the lungs
(DLN) (Figure 2A). As a positive control for a robust T cell response to the same antigens in the
same organ, we infected mice with a recombinant influenza virus engineered to express SIY
(WSN-SIY, Figure 2A). The antigen-specific T cell response to Lenti-LucOS tumors peaked
between 4 and 8 weeks after tumor initiation but declined thereafter in the lungs and DLN,
whereas T cells responding to SIY expressed by influenza expanded and contracted rapidly in
the lungs (within 2-4 weeks) but persisted with increasing frequency in the DLN (Figure 2B, 2C,
and Figure S3, S4A). We hypothesize that the delayed T cell response to tumors was due to
relatively low levels of antigen produced and presented in the lung DLNs early during tumor
development as well as a poor capacity for early-stage tumors to initiate a robust immune
response. To further investigate the early response, we analyzed the T cell response after LentiLucOS introduction into the lungs of mice that lacked the K-rasLSL-G12D allele, and could not form
tumors (Figure S5A-G). We found that SIY and SIN-specific T cells were generated in the
absence of tumor formation, presumably due to antigen expression in normal cells of the lung
110
F=@.P7:!"#$%&S:&'()*+,"
$.07Y:UF'LS:@42/7P.01Z7
!7/012G
α234#IJK
63#2345
α2345IJK
!"#$
!7/012!=AH3
α2345IJK
α234#IJK
@7?A7/0:<B:CD>;E<AD079
!
!"#$
%+-,
./012345
./01234#
%+,%
%+),
%+%%
%
@7?A7/0:<B:CD>;E<AD079
;7;01M7IJK2T@U
F!#
#
;7;01M7IJK2@R
$
!"#$
@7?A7/0:4N#γLO#Nα/7P
()
)%
(%
&
%
%
&
'
()
(*
,
%
%
67789:;<90:0=><?:1/101.01</
&
!"#$
(%:M.D9
,:Q7789
&
'
()
67789:;<90:0=><?:1/101.01</
':Q7789
(%%
(%%
(%%
'%
'%
'%
'%
*%
*%
*%
*%
W:<B:X.G
&%
)%
W:<B:X.G
&%
)%
%
(%
(%
(
(%
)
(%
V
(%
&
(%
(%
(
(%
)
(%
V
(%
&
F!#
)&
./012345
./01234#
%+%&
%+%)
%+%%
%
&
'
()
(*
)%
67789:;<90:0=><?:1/101.01</
)&
F!#
-,
,%
),
%
63#2 !7/012
345 !=AH3
(*:Q7789
UF'Lα2[34#:<?:345\IJK:L
UF'Lα2[34#:<?:345\IJK:/7P
)%
%
%
)%
W:<B:X.G
&%
)%
%
%
(*
(*:Q7789
(%%
W:<B:X.G
&%
(*
(%%
(,
'
()
%+%*
%
),
L:;7;01M7
/<:;7;01M7
(*
'
%+%'
@7?A7/0:4N#γLO#NαL
@7?A7/0:4N#γLO#NαL
!"#$
&
67789:;<90:0=><?:1/101.01</
%
(%
%
(%
(
(%
)
(%
V
(%
&
(%
%
(%
(
(%
)
(%
V
(%
&
@F2(
Figure 2. T cell responses to tumor antigens are generated but not sustained.
(A) Representative analysis of SIY and SIN specific CD8+ T cells in the lung and the draining
mediastinal lymph node (DLN) of Lenti-x or Lenti-LucOS tumor-bearing mice or WSN-SIY
infected mice. FACS plots are gated on PI-negative, CD8+ cells.
(B and C) Percent of lymphocytes specific for SIN or SIY during Lenti-LucOS tumor progression
in the lung or DLN. The percent of anti-SIN or anti-SIY reactive cells (of total lymphocytes) was
determined as shown in Figure S3. n= 2-5 mice per time-point, mean ± SEM.
(D) IFNγ and TNFα cytokine production in SIY and SIN-reactive T cells from the lungs of LentiLucOS tumor-bearing mice at several time points after tumor initiation. The percentage of SIY
and SIN-specific T cells from the lungs that were IFN-γ+TNF-α+ or IFN-γ+TNF-αneg in the
111
absence of stimulation (no peptide) or in the presence of SIY and SIN peptides (+ peptide) is
shown. Percentages were determined by staining duplicate samples for either DimerX (as in A)
or cytokine production as described in Figure S6A. n= 2-7 mice per time-point, mean ± SEM.
(E) The percentage of SIY and SIN-specific T cells in the DLN that were IFN-γ+TNF-α+ 16 weeks
after tumor initiation with Lenti-LucOS or 16 weeks after WSN-SIY infection. n= 4 mice per
group, mean ± SEM.
(F) PD-1 surface expression on anti-SIY/Kb+ and anti-SIN/Kb+ CD8+ cells (open histograms) or
non-specific CD8+ cells (filled histograms) from the lungs of Lenti-LucOS tumor-bearing mice at
several time-points after tumor initiation. n= 2-3 mice per group, mean ± SEM.
(Figure S5H-J). This indicates that while antigen expression is largely restricted to tumors in
this system, the lentiviral infection can stimulate antigen-specific T cells that may contribute to
the ensuing anti-tumor T cell response. Importantly, however, the T cell response in the
absence of transformation was weak and short-lived when compared to mice that generated
tumors (Figure S5A-F). This may indicate that the immune system responds differently to
transformed cells or that tumor proliferation within the first several days following K-rasG12D
activation leads to greater antigen production and, hence, more T cell activation.
Endogenous tumor-reactive T cells are not fully functional and display traits of terminally
differentiated effector cells
To determine whether T cells specific to the tumor antigens were functional, we
measured their capacity to produce IFN-γ and TNF-α at multiple time-points after tumor initiation
(Figure 2D, 2E and Figure S6). Very few tumor-reactive T cells in the lungs had the capacity to
produce both IFN-γ and TNF-α, and this activity was almost completely lost within the first five
weeks of tumor development, whereas T cells maintained production of IFN-γ alone (Figure
2D). This pattern of cytokine production during tumor progression may indicate an increase in
terminally differentiated effector T cells (IFN-γ+TNF-αneg) that is accompanied by a progressive
loss of high-quality, memory-like T cells (IFN-γ+TNF-α+) (Seder et al., 2008; Slifka and Whitton,
112
2000; Wherry et al., 2007). It is possible that persistent, non-productive, T cell engagement with
tumor antigens drives T cells into an exhausted state (Bucks et al., 2009; Mueller and Ahmed,
2009; Redmond and Sherman, 2005). Consistent with this, as tumors progressed, tumorreactive T cells showed increased expression of PD-1, a marker of T cell exhaustion, and low
expression of CD127 (the IL-7Rα chain), a marker of long-lived memory cells (Figure 2F and
not shown) (Wherry et al., 2007). Furthermore, during a productive T cell response against
WSN-SIY, viral clearance was followed by a rapid decline in SIY-specific T cells that were IFNγ+TNF-αneg and the maintenance of IFN-γ+TNF-α+ cells (Figure 2E and Figure S4B).
Adoptively transferred T cells primed by tumors rapidly lose activity
To develop a more detailed view of the dynamics of the T cell response to established
tumors from activation to effector function, we transferred naïve 2C or OT-I TCR transgenic T
cells (which recognize SIY or SIN, respectively) into mice bearing Lenti-x or Lenti-LucOS
tumors. As early as three days after transfer, 2C and OT-I T cells were activated and
proliferated in the lung-draining lymph nodes (Figure 3A and 3B). By 12-14 days after transfer,
the T cells could be detected in the lungs (Figure 3C). However, the activity of these T cells was
significantly impaired in the DLN and lungs compared to T cells responding to the same
antigens expressed in the context of infection with influenza (Figure 3D, Figure S4C). These
results emphasize that although T cells can be stimulated to proliferate and migrate to tumors,
few T cells retain the capacity to produce effector cytokines in response to tumors.
Two potential mechanisms may explain the dysfunction of naïve T cells responding to
antigens expressed by the Lenti-LucOS tumors: the presence of an immunosuppressive
environment in tumor-bearing mice or the suboptimal priming of T cells by antigens expressed
in tumors. To test whether the priming of T cells in the DLN is impaired, we assayed whether T
cell activity in the lungs could be rescued by transferring effector T cells that were activated in
113
B.!</"'!"#$%&Q'%&'()*+,
23
5
23
4
23
4
23
2
23
3
23
BCG
5
P<&"S'>B8TαE@FGHIJT
2
233
233
233
83
83
83
83
73
73
73
9'):';<=
63
9'):';<=
63
43
23 3
5
23
4
23
23
4
23
5
23
6
23
2L3
2
23
6
23
5
23
4
23
23 3
23
2
23
4
23
5
23
6
23
2L3
23 2
23 4
23 5
23 6
2
23 3
23 2
23 4
23 5
23
5
23
6
23
!
23
23
4
2
23 3
23
3
23
2
>B8
23
4
23
5
23
6
23
3
23
2
23
4
23
5
23
23 3
6
233
83
73
73
73
9'):';<=
63
43
43
3
3
3
23
3
23
2
23
4
23
5
23
6
23
3
23
2
23
M3
73
N3
63
53
43
23
3
U@GE@FG C"%&+EC.$K@
4
23
5
23
>B66
F?GγTVG?α'%"/
6
23 2
23 4
23 5
23 6
73
9'):';<=
63
9'):';<=
63
43
F?GγTVG?αT
BCG
23
6
83
43
3
23
3
23
2
23
4
23
5
23
6
23
3
23
>B74C
2
23
4
23
5
23
6
>B7D
V#<%1:"#'<$&+W<&"-'KVEF
!"#$"%&'$(&)*+%"',#)-.$+%/'$"001
23
!"#$"%&'$(&)*+%"',#)-.$+%/'$"001
25LM
23 5
>B6NL2
COGP
23 6
23
5
233
V#<%1:"#'%<+W"'KVEF
>B6NL4T'#"$+,+"%&1
4
83
>?@A
>B6NL2T'KVEF
23
2
233
9'):';<=
63
23 6
3
3
3
83
>B8
"
23
4
233
23 3
23 3
23
2
43
43
3
3
73
9'):';<=
63
9'):';<=
63
43
3
3
!CG
23
23
2
9'):'R<=L
23
αE@FGHIJ
!CG
23
6
3
C"%&+E=
C"%&+EC.$K@
233
9'):'R<=L
BCG
26L5
23
$
C"%&+E=
2LM
23 6
N3
63
53
43
23
3
U@GE@FG C"%&+EC.$K@
COGP
M3
73
N3
63
53
43
23
3
U@GE@FG C"%&+EC.$K@
BCG
!"#$"%&'$(&)*+%"',#)-.$+%/'$"001
C"%&+EC.$K@
23 6
!"#$"%&'$(&)*+%"',#)-.$+%/'$"001
#
N3
63
53
43
23
3
T2'-<(
T24'-<(
C"%&+EC.$K@
COGP
Figure 3. Adoptively transferred T cells are not fully functional in response to established
tumors.
(A) FACS analysis comparing the accumulation of SIN-reactive, OT-I T cells in the mediastinal
draining lymph nodes (DLN) and the inguinal peripheral lymph nodes (PLN) three days after
adoptive transfer of naïve OT-I cells into Lenti-x or Lenti-LucOS tumor-bearing mice 10 weeks
after tumor initiation (similar results were observed when T cells were transferred four weeks
after tumor initiation). Percent of total CD8+ cells that are anti-SIN/Kb+ is indicated. n≥ 3 mice for
Lenti-LucOS or ≥1 mouse for Lent-x tumor-bearing mice in each of two independent
experiments.
(B) Histogram plots of CD8+anti-SIN/Kb+ gated cells for CFSE dilution or surface expression of
CD44, CD62L, or CD69 from the DLN or PLN of Lenti-LucOS (open histogram) or Lenti-x (filled
histogram) tumor-bearing mice. Undivided (CFSEhi) CD8+anti-SIN/Kb+ cells were examined in
CD69 plots to detect initial T cell activation. n≥ 3 mice for Lenti-LucOS or ≥1 mouse for Lent-x
tumor-bearing mice in each of two independent experiments.
(C) FACS analysis for the presence of OT-I T cells in the lungs of Lenti-LucOS tumor-bearing
mice 12 days after adoptive transfer of naïve CD45.1+ OT-I T cells (tumor-bearing mice were
CD45.2+). n≥ 3 mice in at least three independent experiments.
(D) Percent of CD8+CD45.1+ gated OT-I T cells producing IFNγ+TNFα+ or IFNγ+TNFαneg from
the DLN or lung after transfer of naïve OT-I T cells. WSN-SIN represents analysis of OT-I T
cells 7 days after WSN-SIN infection. Lenti-LucOS represents analysis of OT-I T cells 12 days
114
after transfer into Lenti-LucOS tumor-bearing mice at time points between 12 and 16 weeks
after tumor initiation. n= 6-7 mice per group, mean ± SEM.
(E) Percent of CD8+CD45.1+ gated OT-I T cells producing IFNγ+TNFα+ or IFNγ+TNFαneg from the
DLN or lung after transfer of in vitro-activated OT-I T cells. Lenti-LucOS represents analysis one
day (lung only, n = 6 mice) and 12 days after transfer (n= 3 mice) into Lenti-LucOS tumorbearing mice at time points between 10 and 14 weeks after tumor initiation. WSN-SIN in the
DLN represents analysis of OT-I T cells 14 days after WSN-SIN infection (n= 3 mice). Mean ±
SEM.
vitro. Indeed, tumor-specific T cells activated in vitro and then transferred immediately had the
capacity to produce both IFN-γ and TNF-α and maintained this activity in the lungs of LentiLucOS tumor-bearing mice (Figure 3D and 3E). Furthermore, transfer of activated T cells led to
the generation of cells with comparable function to WSN-SIN-primed T cells in the DLN (Figure
3E). T cell activity may also be exacerbated by interactions with cells or secreted factors in the
tumor microenvironment. Foxp3+ T regulatory cells were found to specifically infiltrate LentiLucOS tumors (Figure 2F and Figure S7). Nevertheless, T cell priming by tumor antigens is
clearly suboptimal and contributes to the weak anti-tumor T cell response.
Antigen expression and presentation is maintained in immunogenic tumors
The loss of T cell infiltrates in tumors during tumor progression could also be due to the
selection of tumors that lose antigen expression or antigen presentation to escape the immune
response (Dunn et al., 2002; Khong and Restifo, 2002; Spiotto et al., 2004; Uyttenhove et al.,
1983; Ward et al., 1990; Zhou et al., 2004). To determine whether the T cell response was
driving the outgrowth of tumors lacking antigen, we compared luciferase activity in Lenti-LucOS
tumors generated in immune-competent and immune-compromised mice. At 16 and 20 weeks
post tumor initiation, time-points associated with the dissipation of immune infiltrates in tumors,
luciferase levels were comparable between normal and immune-compromised mice (Figure
4A). The presentation of tumor antigens to T cells in the lung-draining lymph node was also
115
maintained during the latest stages of disease (24 weeks after tumor initiation), because tumorspecific 2C T cells could still proliferate specifically in the draining lymph nodes of Lenti-LucOS
tumor-bearing mice (Figure 4B). Additionally, Lenti-LucOS tumors maintained higher
expression of MHC class I compared to Lenti-LucOS tumors that arose in Rag-2null mice or
compared to Lenti-x tumors (Figure 4C, 4D). However, MHC class I expression decreased on
Lenti-LucOS tumors between 16 and 24 weeks after tumor initiation (Figure 4D, 4E). Reduced
MHC class I expression could contribute to tumor escape and may result from the waning antitumor T cell activity and T cell infiltration into tumors. In support of this hypothesis we used cell
lines derived from the model to show that MHC class I expression and SIN presentation on
Lenti-LucOS tumor cells was regulated by IFNγ and that antigen-specific T cell recognition in
vitro was sufficient to upregulate MHC class I and allow for the specific killing of Lenti-LucOS
tumor cells (Figure S8). Therefore, the rapid decline in T cell infiltration into tumors may result
from the combined effect of a reduced anti-tumor T cell response that results in lower MHC
class I expression on tumor cells, ultimately reducing the potential for T cells to recognize tumor
cells. We found further support for this model in vivo by testing whether in vitro-activated 2C T
cells transferred into Lenti-LucOS tumor-bearing mice 16 weeks after tumor initiation, a time
when few tumors have endogenous lymphocytic infiltrates, could be specifically retained in
tumors. 2C T cells were specifically retained in Lenti-LucOS tumors but not in Lenti-x tumors
(Figure 4F), and activated polyclonal T cells were not retained in Lenti-LucOS tumors (Figure
4F). Furthermore, by staining tumor sections with the epithelial marker cytokeratin 8, we could
observe tumor-infiltrating T cells making direct contacts with the epithelial cancer cells
themselves rather than cytokeratin 8-negative stromal cells (Figure 4G, 4H). Thus, the
dissipating T cell response to immunogenic tumors during the later stages of tumor progression
was not due to the escape of tumors that had lost the capacity to present the tumor antigens.
116
EH8?,1+!"#$%&L+#$%&'()*
!"%
!"$
'()*+µ,+-./0123
!"%
!"&
!"#
(13025(HIJ6L+'?,5#3HNN
(13025(HIJ6L+'?,5#M
(13025@
!"&
!"#
!"!
!"!
!:+O11PQ
(13025(HIJ6
:A;
!4#+B#95C9'D
E(K
!" %
!A$
!4#5567*789
!"
&
!"
&
!"
#
!"
#
!"
!
!"
!
!" "
!" !
!" #
!" &
!" %
:"
"
!" "
!" !
!" #
!" &
!" %
!" "
!" !
9E;
,
'?,5#3HNN
'?,5#M
!:+O11PQ
!""
;"
;"
;"
:"
:"
%"
%"
#"
#"
:"
%"
#"
!" #
!" &
!" %
!""
!" !
!" #
!" &
!" %
!"
"
!"
!
!"
U5#VW
(13025(HIJ6+M+#9
(13025@+M+#9
&RA$+S+%A!
(13025(HIJ6+M+#9
(13025(HIJ6
(13025@
&""
"
!" "
U5#VW
.
-
#""
"
"
#
!" %
#%+O11PQ
!""
!" !
!" &
(13025(HIJ6
(13025@
!""
!" "
!" #
9F6G
<+/=+>?@
<+/=+>?@
(13025(HIJ6
(13025@
;"
#"
!" "
#"+O11PQ
!""
<+/=+>?@
%"
!" "
+
(13025@
!" %
<+/=+>?@
"
#"+O11PQ
U5#VW+B>FZD
'()*+µ,+-./0123
!
:A#+S+"AR
/
#
!"
&
!"
%
"
!:
#%
X11PQ+-/Q0+0HY/.+23202?02/3
(13025(HIJ6+M+-/NTIN/3?N
!A!+S+"A&
(13025(HIJ6+M+#9
117
Figure 4. Antigen expression and presentation is maintained in tumors.
(A) Luciferase activity (RLU/ µg protein) in Lenti-LucOS tumor lysates from K-rasLSL-G12D/+;p53fl/fl
or K-rasLSL-G12D/+;p53fl/fl;Rag-2-/- mice at 16 or 20 weeks after tumor initiation. n= 8-14 tumors per
mouse per group.
(B) FACS analysis comparing the accumulation of SIY-reactive, 2C T cells in the DLN three
days after adoptive transfer of naïve 2C cells into Lenti-x (n= 1) or Lenti-LucOS (n= 3) tumorbearing mice 24 weeks after tumor initiation. CFSE dilution of CD8+1B2+ cells in Lenti-LucOS
(open histogram) or Lenti-x (filled histogram) mice.
(C) Freshly isolated Lenti-LucOS tumors generated in Rag-2null (dashed line) and Rag-2+ mice
(solid line) 20 weeks after tumor initiation were pooled and then compared by FACS analysis for
H-2Kb surface expression after gating-out cells positive for CD45, CD31, Ter119, and IA/IE
(MHCII). Plot is representative of three pooled tumor samples from Rag-2null or Rag-2+ mice.
Filled histogram is a negative control stain for CD8.
(D) Freshly isolated Lenti-LucOS (solid line) and Lenti-x tumors (dashed line) at 16 and 24
weeks after tumor initiation were compared by FACS analysis for H-2Kb surface expression after
gating-out cells positive for CD45, CD31, Ter119, and IA/IE (MHCII). Plots are representative of
pooled and individual tumor samples from two independent experiments. Filled histograms are
negative control stains for H-2Kd.
(E) H-2Kb mean fluorescent intensity on Lenti-LucOS and Lenti-x tumors at 16 and 24 weeks
after tumor initiation. n= 2-3 mice, 6-11 tumor samples per group per time point. Mean ± SEM.
(F) Activated 2C T cells, or α-CD3/CD28 activated polyclonal CD8+ T cells, labeled with CFSE
were transferred into Lenti-LucOS or Lenti-x tumor-bearing mice at 16 weeks post tumor
initiation and tumors were analyzed for the presence of CFSE+ cells (green) 24 hours later.
DAPI counter stained. Tumors are outlined and quantification of the number of CFSE+ cells/
tumor is indicated. Scale = 50µm. n= 2-4 mice, 21-38 tumors, per group, mean ± SEM.
(G) Cytokeratin 8 (red) stained Lenti-LucOS tumors from (F) that received activated CFSE+ 2C
T cells (green) 16 weeks after tumor initiation. Scale = 50µm.
(H) High magnification of box outlined in (G). Scale = 10µm.
118
Immunogenic tumors exhibit delayed tumor progression
Despite the suboptimal anti-tumor T cell response, it remained possible that the immune
response might still affect the course of the disease. To understand the impact of the T cell
response on tumor progression, we first compared the size of immunogenic Lenti-LucOS tumors
to non-immunogenic Lenti-x tumors over the lifespan of tumor-bearing mice. As shown in Figure
5, at each time point examined, immunogenic tumors were smaller than non-immunogenic
tumors, with the differences in size being most significant at 8 and 16 weeks after tumor
initiation, coincident with T cell infiltration into tumors (Figure 5A). Analysis of tumors 8 weeks
after initiation revealed an increase in apoptosis in Lenti-LucOS compared to Lenti-x tumors but
no difference in proliferation (Figure S9A-B). Importantly, the difference in size between LentiLucOS and Lenti-x tumors required the presence of the adaptive immune system, because the
expression of the LucOS antigens had no effect on tumor size when induced in immunecompromised, Rag-2-deficient mice (Figure 5B). We also followed tumor progression by
examining the histological tumor grades at each time-point. Interestingly, at 8 and 16 weeks
after tumor initiation, Lenti-LucOS tumors were of lower grade compared to Lenti-x tumors
(Figure 5C). By 24 weeks post tumor initiation, the primary lung tumor grades were similar;
however, far fewer mice with immunogenic tumors had metastatic disease (Figure 5D). Thus,
early anti-tumor T cell responses can have long-lasting effects on tumor progression that delay
the most deadly phase of the disease. Indeed, analysis of human lung adenocarcinoma gene
expression datasets indicated that high expression of CD3 genes in tumor samples, as well as
other genes expressed specifically in T and B cells, was predictive of better patient outcomes
(Figure 5E and Figure S9C-F).
119
L2F?A-:!"#$%&M:%&'()*+,
"
,-./01,2345
,-./016
OOO
OOO
"#
F->3-./:<G:/2=<>9
,-./01,2345
,-./016
!&
#&
%&
$&
D>?E-:"
D>?E-:$
D>?E-:(
!:C--89
&'"
"#:C--89
,-./01,2345
,-./016
#&
)&
%&
(&
$&
"&
&
N-/?9/?/03:0.E-6
F->3-./:<G:/2=<>9
%&
(&
$&
O
&'!
&'#
&'%
&'$
&'&
"&
$%:C--89
D>?E-:"
D>?E-:$
D>?E-:(
"#:C--89
D>?E-:%
-
#&
'()*+
'()%,
"&&
F->3-./:92>P0P?B
)&
%&
(&
$&
"&
&
&'$
"'&
)&
&
&'(
*&
D>?E-:%
#&
F->3-./:<G:/2=<>9
$
,-./01,2345
,-./016
&'%
$%
&
.2BB
&'&
!
7--89:;<9/:/2=<>10.0/0?/0<.
#
&')::::::::::@?A1$
I2=<>:?>-?:J==$K
&'+
&'!
&'*
&'#
&')
&'%
&'(
&'$
&'"
&'&
,2.A:/2=<>9H:=<29-
I2=<>:?>-?:J==$K
!
!&
O
#&
%&
$&
&
D>?E-:"
D>?E-:$
D>?E-:(
$%:C--89
D>?E-:%
&
"$
$%
(#
%!
#&
I0=-:J=<./Q9K
120
Figure 5. Immunogenic tumors have delayed tumor progression.
(A) Tumor area, mean ± SEM, of Lenti-LucOS and Lenti-x lung tumors at 8, 16, and 24 weeks
after tumor initiation. p-values are ~10-10, ~10-14, and ~0.108, respectively, from n= 9-15 mice,
44-892 tumors, per group.
(B) Tumor area, mean ± SEM, of Lenti-LucOS and Lenti-x lung tumors in K-rasLSLG12D/+
;p53fl/fl;Rag-2-/- mice 16 weeks after tumor initiation. p ~0.561 from n= 6-7 mice, 45-223
tumors, per group.
(C) Tumor grades for Lenti-LucOS and Lenti-x lung tumors at 8, 16 and 24 weeks after tumor
initiation. n= 9-22 mice, 49-1594 tumors, per group. Mean ± SEM.
(D) The number of lung tumors/ mouse and metastatic index (number of mice with detectable
metastases/ total examined, p < 0.02) in Lenti-LucOS and Lenti-x tumor-bearing mice 24 weeks
after tumor initiation. n= 8-12 mice per group, mean ± SEM.
(E) Kaplan-Meyer plot comparing the survival of Stage I/II (lymph node metastases free, N0)
lung adenocarcinoma patients with high or low CD3 expression in tumors. Patients were ranked
from highest to lowest based on CD3 expression in tumors (average expression of CD3δ, CD3γ,
CD3ε, and CD247 (CD3ζ)) and the top versus bottom quartiles were compared, p < 0.02.
121
Loss of antigen expression in transplantable models of lung cancer
Tumor antigen loss has been reported as a mechanism of tumor immune evasion in
transplantable models of cancer (Spiotto et al., 2004; Uyttenhove et al., 1983; Ward et al., 1990;
Zhou et al., 2004). Because the endogenously arising lung tumors described above did not lose
antigen expression or presentation in the face of an anti-tumor immune response, we
hypothesized that this may be the consequence of a weaker immune response against
endogenous tumors compared to transplanted tumors. To compare immune responses in the
context of transplantable or endogenously arising models of cancer, we transduced cell lines
from the K-rasG12D-driven lung tumor model with lentiviral vectors to introduce either SIY fused
to luciferase (LKR10-LucS) or ovalbumin (containing SIN and OVA323-339) and SIY fused to
luciferase (LKR13-LucOS), and transplanted the cell lines subcutaneously into syngeneic
immune-competent mice (Figure S8A). While the parental LKR10 and LKR13 cell lines
produced tumors upon transplantation, the antigen-expressing LKR10-LucS and LKR13-LucOS
were rejected or exhibited delayed growth (Figure 6A and 6B) and generated CD8+ T cells
specific to SIN and/or SIY (Figure S10A-C). In stark contrast to the situation with
autochthonous tumors, all the tumors that grew in immune-competent mice lost expression of
the antigens (Figure 6C). Importantly, the parental and antigen-expressing cell lines were
equally capable of forming tumors in immune-compromised mice (Figure 6A and Figure S10D).
In addition, transferring naïve 2C or OT-I T cells into immune-compromised mice bearing
established LKR10-LucS or LKR13-LucOS tumors led to reductions in tumor volume and
antigen expression in an antigen-specific manner (Figure 6D and 6E). Thus, a T cell response
to a single antigen expressed in transplanted lung tumor cell lines was sufficient to elicit tumor
rejection or induce the selective outgrowth of tumors that had lost antigen expression.
It has been reported that spontaneous tumors can induce a state of systemic tolerance
toward tumor antigens such that T cells no longer respond to transplanted tumors expressing
122
6'L7K<+!"#$%&S+#&'()*+,
"
TGM+;'FG=9+=<U<(;<CE+;G;7@
#7K&/* #7K&/>'@@
%EV
%E/
!"#$4
%EB
!"#$4&!'(,) 4EB
%E3
>MCM
!"#$%
%E$%
!"#$%&!'() /EV
-'FG=+HG@'F<+IFF4J
.
-'FG=+HG@'F<+IFF4J
-
$5%%
!"#$%&!'() !"#$%&!'()
*+,-&.
*+/0
$%%%
A
5%%
%
!"#$4
!"#$4&!'(,)
!"#$4&!'(,)+#7K&/>'@@
4/%
1
$% $2 $3 // /1 4% 42 43 2/ 21
6789+7:;<=+;=7>9?@7>;
#
/3%
!
!"#$4&!'(,)+*+,-.E/0
!"#$%&!'()+*+,-&.
!"#$%&!'()+*+/0
/%%%
"L&!'(,)M4
/2%
B+C789
/%%
$2+C789
$1%
$/%
#7K&/>'@@ #7K&/*
3%
#7K&/>'@@ #7K&/*
2%
%
3
/B
6789+7:;<=+;=7>9?@7>;
$
%
#7K&/*
#7K&/>'@@
#!DE+$%5+(<@@9
$%B
$%1
$%
>'@@
#7K
*,-&.
*/0
5
$%2
$%4
$%/
#!DE+'K+?=G;<N>
/
,=;OG;G?N(+-P
$%5
$%2
$%
$%/
$%$
"L&
"L&
!"#&$4&
!'(,)M4 !'(,)M/ !'(,)
Q<:G=<+-P
R:;<=+-P
4
φ
"L&
"L&
!'(,)M4 !'(,)M/
Figure 6. Transplanted lung tumors induce strong T cell responses that force tumor
elimination or tumor antigen loss.
(A) Number of tumors rejected of the total transplanted subcutaneously into immune-competent
or immune-compromised (Rag-2null) 129S4/SvJae mice (n.d.= not done). Results are cumulative
from three independent experiments and transplantation of from 105 to 106 cells.
(B) Size of LKR13 and LKR13-LucOS tumors eight and 27 days after subcutaneous
transplantation into immune-competent 129S4/SvJae mice or LKR13-LucOS into Rag-2null
129S4/SvJae mice. LKR13-LucOS tumors that were rejected are not included. Results are
cumulative from three independent experiments. n= 2-5 mice per group.
(C) Representative in vivo luciferase activity of LKR10-LucS or LKR13-LucOS tumors
transplanted subcutaneously into Rag-2+ (n= 10 mice: 6 LKR10-LucS, 4 LKR13-LucOS) or Rag2null (n= 13 mice: 7 LKR10-LucS, 6 LKR13-LucOS) 129S4/SvJae mice.
(D) Tumor growth after subcutaneous transplantation of LKR10-LucS or LKR13-LucOS into
Rag-2null mice followed by transfer of naïve OT-I or 2C T cells (arrow) and measurement of
luciferase activity (star) (OTI/2C = OT-I or 2C transfer). n= 3-4 mice per group.
(E) Representative in vivo luciferase activity of LKR10-LucS tumors in Rag-2null mice that
received OT-I or 2C T cells from (D). All LKR13-LucOS tumors lost detectable luciferase activity
with transfer of OT-I or 2C T cells.
123
(F) Representative in vivo luciferase activity 7 and 14 days after subcutaneous transplantation
of a Lenti-LucOS induced lung tumor cell line (KP-LucOS.3) into Rag-2null or Rag-2+
129S4/SvJae mice.
(G) Similar to (D), Lenti-LucOS induced lung tumor cell lines (KP-LucOS.3 and KP-LucOS.2) or
LKR13-LucOS were transplanted subcutaneously into Rag-2null mice, then OT-I, 2C, or no T
cells were transferred and luciferase activity was measured in tumor cell lines generated from
the tumors > 28 days after transplantation.
(H) Lenti-LucOS induced lung tumor cell lines were orthotopically transplanted into 129S 4/SvJae
mice and luciferase activity in the grafted lung tumors was assayed and compared to the activity
before transplantation (φ = not detectable). Mean ± SEM.
the same antigens (Willimsky and Blankenstein, 2005). However, we found that even in the
context of mice bearing autochthonous lung tumors expressing the tumor antigens, T cell
responses were still effective against transplanted tumors expressing the same antigens
(Figure S10E). We suspect that leaky expression of tumor antigens in normal tissues, rather
than tumor-specific expression, may be responsible for the systemic tolerance to tumor antigens
in these transgenic tumor models (Bai et al., 2008; Cheung et al., 2008; Drake et al., 2005;
Willimsky and Blankenstein, 2005).
Finally, we tested whether such T cell responses would also occur if we transplanted cell
lines derived from our Lenti-LucOS-induced autochthonous lung tumors that had escaped the
immune response in situ without losing antigen expression. Indeed, Lenti-LucOS cell lines
transplanted subcutaneously into immune-competent mice lost antigen expression over time in
a T cell-dependent manner (Figure 6F and 6G). Furthermore, Lenti-LucOS lung tumor cell lines
lost antigen expression after orthotopic transplantation into the lung, indicating that
transplantation, even into a tumor’s native site, can induce T cell responses that drive tumor
antigen loss (Figure 6H).
124
Enhancing the anti-tumor T cell response with vaccination causes tumor elimination and
tumor antigen loss
Enhanced T cell priming in the context of transplanted tumors is a likely explanation for
tumor elimination and tumor antigen loss in this setting. Therefore, we reasoned that priming by
vaccination might elicit similar anti-tumor T cell responses against autochthonous tumors. To
examine this possibility, we vaccinated mice prophylactically with a dendritic cell line engineered
to express the model tumor antigens (DC2.4-LucOS) (Cheung et al., 2008) and subsequently
induced Lenti-LucOS or Lenti-x tumors. Vaccination dramatically reduced the number of LentiLucOS tumors but did not affect the number of Lenti-x tumors (Figure 7A and 7B). Furthermore,
the majority of Lenti-LucOS tumors that did develop in vaccinated mice had dramatically
reduced antigen expression (Figure 7C). This was likely due to an enlarged T cell response
early during tumor development, as well as a functionally enhanced response characterized by
increased secretion of effector proteins and the maintenance of CD127+ memory-like cells
(Figure 7D-F).
Despite the dramatic reduction of antigen expression in Lenti-LucOS tumors from
vaccinated mice, there were still significant numbers of tumor-infiltrating lymphocytes 8 weeks
after tumor initiation (not shown). This could be due to the capacity of T cells to recognize small
amounts of antigen expressed in tumors that is below the level of detection by the luciferase
assay. Indeed, T cells specific to the model antigens could still recognize Lenti-LucOS tumors in
DC2.4-LucOS vaccinated mice because in vitro-activated SIY-specific T cells were still retained
in tumors after adoptive transfer (Figure 7G). The maintenance of a T cell response to the
tumor antigens in vaccinated mice led us to ask whether the progression of Lenti-LucOS tumors
was altered in vaccinated compared to unvaccinated mice. Although Lenti-LucOS tumors in
vaccinated mice appeared only modestly smaller 16 weeks after tumor initiation, the differences
in tumor size between vaccinated and unvaccinated mice increased at 24 and 30 weeks after
125
>$.'Z)"!"#$%&["()*+,-./
!"#$%&'()"*$+,%-
64
54
34
74
64
54
34
4
4
34 7
4264
6
34
34
3
34
34 4
6
34 4
34 3
34 5
34 6
34
4
34
3
34
5
34
4
34
7
34 3
34 5
34 6
34 7
34
3
34
4
34
6
34
5
34
3
X&&)(*,%
MGF)
34 4
34 4
34 3
34 5
34 6
34 7
?>34O'
34 3
34 5
34 7
:44
4
34
5
34
3
34
3
34 7
34
6
34
5
34
3
:285
34 4
34 4
34 3
34 5
34 6
34 7
4249
34 7
34 4
34 3
34 5
5
34
3
34 7
5235
34 6
34
34 6
34 4
4
34
3
34
5
34
6
34
7
34 4
34 3
34 5
34 6
34 7
34 6
34 7
34 6
34
6
34
5
34
5
34
3
34
3
4
34
34 7
34 4
5D2:
7627
34 3
34 5
34 6
%'*G,]"32O
34 7
O326
5726
34 4
34 7
34 4
34
6
34 6
34
5
34
5
34
5
34
3
34
3
34
3
7D23
34 4
34 3
34 5
34 6
%'*G,]"42D
34
34 7
3O29
4
34 4
34 3
34 5
34 6
34 3
34 5
%'*G,]"724
34 7
9426
327D
34 4
34 7
34 6
%'*G,]"529
34 7
34 6
34 4
PQ<γ
\">?527@0$(A#
0)/*G@0$(A#
0)/*G@C
8729
O326
34 7
:829
%'*G,]"425
?>34O'
744
34 4
6
5
34 4
0)/*G@0$(A#
R)+,%N
3626
MGF)
34 7
3544
34 5
34 4
54
3D44
6
R2Q2P2
0)/*G@C
3
6
82O6
34
34 5
34
34
34 7
6
(
34
34 7
34
34
α−#PSTUV@.X
64
4
34
34
3
424D
34
<,"B'C >?527@0$(A#
74
34 4
34 7
4
84
82D3
34 5
34 5
58
<,"B'C
>?527@0$(A#
>?527@0$(A#
34 7
4278
6
34 4
0;<=
0)/*G@0$(A#
34
84
>0<
<,"B'C
34 7
R2Q2P2
&
O8
<,"B'C >?527@
0$(A#
<,"B'C >?527@
0$(A#
<,"B'C >?527@0$(A#
34 7
?>35O
!"#$%&'()"*$+,%-
74
84
$
344
0;<=
D4
84
PQ<γ
#
>0<
0)/*G@0$(A#
α−#PSTUV@W.?
"
0)/*G@C
.)%()/*"0$("1,-2"*$+,%-
!
34 4
34 3
34 5
34 6
34 7
%'*G,]"D329
?>77
>?527@0$(A#
628
528
<,"B'C
>?527@0$(A#
Y
Y
324
428
424
3D
57
64
E))F-"1,-*"*$+,%"G/G*G'*G,/
'
.)%()/*",&"*$+,%KG*L"MN+1L,(N*G("G/&GM*%'*)-
<,"B'C
H$+,%"'%)'"I++5J
%
0
<,"B'C
>?527@0$(A#
344
:4
D4
74
54
4
3D
57
64
E))F-"1,-*"*$+,%"G/G*G'*G,/
Figure 7. Vaccination in autochthonous lung tumor model reduces tumor burden and
promotes loss of antigen expression.
(A and B) Tumor number on the lungs of mice with Lenti-x or Lenti-LucOS tumors 16 weeks
after tumor initiation with or without prior DC2.4-LucOS vaccination. p-values are ~0.50 and
~0.0008, respectively.
(C) Freshly explanted Lenti-LucOS tumors were assayed for luciferase activity and the percent
of tumors that were luciferase positive (> 50 RLU/µg protein) in unvaccinated and DC2.4-LucOS
vaccinated mice is plotted. n= 5-6 mice, 27-34 tumors, per group.
126
(D) Representative FACS plots of SIY/Kb-specific T cells (gated on CD8) from the lung and DLN
of unvaccinated or DC2.4-LucOS vaccinated mice, 9 days after tumor initiation with LentiLucOS. n≥ 2 mice per group.
(E) FACS for surface-exposed CD107a and captured IFNγ (gated on CD8) after stimulation of
the DLN with SIY and SIN from unvaccinated or DC2.4-LucOS vaccinated, Lenti-LucOS or
Lenti-x tumor-bearing mice (16 weeks post tumor initiation). Mean fluorescent intensities of
CD107a and IFNγ on reactive T cells from Lenti-LucOS tumor-bearing mice with or without
vaccination. n= 2-4 mice per group, mean ± SEM.
(F) FACS analysis of CD8+ anti-SIY/Kb+- and anti-SIN/Kb+-specific T cells for CD127 and CD44
surface expression from the lungs or DLN of unvaccinated or DC2.4-LucOS vaccinated LentiLucOS or Lenti-x tumor-bearing mice (8 weeks post tumor initiation, n= 2-3 mice per group).
The ratio of CD44+CD127+ (memory-like) to CD44+CD127- (effector-like) is shown below each
FACS plot.
(G) Lenti-LucOS tumors from unvaccinated or DC2.4-LucOS vaccinated mice 8 weeks after
tumor initiation and 24 hours after transfer of SIY-reactive, CFSE-labeled activated T cells
(green). DAPI counter stained. Tumors are outlined. Scale = 50µm.
(H) Mean tumor area of Lenti-LucOS lung tumors at 16, 24 and 30 weeks after tumor initiation
with or without DC2.4-LucOS vaccination. p-values are ~0.050, ~0.038, and ~0.027,
respectively, from n= 3-10 mice, 70-506 tumors, per group at 16 and 24 weeks, and n= 1-2
mice, 15-46 tumors, per group at 30 weeks. Mean ± SEM.
(I) Percent of Lenti-LucOS tumors/ mouse containing infiltrating lymphocytes at 16, 24 and 30
weeks after tumor initiation with or without DC2.4-LucOS vaccination (analysis of mice from part
H). Mean ± SEM.
tumor initiation (Figure 7H). Vaccination had no affect on Lenti-x tumor size (Figure S11A). The
delayed tumor growth in vaccinated mice correlated with a prolonged retention of lymphocytic
infiltrates into tumors (Figure 7I and Figure S11B), again highlighting the potential for tumorinfiltrating lymphocytes to delay tumor progression.
DISCUSSION
Several important conclusions regarding immune-tumor interactions and their role in
regulating tumor progression are evident from our study that could not be directly addressed
using previous cancer models. Thus far, using genetically engineered mouse models, it has not
been possible to compare tumors expressing or lacking tumor-specific antigens (Clark et al.,
2007; Huijbers et al., 2006; Willimsky and Blankenstein, 2005). With the recent sequencing of
127
several cancer genomes, including lung adenocarcinomas, it is clear that human tumors contain
hundreds of mutated proteins and many of these have the potential to be presented on MHC
molecules, thus providing a means for T cells to recognize tumors as distinct from their
surrounding normal tissues (Ding et al., 2008; Segal et al., 2008; Sjoblom et al., 2006). Our
ability to specifically control the induction of tumor formation and tumor antigen expression has
allowed us to closely recapitulate the effects of T cell responses to tumor-specific neoantigens
from inception to malignant transformation.
In contrast to transgenic models of tumor-associated antigen expression, here we
followed endogenous T cell responses against antigens expressed in autochthonous lung
tumors in an attempt to model T cell responses to tumor neoantigens. By tracking the response
of endogenous T cells as well as transferred T cells, we find that suboptimal T cell responses
against tumors are not necessarily due to immune ignorance (Nguyen et al., 2002; Ochsenbein
et al., 1999; Ochsenbein et al., 2001; Speiser et al., 1997; Spiotto et al., 2002). However, the
expansion of T cells specific to the tumor antigens was markedly delayed, potentially allowing
tumor escape (Hanson et al., 2000; Ochsenbein et al., 2001), and the T cell response eventually
declined in both magnitude and quality as tumors progressed. Despite their ability to proliferate
and migrate to tumors, T cells had an altered phenotypic state that was exemplified by a poor
capacity to produce and secrete effector cytokines and a rapid loss of TNFα production in the
lung. The reduced capacity of T cells to produce effector cytokines such as IFNγ and TNFα as
tumors progressed may account for their inefficient migration to tumors over time (Calzascia et
al., 2007; Muller-Hermelink et al., 2008). Our ability to precisely time tumor initiation allowed us
to reveal a dynamic relationship between anti-tumor T cells and tumor cells. We provide
evidence to suggest that a loss of T cell activity could lead to reduced MHC class I expression
on tumor cells and contribute to the rapid decline in tumor-infiltrating lymphocytes. Tumorreactive T cells were also found to express high levels of PD-1 and low levels of CD127,
128
features that resemble T cell responses during chronic infections (Bucks et al., 2009; Seder et
al., 2008; Wherry et al., 2007). Indeed, the T cell response to lung tumors was quite distinct
from the response to the same antigens expressed by recombinant influenza virus after
infection of the lung. It is important to note that blocking the function of the PD-1/PD-L1 pathway
with the use of antibodies or PD-1-deficient T cells did not result in increased activity of tumorreactive T cells (not shown), indicating that signaling through the PD-1/PD-L1 pathway alone is
not responsible for the loss of T cell reactivity against tumors in this model. Instead, T cell
exhaustion might be exacerbated by an ineffective anti-tumor immune response that allows
prolonged T cell exposure to antigens expressed from tumors. Chronic T cell stimulation with
cognate antigen has been shown to be sufficient to drive exhaustion in several settings (Bucks
et al., 2009; Drake et al., 2005; Mueller and Ahmed, 2009; Redmond and Sherman, 2005). We
also found that T regulatory cells were recruited specifically to tumors expressing the LucOS
antigens, and T regulatory cells have been shown to suppress anti-tumor responses in several
cancer models (Clark et al., 2007; Ercolini et al., 2005; Pellegrini et al., 2009). Despite the
evidence of an incomplete anti-tumor T cell response, the immune response still delayed tumor
progression, indicating that suboptimal T cell responses that do not eliminate tumors are
capable of slowing cancer progression (Koebel et al., 2007; Shankaran et al., 2001). This
mirrors observations made here and in several retrospective studies of human cancer patients
in which immune infiltration into tumors significantly correlated with improved patient outcomes
(Buckanovich et al., 2008; Budhu et al., 2006; Dunn et al., 2002; Finak et al., 2008; Galon et al.,
2006; Molldrem et al., 2000; Pages et al., 2005; Piersma et al., 2007).
Prophylactic vaccination provided substantial protection from autochthonous tumors, but
it also led to prolonged immune infiltration into the Lenti-LucOS tumors that grew, delaying
tumor progression even further. Interestingly, this occurred despite dramatically reduced levels
of antigen expression in lung tumors from vaccinated mice. Delayed tumor progression despite
129
antigen loss may occur through direct targeting of tumor cells by T cells or via indirect
elimination of tumor cells during T cell-mediated destruction of stromal cells that cross-present
tumor antigens in the tumor microenvironment (Spiotto et al., 2004). This result has important
implications for immunotherapy because it indicates that generating large numbers of highly
functional, tumor-reactive T cells that persist long-term can effectively inhibit tumor progression
even after the selective outgrowth of tumors that express very low levels of the targeted antigen.
The majority of studies investigating immunity to cancer have focused on transplantable
models of cancer (Dunn et al., 2002; Hanson et al., 2000; Khong and Restifo, 2002; Koebel et
al., 2007; Ochsenbein et al., 2001; Schreiber, 2003; Shankaran et al., 2001; Spiotto et al., 2002;
Street et al., 2002; Thomas and Massague, 2005). These studies have led to the discovery of
several important mechanisms in both T cells and tumor cells that can modulate the
effectiveness of anti-tumor T cell responses or allow for tumor escape. However, these models
may not accurately reflect the human disease (Khong and Restifo, 2002; Ochsenbein et al.,
1999; Ochsenbein et al., 2001). Here we have used comparable models of transplantable and
autochthonous lung cancer to show that transplantation induces anti-tumor T cell responses
capable of driving tumor immunoediting. Even orthotopic transplantation of cell lines derived
from Lenti-LucOS-induced tumors that had escaped immune surveillance in their autochthonous
setting led to an adaptive immune response that forced the outgrowth of tumors with reduced
antigen expression, a phenomenon that never occurred with autochthonous tumors expressing
these antigens. This was due in part to suboptimal priming of T cells in the context of
autochthonous tumors. T cells activated in vitro had enhanced activity compared to T cells
stimulated in vivo and priming of T cells by vaccination led to responses against autochthonous
tumors that were comparable to responses against transplanted tumors. It is possible that
inefficient priming results from a lack of help from CD4+ T cells (Lyman et al., 2004). Indeed,
CD4+ OT-II T cells specific for OVA323-339, proliferated only weakly in response to tumor initiation.
130
From our comparison of autochthonous and transplanted tumors, we conclude that the immune
system recognizes tumors that originate from transformation of somatic cells in an inherently
different way than transplanted tumor cell lines.
Future experiments should address the diversity of anti-tumor immune responses that
likely stem from subtle differences in the models used for the investigations. The development
of models that link the tumor antigen to the driving oncogenic event will be useful for
investigating whether T cell responses to oncogenic antigens, whose expression in tumors
cannot easily be downregulated to evade an immune response, instigate alternative
mechanisms of tumor escape (Willimsky and Blankenstein, 2005). It will also be informative to
develop models in which tumor initiation and expression of tumor-specific antigens are
temporally separable to discern whether T cell responses are different if the induction of antigen
expression occurs in established tumors, where mutated proteins most likely arise in human
cancers. Indeed, using these lentiviral vectors to initiate tumor development and introduce the
antigens has some important limitations. Specifically, these lentiviruses require viral infection
concomitant with tumor initiation, and they cannot completely restrict the expression of the
antigens to tumor cells. However, preliminary experiments with lentiviral systems that can
restrict the expression of the antigens to lung epithelial cells, as well as experiments in which we
delayed antigen expression to several days after the lentiviral infection, indicate that inducing
the expression of these antigens in lung epithelial cells is sufficient to generate antigen-specific
T cell responses. Further investigation utilizing this system will be necessary to determine
whether priming during tumor initiation in the context of the lentiviral infection alters the outcome
of the anti-tumor immune response. Finally, discoveries made in a particular model of cancer
have often been accepted as broadly applicable to all immune-tumor interactions. However, the
immune response and tumor response are likely to vary greatly depending on the originating
tissue and the genetic pathology of the disease. Therefore, it will be important to extend these
131
analyses to other genetically engineered mouse models of cancer to discern whether T cell
responses and tumor development change depending on the context of the disease.
MATERIALS AND METHODS
Mice and tumor induction
129S4/SvJae strains backcrossed 8 generations were used for all tumor transplant experiments
and most autochthonous tumor studies. C57/BL6 mice backcrossed 5 generations were used
when transgenic T cells from C57/BL6 mice were adoptively transferred (Figures 3 and 4B).
Trp53fl mice were provided by A. Berns (Jonkers et al., 2001), K-rasLSL-G12D were generated in
our laboratory (Jackson et al., 2001), and Rag-2-/- mice were purchased from The Jackson
Laboratory (Bar Harbor, ME). Lung tumors were induced in K-rasLSL-G12D/+;p53fl/fl mice by
intratracheal intubation and inhalation of viruses expressing Cre recombinase as reported
previously (DuPage et al., 2009; Jackson et al., 2005). J. Chen provided the 2C, OT-I, and OT-II
TCR transgenic mice that were C57/BL6 and crossed into Rag-2-/-. CCSP-rtTA mice (used in
Figure S5H-J) specifically express the reverse tetracycline transactivator in lung epithelial cells
using the clara cell secretory protein promoter described in (Meylan et al., 2009). All animal
studies and procedures were approved by the Massachusetts Institute of Technology’s
Committee for Animal Care.
Lentiviral production
Lentiviruses were produced by transfection of 293T cells with Δ8.2 (gag/pol), CMV-VSV-G, and
the transfer vector expressing Cre as previously described (Tiscornia et al., 2006). TRELucOS
(used in Figure S5H-J) is similar to Lenti-LucOS but uses a doxycycline-regulated promoter
element (TRE) to control LucOS expression rather than the constitutive UbC promoter (Meylan
et al., 2009).
132
Histology, tumor size, tumor grading
Mice were killed at designated time-points and lung tissues were preserved by inflation and
fixation with 3.6% formaldehyde solution or frozen in O.C.T. for experiments that followed
migration of CFSE-labeled cells in tumors. Bioquant Image Analysis software was used to
measure the area of individual tumors in histological sections from the lungs. Grading was
performed by R.B. as described previously (DuPage et al., 2009). Briefly, Grade 1 lesions
represent hyperplasias and small adenomas, Grade 2 lesions are large adenomas, Grade 3
lesions are adenocarcinomas, Grade 4 lesions are invasive adenocarcinomas. Metastasis to
distant organs was monitored by gross examination of mice during necropsy followed by
histological examination of metastases observed, whereas lung-draining lymph nodes were
always examined by histological analysis for local metastases.
Immunohistochemistry
We performed immunohistochemical analysis on formalin-fixed paraffin sections after heatmediated antigen retrieval in sodium citrate (see Abcam protocol) (Johnson et al., 2001). We
used α-CD3 (1:100, Dako), α-B220 (1:50, RA3-6B2), α-Mac-3 (1:10, M3/84) (BD Pharmingen),
α-Foxp3 (1:25, FJK-16s) (eBioscience), α-cytokeratin 8 (1:25) primary antibodies and
Vectastain ABC secondary reagents followed by DAB substrate detection (Vector Laboratories).
Flow cytometry
Cell suspensions from lymphoid organs were prepared by mechanical disruption between
frosted slides. Cell suspensions from lungs were generated by mincing and digesting the tissues
for ~1 hour at 37°C in 125 U/ml Collagenase Type I (Gibco) and 60 U/ml Hyaluronidase
(Sigma). Single cell suspensions were stained with the specified antibodies for 20-30 min after
treatment with FcBlock (BD Pharmingen). α-CD8α (53-6.7), α-CD45.1 (A20), α-mouse IgG1
133
(X56), α-PD-1 (J43), α-IFNγ (XMG1.2), α-TNFα (MP6-XT22), α-CD107a (ID4B), α-CD44 (IM7),
α-CD62L (MEL-14), α-CD69 (H1.2F3), and DimerX I (Dimeric Mouse H-2Kb:Ig) were from BD
Pharmingen. α-CD127 (A7R34) was from eBioscience. J. Chen provided the 1B2 antibody
recognizing the 2C TCR (Kranz et al., 1984). H. Eisen provided the 25-D1.16 antibody that
recognizes SIN-loaded Kb/MHCI (Porgador et al., 1997). All antibodies were used at 1:200.
Peptide-loaded DimerX reagents were prepared as specified by manufacturer and used at 1:75.
Staining cells with the same ligand labeled with PE or APC for improved sensitivity has been
reported (Stetson et al., 2002; Townsend et al., 2001). Propidium iodide was used to exclude
dead cells when appropriate. Cells were read on a FACSCalibur or LSR II (BD Biosciences) and
analyzed using Flowjo software (Tree Star).
Cytokine production
Single cell suspensions from lungs or lymphoid organs were resuspended in the presence or
absence of SIYRYYGL and/or SIINFEKL peptides in OPTI-MEM I (Gibco) supplemented with
GolgiPlug (BD Pharmingen) for ~4 hours at 37°C, 5% CO2. Cells were then fixed and stained for
intracellular cytokines using the Cytofix/Cytoperm kit (BD Biosciences). Alternatively, for IFNγcapture, cells were stained with the capture and detection antibodies according to the
manufacturer’s protocol (Miltenyi Biotec).
Luciferase detection
Freshly explanted tumors or cell lines were lysed in Cell Culture Lysis Reagent, mixed with
Luciferase Assay Reagent according to manufacturer’s instructions (Promega), and relative light
units (RLU) were detected using the Optocomp I luminometer (MGM Instruments). RLUs were
standardized based on the total number of cells or total protein (Bio-Rad Protein Assay) used in
134
each assay. In vivo bioluminescence images were obtained using the NightOWLII LB983
(Berthold Technologies) after intraperitoneal injection of 1.5 mg Beetle Luciferin (Promega).
CFSE-labeling and adoptive transfer
Naïve and activated T cells were labeled with 5µM CFSE (Molecular Probes) for 10 min at
37°C. Naïve TCR transgenic T cells were harvested from lymphoid organs, stained, and 5x105106 CD8+ cells were transferred intravenously. Alternatively, freshly harvested TCR transgenic T
cells were stimulated in vitro with cognate peptide for 16-20 hours and then grown in completeRPMI supplemented with 5 ng/ml IL-2 (R&D Systems) for 6-7 days to generate activated T cells.
~107 in vitro-activated T cells were then transferred intravenously either unstained or stained
with CFSE depending upon the experiment.
Tumor cell lines and transplantation
LKR10 and LKR13 are cell lines derived from K-rasG12D-driven mouse lung tumors from
129S4/SvJae mice (Johnson et al., 2001; Kumar et al., 2007). LKR10-LucS and LKR13-LucOS
were generated by lentiviral infection with LucS or LucOS constructs (no Cre). KP-LucOS.2, KPLucOS.3 cell lines were generated from Lenti-LucOS-induced lung tumors from 129S4/SvJae
mice. KP-x.1 was generated from a Lenti-x-induced lung tumor from a129S4/SvJae mouse.
From 105-106 cells were used for transplantations either subcutaneously or intravenously
(“orthotopic”) depending on the particular experiment. Immune-competent 129S4/SvJae mice
receiving transplants came from the same mouse strains used to generate the autochthonous
tumors. Immune-compromised recipients were Rag-2-deficient. Subcutaneously transplanted
tumor volumes were calculated by multiplying the (length x width x height) of each tumor.
135
DC2.4-LucOS vaccination
DC2.4-LucOS cells were generated from the dendritic cell line (DC2.4) (Shen et al., 1997) after
infection with lentivirus that expresses LucOS (no Cre). Mice were vaccinated with an
intraperitoneal injection of 5x105 cells >1 month prior to lung tumor initiation.
Influenza
WSN-SIY and WSN-SIN influenza strains were provided by J. Chen (Bai et al., 2008). Mice
were infected with 20 pfu of virus/ mouse by intratracheal intubation and inhalation.
Statistical analyses
P-values from unpaired two-tailed student’s T-tests were used for all statistical comparisons with
the following exceptions: Figure 7F we utilized Welch’s correction because the variance in
vaccinated and unvaccinated mice differed significantly, Figure 5D we utilized the Fischer exact
probability test, and Figure 5E we utilized the logrank test.
ACKNOWLEDGEMENTS
We would like to thank A. Charest and P. Sandy for providing lentiviral vectors, A. Woolfenden
and C.F. Kim for training in the intracheal intubation technique, G. Paradis for assistance with
flow cytometry, E. Vasile for assistance with microscopy, H. Eisen for reagents and
experimental discussions, and A.G. DuPage, D. Feldser, T. Staton, and C. Kim for critical
reading of this manuscript. This work was supported by Grant 1 U54 CA126515-01 from the NIH
and partially by Cancer Center Support (core) grant P30-CA14051 from the National Cancer
Institute and the Margaret A. Cunningham Immune Mechanisms in Cancer Research Fellowship
Award (M.D.) from the John D. Proctor Foundation. M.M.W. was a Merck fellow of the Damon
Runyon Cancer Research Foundation and is a Genentech postdoctoral fellow. T.J. is a Howard
Hughes Investigator and a Daniel K. Ludwig Scholar.
136
!".&/0,1)00
!"#$%
!"#$%
%
-+./*010+!"#$%
-+./*23%
45678%
45678%
0:)>?+'
&'()*%#+",%
0:)>?+0;<=1
&'()*+,-'*
0;<=1%
9.:%
%
-+./*8('+!"#$%
%
./01%2345
9.:%
-+./*8('+!"#$%
-+./*23%
-+./*23%
456!"
456!"
456!"
456!"
!"#$%&&'#"()#*+,-'&".
%&&'#"()#*+,-'&".
2*('.),345%-6)%7$./8939$:4;<=>?@ABC0=C0,&"'8),&"D)0,"C,6'&/#,0'#(,/D)#"+/.+*#"&/5%
"6'%-+./*
010+!"#$
%)77'7'%)77(89%:(;%+6'%<(*=,+,(*)7%'>?;'99,(*%(:%(*<(-'*,<%3@;)9ABCD%:;(E%,+9%'*=(-'*(F9%7(<F9%4G%+6'%'*-,*'';,*-%
(:%)%+;)*9<;,?+,(*)7%)*=%+;)*97)+,(*)7%9+(?%'7'E'*+%:7)*H'=%4G%I(>$%9,+'9%,*%:;(*+%(:%+6'%:,;9+%<(=,*-%'>(*%(:%+6'%EF+)*+%-'*'J%"6'%
45678%)77'7'%2-'*';)+'=%4G%+6'%7)4(;)+(;G%(:%KJ%L';*95%<(*+),*9%I(>$%9,+'9%9F;;(F*=,*-%'>(*9%C@BM%(:%+6'%'*=(-'*(F9%?NO%-'*'J%
Figure
S1. The K-rasLSL-G12D/+;p53fl/fl mouse model of human lung adenocarcinoma.
K:+';%7'*+,P,;)7%,*+;(=F<+,(*%(:%0;'%2I'*+,@>5Q%*(*@,EEF*(-'*,<%+FE(;9%);'%:(;E'=%=F'%+(%+6'%;'E(P)7%(:%+6'%9+(?%'7'E'*+%,*%:;(*+%
The
K-rasLSL-G12D allele
allows for the conditional expression of oncogenic K-rasG12D from its
ABCD
%)77'7'%)*=%+6'%='7'+,(*%(:%*');7G%)77%(:%+6'%?NO%<(=,*-%'>(*9J%1EEF*(-'*,<%+FE(;9%);'%-'*';)+'=%
(:%+6'%(*<(-'*,<%3@;)9
endogenous
locus by the engineering of a transcriptional and translational stop element flanked
):+';%+6'%,*+;(=F<+,(*%(:%*'()*+,-'*9%<(*+),*'=%(*%+6'%7'*+,P,;F9%2I'*+,@IF<#!5J
by LoxP sites in front of the first coding exon of the mutant gene. The p53fl allele (generated by
the laboratory of A. Berns) contains LoxP sites surrounding exons 2-10 of the endogenous p53
gene. After lentiviral introduction of Cre (Lenti-x), non-immunogenic tumors are formed due to
the removal of the stop element in front of the oncogenic K-rasG12D allele and the deletion of
nearly all of the p53 coding exons. Immunogenic tumors are generated after the introduction of
exogenous antigens contained on the lentivirus (Lenti-LucOS).
137
5$6*78;
5$6*785+#9:
"
!"#$%%&'"(()"*+(,-
!
./#0
)44
5$6*785+#9:
5$6*78;
244
4
1
23
<$$=&">,&*"*+(,-"767*7/*7,6
#$%&'()*+,"-..&/0%(/$1)2(/3$42&15*)3&.0'6)7'()/03)8$%89:)$/;$93'73(<)=:"$//73()
$..&/()1(996,"
?@A"5+6B"*+(,-&"76C+#$C"D7*E"5$6*78;",-"5$6*785+#9:"76"!"#$%&'&"()*+,-./0123,23"(7#$"&*/76$C"
!"#$%&'#"()&*+%,"-$#%!,,./%&0#,1%#2341("'"#"&#"4'5%6*&7,%8%9:;3%<"'/,#%=:;3>5%
?FA"G+/6*7H7#/*7,6",H"*E$"(/#-,>E/B$"76H7%*-/*$&"IJ"KLM"?6N")8O"(7#$P"228QR"*+(,-&P">$-"
B-,+>AS
TTU/*E,%,B7#/%"/6/%J&7&",H"LVW"&*/76$C"*+(,-"&$#*7,6&"C7C"6,*"-$X$/%"*E$">-$&$6#$"
,H"/6J",*E$-"*J>$&",H"*+(,-876H7%*-/*76B"7((+6$"#$%%&"&+#E"/&"6$+*-,>E7%&P"I/&,>E7%&P",$,&76,>E7%&S
138
&!
'
&=!"#&.>;//4
#%()
&!!!
%!!
&!
#
&!
'
&!
"
*+,-./01234
?$%$.>;//4
%(%"
&!
$!!
&!
"
&
&!
!
56%
AA5
#!!
&!
"!!
!
!
"!!
#!!
@A5
$!!
%!!
&!!!
!
"!!
@A5
#!!
$!!
%!!
&!!!
&!
&
&!
!
&! #
2;2BCD,F*5
#
*+
NOP8
&!
89:;<
*+,
&!?%#).>;//4
)'(&
*+,-./01234-.56%7
"&&.>;//4
"(&%
'
&!
"
&!
&
&! !
!
"!!
#!!
@A5
$!!
%!!
&!!!
&! !
&! &
&! "
&! '
&! #
2;2BCD,*E
>9/>M/9:;<
G.HI./0123H>0:;4.J.!(&?.K"&&B.&!?%#).>;//4L
!"#$%&'()*.+&%,&-.'/0'12345/,2.&6'#7."-#'4%/,&8$%&*.
Q;2R;4;S:9:TU;.@F5A.2/H:4.IRH1.9.4TSV/;.>;//.4M42;S4THS.HI./MSV.>;//4.IRH1.9.
N;S:T,NM>WA.:M1HR,D;9RTSV.1HM4;.X;1HS4:R9:;4.:3;.V9:TSV.2RH>;XMR;.IHR.>9/>M/9:TSV.
:3;.2;R>;S:.HI.:H:9/./0123H>0:;4.42;>TIT>.IHR.A+Y.HR.A+P(..5;//4.Z;R;.V9:;X.TS.HRX;R.IRH1.
/;I:.:H.RTV3:.:H.VTU;.:3;.:H:9/.SM1D;R.HI.*+,S;V9:TU;-./0123H>0:;7.KD94;X.HS.@A5.9SX.AA5L-.
56%7-.2;2BCD,*E7-.2;2BCD,F*57.>;//4(.[3;.:H:9/.SM1D;R.HI.>;//4.V9:;X.TS.;9>3.2/H:.T4.43HZS.
9DHU;.:3;.2/H:4.K2;R>;S:.V9:;X.TS.2/H:4L(.\;.>9/>M/9:;X.:3;.2;R>;S:.HI./0123H>0:;4.D0.
XTUTXTSV.:3;.SM1D;R.HI.2;2,42;>TIT>.>;//4.D0.:3;.SM1D;R.HI.*+,S;V9:TU;./0123H>0:;4.
K43HZS.D;/HZ.2/H:4L(
139
!"#$
()
%R+
N#01A?BC-L4?@M
N#01A?BC-L0"J3HM
(%
'
%R)
*
&
%R(
)
%
%R%
%
&
'
()
(*
)%
)&
G/H8/#0-37-<IJ.K38I0/6
G/H8/#0-37-<IJ.K38I0/6
!
%R%9
%R&
%R%&
%R+
%R%+
%R)
%R%)
%R(
%R%(
%R%
&%
+%
)%
(%
%
G/H8/#0-8I0351#/-.H32"81#$-8/<<6
%
#
,-./.012/
#3-./.012/
*%
9%
&
'
()
4//56-.360-1#7/8013#
(*
%R%%
%
&
'
()
(*
)%
)&
C&&D3$4-3*$5).&(*5-)A$*EF-'$5)5*5+*5-)
G/H8/#0-BF@γ,E@Fα#/$
G/H8/#0-BF@γ,E@Fα,
!"#$
%R%*
%R9
C&&D3$4-3*$5).&(*5-)A$*EF-'$5)5*5+*5-)
"
O!@
%R*
!"#$
:%
*%
9%
&%
+%
)%
(%
%
%
&
'
()
(*
4//56-.360-1#7/8013#
BF@γ,E@Fα,
BF@γ,E@Fα-#/$
&%
+%
)%
(%
%
;<3#/ =1>/2 ;<3#/ =1>/2
4?@A?BC
!/#01A!"8D?
,DEAB
$%&'()*+,--.*/)00*()12341)*53*6+78+9:-L;M-G/H8/#0-37-<IJ.K38I0/6-6./81718-73H-?BC-1#-0K/-<"#$-N#2-0K/-J/21N601#N<-2HN1#1#$-<IJ.K-#32/-LO!@M-N02/71#/2-01J/6-N70/H-4?@A?BC-1#7/8013#-L63<12-<1#/6P-</70-IAN>16P-#Q-)A+-J18/-./H-01J/A.31#0MR-F3H-H/7/H/#8/P0"J3HA6./81718-N#01A?BC-E-8/<<6-NH/-6K3S#-7H3J-F1$"H/-)-L2300/2-<1#/6P-H1$K0-IAN>16MR!"#$%&'(&)*+,&$-.$/01234&(5.5($6789$(&::3$*;+*$+'&$0<=2>9?=<2@9$-'$0<=2>9?=<2@)&,$9A2$4&4*5B&$+*$
6/T/HN<-01J/6-N70/H-4?@A?BC-1#7/8013#-L#Q-)A+-J18/-./H-01J/A.31#0MRLUM-E3-83#71HJ-0KN0-0K/-2177/H/#8/6-1#-0K/-8N.N810I-03-.H32"8/-8I0351#/6-2"H1#$-1#-T10H3-601J"<N013#-S/H/#30-0K/-H/6"<0-37-N221013#N<-601J"<N03HI-7N803H6-1#-0K/-<"#$6-37-1#7<"/#VN-1#7/80/2-J18/-3H-N221013#N<1#K1W103HI-7N803H6-1#-0K/-<"#$6-37-0"J3HAW/NH1#$-J18/P-S/-601J"<N0/2-8/<<-.3."<N013#6-7H3J-4?@A?BC1#7/80/2-J18/-N#2-!/#01A!"8D?-0"J3HAW/NH1#$-J18/-N<3#/-3H-03$/0K/H-N#2-J3#103H/2-8I0351#/-.H32"8013#RU/<<-6"6./#613#6-7H3J-0K/-<"#$6-37-J18/-1#7/80/2-S10K-4?@A?BC-L,&-S//56M-3H-!/#01A!"8D?-0"J3HAW/NH1#$
J18/-H/8/1T1#$-#NXT/-DEAB-E-8/<<6-L(&-2NI6-.H13HM-S/H/-601J"<N0/2-S10K-./.012/6-N<3#/-3H-1#-83JW1#N013#L=1>/2M-N#2-N#N<IV/2-73H-8I0351#/-.H32"8013#R-E3-2/0/HJ1#/-0K/-./H8/#0-37-8I0351#/-.H32"81#$-8/<<6H/N801T/-03-4?@A?BCP-S/-J3#103H/2-0K/-/#23$/#3"6-E-8/<<6-6./81718-03-?BCP-SK/H/N6-73H-0"J3HA6./81718-E8/<<6P-S/-J3#103H/2-0K/-0HN#67/HH/2-UO&9R(,-DEAB-E-8/<<6-6./81718-03-?B@R-UI0351#/-.H32"8013#-WI-1#7<"/#VNA
H/N801T/-N#2-0"J3HAH/N801T/-E-8/<<6-SN6-#30-N<0/H/2-SK/#-0K/-8/<<-6"6./#613#6-S/H/-601J"<N0/2-1#83JW1#N013#-L=1>/2MP-1#218N01#$-0KN0-0K/-2177/H/#8/6-1#-E-8/<<-7"#8013#-1#-H/6.3#6/-03-1#7<"/#VN-3HN"038K0K3#3"6-0"J3H6-16-2"/-03-1#K/H/#0-3H-N8Y"1H/2-2177/H/#8/6-1#-0K/-7"#8013#N<-60N0/6-37-0K/6/-E-8/<<6R140
/01234/567*
!"
##'"
$'"
!" $
!"
#
!"
#
!"
!
!"
!
!"
!
&$
!" "
!" "
!" !
!" #
!" $
!" %
!#
!" $
!"
#
!8
"'(
#
!" "
"
!"#$%-,-
%
!"
α4*-.,LM
TO/
!" $
!"#$%&'&"()*+,-
%
C=DE&=FG!"$H
!"
)*+,*-.
%
!" !
!" #
!" $
!" %
!" "
!" !
!" #
!" $
!" %
W=?X=YS@>Z?6S20<
/01234/567*
&
B1234*-.U=V5
B1234*-+U=V5
B1234*-.U=)V
B1234*-+U=)V
"'9"
"'#9
"
%
&
!#
!8
#"
#%
&"
8"
8"
%"
%"
%"
#"
#"
#"
"
"
!" %
)00;<=>?<2=25@?A=31323B23?1
!
D[*\
!"
#
!"
$
!"
%
#""
!""
&"(
&'
%
,)*+
-,-
$%
!"#
!"#$%-,:"'!
$=RBS<
"
!" "
!" !
!" #
!" $
!" %
"
!" #
!" $
!" %
!" 9
!" %
$'8
!" $
DE%9'!
!""
&"
!"
$""
7NO=P=/Q*
8"
"
%""
$
!"#
&"
!"
9""
%
$
!"#
!"#$%-,%&'# 9=RBS<
!""
8""
"
-,-
/01234/567*
!"#$%&'&"()*+,:"'% 9=RBS<
!""
+,-
*
"()
&
&'
$%
!"#
"':9
"'""
%
"
DE&
$
&
!" "
!" "
C=α4*-.,α4*-+
/01234/567*
!
!'#
!" $
!"
#
!"
#
!"
!
!"
!
!" "
!" "
!" "
!" !
!" #
!" $
!" %
!" "
!" !
!" #
!" $
!" %
DE&
%
!"#$%&'&"()*+,-
/J+K
!" %
##'%
"'!
!"#$%&'&"()*+,-
!" %
!"#$%-,-
!" %
9:'9
I"'!
!"
$
!"
$
!"
$
!"
$
!"
#
!"
#
!"
#
!"
#
!"
#
!"
!
!"
!
!"
!
!"
!
!"
!
!" "
!" "
!" !
!" #
!" $
!" %
"'9
!" $
!" "
!" "
!" !
!" #
!" %
!" $
I"'!
!" "
!" "
!" %
!" !
!" %
!" #
!" $
!" %
!" !
!" #
!" $
!" %
!" "
I"'!
!" $
!"
$
!"
$
!"
$
#
!"
#
!"
#
!"
#
!"
#
!"
!
!"
!
!"
!
!"
!
!"
!
!" "
!" "
!" "
!" !
DE&
!" #
!" $
!" %
!" "
!" !
!" #
!" $
!" %
!" !
!" #
!" $
!" %
!" #
!" $
!" %
%'!
!" "
!" "
!" "
!" !
!" %
!"
!" "
%&'$
!" "
!" "
!" %
9'#
!"#$%-,-
!" %
$
!" %
E/+
!"#$%-,-
!" %
!"
!" "
DE%9'!
P=9=RBS<=ED#'%4/567*=T??<2
/01234/567*
7NO=P=/Q*
/01234/567*
/01234/567*
/01234/567*
!" "
!" !
!" #
!" $
!" %
!" "
!" !
!" #
!" $
!" %
%'()*+,-./
141
!
ATU.B.)60
!"#$%-,!!&E
$.FGHI
P.CQ.RGD
!""
):;<=/)?@A0
):;<=/)?@A0
!""
!"#$%&'&"()*+,EN&O
E.FGHI
2"
2"
2"
'"
'"
'"
%"
%"
%"
#"
#"
#"
"
!"
#
!"
$
!"
%
!"
E
()*
6)*
"
"
"
!"#$%-,N"&O
E.FGHI
!""
!" "
!" !
!" #
!" $
!" %
!"
"
!"
!
!"
#
!"
$
!"
%
KS0>
):;<=/47>)?@A0
):;<=/47>)?@A0
B.(CD."/E.FGHI
!"#$%
!"
$
"&"#
!" $
α/01*859
!"
%
!"
#
!"
#
!"
!
!"
!
!"
"
!"
"
!" "
!" !
!" #
!" $
!" %
)+*,
B.(CD."/E.FGHI
&'&"()*+,-
.//'0"#123
!#""
!"""
"&!'
-.α/01*
!"
!"#$%
%
-,-
2""
'""
%""
#""
!" "
!" !
!" #
!" $
"
!" %
α/01*85
9
):;<=/47>)?@A0
KC;<LCM
!"#$%-,-
!" %
α/01*859
!"#$%&'&"()*+,-.//'0"#123
!" %
"&!'
!" $
#&2$
!" $
!"
#
!"
#
!"
!
!"
!
"
!"
!
!"
#
α/01*859
!"
$
!"
%
!#""
!"""
!" !
!" #
!" $
!" %
()*
%"""
#"""
!" "
$
!'""
$"""
!" "
!" "
!"
)+*,
B.(CD.'/!$.FGHI
-.α/01*
#
34 567
-.α/01*
"
"
34 567
2""
%""
"
B.(CDJ
34 567 567
"/EFHI
'/!$FHI
Figure S5. Lenti-LucOS infection initiates a small and transient T cell response against
the OVA and SIY antigens.
(A) FACS analysis for the presence of anti-SIY T cells from bronchoalveolar lavage (BAL) 7
days after infection of K-rasLSL-G12D/+ or K-ras+/+ mice with WSN/SIY or Lenti-LucOS (gated on PI
negative cells). Percent of CD8+ T cells specific for SIY is shown. Data is representative of
infections from at least three different lentiviral preparations.
(B) Quantification of the number of CD8+ T cells in the BAL of K-rasLSL-G12D/+ or K-ras+/+ mice
infected with Lenti-LucOS or Lenti-LucS. Representative of infections from at least three
different lentiviral preparations.
(C) Quantification of the number of anti-SIY or anti-SIN T cells in the BAL of K-rasLSL-G12D/+ or Kras+/+ mice infected with Lenti-LucOS or Lenti-LucS. Data is representative of infections from at
least three different lentiviral preparations.
(D) Percent of lymphocytes specific for SIN or SIY in the lungs of mice after infection with LentiLucOS. Solid lines show the percent of lymphocytes in K-rasLSL-G12D/+ mice that form tumors (Tu)
while dashed lines show the percent of lymphocytes in K-ras+/+ mice that do not form tumors
142
(WT). Percent of lymphocytes is calculated as shown in Figure S3. n= 2-5 mice per time-point,
mean ± SEM.
(E) FACS analysis comparing CFSE dilution (top) in naïve 2C or OT-I cells (gated:
CD8+CD45.1+) from the mediastinal DLN five days after transfer of 2x105 CFSE-labeled naïve
2C cells into K-rasLSL-G12D/+;p53fl/fl or K-ras+/+ mice followed by intratracheal administration of
Lenti-LucOS lentivirus or three days after transfer of 2x105 CFSE-labeled naïve OT-I T cells
followed by intratracheal administration of 10-5 µmol Ovalbumin + 1µg LPS (as a positive
control). The percent of cells fully diluted of CFSE is shown. Grey histograms represent DLN
cells transferred into uninfected control mice. FACS analysis comparing the percentage of
CD8+CD45.1+ transferred 2C cells of the total CD8+ in the DLN of K-rasLSL-G12D/+;p53fl/fl or K-ras+/+
mice five days after Lenti-LucOS infection (bottom). FACS plots are representative of results
from 2-3 mice per group.
(F) FACS analysis for the persistence of transferred naïve OT-I T cells after infection with LentiLucOS in K-rasLSL-G12D/+ or K-ras+/+ mice. 5x105 naïve OT-I T cells were transferred into mice
prior to infection with Lenti-LucOS and the presence of OT-I T cells (gate: CD8+CD45.1+) was
analyzed after 6 weeks (n= 2 mice per group) or 5 days after i.p. injection of 5x105 DC2.4LucOS cells (8 weeks after infection, n= 2 mice per group). As a positive control, mice were
inoculated intratracheally with 10-5 µmol Ovalbumin + 1µg LPS.
(G)
FACS analysis comparing CFSE dilution in OT-II cells (gated: CD4+CD45.1+) from the
mediastinal DLN or inguinal PLN three or five days after transfer of 2x105 CFSE-labeled naïve
OT-II cells into K-rasLSL-G12D/+ or K-ras+/+ mice followed by intratracheal administration of 10-5
µmol Ovalbumin + 1µg LPS (as a positive control) or Lenti-LucOS lentivirus. The percent of
undivided cells in the DLN is shown. Representative results from three mice per group.
(H) FACS analysis for the presence of anti-SIN T cells from the lungs of wild-type or K-rasLSLG12D/+
;p53fl/fl;CCSP-rtTA+ (KPR) mice five days after infection with TRELucOS and Dox feeding.
Percent of PI-negative, CD8+ cells is shown in the FACS plots and the total number of anti-SIN
cells per lung from 3 mice in each group is shown (the approximate limit of detection is indicated
by the dashed line).
(I) FACS analysis for the presence of anti-SIN T cells from the lungs of K-rasLSLG12D/+
;p53fl/fl;CCSP-rtTA+ (KPR) mice infected with TRELucOS and then fed Dox food for 7 days
beginning 5 days after infection (days 6-13). Percent of PI-negative, CD8+ cells is shown in the
FACS plots and the total number of anti-SIN cells per lung from 3 mice is shown (the
approximate limit of detection is indicated by the dashed line). Three uninfected wild-type mice
served as controls.
(J) The total number of anti-SIN cells per DLN from the experiments described in panels H and I
(the approximate limit of detection is indicated by the dashed line).
143
U5-)a$JKL&
C)0-/=G
M$:):
+
*+ ,
$#"α
CZ3U
*+ -
0'$:):
*+ -
[FXL
*+ ,
*+ -
S
*+
S
*+
S
*+
*
*+
*
*+
*
[F[X
*+ +
*+ *
*+ S
*+ ,
[FX\
*+ +
*+ -
*+ +
*+ *
*+ S
*+ ,
[FWY
*+ +
*+ -
*+ +
*+ *
*+ S
*+ ,
?/>)$()??,
*+ -
[FWW
*+ ,
*+
*+ +
C)0-/=C+(D;
M$:):
:):^bE="%J
<
*+
,
*+
S
*+
*
SFXT
*+ +
*+ -
*+ +
!"#γ
*+ *
*+ S
*+ ,
*+ -
:):^b =%V
E
J5?(+?5-)a !"#γ $#"α .%![FWW$=$[FXL#^SFXT$R$XTFX_
&
&
!"#γ&$#"α'().%![FWY$=$[FX\#^SFXT$R$X`FL_
=
U5-)a$JKL&
SFW$H))9,
*+ -
[FYW
CZ3U
*+ ,
*+
S
*+
S
*+
*
*+
*
*+
*
[FYT
KC3
*+
-
*+
,
*+
S
*+
*
*+ S
[F[W
*+ +
*+ *
*+ S
*+ -
,
*+
S
*+
*
*+ ,
*+ +
*+ *
*+ S
*+ ,
*+
*+
*+
,
*+
,
*+
S
*+
S
*+
*
*+
*
[FXL
*+ +
*+ +
*+ *
*+ S
*+ -
*+ ,
[F[]
[FXY
*+ +
*+ *
*+ S
*+ ,
*+ -
*+ *
*+ S
*+ ,
*+ -
[F[\
[FSL
*+ +
*+ -
*+ ,
*+ +
*+ +
-
[FX]
[F]W
*+ +
*+ -
-
*+ -
[FXS
[FXS
*+ +
*+ -
[F[]
*+ +
*+
*+ ,
[F[X
*+ ,
S
*+ *
XY$H))9,
*+ -
*+
*+ +
;%CVV3
[F[W
*+ ,
*+ +
$#"α
W$H))9,
*+ -
*+ +
*+ *
*+ S
*+ -
*+ ,
*+ -
[F[S
*+ ,
*+
S
*+
S
*+
*
*+
*
[FX[
*+ +
*+ +
*+ *
*+ S
*+ ,
*+ -
[FWS
*+ +
*+ +
*+ *
*+ S
*+ ,
*+ -
!"#γ
!"#$%&'()*%+,-./"0&'1%.2$3-".0'4%.5'6&0-"76$38('-$5.%791&3"4"3':'3&;;9*
!"#$%&'()*+&)$+,)*$-'$*)-)&./0)$1234$50*$6327$(8-'9/0)$:&'*+(-/'0$/0$;1<$50*$;13=&)5(-/>)$
6$()??,$@&'.$-A)$?+0B,$'@$C)0-/=C+(D;$-+.'&=E)5&/0B$./()F$C)0-/=G$-+.'&=E)5&/0B$./()$,)&>)*$
5,$5$0)B5-/>)$('0-&'?$50*$H)&)$('.:?)-)?8$0)B5-/>)$/0$-A)$123=4&632=7&$'&$123=4&632=7'()%
I+5*&50-,$5@-)&$;1<$50*$;13$:):-/*)$,-/.+?5-/'0F$6A)$:)&()0-5B)$'@$;1<$50*$;13=,:)(/@/($6$()??,$
@&'.$-A)$?+0B,$-A5-$H)&)$123=4&632=7&$'&$123=4&632=7'()$H5,$*)-)&./0)*$E8$505?8,/,$'@$(8-'9/0)$
:&'*+(-/'0$@&'.$JKL&$()??,$/0$-A)$5E,)0()$'@$,-/.+?5-/'0$!0'$:):#$'&$/0$-A)$:&),)0()$'@$;1<$'&$;13$
:):-/*),$!M$:):#F$6A)$:)&()0-$'@$JKL&$(8-'9/0)$:&'*+(/0B$()??,$H5,$-A)0$*/>/*)*$E8$-A)$:)&()0-$'@$
;1<$'&$;13=,:)(/@/($JKL&$()??,$@&'.$*+:?/(5-)$,5.:?),$,-5/0)*$H/-A$K/.)&NF
!O#$P):&),)0-5-/>)$1234$50*$6327$(8-'9/0)$:&'*+(-/'0$E8$JKL&$()??,$@&'.$-A)$?+0BQ$KC3Q$50*$
,:?))0$'@$C)0-/=C+(D;$-+.'&=E)5&/0B$./()$,-/.+?5-)*$H/-A$;1<$50*$;13$:):-/*),$5-$,)>)&5?$-/.)$
:'/0-,$*+&/0B$-+.'&$*)>)?':.)0-$!0R$S=T$./()$:)&$-/.)=:'/0-#F$U5-)*$'0$JKL&$()??,F
144
!"#$%&'
!"#$%&!()*+
"
=9)">>8?9::49$(:-;
!
,-'./
233
533
!"#$&!()*+
!"#$%&'
433
3
0
12
6""789.-8$9$(:-;9%#%$%<$%-#
#$%&'()*+,9-..&/0%(/$1)2(/3$42&15*)3&.0'6)7'()$/8$93'73(:);<)#0=>?@)A)'(%&9730'<)1(996,9
@AB9!(#C9$(:-;89%#D()"D9E%$F9!"#$%&'9-;9!"#$%&!()*+9%#9!"#$%&'&"()*+,-./0123,239:%)"98$<%#"D9E%$F9
!"#$%&'()*+,$-.#+/,,01+!2#,3+#45'3%$"$#$!#$'"6+78!9,+:+;<=5+>$"1,#+?<=5@6+
@GB9H(<#$%I%)<$%-#9-I9$F"9J9;"C(><$-;K9)">>9%#I%>$;<$"89LK9MNO9@#P94&Q9:%)"R91S&1319$(:-;8R9
.";9C;-(.BT
145
!
%
$8"9:;<-5=
4>
4>?
4>=
4
4>
@>
@>
@>
6>
6>
6>
?>
?>
?>
#>
#>
#>
>
>
4
4>
#
4>
=
4>
?
&'()'*+,
#
4>
=
4>
?
4>
>
4>
4
4>
#
4>
=
4>
?
%
4>>
4>>
@>
@>
6>
N&$LMNγ
Y$LMNγ
6>
&'()'*+,
?>
&'()'*+,
?>
#>
$8",54
4>>
4#>
4>
#>
>
>
4>
>
4>
4
4>
#
4>
=
4>
?
4>
>
4>
4
4>
#
4>
=
4>
?
-./0$ 1#2"345467
%'
:DEHA+EAF
BTALF$J$>*
BTALF$EK$>*,
>
4>
>
4>
4
4>
#
4>
=
4>
?
4> >
4> 4
4> #
4> =
4> ?
%
Q))A;E(HW'U+HTAE
4>W4
4W4
>
#P
4>>
@>
6>
?>
#>
$
-:HLGL+C'1&7
&'()'*+,
$8"9:;<-54
4>>
4>
>
>
!"#$
$8"9:;<-5=
>
#>
4>
4>>
-:HLGL+C'1&7
6>
&'()'*+,
?>
#>
4>
!"#$
4#>
Q))A;E(HW'U+HTAE
4>W4
4W4
>
<U".
4>>
@>
6>
?>
#>
>
>
$8"9:;<-5= $8"9:;<-54
@>
6>
>
9
#
@>
#
: - - 5= 54 5#
:
-'U EG",'U "9:; :;< ;<- ;<- ;<<
>
D
;
"9 : : :
4
"9: 9A9$M M4=$8"9 $8"9$8"9
G
E
AD
9$
4>
4>>
&'()'*+,
?>
4>4
"
$8",54
4>>
2
&'()'*+,
M9X0'µT'BH(EAGD
4>6
$8",54
$8"9:;<-5= $8"9:;<-54
$8",54
!"#$%&'()*'+,-'./011'2'3%&1&4505"64'67'4&6045"#&41'"1'"48$.&8'9:'2!;<'048'='.&//'%&.6#4"5"64*'
!"#$%&'()*+,&-$&.$-/&)-0+1/-$/2(*/,,+&-$34$),,)4+-1$567+./*),/$)70+8+04$!9:;<$=1$(*&0/+-#$+-$.*/,>54$
A,BC+DEAF'9ADEG"9:;<-'(H'9ADEG",'E:I(HJK'(H';ACC'CGDAJ'FAHGLAF')H(I'9ADEG"9:;<-'E:I(HJ'1$8"9:;<-54K'
?@A:67BCDEF$?@A:67BCDG#$&*$7/55$5+-/,$+-$H>+7>$:67C$&*$:67BC$H/*/$+-0*&I67/I$34$5/-0+8+*)5$+-./70+&-$
19$M4>"9:;-'+DF'9$M4="9:;<-75
1N7'!"#$%'+DF'-./0$%'1#234546'+DEG%(FO7'A,BHAJJG(D'GD'+'HABHAJADE+EGLA'9ADEG"9:;<-';ACC"CGDA'
!?@A:67BCDG#$)-I$:/-0+A2$7/55A5+-/$!?@A2DJ#$).0/*$0*/)0'/-0$H+0>$JKK$;<':$LMNO$.&*$PQ$>&6*,$!,&5+I$5+-/#$
&*$-&$0*/)0'/-0$!I&00/I$5+-/#D$N&0/R$,+'+5)*$*/,650,$H+0>$JK$;<'5$LMNOD$M+55/I$>+,0&1*)',$)*/$-/1)0+8/$
;(DEH(C'+DEG%(FO'JE+GDJ')(H'P3@5
!%#$SAE?3$/2(*/,,+&-$&-$?@A:67BCDGF$?@A:67BCDJF$)-I$?@A2DJ$).0/*$76506*/$H+0>$!"#$!%&''+;EGL+EAF'
BTAL$T$7/55,D$T6'&*$7/55$5+-/,$H/*/$(5)0/I$)-I$0>/$.&55&H+-1$I)4$BTAL$T$7/55,$H/*/$)II/I$)0$)$5&H$I/-,+04$
!UE$06'&*$7/55,$0&$J$T$7/55#D$".0/*$JF$EF$GF$P$&*$EK$>&6*,F$BTAL$7/55,$H/*/$H),>/I$)H)4$.*&'$)I>/*/-0$06'&*$
7/55,D$SAE?3$/2(*/,,+&-$H),$)-)54V/I$&-$)55$06'&*$7/55$5+-/$,)'(5/,$34$M"%,$EK$>&6*,$).0/*$+-+0+)5$T$7/55$
)II+0+&-$!T$7/55,$H/*/$/2756I/I$6,+-1$%WPXDJY#D$",$5+005/$),$J$>&6*$7&A76506*/$&.$BTAL$T$7/55,$H+0>$06'&*$
7/55,$H),$,6..+7+/-0$0&$6(*/165)0/$ZS%L$)-I$&-54$&776**/I$&-$7/55$5+-/,$0>)0$/2(*/,,/I$CLND$ZS%L$
/2(*/,,+&-$+-7*/),/I$H+0>$+-7*/),+-1$76506*/$0+'/$!&-54$J$)-I$EK$>&6*,$,>&H-#D
13'+DF'Q7'*(FG)GAF'RGCC'+JJ+OJ'SAHA'BAH)(HIAF'%O'BC+EGDT'4>K'4K'(H'>'!"#$!%&''+;EGL+EAF'#P'137'(H'<U".'U'
7/55,$![#$0&$06'&*$7/55,$![../70&*R$T)*1/0#$&-$W)4$JF$*/'&8+-1$0>/$T$7/55,$H+0>$0>*//$@\C$H),>/,$&-$)55$
,)'(5/,$&-$W)4$EF$)-I$0>/-$7&6-0+-1$0>/$0&0)5$-6'3/*$&.$06'&*$7/55,$&-$W)4$PD$C6*8+8)5$(/*7/-0)1/,$
H/*/$I/0/*'+-/I$34$I+8+I+-1$0>/$-6'3/*$&.$7/55,$).0/*$7&A76506*/$H+0>$T$7/55,$34$0>/$-6'3/*$&.$7/55,$
H+0>&60$7&A76506*/D$C(/7+.+7$]+55+-1$&-54$&776**/I$+-$?@A:67BC$7/55$5+-/,$)-I$-&0$)$?@A2$7&-0*&5$7/55$5+-/D$
T6'&*$7/55,$H/*/$-&0$0*/)0/I$H+0>$LMN$(*+&*$0&$7&A76506*/F$+-I+7)0+-1$T$7/55,$H/*/$7)()35/$&.$,(/7+.+7)554$
HA;(TDGVGDT'$8"9:;<-';ACC'CGDAJ'FAJBGEA'C(S'CALACJ'()'*!P.'J:H)+;A'A,BHAJJG(D5
146
BK
FK
SK
K
QKK
BKK
FKK
SKK
K
O!.//01
!"#$%
!"#&'
OK
RK
QK
FK
K
K
SF
FQ
BR
QO
!"+
,*+,
,*+,
!",
86.
O!.//01
<
86.
!!"" #$""
#%"" !&""
OK
RK
QK
FK
K
"
!"()*$%
!"()*&'
SKK
RK
X*5/!"56%-,1(
=
:/%-*;:49<=
:/%-*;>
=43@*@$8!"A/39/%-(
QK
NKK
K
SF
FQ
BR
QO
RK
X*5/!"56%-,1(
!",$%!"+$%
!",&'!"+&'
SKK
=43@*@$8!"A/39/%-(
NK
SKK
=43@*@$8!"A/39/%-(
;
CDDBV!9/881W!55F!-4563
:
??BV!9/881W!55F!-4563
!
OK
RK
QK
FK
K
K
SF
FQ
BR
QO
X*5/!"56%-,1(
RK
"#$%&'()*+!,-.&'/0'1(/232430#0(#-(5'-4#65%.7)(4%83&0(.382/&'1(43(5'-4#69(4%83&0+!
"#!$%&!'(!)*+,-!.//01!$2-/3!-4563!*%*-*$-*6%7!84%+!-45631!*%&49/&!.*-,!:/%-*;:49<=!63!:/%-*;>!./3/!
1-$*%/&!.*-,!$%-*;?8/$@/&!9$1A$1/;B!"??B(!263!$A6A-61*1!"#(!63!$%-*;C,61A,6;D*1-6%/!DB!"CDDB(!
263!A368*2/3$-*6%!"'(!$%&!-,/!%45E/3!62!A61*-*@/!9/881!A/3!55F!62!-4563!$3/$!.$1!G4$%-*2*/&H!I63!"#(7!
AJ!KHKKKL7!%M!F;N!5*9/7!BO;PQ!-456317!A/3!+364AH!I63!"'(7!AJ!KHQR7!%M!Q;N!5*9/7!SR;QO!-456317!A/3!
+364AH!T/$%!U!=)TH
>#$?('92&'00#3-(3@(./-3-#./A(AB82?3.B4'($'-'0(.3&&'A/4'0(C#4?(D'44'&(2/4#'-4(2&3$-30#0+
"?!$%&!Y(!Z$A8$%;T/[/3!143@*@$8!943@/1!263!=-$+/!\W\\!"8[5A,%6&/;5/-$1-$1/1!23//!]K(!84%+!
$&/%69$39*%65$!A$-*/%-1!965A$3*%+!-,/!-6A!@/3141!E6--65!G4$3-*8/!"OO!A$-*/%-1!/$9,(!263!-,/!*%&*9$-/&!
+/%/1H!C$-*/%-!143*@*@$8!E$1/&!6%!?YF!/>A3/11*6%!63!?YLP#!/>A3/11*6%7!A;@$84/1!J!KHKKB!$%&!^!KHKB7!
3/1A/9-*@/8[H!?YF!":I#;F(!*1!$%!$&,/1*6%!568/948/!*%@68@/&!*%!X!9/88!$9-*@$-*6%H!?YLP#!/%96&/1!-,/!\+α!
1*+%$88*%+!9,$*%!62!-,/!';9/88!3/9/A-63H
")(!?6%-*%+/%9[!-$E8/!62!?YQ!$%&!?YO!/>A3/11*6%!"$@/3$+/!62!?YOα!$%&!?YOβ(!965A$3*%+!-6A!,$82!
@/3141!E6--65!,$82!263!/$9,!+/%/H!I*1,/3!A;@$84/!^SK;SQH!
"I(!C$-*/%-!143*@$8!E$1/&!6%!-,/!/>A3/11*6%!62!?YQ!$%&!?YO!"2365!)(H!A;@$84/!^KHKLH!
Y$-$!*1!2365!=,/&&/%!/-!$87!]$-43/!T/&*9*%/7!FKKO!2365!1$5A8/1!9688/9-/&!$%&!A369/11/&!$-!2643!*%1-*-4-/1H
147
$
α.2@AB,C
!?$&
+,-#!.+/02
#! &
#! %
#!
$
#!
#
%?"
L+D
3+D
#!
$
#!
#
&?!
#! !
#! !
#! #
#! $
#! &
α.=:=9HN:B,C
"
#! %
#! !
+,-#%.+/012
α.2@DB,C
!?!$
#! &
#! %
#! &
#! !
#! #
#! $
#! &
!?$!
#!
%
#!
%
#!
$
#!
$
#!
#
#!
#
%?&
#! %
#! &
!?#(
#
E/FG;
%?(
#! !
#! !
#! !
#! #
#! $
#! %
#! !
#! &
#! #
#! $
#! %
#! &
!
E/FG;76HI:7JFF$K
M3)
*!
)!
(!
'!
"!
&!
%!
$!
#!
!
+,-#%
+,-#%.+/012
!""#
!
"
#!
#"
$!
$"
34567489:;79;4<6=>4<9
B
%!
+/<Q79/FG;.C:4;H<Q]
+:<9H.V
+,-.#%7G;
+,-.#%.+/012
E;4<6=>4<9
+:<9H.+/012
DG?79/FG;67;:[:09:NB79G94>
+:<9H.
+:<9H.V +/012 -4Q.$</>>
!B%
!B(
!B%
+,-#%
$B(
!B%
+,-#%.+/012 #B%
DG?79/FG;67>/0H8:;46:7=G6?B79G94>
+:<9H.
+:<9H.V +/012 -4Q.$</>>
!B&
$B$
+,-#%.+/012 !B#
+:<9H.+/012\7$%&'!()**
%&'()*+,-./70+1*22+)*34563*3+75+89:-.;89:-<+7)=6342=67=>2*+?5@*23+5A+2(6'+1=61*)/
JOK72=>:<G059:678;GF7FH0:7P4;CG;H<Q76/C0/94<:G/67+,-#!.+/027G;7+,-#%.+/01279/FG;67
R:;:7694H<:N78G;7M3)74<N72@A7G;72@D7>G4N:N7STM@B,C73HF:;U7J5.4VH6K?7
JW+4<N+MK7TXY7694H<:N7H<Q/H<4>7>5F=P7<GN:676PGR7PHQP>57;:409HZ:7JF/>9H=>:7Q:;FH<4>70:<9:;6K7
H=6H>49:;4>7>5F=P7<GN:67N;4H<H<Q79P:76/C0/94<:G/6>579;4<6=>4<9:N79/FG;67J3+DK70GF=4;:N7
9G79P:70G<9;4>49:;4>7>5F=P7<GN:67JL+DK?7+,-#%.+/01279/FG;7RH9P7H<9;49/FG;4>7>5F=PG059:6?
J3K7+,-#%74<N7+,-#%.+/0127Q;GR74970GF=4;4C>:7;49:67H<7-4Q.$.B.7FH0:?7
JYK7Y<NGQ:<G/67>/<Q79/FG;.C:4;H<Q7FH0:7NG7<G97Q:<:;49:76569:FH079G>:;4<0:79GR4;N79;4<6=>4<9:N
+,-.#%.+/01279/FG;6?7D/FC:;7G879/FG;679P497R:;:7;:[:09:N7G;7;:F4H<:N7>/0H8:;46:7=G6H9HZ:
G879P:79G94>74<4>5I:N7489:;79;4<6=>4<949HG<7H<9G7HFF/<:.0GF=:9:<97G;7-4Q.$</>>7FH0:7C:4;H<Q7
148
+:<9H.V7G;7+:<9H.+/012.H<N/0:N7>/<Q79/FG;6?
9@G
/&()012
Z
6G 9
6G
9
8P[11#R?
6
6G G
9:
_&&<5!E,5)!)*+,-!0(0)0')0,(
CR/SSY
6G
67
6G
6@L6
8P[11#R?
6G G
G@L
:
Z
6G
6@G
6G
G@ZL
W!>?9@:1/*ABC
/&()01/*ABC
/&()012
6G
6G
6@L
G
/&()01/*ABC
6G :
Y,!^'2
>?9@:1/*ABC
>/Y
9@L
?
6G 6
6G 9
6G Z
6G
6
6G
G
6G :
:
G@6H
6G Z
α1CTU\]Q1#R?
X*+,-!'-&'!"++9$
>
:
6@V:
6G Z
6G
9
8P[11#R?
6G 6
6G G
6G
6G
:
6G
Z
G
6G
6
6G
9
6G
Z
6G
:
6G
G
:@::
6G 9
6G 9
8P[11#R?
8P[11#R?
6G 6
6G 6
6G
G
6G 6
6G 9
6G Z
6G :
9
6G
Z
6G
:
6G
:
6G G
6G G
6G G
6G
:@6H
6G 9
6
6
6G Z
8P[11#R?
6G
6G
6G :
6G G
6G 6
6G 9
6G Z
6G :
6G
G
6G
6
6G
9
6G
Z
α1CTU\]Q1RS
!"#$%&'())*!+,-*./0$12('3411"546"75'8&49:'67'4':$:64"5&9'456"/6$;7%'<'1&88'%&:=75:&*!
"#$!%&'(!)*+,-!'-&'!,.!/&()012!3*(4!)*+,-5!')!67!'(8!9:!;&&<5!'.)&-!)*+,-10(0)0')0,(!;0)=!,-!
;0)=,*)!>?9@:1/*ABC!D'AA0(')0,(!"E1D'3*&5!'-&!FG@G7!'(8!FG@HGI!-&5E&A)0D&3JI!.-,+!(K!91L!
+0A&I!771HM!)*+,-5I!E&-!4-,*E$@
"N$!O#?C!'('3J505!;0)=!);,!E&E)08&13,'8&8!>0+&-P!-&'4&()5!3'Q&3&8!;0)=!RS!"21'205$!,-!#R?!
"J1'205$!),!8&)&A)!CTU15E&A0.0A!?>VW!X!A&335!0(!)=&!>/Y!'(8!5E3&&(!,.!/&()01/*ABC!,-!/&()012!
)*+,-1Q&'-0(4!+0A&!;0)=!,-!;0)=,*)!>?9@:1/*ABC!D'AA0(')0,(!E-0,-!),!)*+,-10(0)0')0,(!"6H!;&&<5
'.)&-!)*+,-10(0)0')0,($@!O#?C!E3,)5!'-&!4')&8!,(!?>VW!A&335!'(8!)=&!E&-A&()!,.!?>V!A&335!05!5=,;(@
149
REFERENCES
Bai, A., Higham, E., Eisen, H.N., Wittrup, K.D., and Chen, J. (2008). Rapid tolerization of virusactivated tumor-specific CD8+ T cells in prostate tumors of TRAMP mice. Proc Natl Acad Sci U
S A 105, 13003-13008.
Balkwill, F., and Mantovani, A. (2001). Inflammation and cancer: back to Virchow? Lancet 357,
539-545.
Buckanovich, R.J., Facciabene, A., Kim, S., Benencia, F., Sasaroli, D., Balint, K., Katsaros, D.,
O'Brien-Jenkins, A., Gimotty, P.A., and Coukos, G. (2008). Endothelin B receptor mediates the
endothelial barrier to T cell homing to tumors and disables immune therapy. Nat Med 14, 28-36.
Bucks, C.M., Norton, J.A., Boesteanu, A.C., Mueller, Y.M., and Katsikis, P.D. (2009). Chronic
antigen stimulation alone is sufficient to drive CD8+ T cell exhaustion. J Immunol 182, 66976708.
Budhu, A., Forgues, M., Ye, Q.H., Jia, H.L., He, P., Zanetti, K.A., Kammula, U.S., Chen, Y., Qin,
L.X., Tang, Z.Y., and Wang, X.W. (2006). Prediction of venous metastases, recurrence, and
prognosis in hepatocellular carcinoma based on a unique immune response signature of the
liver microenvironment. Cancer Cell 10, 99-111.
Calzascia, T., Pellegrini, M., Hall, H., Sabbagh, L., Ono, N., Elford, A.R., Mak, T.W., and
Ohashi, P.S. (2007). TNF-alpha is critical for antitumor but not antiviral T cell immunity in mice.
J Clin Invest 117, 3833-3845.
Cheung, A.F., Dupage, M.J., Dong, H.K., Chen, J., and Jacks, T. (2008). Regulated expression
of a tumor-associated antigen reveals multiple levels of T-cell tolerance in a mouse model of
lung cancer. Cancer Res 68, 9459-9468.
Clark, C.E., Hingorani, S.R., Mick, R., Combs, C., Tuveson, D.A., and Vonderheide, R.H.
(2007). Dynamics of the immune reaction to pancreatic cancer from inception to invasion.
Cancer Res 67, 9518-9527.
Ding, L., Getz, G., Wheeler, D.A., Mardis, E.R., McLellan, M.D., Cibulskis, K., Sougnez, C.,
Greulich, H., Muzny, D.M., Morgan, M.B., et al. (2008). Somatic mutations affect key pathways
in lung adenocarcinoma. Nature 455, 1069-1075.
Drake, C.G., Doody, A.D., Mihalyo, M.A., Huang, C.T., Kelleher, E., Ravi, S., Hipkiss, E.L.,
Flies, D.B., Kennedy, E.P., Long, M., et al. (2005). Androgen ablation mitigates tolerance to a
prostate/prostate cancer-restricted antigen. Cancer Cell 7, 239-249.
Dunn, G.P., Bruce, A.T., Ikeda, H., Old, L.J., and Schreiber, R.D. (2002). Cancer
immunoediting: from immunosurveillance to tumor escape. Nat Immunol 3, 991-998.
DuPage, M., Dooley, A.L., and Jacks, T. (2009). Conditional mouse lung cancer models using
adenoviral or lentiviral delivery of Cre recombinase. Nat Protoc 4, 1064-1072.
150
Ercolini, A.M., Ladle, B.H., Manning, E.A., Pfannenstiel, L.W., Armstrong, T.D., Machiels, J.P.,
Bieler, J.G., Emens, L.A., Reilly, R.T., and Jaffee, E.M. (2005). Recruitment of latent pools of
high-avidity CD8(+) T cells to the antitumor immune response. J Exp Med 201, 1591-1602.
Finak, G., Bertos, N., Pepin, F., Sadekova, S., Souleimanova, M., Zhao, H., Chen, H.,
Omeroglu, G., Meterissian, S., Omeroglu, A., et al. (2008). Stromal gene expression predicts
clinical outcome in breast cancer. Nat Med 14, 518-527.
Finn, O.J. (2008). Immunological weapons acquired early in life win battles with cancer late in
life. J Immunol 181, 1589-1592.
Forster, I., Hirose, R., Arbeit, J.M., Clausen, B.E., and Hanahan, D. (1995). Limited capacity for
tolerization of CD4+ T cells specific for a pancreatic beta cell neo-antigen. Immunity 2, 573-585.
Frese, K.K., and Tuveson, D.A. (2007). Maximizing mouse cancer models. Nat Rev Cancer 7,
645-658.
Galon, J., Costes, A., Sanchez-Cabo, F., Kirilovsky, A., Mlecnik, B., Lagorce-Pages, C.,
Tosolini, M., Camus, M., Berger, A., Wind, P., et al. (2006). Type, density, and location of
immune cells within human colorectal tumors predict clinical outcome. Science 313, 1960-1964.
Getnet, D., Maris, C.H., Hipkiss, E.L., Grosso, J.F., Harris, T.J., Yen, H.R., Bruno, T.C., Wada,
S., Adler, A., Georgantas, R.W., et al. (2009). Tumor recognition and self-recognition induce
distinct transcriptional profiles in antigen-specific CD4 T cells. J Immunol 182, 4675-4685.
Hanson, H.L., Donermeyer, D.L., Ikeda, H., White, J.M., Shankaran, V., Old, L.J., Shiku, H.,
Schreiber, R.D., and Allen, P.M. (2000). Eradication of established tumors by CD8+ T cell
adoptive immunotherapy. Immunity 13, 265-276.
Huijbers, I.J., Krimpenfort, P., Chomez, P., van der Valk, M.A., Song, J.Y., Inderberg-Suso,
E.M., Schmitt-Verhulst, A.M., Berns, A., and Van den Eynde, B.J. (2006). An inducible mouse
model of melanoma expressing a defined tumor antigen. Cancer Res 66, 3278-3286.
Jackson, E.L., Olive, K.P., Tuveson, D.A., Bronson, R., Crowley, D., Brown, M., and Jacks, T.
(2005). The differential effects of mutant p53 alleles on advanced murine lung cancer. Cancer
Res 65, 10280-10288.
Jackson, E.L., Willis, N., Mercer, K., Bronson, R.T., Crowley, D., Montoya, R., Jacks, T., and
Tuveson, D.A. (2001). Analysis of lung tumor initiation and progression using conditional
expression of oncogenic K-ras. Genes Dev 15, 3243-3248.
Johnson, L., Mercer, K., Greenbaum, D., Bronson, R.T., Crowley, D., Tuveson, D.A., and Jacks,
T. (2001). Somatic activation of the K-ras oncogene causes early onset lung cancer in mice.
Nature 410, 1111-1116.
Jonkers, J., Meuwissen, R., van der Gulden, H., Peterse, H., van der Valk, M., and Berns, A.
(2001). Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model
for breast cancer. Nat Genet 29, 418-425.
151
Khong, H.T., and Restifo, N.P. (2002). Natural selection of tumor variants in the generation of
"tumor escape" phenotypes. Nat Immunol 3, 999-1005.
Koebel, C.M., Vermi, W., Swann, J.B., Zerafa, N., Rodig, S.J., Old, L.J., Smyth, M.J., and
Schreiber, R.D. (2007). Adaptive immunity maintains occult cancer in an equilibrium state.
Nature 450, 903-907.
Kranz, D.M., Tonegawa, S., and Eisen, H.N. (1984). Attachment of an anti-receptor antibody to
non-target cells renders them susceptible to lysis by a clone of cytotoxic T lymphocytes. Proc
Natl Acad Sci U S A 81, 7922-7926.
Kumar, M.S., Lu, J., Mercer, K.L., Golub, T.R., and Jacks, T. (2007). Impaired microRNA
processing enhances cellular transformation and tumorigenesis. Nat Genet 39, 673-677.
Lee, P.P., Yee, C., Savage, P.A., Fong, L., Brockstedt, D., Weber, J.S., Johnson, D., Swetter,
S., Thompson, J., Greenberg, P.D., et al. (1999). Characterization of circulating T cells specific
for tumor-associated antigens in melanoma patients. Nat Med 5, 677-685.
Lyman, M.A., Aung, S., Biggs, J.A., and Sherman, L.A. (2004). A spontaneously arising
pancreatic tumor does not promote the differentiation of naive CD8+ T lymphocytes into effector
CTL. J Immunol 172, 6558-6567.
Meylan, E., Dooley, A.L., Feldser, D.M., Shen, L., Turk, E., Ouyang, C., and Jacks, T. (2009).
Requirement for NF-kappaB signalling in a mouse model of lung adenocarcinoma. Nature 462,
104-107.
Molldrem, J.J., Lee, P.P., Wang, C., Felio, K., Kantarjian, H.M., Champlin, R.E., and Davis,
M.M. (2000). Evidence that specific T lymphocytes may participate in the elimination of chronic
myelogenous leukemia. Nat Med 6, 1018-1023.
Mueller, S.N., and Ahmed, R. (2009). High antigen levels are the cause of T cell exhaustion
during chronic viral infection. Proc Natl Acad Sci U S A 106, 8623-8628.
Muller-Hermelink, N., Braumuller, H., Pichler, B., Wieder, T., Mailhammer, R., Schaak, K.,
Ghoreschi, K., Yazdi, A., Haubner, R., Sander, C.A., et al. (2008). TNFR1 signaling and IFNgamma signaling determine whether T cells induce tumor dormancy or promote multistage
carcinogenesis. Cancer Cell 13, 507-518.
Nguyen, L.T., Elford, A.R., Murakami, K., Garza, K.M., Schoenberger, S.P., Odermatt, B.,
Speiser, D.E., and Ohashi, P.S. (2002). Tumor growth enhances cross-presentation leading to
limited T cell activation without tolerance. J Exp Med 195, 423-435.
Novellino, L., Castelli, C., and Parmiani, G. (2005). A listing of human tumor antigens
recognized by T cells: March 2004 update. Cancer Immunol Immunother 54, 187-207.
Ochsenbein, A.F., Klenerman, P., Karrer, U., Ludewig, B., Pericin, M., Hengartner, H., and
Zinkernagel, R.M. (1999). Immune surveillance against a solid tumor fails because of
immunological ignorance. Proc Natl Acad Sci U S A 96, 2233-2238.
152
Ochsenbein, A.F., Sierro, S., Odermatt, B., Pericin, M., Karrer, U., Hermans, J., Hemmi, S.,
Hengartner, H., and Zinkernagel, R.M. (2001). Roles of tumour localization, second signals and
cross priming in cytotoxic T-cell induction. Nature 411, 1058-1064.
Pages, F., Berger, A., Camus, M., Sanchez-Cabo, F., Costes, A., Molidor, R., Mlecnik, B.,
Kirilovsky, A., Nilsson, M., Damotte, D., et al. (2005). Effector memory T cells, early metastasis,
and survival in colorectal cancer. N Engl J Med 353, 2654-2666.
Parmiani, G., De Filippo, A., Novellino, L., and Castelli, C. (2007). Unique human tumor
antigens: immunobiology and use in clinical trials. J Immunol 178, 1975-1979.
Pellegrini, M., Calzascia, T., Elford, A.R., Shahinian, A., Lin, A.E., Dissanayake, D., Dhanji, S.,
Nguyen, L.T., Gronski, M.A., Morre, M., et al. (2009). Adjuvant IL-7 antagonizes multiple cellular
and molecular inhibitory networks to enhance immunotherapies. Nat Med 15, 528-536.
Piersma, S.J., Jordanova, E.S., van Poelgeest, M.I., Kwappenberg, K.M., van der Hulst, J.M.,
Drijfhout, J.W., Melief, C.J., Kenter, G.G., Fleuren, G.J., Offringa, R., and van der Burg, S.H.
(2007). High number of intraepithelial CD8+ tumor-infiltrating lymphocytes is associated with the
absence of lymph node metastases in patients with large early-stage cervical cancer. Cancer
Res 67, 354-361.
Porgador, A., Yewdell, J.W., Deng, Y., Bennink, J.R., and Germain, R.N. (1997). Localization,
quantitation, and in situ detection of specific peptide-MHC class I complexes using a
monoclonal antibody. Immunity 6, 715-726.
Redmond, W.L., and Sherman, L.A. (2005). Peripheral tolerance of CD8 T lymphocytes.
Immunity 22, 275-284.
Rubio, V., Stuge, T.B., Singh, N., Betts, M.R., Weber, J.S., Roederer, M., and Lee, P.P. (2003).
Ex vivo identification, isolation and analysis of tumor-cytolytic T cells. Nat Med 9, 1377-1382.
Savage, P.A., Vosseller, K., Kang, C., Larimore, K., Riedel, E., Wojnoonski, K., Jungbluth, A.A.,
and Allison, J.P. (2008). Recognition of a ubiquitous self antigen by prostate cancer-infiltrating
CD8+ T lymphocytes. Science 319, 215-220.
Schreiber, H. (2003). Tumor Immunology. In Fundamental Immunology, W.E. Paul, ed.
(Philadelphia: Lippincott Williams & Wilkins), pp. 1557-1592.
Seder, R.A., Darrah, P.A., and Roederer, M. (2008). T-cell quality in memory and protection:
implications for vaccine design. Nat Rev Immunol 8, 247-258.
Segal, N.H., Parsons, D.W., Peggs, K.S., Velculescu, V., Kinzler, K.W., Vogelstein, B., and
Allison, J.P. (2008). Epitope landscape in breast and colorectal cancer. Cancer Res 68, 889892.
Shankaran, V., Ikeda, H., Bruce, A.T., White, J.M., Swanson, P.E., Old, L.J., and Schreiber,
R.D. (2001). IFNgamma and lymphocytes prevent primary tumour development and shape
tumour immunogenicity. Nature 410, 1107-1111.
153
Shen, Z., Reznikoff, G., Dranoff, G., and Rock, K.L. (1997). Cloned dendritic cells can present
exogenous antigens on both MHC class I and class II molecules. J Immunol 158, 2723-2730.
Sjoblom, T., Jones, S., Wood, L.D., Parsons, D.W., Lin, J., Barber, T.D., Mandelker, D., Leary,
R.J., Ptak, J., Silliman, N., et al. (2006). The consensus coding sequences of human breast and
colorectal cancers. Science 314, 268-274.
Slifka, M.K., and Whitton, J.L. (2000). Activated and memory CD8+ T cells can be distinguished
by their cytokine profiles and phenotypic markers. J Immunol 164, 208-216.
Smith, K.M., Olson, D.C., Hirose, R., and Hanahan, D. (1997). Pancreatic gene expression in
rare cells of thymic medulla: evidence for functional contribution to T cell tolerance. Int Immunol
9, 1355-1365.
Speiser, D.E., Miranda, R., Zakarian, A., Bachmann, M.F., McKall-Faienza, K., Odermatt, B.,
Hanahan, D., Zinkernagel, R.M., and Ohashi, P.S. (1997). Self antigens expressed by solid
tumors Do not efficiently stimulate naive or activated T cells: implications for immunotherapy. J
Exp Med 186, 645-653.
Spiotto, M.T., Rowley, D.A., and Schreiber, H. (2004). Bystander elimination of antigen loss
variants in established tumors. Nat Med 10, 294-298.
Spiotto, M.T., Yu, P., Rowley, D.A., Nishimura, M.I., Meredith, S.C., Gajewski, T.F., Fu, Y.X.,
and Schreiber, H. (2002). Increasing tumor antigen expression overcomes "ignorance" to solid
tumors via crosspresentation by bone marrow-derived stromal cells. Immunity 17, 737-747.
Stetson, D.B., Mohrs, M., Mallet-Designe, V., Teyton, L., and Locksley, R.M. (2002). Rapid
expansion and IL-4 expression by Leishmania-specific naive helper T cells in vivo. Immunity 17,
191-200.
Street, S.E., Trapani, J.A., MacGregor, D., and Smyth, M.J. (2002). Suppression of lymphoma
and epithelial malignancies effected by interferon gamma. J Exp Med 196, 129-134.
Thomas, D.A., and Massague, J. (2005). TGF-beta directly targets cytotoxic T cell functions
during tumor evasion of immune surveillance. Cancer Cell 8, 369-380.
Tiscornia, G., Singer, O., and Verma, I.M. (2006). Production and purification of lentiviral
vectors. Nat Protoc 1, 241-245.
Townsend, S.E., Goodnow, C.C., and Cornall, R.J. (2001). Single epitope multiple staining to
detect ultralow frequency B cells. J Immunol Methods 249, 137-146.
Uyttenhove, C., Maryanski, J., and Boon, T. (1983). Escape of mouse mastocytoma P815 after
nearly complete rejection is due to antigen-loss variants rather than immunosuppression. J Exp
Med 157, 1040-1052.
Ward, P.L., Koeppen, H.K., Hurteau, T., Rowley, D.A., and Schreiber, H. (1990). Major
histocompatibility complex class I and unique antigen expression by murine tumors that
escaped from CD8+ T-cell-dependent surveillance. Cancer Res 50, 3851-3858.
154
Wherry, E.J., Ha, S.J., Kaech, S.M., Haining, W.N., Sarkar, S., Kalia, V., Subramaniam, S.,
Blattman, J.N., Barber, D.L., and Ahmed, R. (2007). Molecular signature of CD8+ T cell
exhaustion during chronic viral infection. Immunity 27, 670-684.
Willimsky, G., and Blankenstein, T. (2005). Sporadic immunogenic tumours avoid destruction by
inducing T-cell tolerance. Nature 437, 141-146.
Zhou, G., Lu, Z., McCadden, J.D., Levitsky, H.I., and Marson, A.L. (2004). Reciprocal changes
in tumor antigenicity and antigen-specific T cell function during tumor progression. J Exp Med
200, 1581-1592.
155
CHAPTER 3
Tumor antigen expression is required for immunoediting against sarcomas
Michel DuPage1, Claire Mazumdar1, Ann F. Cheung1, Tyler Jacks1,2
1
Koch Institute for Integrative Cancer Research and Department of Biology, Massachusetts
Institute of Technology, Cambridge, MA 02139, USA.
2
Howard Hughes Medical Institute, Massachusetts Institute of Technology, Cambridge, MA
02139, USA.
The author performed all experiments with the following exceptions. C. Mazumdar, an
undergraduate student supervised by the author, performed 5-aza-2’-deoxycytidine experiments
and assisted with transplantation and the monitoring of tumor growth. A. Cheung generated the
R26LSL-LSIY/+ mice. Experiments were performed in the laboratory of Tyler Jacks.
156
ABSTRACT
Cancer immunosurveillance and immunoediting have largely been defined using carcinogendriven models of sarcomagenesis. However, the heterogeneity and tractability of these models
is often complex, making it difficult to pinpoint key aspects of these processes. Here we use a
genetically engineered mouse model of sarcomagenesis to investigate cancer immunoediting.
This system allowed us to conditionally induce tumors harboring similar genetic and
histopathological characteristics that either did or did not express potent antigens recognizable
by endogenous T cells. We show that the process of immunoediting requires the presence of
potent T cell antigens and that lymphocytes drive the evolution of less immunogenic, or “edited,”
tumors by selecting for tumor cells that epigenetically silence the expression of the targeted
antigens.
157
INTRODUCTION
Cancer immunosurveillance is a mechanism by which immune cells, particularly
lymphocytes of the adaptive immune system, protect the host from the de novo development
and progression of cancer (Dunn et al., 2002; Schreiber, 2003). To understand this process,
investigations have relied heavily on the use of mouse models of cancer in which carcinogens,
primarily methylcholanthrene (MCA), are used to induce the development of tumors (Dunn et
al., 2006). Using these models, various immune-compromised mice lacking particular
populations of immune cells or harboring deficiencies in key immune effector pathways have
been shown to have an increased susceptibility to MCA-induced sarcoma formation, providing
experimental evidence to support cancer immunosurveillance (Kaplan et al., 1998; Shankaran
et al., 2001; Swann et al., 2008). Furthermore, as a consequence of the selective pressures
imposed on cancers during immunosurveillance, anti-tumor immune responses can also drive
the outgrowth of tumor cells with decreased sensitivity to immune attack, or decreased
“immunogenicity,” through a process called cancer immunoediting (Dunn et al., 2002;
Shankaran et al., 2001). Cancer immunoediting by the adaptive immune system was revealed
experimentally when tumors induced with MCA in lymphocyte-deficient mice were demonstrated
to be more immunogenic in transplantation assays than tumors induced in immune-competent
mice (Engel et al., 1997; Koebel et al., 2007; Shankaran et al., 2001; Svane et al., 1996).
Therefore, tumors that arise in immune-competent animals are susceptible to immune-mediated
destruction and those tumors that are not eradicated by immune attack, represent escape
variants that were shaped, or “edited,” by the immune system during tumor evolution. Despite
the evidence that cancer immunoediting occurs in these settings, as well as in human cancers,
we still lack a clear understanding of the mechanisms by which tumors are edited by immune
cells during this process.
158
Recognition of tumor-specific antigens (TSAs) by lymphocytes may be critical for
immunosurveillance of cancers. Evidence to support the role of TSAs in cancer immunity comes
from the fact that effective immunization against MCA-induced sarcomas is often restricted to an
individual tumor, that is, immunization with one tumor rarely provides cross-reactive protection
against other independently derived tumors (Prehn, 1968; Prehn and Main, 1957). However,
these experiments only reveal an important role for TSAs in the context of prior immunization
and transplantation, not in the context of primary tumor development. In addition, the molecular
identification of TSAs from carcinogen-induced tumors can be laborious and difficult owing to
their unique existence in individual tumors. In a few studies of long-term-cultured tumor cell
lines, TSAs capable of mediating tumor regression were characterized, and the antigens were
found to be somatically mutated normal genes (Dubey et al., 1997; Monach et al., 1995; Stauss
et al., 1986). Therefore, a plausible mechanism by which the process of immunoediting may
occur is by enforcing the selective outgrowth of tumors that do not express specific antigens that
can be targeted by T or B lymphocytes. Immunoediting by selective antigen loss is supported by
the observation that serial passage of immunogenic tumors in immune-competent animals can
lead to the reduced immunogenicity of tumors (Prehn, 1968). In several examples, reduced
immunogenicity upon transplantation of tumor cell lines has been demonstrated to occur by the
selective outgrowth of tumor cells that no longer express the TSAs targeted by lymphocytes
(Stauss et al., 1986; Urban et al., 1982; Uyttenhove et al., 1983; Zhou et al., 2004). However,
these types of investigations, while feasible with transplantable cell line models of cancer, are
extremely difficult to perform in the context of spontaneous or carcinogen-induced primary
tumorigenesis because the antigens targeted by T cells are unknown. Therefore, it is still
unclear whether T cell responses directed against TSAs and subsequent tumor antigen loss
during the evolution of primary tumors in their natural setting are critical features of the
immunoediting process. The investigation of these key issues would be simplified with a more
159
tractable system in which primary tumors could be initiated that harbored defined tumor
antigens.
To investigate the role of tumor-specific antigens targeted by endogenous T cells in the
process of immunoediting, we utilized a genetically engineered mouse model of
sarcomagenesis (Kirsch et al., 2007). This model allowed us to induce the in situ transformation
of genetically and histopathologically defined sarcomas that could be engineered to express
model tumor antigens. The capacity to induce sarcoma formation with defined tumor antigens
provided a significant improvement over carcinogen-induced primary sarcomas because it
allowed us to monitor antigen-specific T cell responses to tumors, track the expression of the
antigens in tumors, and compare the immunogenicity of tumors in the presence or absence of
these tumor antigens. We find that the process of cancer immunoediting requires the presence
of TSAs that can be recognized by T lymphocytes. Furthermore, cancer immunoediting can
occur through the selective outgrowth of tumor cells that have epigenetically silenced TSA
expression.
RESULTS
Sarcoma formation with increased penetrance and reduced latency in Rag-2-/- immunecompromised mice
To investigate whether T lymphocytes could prevent the formation of tumors in an
autochthonous mouse model of sarcomagenesis, we utilized a recently described model to
induce the formation of sarcomas by intramuscular injection of viral vectors expressing Cre in KrasLSL-G12D/+;p53fl/fl;Rag-2+/- (KP) or K-rasLSL-G12D/+;p53fl/fl;Rag-2-/- (KPR) mice (Kirsch et al., 2007).
To induce sarcomas with potentially immunogenic antigens we used lentiviral vectors that
express Cre alone (Lenti-x) or vectors that, in addition to Cre, express the T cell antigens
SIYRYYGL (SIY) and two antigens from ovalbumin – SIINFEKL (SIN, OVA257-264) and OVA323-339
160
– fused to the C-terminus of luciferase (Lenti-LucOS), as recently described (DuPage et al.,
Chapter 2). Fusion of the T cell antigens to luciferase provided a means to assay for the
maintenance of antigen expression in sarcomas.
KP mice that were induced to express strong T cell antigens with Lenti-LucOS formed
tumors later (85.7 ± 6.2 days) than KPR mice (73.6 ± 4.3 days) (Figure 1A,B). In addition,
whereas only 28% of KP mice developed sarcomas within the first 180 days after tumor
induction via Lenti-LucOS injection, 100% of KPR mice developed sarcomas after Lenti-LucOS
injection (Figure 1A,C). A comparable difference in sarcoma latency and penetrance was also
observed in KP compared to KPR mice injected with Lenti-x (Figure 1B,C). Therefore, tumor
immunosurveillance in this model of sarcomagenesis may act independently of the presence of
exogenous antigens. Alternatively, it is possible that Cre is immunogenic in the context of
sarcoma development, although Cre was not found to be immunogenic in previous studies
investigating lung adenocarcinoma development (DuPage, et al., Chapter 2). Interestingly,
sarcomas also developed with significantly delayed onset (beyond 180 days) in immunecompetent mice. In KP mice injected with Lenti-LucOS, 4/13 mice monitored beyond 180 days
developed tumors at 186, 278, 359, or 525 days after intramuscular injection. Only 1/6 mice
injected with Lenti-x developed a tumor after 180 days. It is possible that the late onset of
tumors seen in KP mice, but not KPR mice, is due to the activity of adaptive immune responses
that maintain small tumors in an equilibrium state (Dunn et al., 2002; Koebel et al., 2007).
To further investigate the importance of the expression of experimentally delivered T cell
antigens in mediating these phenotypes, we used K-rasLSL-G12D/+;p53fl/fl;R26LSL-LSIY/+ (KPL-SIY)
mice which are tolerant to luciferase and the SIY antigen due to thymic deletion of reactive T
cells (Supplementary Figure 1 and (Cheung et al., 2008)). We compared sarcoma
development KP versus KPL-SIY mice after the injection of Lenti-LucS, a lenti-vector that
expresses SIY fused to luciferase. Thus, we were able to directly compare tumor induction
161
,
>
455
9:;
9:
=5
E%("0F%$G0A?/(CD?A0HIJ
E%("0F%$G0A?/(CD?A0HIJ
455
65
<5
25
5
5 <5
425
65
=5
455
>?@A0BCA$0$'DC/0%#%$%?$%C#
K3
#0106
35
23
#0103
5
!"#$%$
!"#&'&(&')*%$
8
*?/(CD?0C#A"$0H-?@AJ
425
9:
9:;
455
=5
65
<5
!"#$%&!'()*
.
,-&./"
#01044
455
E%("0F%$G0A?/(CD?A0HIJ
!"#$%&+
#01045 #0106
K3
#0107
9:
9:;
#01024
35
#01023
23
5
!"#$%&!'()*
!"#$%&+
,-&./"
Figure 1. Sarcoma formation in Rag-2-/- mice occurs with reduced latency and increased
penetrance.
(A) 25 K-rasLSL-G12D/+;p53fl/fl;Rag-2+/- mice (KP) and 11 K-rasLSL-G12D/+;p53fl/fl;Rag-2-/- mice (KPR)
were injected intramuscularly with equivalent doses of Lenti-LucOS and the development of
palpable sarcomas was monitored twice a week for at least 180 days. The percentage of mice
in each group with sarcomas is depicted over the first 120 days (no tumors developed between
120-180 days). Cumulative data from at least three independent experiments is shown (A-C).
(B) Time after intramuscular injection of Lenti-LucOS, Lenti-x, or Ad-Cre in which sarcomas first
became palpable in KP (triangles) or KPR (circles) tumor-bearing mice. Data only includes
162
sarcomas that arose within the first 180 days, see text for data relating to tumors arising after
180 days.
(C) Percentage of KP and KPR mice with sarcomas of the total (n) injected intramuscularly with
Lenti-LucOS, Lenti-x, or Ad-Cre within the first 180 days (see text for data relating to tumors
arising after 180 days).
(D)
5 KP and 6 KP-LSIY (K-rasLSL-G12D/+;p53fl/fl;R26LSL-LSIY/+) littermates received intramuscular
injection of Lenti-LucS and the percentage of mice with sarcomas within the first 180 days is
plotted. See text for a comparison of the latency of tumor development in these mice.
utilizing the same lentiviral vector (Lenti-LucS) in immune-competent mice versus mice that
were specifically tolerant only to the LucS antigens introduced into the tumors. KPL-SIY mice
had an enhanced susceptibility to sarcoma formation with Lenti-LucS and the sarcomas became
palpable much sooner than in KP littermates (89.5 ± 17.4 versus 139 days) (Figure 1D). Thus,
protection from sarcoma formation depends on the presence of strong antigens in tumors that
can be recognized by T cells. Furthermore, although Cre may be immunogenic in this setting,
expression of Luciferase and SIY are clearly influential in preventing sarcoma formation.
Functional T cell responses are generated against antigens expressed in sarcomas
An advantage of this genetically engineered and conditional cancer model is that we can
examine the endogenous antigen-specific T cell response to engineered tumor antigens. We
used MHCI/Kb DimerX reagents loaded with SIY and SIN peptides to track tumor-reactive CD8+
T cells in Lenti-LucOS-induced sarcoma-bearing mice by flow cytometry. To improve the
sensitivity of the DimerX reagent, we utilized both PE- and APC-labeled dimers to co-stain CD8+
T cells. Mice bearing Lenti-LucOS sarcomas, but not mice bearing Lenti-x sarcomas, harbored
CD8+ T cells specific to SIY and SIN in the inguinal lymph nodes nearest to the tumors as well
as in the spleen (Figure 2A,B). To look at the functional activity of these T cells, we
restimulated lymphocytes harvested from the lymphoid organs of tumor-bearing mice with SIY
and/or SIN peptides in vitro and performed intracellular cytokine staining (ICCS) for IFN-γ and
163
TNF-α production. CD8+ T cells specific for the tumor antigens were completely functional as
evidenced by the fact that all of the cells could produce large quantities of both IFN-γ and TNF-α
upon in vitro restimulation (Figure 2A-D). This is in complete contrast to the dramatically
reduced activity of T cells responding to lung adenocarcinomas that expressed the same
antigens (DuPage et al., Chapter 2). Interestingly, we consistently found slightly increased
percentages of antigen-specific T cells in the inguinal lymph node closest to the tumor
compared to the contralateral lymph node (Figure 2A). This may indicate an increased
concentration of the antigens in the proximal lymph node due to ongoing antigen production in
the tumors.
We also analyzed KP mice injected with Lenti-LucOS that did not develop sarcomas for
antigen-specific T cells. We hypothesized that such T cell responses could be protecting these
mice from sarcoma formation. Indeed, we detected fully functional antigen-specific T cells in
these mice as well (Figure 2C,D). Importantly, intramuscular lentiviral injection of p53fl/fl mice
with Lenti-LucOS that cannot initiate tumor formation also generated fully functional antigenspecific T cells (Supplementary Figure 1). Therefore, antigen-specific T cell responses
induced after intramuscular injection of Lenti-LucOS are functional and likely provide protection
from the formation of sarcomas that express these antigens.
Cancer immunoediting phenotypes require the presence of potent T cell antigens
To assay whether autochthonous sarcomas become “edited” (i.e. less susceptible to
rejection) when derived in immune-competent animals, we compared the growth of multiple
sarcomas independently derived in KP or KPR mice after transplantation into either wildtype or
Rag-2-/- immune-compromised mice. While Lenti-LucOS-induced tumors derived in KP mice
grew similarly upon transplantation into both wildtype or Rag-2-/- mice, many of the Lenti-LucOSinduced tumors derived in KPR mice were completely rejected (1/7) or had stunted growth (4/7)
164
!"
$
!"
#
!"
!
134567<
134567189:;
"&"$
!" "
!"
"
!"
!
!"
#
;,=>;,.?@ 72B
!"
$
!"
%
A
!" %
/.-α
!"
#
!"
!
"&"
>;,=
>;,.
!" $
"&"(
!" !
%
!"
%
!"
$
!"
$
!"
#
!"
#
!"
!
!"
!
!" "
!" "
!" #
!" $
!" !
!" #
;,=>;,.?@ 72B
!" %
A
!" $
!" %
"&#%
>;,=
>;,.
!" "
!" %
!"
$
#
!"
#
!"
!
!"
!
"&"+
!" "
!" %
!" "
!" !
!" #
!" $
"&!
!" "
!"
,-.γ
N
"&!'
!" $
!" "
!" "
!"
21.
%
;,=>;,.?@A7C2D
!"
01.
1.
;,=>;,.?@A7C2D
TJ53UFD0*>
/.-α
C
!" !
!" #
!" $
"&"*
>;,=
>;,.
"&"%
!" "
!" %
!" %
!" "
!" !
!" #
!" $
!" %
,-.γ
TJ53UFD0*>
134567189:;
134567<
$
!"
#
!"
!
!"
"
"&"(
!" "
!" !
!" #
!" $
%
!"
#
!"
!
!" "
/.-α
!" $
!"
#
!"
!
!"
"&!
!" "
!" !
!" #
!" $
!"
#
!"
!
!" "
!#+
!""
(+
+"
#+
"
#
!"
!
!" "
!" "
!" !
!" #
!" $
!" %
"&*'
')&'
!" !
!" #
!" $
!" %
,-.γ
0
;PS334
23E9345F,-.γ>/.-α>
23E9345F,-.γ>/.-α>
1OLPQF4KR3I
!+"
!" %
!"
!&))
>;,=
>;,.
!" "
!" %
,-.γ
D
!" $
"&(!
;,.?@A72B
%
!" $
!" "
!" #
;,=?@A72B
"&"#
>;,=
>;,.
!" !
%
!" $
!" "
!" %
/.-α
!"
!&$
!" $
;,=>;,.?@A72B
%
!"
;,.?@A7C2D
!"
!"
;,=?@A7C2D
;,=>;,.?@A7C2D
;21BB.
!" %
#+"
#""
!+"
!""
G;.7;,=
134567189:;HFIJE9KLJ
134567189:;HF58LKE7ME33
+"
"
Figure 2. Functional T cell responses are generated against antigens expressed in
sarcomas.
(A) Top: FACS analysis using SIY- and SIN-loaded DimerX reagents to detect tumor antigen
specific CD8+ T cells in the inguinal lymph nodes either draining (DLN) or peripheral to (PLN)
the tumors of Lenti-x-induced or Lenti-LucOS-induced tumor-bearing mice. FACS plots are
gated on PI-negative, CD8+ cells. Bottom: IFN-γ and TNF-α cytokine production in CD8+ T cells
165
from the lymph nodes of mice depicted in (A) after restimulation in vitro with both SIY and SIN
peptides. FACS plots are representative of at least three mice per group.
(B) Similar to analyses in (A) except with cells harvested from spleens of tumor-bearing mice.
To detect tumor-specific CD8+ T cells in the spleens of Lenti-LucOS tumor bearing mice FACS
was performed with DimerX reagents loaded either with SIY or SIN.
(C and D) The percent of SIY and SIN-specific T cells that were IFN-γ+TNF-α+ from lymph
nodes (C) or spleens (D) of KP mice infected with Lenti-LucOS that developed a “sarcoma” or
were “tumor-free” at the time of analysis. Percentages were determined by comparing the
fraction of CD8+ cells in duplicate samples stained for DimerX or cytokine production. As a
positive control, mice were infected with an influenza strain engineered to express the SIY
antigen (WSN-SIY) and endogenous T cells reactive to SIY were analyzed after four months.
Data is the mean ± SEM from 3-4 mice per group.
upon transplantation into wildtype mice compared to Rag-2-/- mice (Figure 3A,B). These results
closely recapitulate previous experiments performed in a similar manner using a MCA-induced
model of sarcomagenesis (Shankaran et al., 2001).
Next we wanted to determine whether similar results could be observed after the
transplantation of Lenti-x-induced sarcomas that lacked the strong T cell antigens. Surprisingly,
Lenti-x-induced sarcomas derived in KP and KPR mice grew equally well when transplanted
into wildtype and Rag-2-/- mice, regardless of their origin (Figure 3C,D). Therefore, unlike LentiLucOS-induced sarcomas derived in KPR mice, Lenti-x-induced sarcomas derived in KPR mice
were not susceptible to lymphocyte-mediated anti-tumor responses. We hypothesize that a lack
of potent T cell antigens in Lenti-x-induced sarcomas may be responsible for the failure of
lymphocytes to impede tumor growth upon transplantation, and that tumor antigen expression is
a prerequisite for the lymphocyte-mediated immunoediting process. The previously reported
immunogenicity of MCA-induced sarcomas derived in immune-compromised mice may be due
to the de novo generation of potent tumor neoantigens as a consequence of the mutagenic
activity of MCA (Khong and Restifo, 2002; Prehn and Main, 1957). However, we found that
MCA treatment of non-immunogenic Lenti-x-induced sarcoma cell lines in vitro, very rarely
produced clones with increased immunogenicity (Supplementary Figure 2). This may not be
166
!"#$%&!'()*
=
@AAA
2'3405647'3"58339:
2'3405647'3"58339:
,-.&/"0%1"/
;
!"#$%&+
EAA
DAA
CAA
BAA
A
B?AA
BAAA
@?AA
@AAA
?AA
A
?
@A
@?
BA
<FGH5FI$"05$0F#HJ7F#$
>
<
2'3405647'3"58339:
2'3405647'3"58339:
,-&/"0%1"/
B?AA
BAAA
@?AA
@AAA
?AA
A
?
@A
@?
BA
?
@A
@?
BA
<FGH5FI$"05$0F#HJ7F#$
B?AA
BAAA
@?AA
@AAA
?AA
A
?
@A
@?
BA
<FGH5FI$"05$0F#HJ7F#$
<FGH5FI$"05$0F#HJ7F#$
Figure 3. Cancer immunoediting phenotypes require the presence of potent T cell
antigens.
Results plotted in A-D are the mean tumor volume ± S.E.M. of tumors transplanted into 3-5
wildtype or 1-2 Rag-2-/- recipients. Transplant experiments were repeated for many of the tumor
lines in A-D with similar results.
(A) Lenti-LucOS-induced sarcomas derived independently in KPR mice (n = 6, different color
lines) were transplanted into wildtype mice (solid lines) or Rag-2-/- mice (dashed lines) and
tumor growth was measured. One additional tumor line exhibited delayed growth in wildtype
mice but was excluded from this plot due to its rapid growth beyond the scale of the tumors
depicted.
(B) Similar to (A) except Lenti-LucOS-induced sarcomas were derived in KP mice.
(C) Similar to (A) except sarcomas were induced in KPR mice with Lenti-x.
(D) Similar to (C) except sarcomas were induced in KP mice with Lenti-x.
167
surprising since experiments with fibroblasts transformed with MCA in vitro have shown that
immunogenicity and tumorigenicity are linked, meaning that the immunogenicity of tumor cell
lines was the result of mutations that were likely to be both antigenic and oncogenic (Embleton
and Heidelberger, 1972; Mondal et al., 1971).
Immunoediting occurs by antigen loss
If it is true that cancer immunoediting by lymphocytes requires the presence of potent
tumor antigens, we hypothesized that Lenti-LucOS-induced tumors that appear edited after
forming in KP mice, may result from the selection of tumor cells that have lost the expression of
these potent antigens. We analyzed the expression of the tumor antigens by measuring the
levels of luciferase activity in tumors. Whereas tumors derived in KPR mice were universally
luciferase positive, tumors derived in KP mice had reduced luciferase activity in all but one out
of six sarcomas tested (Figure 4A,B). Interestingly, we discovered that the sarcoma that had
retained luciferase activity also had significantly reduced H-2Kb expression, the MHC class I
allele responsible for presenting the SIY and SIN antigens, when compared to a panel of other
sarcomas, even after prior treatment with IFN-γ (Supplementary Figure 3). Therefore, cancer
immunoediting appears to proceed by the selective outgrowth of tumors that either do not
express potent antigens or cannot present the antigens on their cell surface.
Lenti-LucOS-induced sarcomas derived in KPR mice that grew out after transplantation
into wildtype mice, also lost the expression of tumor antigens (Supplementary Figure 4).
Importantly, the growth of tumors that lost antigen expression after being passaged through
wildtype mice grew comparably upon secondary transplantation into wildtype and Rag-2-/- mice,
whereas sarcomas passaged through Rag-2-/- mice did not (Supplementary Figure 4), thus
mirroring the results of de novo sarcomagenesis in this transplantable system. To test whether
antigen loss in edited tumors was sufficient to provide a means of escape for Lenti-LucOS-
168
7
8
()
66
!"#$%&'(
+,-./0
+1234
+,-./05
+,-./05
!"
'
!"
&
!"
%
!"
$
!"
#
!"
!
:IH&
!"!%$
$::H
%"H"
$I!I
%#$#
:IH'
:$HH
:I%%
:I%H
II!#
!""HH
!""!&
*+MNEµAEO;P.,/-
()*
+,-./0+1234
()*
()
7A+PGG
7A+PGGQE*=A#-1BB
$I!I
!&""
!"""
&""
"
& !" !& #" #& $" $&
L=>GE=F.,;E.;=-GOB=-.
#&""
%"H"
#"""
!&""
!"""
&""
"
& !" !& #" #& $" $&
L=>GE=F.,;E.;=-GOB=-.
L
*,B=./D,EB12/F,;=G,
?1<P;ESPB1<,ET<<$U
#"""
?1<P;ESPB1<,ET<<$U
7A+PGG0+1234
7A+PGG0+1234QE*=A#-1BB
R
);/<=;>
?50@*=A-1BB
?50@*=AC
!9#
!9"
"9:
"9'
"9%
"9#
"9" 0 0 0
C C C
%"H"
0 C0C 0C 0C
$I!I
0 C 0 C 0 C 0 C J7K=
$::H
Figure 4. Immunoediting occurs by the selection of tumor cells that have epigenetically
silenced antigen expression.
(A) Representative in vivo luciferase activity of Lenti-x-induced or Lenti-LucOS-induced
sarcomas derived in K-rasLSL-G12D/+;p53fl/fl;Rag-2+/- (KP) or K-rasLSL-G12D/+;p53fl/fl;Rag-2-/- (KPR)
mice demonstrates antigen loss in KP-derived sarcomas. Note: “POS CTRL” sarcoma was
luciferase positive because it was induced in a K-rasLSL-G12D/+;p53fl/fl;R26LSL-LSIY/+ (KPL-SIY)
mouse.
(B) Luciferase expression from Lenti-LucOS-induced sarcoma cell lines derived in KPR or KP
mice.
(C) Two independent tumors (3919 and 4070) that lost antigen expression (AgLoss) upon
passage through immune-competent mice were transplanted into wildtype (solid blue line) or
Rag-2-/- (dashed blue line) 129S4/SvJae mice and tumor growth was monitored. Alternatively,
the antigens were reintroduced by lentiviral infection with a LucOS-expressing construct (AgLossLucOS) and then transplanted into wildtype (solid red line) or Rag-2-/- (dashed red line)
129S4/SvJae mice and tumor growth was monitored. Tumor volumes represent the mean ±
SEM of transplants into three wildtype mice or one Rag-2-/- mouse as a control.
(D) Cell lines from three independent Lenti-LucOS sarcomas induced in KPR mice (Primary,
black columns), and then passaged through Rag-2-/- mice (Tx->Ragnull, grey columns), or
passaged through wildtype mice (Tx->Rag+) were treated for three days with 1µM of 5-aza-2’deoxycytidine (Aza) and then analyzed for luciferase activity (luciferase activity from untreated
cell lines is also shown). Mean ± SEM from two experiments.
169
induced sarcomas derived in KP mice, we reintroduced the LucOS antigens into sarcomas that
had lost expression of the antigens after passage in wildtype mice (called AgLoss tumors). Reexpression of LucOS in these tumors led to severely stunted tumor growth, indicating that loss
of antigen expression was the primary means of tumor escape in this setting (Figure 4C).
Tumor antigens expression is impeded by hypermethylation
To test whether epigenetic silencing of tumor antigen expression via DNA methylation
was responsible for antigen loss and tumor escape, we treated cell lines that had lost luciferase
expression with 5-aza-2’-deoxycytidine (Aza) and monitored the cells for the induction of
luciferase activity. In several independent lines that lost luciferase expression after
transplantation into immune-competent mice, luciferase activity could be completely restored
with Aza treatment (Figure 4D). Therefore, in the context of T cell responses targeting potent
antigens that are not required for tumor maintenance, tumors can escape through the selection
of cells that have epigenetically silenced the expression of the tumor antigens.
DISCUSSION
Using an autochthonous, genetically engineered model of sarcomagenesis we show that
T lymphocyte-driven tumor antigen loss is a critical means by which cancer immunoediting
occurs in a primary tumor setting. While tumor antigen loss has been shown in multiple
transplantable models of cancer, it has not been demonstrated in tumors that arise
endogenously. This is primarily due to the fact that carcinogen-induced models of primary
sarcomagenesis generate different antigens in each individual tumor, making it impossible to
track their expression during primary tumor development. Furthermore, identifying such
antigens from progressively growing tumors may not be possible if antigen expression is lost
during the natural progression of these tumors due to the pressures of immunoediting. The
170
presence of such antigens could have been inferred by the fact that sarcomas derived in
lymphocyte-deficient mice are often rejected upon transplantation into immune-competent mice,
but such antigens have yet to be identified. In addition, a multitude of alternative mechanisms,
besides antigen loss, could provide a means for the escape of carcinogen-induced sarcomas in
immune-competent mice (Dunn et al., 2006; Smyth et al., 2001; Zitvogel et al., 2006). By
inducing sarcomas that express model tumor antigens, we were able to obviate these
complications, enabling us to track tumor-reactive T cells as well as the expression of model
tumor antigens in tumors.
Just as MCA-induced sarcomas do not form as readily in immune-competent compared
to immune-compromised mice, we found that autochthonous tumors arose less frequently in
immune-competent mice (KP) compared to Rag-2-/- mice (KPR). Protection seemed to be at
least partially dependent on the expression of tumor-specific antigens (TSAs) in the primary
tumors. Only tumors induced with integrating lentiviruses that provide stable gene expression in
infected cells showed differences in tumor penetrance in KP compared to KPR mice. Tumors
induced in wildtype mice with Ad-Cre viruses, that only express Cre protein transiently, did not
have reduced susceptibility to tumor formation (as demonstrated here and previously in (Kirsch
et al., 2007)). However, the titer of Ad-Cre and lentiviral-Cre vectors is difficult to compare
directly, and it is possible that the dose of Ad-Cre administered causes the transformation of a
much greater number of tumor-initiating cells, thus overwhelming the capacity of the immune
system to regulate the development of sarcomas. In addition, we found that tumor-specific
antigen expression was critical to immune surveillance when R26LSL-LSIY/+ mice, that are tolerant
to the luciferase and SIY antigens, were more susceptible to tumor formation with Lenti-LucS
than R26+/+ littermates. Although this model may be complicated by the use of viral vectors to
initiate the tumorigenesis process, all inducible models of cancer may be problematic due to
effects on the immune system during tumor induction. The effect of MCA-injection on the
171
immune system and tumor development has also been debated (Hewitt et al., 1976; Qin and
Blankenstein, 2004; Scott, 1991). Nevertheless, using a new model of sarcomagenesis, we find
evidence to support the hypothesis that lymphocyte responses can act early to eliminate
sarcomas prior to their progression to palpable tumors.
Using this autochthonous model of sarcomagenesis we also found evidence of an
equilibrium state waged between lymphocytes and tumors during the process of cancer
immunoediting. Although we have not provided direct evidence that lymphocytes actively
restrain tumor progression in this model, this was elegantly shown previously with the use of
antibodies to deplete specific immune cell populations (Koebel et al., 2007), it is evident that the
onset of overt disease is delayed in immune-competent KP mice compared to lymphocytedeficient KPR mice. It is possible that the equilibrium state may be partially dependent upon the
presence of tumor-specific antigens, because significantly delayed tumor onset (after 180 days)
occurred more frequently in mice with Lenti-LucOS-induced sarcomas (30.7%, n = 13) than
mice with Lenti-x-induced sarcomas (16.7%, n = 6). It would be interesting to detect very early
stage tumors that are held in an equilibrium state to determine whether the maintenance of
antigen expression is required for this phase of immunoediting, however, it has not been
possible to detect such early lesions in this model. This may indicate that the equilibrium phase
occurs at a stage of tumor development that is not detectable by histological analysis. Depleting
CTLs with anti-CD8 antibodies following tumor-induction could serve as an alternative means to
examine tumors in equilibrium. CD8-depletion may release tumors to grow progressively so that
they could be harvested and analyzed for antigen expression (Koebel et al., 2007).
Ultimately, autochthonous tumors escape immune regulation by losing expression of
TSAs. The absence of TSAs in sarcomas arising in immune-competent mice coincides with the
capacity of these tumors to grow progressively upon serial transplantation into other immunecompetent mice. Re-introduction of the same tumor antigens into sarcomas that had escaped
172
immunosurveillance, was sufficient to make the sarcomas susceptible to lymphocyte-mediated
regulation, making it unlikely that alternative means of immune evasion were acquired in these
tumors. The mechanism of antigen loss in these tumors was epigenetic in nature because
antigen expression in tumor cell lines could be restored by blocking the incorporation of
methylated DNAs with 5-aza-2’-deoxycytidine treatment. However, it is possible that the
expression of model TSAs from genomically integrated lentiviral vectors might make the
antigens more susceptible to gene silencing by methylation. It will be important to examine
whether TSAs expressed from endogenous loci are silenced by a similar mechanism to evade
immunosurveillance.
Most importantly, we found that inducing sarcomas in the absence of potent TSAs (using
Lenti-x), significantly reduced the immunogenicity of transplanted sarcomas. Lenti-x-induced
sarcomas derived in Rag-2-/- mice were able to grow without inhibition after transplantation into
immune-competent hosts. This may indicate that the immunogenicity of MCA-induced sarcomas
is a consequence of the generation of carcinogen-mediated mutations that serve as TSAs in
these tumors (Embleton and Heidelberger, 1975; Khong and Restifo, 2002; Prehn and Main,
1957). In contrast to carcinogen-induced cancer models, cancers that arise spontaneously in
mice have typically been found to be weakly immunogenic (Embleton and Heidelberger, 1972,
1975; Hewitt et al., 1976; Prehn and Main, 1957; Scott, 1991). Therefore, expression of TSAs
may be a critical prerequisite for the establishment of immunosurveillance and the process of
immunoediting. The identification and characterization of such TSAs in human cancers may
also be a critical prerequisite for the generation of effective anti-cancer immunotherapies in
patients suffering with this disease.
173
MATERIALS AND METHODS
Mice and tumor induction
129S4/SvJae strains backcrossed 8 generations were used for all experiments. Trp53fl mice
were provided by A. Berns (Jonkers et al., 2001), K-rasLSL-G12D were generated in our laboratory
(Jackson et al., 2001), and Rag-2-/- mice were purchased from The Jackson Laboratory (Bar
Harbor, ME). Sarcomas were induced in KP and KPR mice by intramuscular injection of the left
hind limb with replication-incompetent viruses expressing Cre recombinase as reported
previously (Kirsch et al., 2007). Mice were monitored twice a week for palpable sarcoma
formation beginning 50 days after intramuscular injection. All animal studies and procedures
were approved by the Massachusetts Institute of Technology’s Committee for Animal Care.
Lentiviral production
Lentivirus production by transfection of 293T cells with Δ8.2 (gag/pol), CMV-VSV-G, and various
transfer vectors expressing Cre was performed as previously described (Tiscornia et al., 2006).
Preparation, culture, and transplantation of primary sarcomas
Primary sarcomas were explanted and single cell suspensions were generated by mincing and
digesting the tissues for ~1 hour at 37°C in 125 U/ml Collagenase Type I (Gibco), 60 U/ml
Hyaluronidase (Sigma), and 2 mg/ml Collagenase/Dispase (Roche), followed by passage
through a 70 µm filter. In most experiments, 2x105 cells were used for immediate subcutaneous
transplantation of the freshly isolated single cell suspensions. Alternatively, cell lines were
generated from the primary autochthonous tumors and subsequently 2x105 cells were
transplanted after trypsinization and three washes in plain DME medium. Immune-competent or
Rag-2-/- mice on the 129S4/SvJae background were used for all transplants. These recipient
mice came from the same mouse strains used to generate the autochthonous tumors.
174
Subcutaneously transplanted tumor volumes were calculated by multiplying the length x width x
height of each tumor.
Flow cytometry
Cell suspensions from lymphoid organs were prepared by mechanical disruption between
frosted slides. Single cell suspensions were stained with the specified antibodies for 20-30 min
after treatment with Fc Block (BD Pharmingen). α-CD8α (53-6.7), α-IFNγ (XMG1.2), α-TNFα
(MP6-XT22), and DimerX I (Dimeric Mouse H-2Kb:Ig) were from BD Pharmingen. All antibodies
were used at 1:200. Peptide-loaded DimerX reagents were prepared as specified by
manufacturer and used at 1:75. Staining cells with the same ligand labeled with PE or APC for
improved sensitivity has been previously reported (Stetson et al., 2002; Townsend et al., 2001).
Propidium iodide was used to exclude dead cells when cell were not fixed. Cells were read on a
FACSCalibur and analyzed using Flowjo software (Tree Star).
Cytokine production
Single cell suspensions from lymphoid organs were resuspended in the presence or absence of
SIYRYYGL and SIINFEKL peptides in OPTI-MEM I (Gibco) supplemented with GolgiPlug (BD
Pharmingen) for ~4 hours at 37°C, 5% CO2. Cells were then fixed and stained for intracellular
cytokines using the Cytofix/Cytoperm kit (BD Biosciences).
Luciferase detection
Freshly explanted tumors or cell lines were lysed in Cell Culture Lysis Reagent, mixed with
Luciferase Assay Reagent according to manufacturer’s instructions (Promega), and relative light
units (RLU) were detected using the Optocomp I luminometer (MGM Instruments). RLUs were
standardized based on the total number of cells or total protein (Bio-Rad Protein Assay) used in
175
each assay. In vivo bioluminescence images were obtained using the NightOWLII LB983
(Berthold Technologies) or the IVIS Spectrum (Xenogen Corp.) bioluminescent imaging systems
after intraperitoneal injection of 1.5 mg Beetle Luciferin (Promega).
5-aza-2’-deoxycytidine treatment
Tumor cell lines were plated at low confluency (2x105 cells/ well of 6-well plate), and treated with
1µM of 5-aza-2’-deoxycytidine replaced daily for three consecutive days. Cells were then
analyzed for luciferase activity.
Influenza
The WSN-SIY influenza strain was provided by J. Chen. Mice were infected with 20 pfu of virus/
mouse by intratracheal intubation and inhalation. FACS analyses presented here were
performed four months after infection.
Statistical analyses
P-values for statistical comparisons were generated using unpaired two-tailed Fisher exact
probability tests or student’s T-tests.
ACKNOWLEDGEMENTS
We would like to thank A. Charest and P. Sandy for providing lentiviral vectors, G. Paradis for
assistance with flow cytometry, H. Eisen for reagents and experimental discussions, and M.M.
Winslow and A.G. DuPage for critical reading of this manuscript. This work was supported by
Grant 1 U54 CA126515-01 from the NIH and partially by Cancer Center Support (core) grant
P30-CA14051 from the National Cancer Institute and the Margaret A. Cunningham Immune
Mechanisms in Cancer Research Fellowship Award (M.D.) from the John D. Proctor
Foundation. T.J. is a Howard Hughes Investigator and a Daniel K. Ludwig Scholar.
176
7
=->./4=?@AB
=->./4=?@AB
25
#&'+
!" $
#
!"
!
!&+%
!" $
!" "
!"
#
!"
!
!" "
!" "
!" !
!" #
!" $
!" "
!" %
B9CEB9;123456
!" %
!" $
!"
#
!"
#
!"
!
!"
!
"&'*
!" "
<;:α
#
!"
!
!" #
!" $
!"
"&!*
!" "
%
"&'!
!" "
!" !
!" #
!" $
!"
#
!"
!
#&!$
,-,./0-1234758
!" $
!" !
!" #
!" $
!" %
!"
$
#
!"
#
!"
!
!"
!
!" "
!"#$%&-,B9C
!" %
!"
!" !
!" #
!" $
!" %
"&%'
!" "
!" !
!" #
!" $
!"
$
!"
!"
'&(!
#
!"
#
!
!"
!
!" "
!" %
!" !
!" #
!" $
!" %
!" "
!" %
!" %
"&%)
!"&#
$
!"
$
!"
$
!"
$
!"
#
!"
#
!"
#
!"
#
!"
!
!"
!
!"
!
!"
!
!&#(
9:;γ
!" !
!" #
!" #
!" $
!" %
!"
!" "
!" !
!" %
,-,./0-12 456
3
)&"*
!" "
"&"(
!" "
!" "
!" $
!" %
"&!+
!" "
!" "
!" !
!" #
!" $
!" %
<;:α
<;:α
!" %
$
,-,./0-12 456
!"#$%&'(')'(*+,B9C
!"
3
!" %
=->./4=?@AB
B9;
!" %
!" "
!" "
!" %
"&"+
9:;γ
=->./4=?@AB
B9;
!" %
!" $
"&#
9:;γ
D
!" #
!" "
!" "
!" %
!" !
;F
,-,&
!" $
!" "
!" %
!&!$
!" "
!" %
"&"*
;F
,-,&
!" $
!"
!" !
<;:α
!"
%
!" $
EB9C
EB9;
!" $
!" "
!" #
!" %
!&++
EB9C
EB9;
!" !
B9CEB9;123456
,-,./0-1234758
!"
G/H0.I,-
!" %
B9CEB9;1234758
B9CEB9;1234758
!" %
!&'
!" "
!" "
!" !
!" #
!" $
!" %
"&"
"&"'
!" "
!" "
!" !
!" #
!" $
!" %
9:;γ
177
Supplementary Figure 1. T cell responses are specific to exogenously introduced
neoantigens expressed in lentiviruses and tumors.
(A) KP (left) or wildtype mice (right) were infected intramuscularly with Lenti-LucOS and
analyzed five weeks later (no tumors were apparent in KP infected mice). Top: FACS analysis
using SIY- and SIN-loaded DimerX reagents to detect antigen specific CD8+ T cells in the
spleen. FACS plots are gated on PI-negative, CD8+ cells. Bottom: IFN-γ and TNF-α cytokine
production in CD8+ T cells from the spleens of mice depicted in (A) after restimulation in vitro
with both SIY and SIN peptides or left unstimulated with peptides. Representative of the
analysis of three mice per group.
(B) Comparison of KP (left) or KPL-SIY mice (right) infected intramuscularly with Lenti-LucOS
shows a lack of SIY-specific T cells in mice with the R26LSL-LSIY allele. Top: FACS analysis using
SIY- or SIN-loaded DimerX reagents to detect antigen specific CD8+ T cells in the spleen. FACS
plots are gated on PI-negative, CD8+ cells. Bottom: IFN-γ and TNF-α cytokine production in
CD8+ T cells from the spleens of mice depicted in (A) after restimulation in vitro with either SIY
or SIN peptides. Representative of the analysis of at least 2 mice per group.
178
9#99
9#99:);&<$.200
9#99=>?@
9#99=>?@:);&<$.200
1234-)54023,)63378
1234-)54023,)63378
$!##
9#9$=>?@
9#9$=>?@:);&<$.200
9#9$=>?@
9#9$=>?@:);&<$.200
$###
"!##
"###
!##
#
$!##
$###
"!##
"###
!##
#
!
"#
"!
$#
%&'()&*++,-)+-&.(/0&.+
!
"#
"!
$#
%&'()&*++,-)+-&.(/0&.+
Supplementary Figure 2. MCA treatment in vitro may enhance the immunogenicity of
Lenti-x-induced sarcomas.
Examples from two Lenti-x-induced sarcoma cell lines treated with 40 µg/ml of MCA for six days
before single clones were picked and grown by limiting-dilution. Only two clones of 20 total
MCA-treated clones that were transplanted, showed a modest increase in immunogenicity
compared to the parent (4044, left) or other MCA-treated clones (4042, right). However,
because it was such a rare occurrence to find clones that had increased immunogenicity after
treatment with MCA, it is possible that these clones represent variants already present within
the primary tumor cell population that were naturally more immunogenic, and their
immunogenicity may not be a direct effect of MCA treatment.
179
()*+)2-.
!"!%$
//!#
'$11
!""11
!""!0
!""
!""
!""
!""
!""
'"
'"
'"
'"
'"
&"
&"
&"
&"
%"
%"
#"
#"
#"
#"
"
"
"
"
!"
"
!"
!
34#5
!"
#
!"
$
!"
%
()*+),-.
%"
!"
"
!"
!
!"
#
!"
$
!"
%
!"
"
&"
()*+),-.
%"
%"
!"
!
!"
#
!"
$
!"
%
#"
"
!"
"
!"
!
!"
#
!"
$
!"
%
!"
"
!"
!
!"
#
!"
$
!"
%
6
Supplementary Figure 3. Reduced expression of H-2Kb on the surface of a KP-derived
Lenti-LucOS-induced sarcoma that maintained expression of the LucOS antigens.
Freshly harvested sarcomas were cultured 20-24 hours in the presence of 10 units of IFN-γ
(solid line) or left untreated (dashed line) and then stained with anti-H-2Kb antibodies and
analyzed by flow cytometry (shaded histogram are negative controls stained with anti-CD8
antibodies). All sarcomas were induced with Lenti-LucOS. 10143 was derived in a KPR mouse
and retained luciferase expression, while 8377, 9912, and 10077 were derived in KP mice and
were luciferase negative (see Figure 4B). 10015 was derived in a KP mouse but retained
luciferase expression.
180
M
?4-2+4@/J?(EC)4-.)C/,1234A
D):,-EF
D):,-ED1GHI
()*+,-.)/01234/53*12)/6(+789(+7:1**;
L
'"#
'"!
!"&
!"%
#:C/04+:AB*+:,
!"$
!"#
!"!
(+78 (+7:1** (+78 (+7:1**
<='=
$!>!
'A,/04+:AB*+:,
?4-2+4@
'A,/04+:AB*+:,/D):,-ED1GHI
(+7E#E9E
K-*C,@B)
Supplementary Figure 4. Reduced antigen expression and immunogenicity of KPRderived tumors passaged through immune-competent mice.
(A) Antigen expression in primary Lenti-LucOS-induced tumors derived in KPR mice was lost
after transplantation into wildtype mice but not Rag-2-/- mice (as determined by luciferase activity
in vivo). Luciferase-negative tumor in wildtype recipient is outlined. Results are representative of
the transplantation of four independent Lenti-LucOS-induced sarcomas derived in KPR mice.
(B) Comparison of the relative tumor growth (mean final tumor volume in wildtype versus
Rag-2-/-) of Lenti-LucOS-induced sarcomas derived in KPR mice after passage through wildtype
(Rag+) or Rag-2-/- mice (Ragnull). Note that upon the 2nd transplantation, tumors passaged
through wildtype animals grew more similarly in wildtype and Rag-2-/- mice than tumors
passaged through Ragnull mice, indicating a reduction in immunogenicity coincident with a loss
in tumor antigen expression (A).
181
REFERENCES
Cheung, A.F., Dupage, M.J., Dong, H.K., Chen, J., and Jacks, T. (2008). Regulated expression
of a tumor-associated antigen reveals multiple levels of T-cell tolerance in a mouse model of
lung cancer. Cancer Res 68, 9459-9468.
Dubey, P., Hendrickson, R.C., Meredith, S.C., Siegel, C.T., Shabanowitz, J., Skipper, J.C.,
Engelhard, V.H., Hunt, D.F., and Schreiber, H. (1997). The immunodominant antigen of an
ultraviolet-induced regressor tumor is generated by a somatic point mutation in the DEAD box
helicase p68. J Exp Med 185, 695-705.
Dunn, G.P., Bruce, A.T., Ikeda, H., Old, L.J., and Schreiber, R.D. (2002). Cancer
immunoediting: from immunosurveillance to tumor escape. Nat Immunol 3, 991-998.
Dunn, G.P., Koebel, C.M., and Schreiber, R.D. (2006). Interferons, immunity and cancer
immunoediting. Nat Rev Immunol 6, 836-848.
Embleton, M.J., and Heidelberger, C. (1972). Antigenicity of clones of mouse prostate cells
transformed in vitro. Int J Cancer 9, 8-18.
Embleton, M.J., and Heidelberger, C. (1975). Neoantigens on chemically transformed cloned
C3H mouse embryo cells. Cancer Res 35, 2049-2055.
Engel, A.M., Svane, I.M., Rygaard, J., and Werdelin, O. (1997). MCA sarcomas induced in scid
mice are more immunogenic than MCA sarcomas induced in congenic, immunocompetent mice.
Scand J Immunol 45, 463-470.
Hewitt, H.B., Blake, E.R., and Walder, A.S. (1976). A critique of the evidence for active host
defence against cancer, based on personal studies of 27 murine tumours of spontaneous origin.
Br J Cancer 33, 241-259.
Jackson, E.L., Willis, N., Mercer, K., Bronson, R.T., Crowley, D., Montoya, R., Jacks, T., and
Tuveson, D.A. (2001). Analysis of lung tumor initiation and progression using conditional
expression of oncogenic K-ras. Genes Dev 15, 3243-3248.
Jonkers, J., Meuwissen, R., van der Gulden, H., Peterse, H., van der Valk, M., and Berns, A.
(2001). Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model
for breast cancer. Nat Genet 29, 418-425.
Kaplan, D.H., Shankaran, V., Dighe, A.S., Stockert, E., Aguet, M., Old, L.J., and Schreiber, R.D.
(1998). Demonstration of an interferon gamma-dependent tumor surveillance system in
immunocompetent mice. Proc Natl Acad Sci U S A 95, 7556-7561.
Khong, H.T., and Restifo, N.P. (2002). Natural selection of tumor variants in the generation of
"tumor escape" phenotypes. Nat Immunol 3, 999-1005.
Kirsch, D.G., Dinulescu, D.M., Miller, J.B., Grimm, J., Santiago, P.M., Young, N.P., Nielsen,
G.P., Quade, B.J., Chaber, C.J., Schultz, C.P., et al. (2007). A spatially and temporally
restricted mouse model of soft tissue sarcoma. Nat Med 13, 992-997.
182
Koebel, C.M., Vermi, W., Swann, J.B., Zerafa, N., Rodig, S.J., Old, L.J., Smyth, M.J., and
Schreiber, R.D. (2007). Adaptive immunity maintains occult cancer in an equilibrium state.
Nature 450, 903-907.
Monach, P.A., Meredith, S.C., Siegel, C.T., and Schreiber, H. (1995). A unique tumor antigen
produced by a single amino acid substitution. Immunity 2, 45-59.
Mondal, S., Embleton, M.J., Marquardt, H., and Heidelberger, C. (1971). Production of variants
of decreased malignancy and antigenicity from clones transformed in vitro by
methylcholanthrene. Int J Cancer 8, 410-420.
Prehn, R.T. (1968). Tumor-specific antigens of putatively nonviral tumors. Cancer Res 28, 13261330.
Prehn, R.T., and Main, J.M. (1957). Immunity to methylcholanthrene-induced sarcomas. J Natl
Cancer Inst 18, 769-778.
Qin, Z., and Blankenstein, T. (2004). A cancer immunosurveillance controversy. Nat Immunol 5,
3-4; author reply 4-5.
Schreiber, H. (2003). Tumor Immunology. In Fundamental Immunology, W.E. Paul, ed.
(Philadelphia: Lippincott Williams & Wilkins), pp. 1557-1592.
Scott, O.C. (1991). Tumor transplantation and tumor immunity: a personal view. Cancer Res 51,
757-763.
Shankaran, V., Ikeda, H., Bruce, A.T., White, J.M., Swanson, P.E., Old, L.J., and Schreiber,
R.D. (2001). IFNgamma and lymphocytes prevent primary tumour development and shape
tumour immunogenicity. Nature 410, 1107-1111.
Smyth, M.J., Godfrey, D.I., and Trapani, J.A. (2001). A fresh look at tumor immunosurveillance
and immunotherapy. Nat Immunol 2, 293-299.
Stauss, H.J., Van Waes, C., Fink, M.A., Starr, B., and Schreiber, H. (1986). Identification of a
unique tumor antigen as rejection antigen by molecular cloning and gene transfer. J Exp Med
164, 1516-1530.
Stetson, D.B., Mohrs, M., Mallet-Designe, V., Teyton, L., and Locksley, R.M. (2002). Rapid
expansion and IL-4 expression by Leishmania-specific naive helper T cells in vivo. Immunity 17,
191-200.
Svane, I.M., Engel, A.M., Nielsen, M.B., Ljunggren, H.G., Rygaard, J., and Werdelin, O. (1996).
Chemically induced sarcomas from nude mice are more immunogenic than similar sarcomas
from congenic normal mice. Eur J Immunol 26, 1844-1850.
Swann, J.B., Vesely, M.D., Silva, A., Sharkey, J., Akira, S., Schreiber, R.D., and Smyth, M.J.
(2008). Demonstration of inflammation-induced cancer and cancer immunoediting during
primary tumorigenesis. Proc Natl Acad Sci U S A 105, 652-656.
183
Tiscornia, G., Singer, O., and Verma, I.M. (2006). Production and purification of lentiviral
vectors. Nat Protoc 1, 241-245.
Townsend, S.E., Goodnow, C.C., and Cornall, R.J. (2001). Single epitope multiple staining to
detect ultralow frequency B cells. J Immunol Methods 249, 137-146.
Urban, J.L., Holland, J.M., Kripke, M.L., and Schreiber, H. (1982). Immunoselection of tumor cell
variants by mice suppressed with ultraviolet radiation. J Exp Med 156, 1025-1041.
Uyttenhove, C., Maryanski, J., and Boon, T. (1983). Escape of mouse mastocytoma P815 after
nearly complete rejection is due to antigen-loss variants rather than immunosuppression. J Exp
Med 157, 1040-1052.
Zhou, G., Lu, Z., McCadden, J.D., Levitsky, H.I., and Marson, A.L. (2004). Reciprocal changes
in tumor antigenicity and antigen-specific T cell function during tumor progression. J Exp Med
200, 1581-1592.
Zitvogel, L., Tesniere, A., and Kroemer, G. (2006). Cancer despite immunosurveillance:
immunoselection and immunosubversion. Nat Rev Immunol 6, 715-727.
184
CHAPTER 4
Therapeutic vaccinations with influenza enhance anti-tumor T cell responses
against mouse lung cancer via antigen-dependent and independent mechanisms
Michel DuPage1 and Tyler Jacks1,2
1
Koch Institute for Integrative Cancer Research and Department of Biology, Massachusetts
Institute of Technology, Cambridge, MA 02139, USA.
2
Howard Hughes Medical Institute, Massachusetts Institute of Technology, Cambridge, MA
02139, USA.
The author performed all of the experiments in the laboratory of Tyler Jacks.
185
ABSTRACT
Immunotherapeutic strategies to boost anti-tumor T cell responses have been largely ineffective
in treating cancer patients. New experimental models of cancer in mice more accurately
recapitulate the human disease and provide new opportunities to test immunotherapeutic
strategies and understand these inconsistent clinical results. Several investigations have shown
that therapeutic vaccination against tumor-associated antigens can provoke anti-tumor T cell
responses, but that T cells rapidly lose activity within the context of the tumor microenvironment.
Here we show that vaccination against specific antigens expressed in autochthonous mouse
lung adenocarcinomas using engineered strains of influenza can stimulate anti-tumor T cell
responses that retain activity and can eradicate recently initiated tumors. In addition, therapeutic
vaccination can act to stimulate anti-tumor T cell responses in an antigen-independent fashion,
enhancing the expansion and the activity of T cells that recognize antigens whose expression is
confined to tumors. Understanding the mechanism by which the T cell response is enhanced
under the conditions of live viral infection could reveal new strategies to boost anti-tumor T cell
responses to treat human lung cancer.
186
INTRODUCTION
Vaccination strategies can generate strong memory responses to specific antigens that
can protect individuals from subsequent infection by pathogens that produce the antigen
(Dermime et al., 2004; Pardoll, 2002). Such strategies are effective because the vaccine
preparations are formulated to provide optimal conditions for immune cell activation, providing
signals to facilitate enhanced antigen presentation and co-stimulation to responding T cells. A
critical issue with cancer vaccination, however, is that it is rarely prophylactic (Lollini and Forni,
2002). Patients with cancer have instead been treated therapeutically with “vaccines” when the
cancers have already formed and frequently after the disease has already significantly
progressed (Goldman and DeFrancesco, 2009). Therefore, cancer vaccines are really a form of
active immunotherapy to boost immune responses against tumors rather than protect patients
from the development of cancer. As such, cancer vaccines more closely resemble
immunotherapies against persistent infections (Kim and Ahmed, 2010; Pardoll, 2002).
Nevertheless, cancer immunotherapy is an attractive strategy to enhance immunity to cancer by
stimulating new, or greater numbers of T cells specific for tumor antigens. In these techniques,
tumor antigens, in the form of DNA, peptides, or killed tumor cells, can be administered with
adjuvants, such as cytokines or Toll-like receptor agonists, or antigens can be engineered into
attenuated pathogens, such as bacteria or viruses, to enhance immune cell activation towards
antigens expressed in tumors (Bocchia et al., 2000). The use of adjuvants or pathogens in
vaccines may also facilitate the effector function of T cells by promoting an inflammatory
environment at the tumor site (Laheru et al., 2005). Some of these methods have demonstrated
efficacy in the clinic, but the results are highly variable (Goldman and DeFrancesco, 2009;
Morse et al., 2005; Rosenberg et al., 2004). With the development of new experimental models
of cancer in mice that more accurately recapitulate the human disease, we may gain new
insights into the reasons behind this inconsistent clinical data.
187
We previously explored the role of endogenous immune responses to autochthonous
lung adenocarcinomas that expressed potent T cell antigens (DuPage et al., Chapter 2). T cell
responses against these tumors could delay the malignant progression of the cancer, but
ultimately the anti-tumor T cells lost activity and the tumors escaped this immune response.
Prophylactic vaccination with a dendritic cell line could provide protection from lung tumor
development and further delay malignant progression of the cancers. However, it is likely
impossible to recapitulate such results in human cancer patients because the tumor antigens to
target for prophylactic vaccination cannot be predicted (Bocchia et al., 2000; Lollini and Forni,
2002). In this study we wanted to understand how autochthonous lung tumors respond to
therapeutically generated T cell responses as these types of responses might mimic immunetumor interactions in cancer patients after immunotherapy.
Several previous investigations using endogenous mouse models of cancer have shown
that various forms of vaccination against tumor-associated antigens can either generate or
enhance anti-tumor T cell responses to pancreatic, mammary, and prostate cancers (Anderson
et al., 2007; Ercolini et al., 2005; Speiser et al., 1997; Verdeil et al., 2008). Here we describe a
vaccination approach against endogenous mouse lung cancers that utilizes a recombinant
influenza virus engineered to express the SIY peptide, WSN-SIY (Bai et al., 2008; Cheung et
al., 2008). Influenza is an ideal candidate vaccine against lung cancer because it naturally
infects the lung, generating a robust cytotoxic T cell response within the same environment that
tumors develop. We reveal that vaccination can stimulate anti-tumor T cell responses both
directly and indirectly by modifying the expansion and the activity of tumor-specific T cells.
Vaccination can be effective early during tumor progression, but as tumors progress,
vaccination is less effective because it does not induce as many anti-tumor T cells and tumor
growth may allow tumors to escape more readily via antigen loss. These data have clear
188
implications for understanding the pitfalls of vaccination therapies and provide important insights
into additional ways that vaccines may be harnessed to treat human lung cancer.
RESULTS
Infection with an influenza engineered to express tumor antigens at tumor initiation
protects mice from tumor development
We previously showed that prophylactic vaccination of mice with a dendritic cell line
(DC2.4-LucOS) engineered to express the same antigens expressed in lung adenocarcinomas
induced with Lenti-LucOS could generate memory T cells that provided substantial protection
from the development of these lung tumors (DuPage, et al., Chapter 2). Here we wanted to
examine the effect of therapeutic vaccination with strains of influenza engineered to express the
SIY antigen (WSN-SIY) that is also expressed in Lenti-LucOS tumors. To first determine the
effect of a strong effector T cell response directed against a single antigen (SIY) at tumor
initiation (as opposed to a memory T cell response), we infected mice with a control influenza
that does not express SIY (WSN) or the SIY-expressing influenza (WSN-SIY) by intratracheal
inhalation of 20 pfu of virus concomitantly with the inhalation of Lenti-LucOS to initiate lung
tumor formation. As shown in Figure 1A, simultaneous influenza infection had a significant
effect on tumor burden, but only when the influenza expressed the SIY antigen also present in
the initiated tumors. These data suggest that the expression of antigens in lung tumors
themselves do not sufficiently prime T cells to respond to and eliminate tumors. In addition,
simply creating an inflammatory environment with influenza infection during tumor initiation is
not sufficient to enhance the priming of tumor antigen-specific T cell responses that can control
tumor development.
The reduced tumor burden after WSN-SIY infection at tumor initiation correlated with an
increased frequency of SIY-specific CD8+ T cells in the lymph nodes draining the lungs (DLN)
189
α/.01&!&*),,L&MN&$O&@3:P
???
85
75
65
45
5
*+%,
-.' -.'/.01
8
7
6
4
5
-.'/ -.'/ -.' *+%,
.01 .01
D)G+H/I
C
59:
59=
-.'/ -.'/ -.' *+%,
.01 .01
D)G+H/I
D)G+H/D"*K.
6
45
4
45
45
45 5
45 4
45 6
45 7
6
45
4
45
689=
45 5
45 8
@3:
45 7
45 5
45 8
45 4
45 6
45 7
6
45
4
45 5
@3:
45 8
45 8
45
6
45
6
45
6
45
6
45
4
45
4
45
=494
45 5
45 6
45 7
.01AB(
45 7
4
45 8
45 5
45 4
4
45 5
45 6
45 7
45 5
45 8
@3:
45 8
59:
45
7
45
6
45
6
45
6
45
4
4
45
;;9>
45 5
45 4
45 6
45 7
45 8
45 5
F
<5
85
75
65
45
*+%,
-.'/.01
59:
45 8
45 7
45 8
45 7
45 8
779;
45 5
45 5
45 4
45 6
'939
'939
4
45 5
45 6
45 7
@3:
@3:
=5
45 4
α/.01&!&*),,L&MN&$O&@3:P
45 5
.01AB(
7
@346;
45
45 7
@3:
45 8
6694
7
45 5
45 6
@3:
45
45
45 4
45 6
==97
49=
7
45 4
45 4
@3:
45
45 5
4;95
45 5
45 8
7
@346;
.01AB(
4
45
45
45 8
.01AB(
6
45 7
45
:795
45 7
45 5
45 6
7:9>
@3:
5
45 4
45 8
49:
45 5
D)G+H/D"*K.
*+%,
45 8
@3:
45 8
D)G+H/D"*K.
J-.'/.01
8
45 7
45 5
45 5
D)G+H/D"*K.
896
.01AB(
45
;<98
45 7
@346;
.01AB(
D)G+H/I
J-.'/.01
45
@346;
695
'939
598
.EDFF'
45 8
45 7
!"#$%&'"#()%
495
3D'
45 8
3
496
@346;
!"#$%&'"#()%
@
?
<5
α/.0'&!&*),,L&MN&$O&@3:P
2
45 8
45 5
45 4
45 6
45 7
45 8
@3:
?
59=
598
596
595
-.'/.01 *+%,
190
Figure1. Infection with an influenza engineered to express tumor antigens at tumor
initiation protects mice from tumor development.
(A) The number of tumors on the surface of the lungs of mice bearing Lenti-LucOS tumors
analyzed 16 weeks after tumor initiation that received no influenza (ctrl), WSN, or WSN-SIY
simultaneously with tumor initiation. Each data point represents a mouse. *p<0.05, ***p<0.001.
(B) Representative FACS analysis for SIY-specific CD8+ T cells in the lung tumor draining lymph
nodes (DLN) and spleens of Lenti-x or Lenti-LucOS tumor-bearing mice from (A) that received
WSN-SIY at the time of tumor initiation (ctrl did not receive influenza). FACS plots are gated on
PI-negative cells and the percent of SIY/Kb+ cells of total CD8+ cells is indicated. Adjacent plots
to right are gated on the SIY/Kb+CD8+ cells to show CD127 expression on SIY-specific cells.
The percent of cells that are CD127+CD8+ and CD127-CD8+ is indicated. N.D. (None Detected)
indicates that the percent of SIY/Kb+ cells is below the reliable limit of detection with the DimerX
reagent.
(C) Graphs of the percent of SIY-specific (left) or SIN-specific (right) CD8+ T cells detected in
the DLN from mice in (A). n = 3-6 mice per group, mean ± SEM.
(D) The number of tumors on the surface of the lungs of mice bearing Lenti-LucOS tumors
analyzed 16 weeks after tumor initiation that received no influenza (ctrl) or WSN-SIY eight
weeks after tumor initiation.
(E) Graph of the percent of SIY-specific CD8+ T cells detected in the DLN from mice in (D). n =
4-7 mice per group, mean ± SEM. *p<0.05.
191
(Figure 1B,C). In contrast, the SIN-specific T cell response was unaffected or potentially even
reduced by WSN-SIY vaccination, and SIY-specific T cells were only detected in the spleens of
mice that had received WSN-SIY (Figure 1B,C). We examined the expression of CD127 (IL7Rα) on the surface of SIY-specific T cells to determine the differentiated state of the T cells
(Kaech et al., 2003). We found that while Lenti-x tumor-bearing mice infected with WSN-SIY
had high expression of CD127 on the majority of SIY-specific cells, indicative of a resting central
memory phenotype, SIY-specific T cells in Lenti-LucOS tumor-bearing mice were largely CD127
negative, indicative of effector T cells (Figure 1B). Lenti-LucOS tumor-bearing mice that
received WSN-SIY had SIY-specific T cells with an intermediate CD127 phenotype, indicating
that while memory cells may have been generated, an ongoing effector response was still
engaged due to the expression of the SIY antigen in the lung tumors (compare the fraction of
CD127 positive cells in the DLN versus the spleen) (Figure 1B). Regardless of WSN-SIY
infection, between 70-80% of SIN-specific T cells in Lenti-LucOS tumor-bearing mice had low
expression of CD127, similar to SIY-specific T cells in the unvaccinated Lenti-LucOS tumorbearing mice (not shown). Therefore, effective vaccination at tumor initiation requires viruses
that express shared antigens also expressed in tumors to induce the differentiation of antigenspecific memory T cells against the virus that can cross-react with tumors.
Therapeutic vaccination against established tumors does not reduce the lung tumor
burden
We have established that early infection at tumor initiation can protect mice from tumor
development, next we wanted to determine whether established tumors could be eradicated by
WSN-SIY vaccination. In this regimen we allowed Lenti-LucOS tumors to develop for eight
weeks before challenging mice with WSN-SIY. Unlike, the results with WSN-SIY infection at
tumor initiation, vaccination at eight weeks after tumor initiation did not reduce the tumor
192
burden, despite significantly increasing the fraction of SIY-specific T cells in the DLN and spleen
(Figure 1D,E, and not shown). Therefore, tumor elimination with WSN-SIY vaccination is only
possible when challenging mice very early during tumor development.
A narrow window of time exists after tumor initiation for effective therapeutic vaccination
with influenza
Next we wanted to determine if therapeutic vaccination with WSN-SIY could be effective
in reducing the tumor burden at any point after tumors had been initiated. We vaccinated mice
with WSN-SIY simultaneously, two weeks, or six weeks after tumor initiation and assessed the
tumor burden as well as the magnitude of the SIY-specific T cell response 16 weeks after the
tumors were induced (Figure 2A). Vaccination simultaneously or at two weeks post tumor
initiation reduced the number of tumors in Lenti-LucOS tumor-bearing mice, but vaccination at
six weeks after tumor initiation did not reduce the tumor burden (Figure 2B). WSN-SIY
vaccination at any time during the course of Lenti-x tumor progression did not affect the tumor
burden, indicating that the effectiveness of influenza vaccination depended upon shared antigen
expression in the viruses and tumors (Figure 2C). We also analyzed luciferase activity in
explanted tumors to assess the maintenance of LucOS antigen expression. We found that LentiLucOS tumors in WSN-SIY vaccinated mice selectively lost antigen expression, but in a manner
dependent on the time of vaccination (Figure 2D). Mice vaccinated simultaneously or at two
weeks after tumor initiation lost antigen expression in 80-90% of tumors, whereas only 30% of
tumors lost antigen expression in mice vaccinated six weeks after tumor initiation. In addition,
the level of luciferase activity in tumors that did not completely lose antigen expression was
quantifiably reduced compared to unvaccinated mice (not shown). Therefore, although there
was no reduction in tumor number in mice vaccinated six weeks after tumor initiation, there was
a reduction in the number of tumors that maintained high antigen expression.
193
!
6'78(98:;'
:/.,.0,.%/ 1'2**34'
5'2**34'
N5'2**34'
"#$%&'./.,.0,.%/'
=*/,.9=#+A8'
=*/,.9@'
!/0-L4.4''
!'"#$%&')#&U*/'
!'!/,.V*/'*@>&*44.%/'
!'"'+*--'G#/+,.%/'
!"#$%&%'()*+,-./012312-
5M
SM
1M
T
α98:;'"'+*--4'EF'%G'CDHI
M
O
5M
SM
1M
+,&- ./.,.0,.%/ 1234 5234
6'78(98:;
PO
OM
1O
M
J
WWW
WWW W
=*/,.9@ +,&- ./.,.0,? 1234 5234
6'78(98:;
=*/,.9=#+A8
NPO
NOM
N1O
X
NMM
1
M
HM
M
+,&- ./.,.0,.%/ 1234 5234
6'78(98:;
S
N
NMM
NMM
HM
<*&+*/,'=#+'>%4',#$%&4
"#$%&'(#$)*&
NMM
D
=*/,.9@
N1M
"#$%&'(#$)*&
N1M
C
=*/,.9=#+A8
W
WW
"#$%&'(#$)*&
B
WWW
(?D?
=*/,.9@ ./.,.0,? 1'234 5'234 +,&-
=*/,.9=#+A8
6'78(98:;
Q1'R'M?S5O
PO
OM
1O
M
M?N
N
J&*K#*/+L'%G'α98:;'"'+*--4
NM
Figure 2. Therapeutic vaccination early after tumor initiation can reduce the lung tumor
burden.
(A) Scheme of therapeutic vaccination regimen.
(B) The number of tumors counted in ten histological sections of lungs (two sections per lobe) of
mice bearing Lenti-LucOS tumors analyzed 16 weeks after tumor initiation that received no
influenza (ctrl) or WSN-SIY at tumor initiation or two weeks or six weeks after tumor initiation.
Each data point represents the tumor number from one mouse. *p<0.05, **p<0.01.
(C) Same scheme as in (B) but with Lenti-x tumors. Tumor numbers were not significantly
different between any of the groups.
194
(D) Freshly explanted tumors were assayed for luciferase activity and the percent of tumors that
were luciferase positive (> 50 RLU/µg protein) in unvaccinated and WSN-SIY vaccinated mice is
plotted. n = 2-7 mice, 8-28 tumors, per group.
(E) Graph of the percent of SIY-specific CD8+ T cells detected in the spleen from mice in (B)
and (C). Lenti-x + WSN-SIY represents combined data from WSN-SIY infection at different time
points because the percent of SIY-specific cells was unchanged. n = 4-13 mice per group, mean
± SEM. *p<0.05, ***p<0.001.
(F) Comparison of tumor number versus the percent of SIY-specific T cells in the spleens (log
scale) from Lenti-LucOS tumor-bearing mice vaccinated with WSN-SIY.
To understand why vaccination as early as six weeks after tumor initiation did not lead to
significant tumor eradication, we assessed the capacity of the differentially timed vaccinations to
generate memory T cells. The fraction of SIY-specific T cells found in the spleens of mice
vaccinated six weeks after tumor initiation was significantly reduced compared to mice
vaccinated at the earlier time points (Figure 2E). Indeed, we could observe an inverse
correlation between the fraction of SIY-specific T cells found in the spleens of WSN-SIY
vaccinated mice and their lung tumor burden (Figure 2F). Thus, the capacity to generate
memory cells in mice bearing six week-old established Lenti-LucOS tumors is reduced and may
account for the lack of a reduced tumor burden in these mice.
Epitope spreading to the tumor-expressed SIN antigen with influenza vaccination
During the various therapeutic vaccinations with WSN-SIY, in addition to following the
SIY-specific response, we continued to monitor the response of T cells specific for SIN.
Interestingly, we noticed that unlike infection at initiation, in which the SIN-specific T cell
response seemed unaltered, the SIN-specific T cell response appeared to be enhanced with
WSN-SIY infection at time points well after tumors were initiated. Eight days after WSN-SIY
infection of mice bearing established Lenti-LucOS tumors, most SIY-specific T cells in the DLNs
and lungs had the capacity to produce IFN-γ and TNF-α (Figure 3A-C). This activity was much
195
)*+,-.)6782
α.243=>?
α.245=>?
!" %
!" %
#&$!
!" $
!"
#
!"
#
!"
!
!"
!
!" "
"&NN
;*;,-<*=>?.B@C
!" $
!" "
!" "
!" !
!" #
!" $
!" %
!" "
!" !
!" #
!" $
!" $
!"
#
!"
!
!" "
!" %
!" "
$
!"
!"
0243
!"
$
#
!"
!
!"
!&'!
!" "
!" %
P3Fα
!" %
"&'#
!"
$
!"
#
!"
!
!" "
!" !
!" #
!" $
0245
!"
$
#
!"
#
!
!"
!
!&'N
!" "
!" %
!" "
!" !
!" #
!" $
!" %
!" %
!"
$
!"
#
!"
"&"%
!" "
!" "
!" !
!" #
!" $
!" %
"&ON
!" "
!" !
!" #
!" $
!" %
"&"!
+G
;*;&
!
"&'#
!" "
!" !
!" #
!" $
!" %
4F3γ
D
A
D)3
α.245=>
α.243=>
?
?
!" %
!" %
"&$'
!" $
)*+,-./
0123.245
"&""
!" $
!"
#
!"
#
!"
!
!"
!
!" "
!" "
!" "
!" !
!" #
!" $
!" %
!" %
!" "
!" !
!" #
M"
O"
D)3
243
245
)*+,-.)6782
245
1P
)U3V
'"
%"
$"
#"
!"
"
243
245
)*+,-.)6782
@E-9
FG:<H-+79*EI*H-+Hα.243=>?
J-,KH123.245H-+L*7,-G+
!
#
$
%
'
#&!$
#&$(
#&%'
"&('
"&##
245
1P
"&$(
!"
$
!"
$
!"
#
!"
#
!"
!
!"
!
!" "
!" "
!" "
!" !
!" #
!" $
!" %
!" %
;*;,-<*=>?.B@C
!" %
("
N"
M"
O"
'"
%"
$"
#"
!"
"
!" %
"&'
)*+,-.)6782
0123.245
!" $
T
4F3γ 0HP3Fα0
4F3γ 0HP3FαH+*Q
C
!" "
!" %
4F3γ
)*+,-.)6782
7,9:
!" $
"&O"
0245
!" "
"&""
+G
;*;&
!" #
!" %
"&M%
P3Fα
!"
!" !
;*;,-<*=>?.@A
;*;,-<*=> .@A
?
!" %
1P
α.245=>?
!" %
#&#$
;*;,-<*=>?.B@C
123.245
@*97*+,H7R,GS-+*H;9G<67-+QH7*::I
123.245
@*97*+,H7R,GS-+*H;9G<67-+QH7*::I
B
!" "
!" !
!" #
!" $
!" %
!" %
"&#
!" $
"&!$
!" $
!"
#
!"
#
!"
!
!"
!
!" "
!" "
!" "
!" !
!" #
!" $
!" %
!" "
!" !
!" #
!" $
!" %
;*;,-<*=> .@A
?
196
Figure 3. Influenza vaccination with WSN-SIY can enhance T cell responses to tumor
antigens not expressed in the virus.
(A) Top: FACS analysis using SIY- or SIN-loaded DimerX reagents to detect tumor antigen
specific CD8+ T cells in the lung-draining lymph nodes (DLN) of Lenti-LucOS tumor-bearing or
wild-type mice infected with WSN-SIY. FACS plots are gated on PI-negative, CD8+ cells.
Bottom: IFN-γ and TNF-α cytokine production in CD8+ T cells from the DLN of the same mice
shown above after in vitro stimulation with SIY, SIN, or no peptides. Tumor-bearing mice were
infected with WSN-SIY 15 weeks after tumor initiation. FACS plots are representative of two
mice per group.
(B and C) Graphs of the percent of SIN-specific or SIY-specific T cells from the mice depicted in
(A) capable of producing both IFN-γ and TNF-α (black bars) or IFN-γ alone (white bars) in the
DLN (B) or the lung (C).
(D) FACS analysis using SIY- (left) or SIN-loaded (right) DimerX reagents to detect tumor
antigen specific CD8+ T cells in the DLN of Lenti-LucOS or Lenti-x tumor-bearing mice left
untreated or infected with WSN-SIY. FACS plots are gated on PI-negative, CD8+ cells. The
arrow indicates the concomitant increase in SIN-specific T cells in the DLN after WSN-SIY
infection of Lenti-LucOS tumor-bearing mice.
(E) To determine whether the population of SIN-specific T cells was increased in mice receiving
WSN-SIY, we divided the percent of SIN-specific T cells in vaccinated mice by the percent of
SIN-specific T cells in unvaccinated littermate pairs. The fold increase in SIN-specific T cells
with WSN-SIY infection is shown for five different littermate pairs.
197
greater than that of endogenous T cells responding to SIY when produced from Lenti-LucOS
tumors in the absence of WSN-SIY infection (DuPage, et al., Chapter 2). This indicates that the
failure of therapeutic vaccination beyond six weeks after tumor initiation is not due to SIYspecific T cells lacking effector function. The surprising finding, however, was that SIN-specific T
cells also had enhanced activity similar to that of the SIY-specific T cells (Figure 3A-C). This
was unexpected because only SIY-specific T cells could be directly primed by the engineered
influenza, and therefore, the increased activity of SIN-specific T cells must have been through
some different and indirect means. We hypothesized that the enhanced activity of SIY-specific T
cells primed by WSN-SIY would increase anti-tumor cytotoxicity, and in the context of a viral
infection in the lung, this could lead to increased presentation of the SIN antigen in the DLN by
virally matured antigen-presenting cells. This phenomenon by which a productive response to
one antigen leads to a stronger response to a second antigen, is known as “epitope spreading”
(Lally et al., 2001; Markiewicz et al., 2001; Vanderlugt and Miller, 2002). By comparing
littermates that either did or did not receive WSN-SIY, we found that in four of five pairs
examined, mice that received WSN-SIY had equal or greater numbers of SIN-specific T cells in
their DLN (Figure 3D,E). Thus, the expansion of SIN-specific T cells with increased functional
activity may be due to epitope spreading after a cytotoxic anti-tumor response by SIY-specific T
cell primed directly by WSN-SIY.
Influenza infection without shared tumor antigen expression can boost anti-tumor T cell
responses against established tumors
To further investigate whether epitope spreading was boosting the SIN-specific T cell
response after WSN-SIY vaccination, we examined the response of naïve OT-I T cells that were
transferred a day after infecting mice harboring established Lenti-LucOS tumors with influenza
viruses that either did (WSN-SIY) or did not (WSN) express common antigens present in the
198
lung tumors (Figure 4A). Indeed, OT-I T cells had increased activity when transferred into mice
receiving WSN-SIY compared to mice that did not receive virus (Figure 4B,C). However, we
were surprised to find that mice infected with the WSN influenza, that does not express any
common tumor antigens, also resulted in the increased activity of the transferred OT-I T cells
(Figure 4B,C). In addition, mice receiving the WSN infection had an enhanced recruitment of
OT-I T cells into the lungs compared to mice that didn’t receive any virus (Figure 4D). Most
importantly, however, mice receiving naïve OT-I T cells after WSN infection had reduced
luciferase activity in their lungs that was nearly equivalent to the luciferase activity of WSN-SIY
infected mice (Figure 4E). This indicates that the combination of OT-I transfer and WSN
infection had the functional consequence of either inducing the death of antigen-expressing
tumor cells or selecting for tumor cells with reduced expression of the antigens. To look more
specifically at the tumors themselves, we repeated the therapeutic regimen shown in Figure 4A,
but waited several months to allow tumors to grow large enough for individual tumor analysis.
Whereas the transfer of naïve OT-I T cells or WSN infection alone did not result in the outgrowth
of tumors that had lost antigen expression, the combination of OT-I T cell transfer and WSN
infection led to a significant number of tumors that lost antigen expression, though not to the
extent obtained with WSN-SIY infection (Figure 4F). Further investigation will be needed to
determine how WSN infection enhances the response of naïve OT-I T cells. Understanding the
mechanism by which the T cell response is enhanced under these conditions will be important
for therapeutic strategies.
DISCUSSION
Several important insights can be gleaned from these experiments that may help to
guide better strategies for the treatment of patients with cancer using T cell-based
immunotherapies. Although it is clear that early vaccination of the tumor site with influenza
199
-
+8*9:+8*9&*;<:+=5(>+
64.#17"4+)6&;+?@A+B.0C+
!"#$%&!'()*+
!"#$%&,+
-#./01%1++
!+-#$%2"#+",34"11%5#+
!+6+("//+7'#($%5#+
!"#$ %&%'()*+,-./012013
;W9γ @+69Wα@
;W9γ @+69Wα+#"2
HF
DF
AF
($4/
)6&;
L
)6&;+3"4("#$+57+/0=3R5(0$"1
DQF
8*9&
8*9&
*;<
*;<
!"#$%&!'()* DJ !"#$%&,
8*9
)6&;
)6&;@8*9
L!9
FQP
FQF
DF
AP
AF
DF
AF
F
!M9N
8*9&
8*9&
*;<
*;<
!"#$%&!'()* DJ !"#$%&,
($4/
8*9
)6&;
W
DFF
AQF
;W9γ @+69Wα@
;W9γ @+69Wα+#"2
HF
V
AQP
DP
!M9N
K"4("#$+!'(+351+$'=541
F
K"4("#$+(0$5>%#"+345B'(%#2+("//1
IF
AD+B.01+
J
L!9
X!MY+µ2+345$"%#
K"4("#$+(0$5>%#"+345B'(%#2+("//1
O
E&AF+G"">1+
APF
AFF
PF
F
8*9& 8*9& 8*9 ($4/
*;< *;<
DJ
)6&;
#TDF
AFF
#TAU
SP
PF
#TDH #TDD
#TAF #TAF
DP
F
)6&;Z
W/'Z
&
@
@
&
8*9& 8*9&
8*9 8*9
*;< *;<
@
&
&
&
P
F
Figure 4. Influenza vaccination boosts anti-tumor T cell responses against established
tumors even without expressing shared tumor antigens.
(A) Scheme of therapeutic vaccination and adoptive T cell transfer regimen.
(B and C) Graphs of the percent of transferred OT-I T cells (identified as CD8+CD45.1+) capable
of producing both IFN-γ and TNF-α (black bars) or IFN-γ alone (white bars) in the DLN (B) or the
lung (C) after transfer into Lenti-LucOS tumor-bearing mice that had not received an influenza
200
vaccine (ctrl) or had received WSN or WSN-SIY one day prior to the adoptive transfer of naïve T
cells. As a control, 2C T cells were transferred into Lenti-x tumor-bearing mice following
infection with WSN-SIY. n = 3 mice per group, mean ± SEM.
(D) The size of the OT-I population in the DLN (top) and the lung (bottom) in the absence (white
bars) or presence (grey bars) of WSN infection prior to the adoptive transfer of naïve T cells.
The size of the OT-I population is represented as the fraction of total lymphocytes (based on
FSC/SSC characteristics) that were CD8+CD45.1+. n = 3 mice per group, mean ± SEM.
(E) Luciferase activity (RLU/µg protein) from whole lung lysates of mice 12 days after receiving
naïve 2C or OT-I T cells after prior infection with WSN or WSN/SIY or without prior influenza
infection (ctrl) as depicted in (A). Lung tissues samples came from mice shown in B-D. n = 3
mice per group, mean ± SEM.
(F) Freshly explanted tumors were assayed for luciferase activity and the percent of tumors that
were luciferase positive (> 50 RLU/µg protein) of the total in each group is plotted. The total
number of tumors (n) analyzed from at least two mice per group is indicated.
viruses that express antigens shared with the tumors can be effective in reducing the
subsequent tumor burden, it is unlikely that this can have any direct application in a clinical
setting because of the need to identify the tumor antigens and begin treatment so early.
However, despite the fact that viruses that did not express shared tumor antigens were not
efficacious when given at tumor initiation, the treatment of established tumors with such viruses
did markedly improve the recruitment and function of tumor-specific T cells. This indicates that
direct priming against specific tumor antigens may not be necessary for effective
immunotherapy.
Therapeutic vaccination two weeks after tumor initiation with WSN-SIY could also
significantly reduce the overall tumor burden in mice. However, WSN-SIY infection as early as
six weeks after tumor initiation did not reduce the lung tumor burden. Previous studies have
indicated that advanced tumors can create an immunosuppressive microenvironment that
directly suppresses anti-tumor T cell responses (Rabinovich et al., 2007; Zou, 2005). While
vaccination in several mouse cancer models could produce large numbers of T cells specific to
tumor antigens, the cells rapidly became unresponsive to the antigens after entering the tumor
microenvironment (Anderson et al., 2007; Bai et al., 2008; Cheung et al., 2008; Willimsky et al.,
201
2008). This did not appear to be the case in this system, however, because SIY-specific (and
even SIN-specific) T cells retained activity in the lungs following vaccination. In addition,
reduced levels of antigen expression in the tumors of vaccinated mice indicated that a
productive T cell response was generated that could drive the outgrowth of tumor cells with
reduced antigen expression. Similar results have been documented after the treatment of
patients with melanoma in which tumors had reduced or lost expression of antigens targeted by
the vaccines (Khong and Restifo, 2002).
A key difference in mice vaccinated with WSN-SIY at later time points after tumor
initiation was the size of the responding SIY-specific T cell population detected in the spleen.
Mice vaccinated six weeks after tumor initiation had significantly fewer SIY-specific T cells
compared to mice vaccinated earlier with WSN-SIY. Not surprisingly, smaller populations of
SIY-specific T cells in the spleen correlated with increased numbers of tumors in the lungs. The
persistence of greater numbers of tumor-reactive T cells in patients receiving adoptively
transferred tumor-specific T cells has also been correlated with improved clinical responses
(Morgan et al., 2006). Therefore, quantitatively weaker T cell responses with vaccination at the
later time points during tumor progression may underlie their reduced efficacy. It is not clear
what is responsible for the reduced T cell response in the context of established tumors, but we
speculate that it could paradoxically be the result of the endogenous SIY-specific T cells
responding to SIY expressed from the tumors that helps to more rapidly eradicate virus-infected
cells, consequently reducing the effectiveness of the vaccine. In support of this, we find that the
number of bulk CD8+ T cells recruited to the lungs of mice infected with WSN-SIY compared to
WSN is reduced in Lenti-LucOS tumor-bearing mice (not shown). In addition to the generation
of a smaller pool of tumor-reactive T cells, the greater tumor burden encountered at progressive
states of the disease may present too large of a burden for cytotoxic T cells to completely
eradicate entire tumor masses.
202
Interestingly, we found that infection with WSN-SIY or even WSN, can enhance the
expansion and activity of T cells responding to the SIN tumor antigen. We suspect that influenza
infection may improve the priming of T cells against tumor antigens by inducing tissue damage
in the lungs that increases tumor antigen release and enhances the activity of the antigenpresenting cells that stimulate naïve T cells in the tumor draining lymph nodes (Vanderlugt and
Miller, 2002). However, influenza infection of the lung may also modulate the anti-tumor T cell
response by changing the tumor microenvironment in the lung such that T cells are recruited
more efficiently and retain activity upon encounter with tumors expressing cognate antigen
(Laheru et al., 2005). Understanding the mechanism by which the T cell response is enhanced
under these conditions could reveal new strategies to boost anti-tumor T cell responses. We
have begun to further examine how influenza infection may indirectly boost anti-tumor T cell
responses by trying to recapitulate these results using various TLR ligands in place of influenza
vaccination. Other investigations have shown that the administration of TLR ligands, such as
poly(I:C), can increase the accumulation and function of anti-tumor T cells in tumors (Verdeil et
al., 2008). Preliminary experiments indicate that the intratracheal introduction of poly(I:C), but
not LPS, can significantly improve the recruitment of CD8+ T cells to the lungs of tumor-bearing
mice (not shown). Vaccination with influenza in different anatomical locations in tumor-bearing
mice could potentially separate the role of priming versus augmentation of the tumor
microenvironment in boosting the anti-tumor T cell response. Although the results presented
here are sobering because we define only a very short window of time for effective immune
therapy, we have also provided hope that continued investigation of autochthonous mouse
models of cancer may provide for the discovery of new and unanticipated approaches to
effectively treat cancer by harnessing the immune system.
203
MATERIALS AND METHODS
Mice and tumor induction
129S4/SvJae strains backcrossed 8 generations were used for the experiments in Figures 1-3.
C57/BL6 mice backcrossed 11 generations were used for experiments in Figure 4 in which
transgenic T cells from C57/BL6 mice were adoptively transferred. Trp53fl mice were provided
by A. Berns (Jonkers et al., 2001), K-rasLSL-G12D were generated in our laboratory (Jackson et al.,
2001), and Rag-2-/- mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Lung
tumors were induced in K-rasLSL-G12D/+;p53fl/fl mice by intratracheal intubation and inhalation of
viruses expressing Cre recombinase as reported previously (DuPage et al., 2009; Jackson et
al., 2005). J. Chen provided the 2C, OT-I, and OT-II TCR transgenic mice and transgenic T cells
for adoptive transfers came from 2C;Rag-2-/- or OT-I;Rag-2-/- mice on a pure C57/BL6
background. All animal studies and procedures were approved by the Massachusetts Institute of
Technology’s Committee for Animal Care.
Lentiviral production
Lentivirus production by transfection of 293T cells with Δ8.2 (gag/pol), CMV-VSV-G, and various
transfer vectors expressing Cre was performed as previously described (Tiscornia et al., 2006).
Influenza
The WSN-SIY and WSN influenza strains were provided by J. Chen. Mice were infected with 20
pfu of virus/ mouse by intratracheal intubation and inhalation.
Flow cytometry
Cell suspensions from lymphoid organs were prepared by mechanical disruption between
frosted slides. Cell suspensions from lungs were generated by mincing and digesting the tissues
204
for ~1 hour at 37°C in 125 U/ml Collagenase Type I (Gibco) and 60 U/ml Hyaluronidase
(Sigma). Single cell suspensions were stained with the specified antibodies for 20-30 min after
treatment with Fc Block (BD Pharmingen). α-CD8α (53-6.7), α-CD45.1 (A20), α-mouse IgG1
(X56), α-IFNγ (XMG1.2), α-TNFα (MP6-XT22), and DimerX I (Dimeric Mouse H-2Kb:Ig) were
from BD Pharmingen. α-CD127 (A7R34) was from eBioscience. All antibodies were used at
1:200. Peptide-loaded DimerX reagents were prepared as specified by manufacturer and used
at 1:75. Staining cells with the same ligand labeled with PE or APC for improved sensitivity has
been previously reported (Stetson et al., 2002; Townsend et al., 2001). Propidium iodide was
used to exclude dead cells when cell were not fixed. Cells were read on a FACSCalibur and
analyzed using Flowjo software (Tree Star).
Cytokine production
Single cell suspensions from lungs or lymphoid organs were resuspended in the presence or
absence of SIYRYYGL and SIINFEKL peptides in OPTI-MEM I (Gibco) supplemented with
GolgiPlug (BD Pharmingen) for ~4 hours at 37°C, 5% CO2. Cells were then fixed and stained for
intracellular cytokines using the Cytofix/Cytoperm kit (BD Biosciences).
Luciferase detection
Freshly explanted tumors or whole lungs were lysed in Cell Culture Lysis Reagent, mixed with
Luciferase Assay Reagent according to manufacturer’s instructions (Promega), and relative light
units (RLU) were detected using the Optocomp I luminometer (MGM Instruments). RLUs were
standardized based on the total number of cells or total protein (Bio-Rad Protein Assay) used in
each assay.
205
Adoptive T cell transfer
Naïve TCR transgenic T cells from OT-I or 2C mice were harvested from lymphoid organs, then
counted and enumerated by FACS analysis before 2x105 CD8+ cells were transferred
intravenously.
Statistical analyses
P-values for statistical comparisons were generated using unpaired two-tailed student’s T-tests.
ACKNOWLEDGEMENTS
We would like to thank G. Paradis for assistance with flow cytometry, H. Eisen for reagents and
experimental discussions, and A.G. DuPage for critical reading of this manuscript. This work
was supported by Grant 1 U54 CA126515-01 from the NIH and partially by Cancer Center
Support (core) grant P30-CA14051 from the National Cancer Institute and the Margaret A.
Cunningham Immune Mechanisms in Cancer Research Fellowship Award (M.D.) from the John
D. Proctor Foundation. T.J. is a Howard Hughes Investigator and a Daniel K. Ludwig Scholar.
206
REFERENCES
Anderson, M.J., Shafer-Weaver, K., Greenberg, N.M., and Hurwitz, A.A. (2007). Tolerization of
tumor-specific T cells despite efficient initial priming in a primary murine model of prostate
cancer. J Immunol 178, 1268-1276.
Bai, A., Higham, E., Eisen, H.N., Wittrup, K.D., and Chen, J. (2008). Rapid tolerization of virusactivated tumor-specific CD8+ T cells in prostate tumors of TRAMP mice. Proc Natl Acad Sci U
S A 105, 13003-13008.
Bocchia, M., Bronte, V., Colombo, M.P., De Vincentiis, A., Di Nicola, M., Forni, G., Lanata, L.,
Lemoli, R.M., Massaia, M., Rondelli, D., et al. (2000). Antitumor vaccination: where we stand.
Haematologica 85, 1172-1206.
Cheung, A.F., Dupage, M.J., Dong, H.K., Chen, J., and Jacks, T. (2008). Regulated expression
of a tumor-associated antigen reveals multiple levels of T-cell tolerance in a mouse model of
lung cancer. Cancer Res 68, 9459-9468.
Dermime, S., Gilham, D.E., Shaw, D.M., Davidson, E.J., Meziane el, K., Armstrong, A.,
Hawkins, R.E., and Stern, P.L. (2004). Vaccine and antibody-directed T cell tumour
immunotherapy. Biochim Biophys Acta 1704, 11-35.
DuPage, M., Dooley, A.L., and Jacks, T. (2009). Conditional mouse lung cancer models using
adenoviral or lentiviral delivery of Cre recombinase. Nat Protoc 4, 1064-1072.
Ercolini, A.M., Ladle, B.H., Manning, E.A., Pfannenstiel, L.W., Armstrong, T.D., Machiels, J.P.,
Bieler, J.G., Emens, L.A., Reilly, R.T., and Jaffee, E.M. (2005). Recruitment of latent pools of
high-avidity CD8(+) T cells to the antitumor immune response. J Exp Med 201, 1591-1602.
Goldman, B., and DeFrancesco, L. (2009). The cancer vaccine roller coaster. Nat Biotechnol
27, 129-139.
Jackson, E.L., Olive, K.P., Tuveson, D.A., Bronson, R., Crowley, D., Brown, M., and Jacks, T.
(2005). The differential effects of mutant p53 alleles on advanced murine lung cancer. Cancer
Res 65, 10280-10288.
Jackson, E.L., Willis, N., Mercer, K., Bronson, R.T., Crowley, D., Montoya, R., Jacks, T., and
Tuveson, D.A. (2001). Analysis of lung tumor initiation and progression using conditional
expression of oncogenic K-ras. Genes Dev 15, 3243-3248.
Jonkers, J., Meuwissen, R., van der Gulden, H., Peterse, H., van der Valk, M., and Berns, A.
(2001). Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model
for breast cancer. Nat Genet 29, 418-425.
Kaech, S.M., Tan, J.T., Wherry, E.J., Konieczny, B.T., Surh, C.D., and Ahmed, R. (2003).
Selective expression of the interleukin 7 receptor identifies effector CD8 T cells that give rise to
long-lived memory cells. Nat Immunol 4, 1191-1198.
Khong, H.T., and Restifo, N.P. (2002). Natural selection of tumor variants in the generation of
"tumor escape" phenotypes. Nat Immunol 3, 999-1005.
207
Kim, P.S., and Ahmed, R. (2010). Features of responding T cells in cancer and chronic
infection. Curr Opin Immunol 22, 223-230.
Laheru, D.A., Pardoll, D.M., and Jaffee, E.M. (2005). Genes to vaccines for immunotherapy:
how the molecular biology revolution has influenced cancer immunology. Mol Cancer Ther 4,
1645-1652.
Lally, K.M., Mocellin, S., Ohnmacht, G.A., Nielsen, M.B., Bettinotti, M., Panelli, M.C., Monsurro,
V., and Marincola, F.M. (2001). Unmasking cryptic epitopes after loss of immunodominant tumor
antigen expression through epitope spreading. Int J Cancer 93, 841-847.
Lollini, P.L., and Forni, G. (2002). Antitumor vaccines: is it possible to prevent a tumor? Cancer
Immunol Immunother 51, 409-416.
Markiewicz, M.A., Fallarino, F., Ashikari, A., and Gajewski, T.F. (2001). Epitope spreading upon
P815 tumor rejection triggered by vaccination with the single class I MHC-restricted peptide
P1A. Int Immunol 13, 625-632.
Morgan, R.A., Dudley, M.E., Wunderlich, J.R., Hughes, M.S., Yang, J.C., Sherry, R.M., Royal,
R.E., Topalian, S.L., Kammula, U.S., Restifo, N.P., et al. (2006). Cancer regression in patients
after transfer of genetically engineered lymphocytes. Science 314, 126-129.
Morse, M.A., Chui, S., Hobeika, A., Lyerly, H.K., and Clay, T. (2005). Recent developments in
therapeutic cancer vaccines. Nat Clin Pract Oncol 2, 108-113.
Pardoll, D.M. (2002). Spinning molecular immunology into successful immunotherapy. Nat Rev
Immunol 2, 227-238.
Rabinovich, G.A., Gabrilovich, D., and Sotomayor, E.M. (2007). Immunosuppressive strategies
that are mediated by tumor cells. Annu Rev Immunol 25, 267-296.
Rosenberg, S.A., Yang, J.C., and Restifo, N.P. (2004). Cancer immunotherapy: moving beyond
current vaccines. Nat Med 10, 909-915.
Speiser, D.E., Miranda, R., Zakarian, A., Bachmann, M.F., McKall-Faienza, K., Odermatt, B.,
Hanahan, D., Zinkernagel, R.M., and Ohashi, P.S. (1997). Self antigens expressed by solid
tumors Do not efficiently stimulate naive or activated T cells: implications for immunotherapy. J
Exp Med 186, 645-653.
Stetson, D.B., Mohrs, M., Mallet-Designe, V., Teyton, L., and Locksley, R.M. (2002). Rapid
expansion and IL-4 expression by Leishmania-specific naive helper T cells in vivo. Immunity 17,
191-200.
Tiscornia, G., Singer, O., and Verma, I.M. (2006). Production and purification of lentiviral
vectors. Nat Protoc 1, 241-245.
Townsend, S.E., Goodnow, C.C., and Cornall, R.J. (2001). Single epitope multiple staining to
detect ultralow frequency B cells. J Immunol Methods 249, 137-146.
208
Vanderlugt, C.L., and Miller, S.D. (2002). Epitope spreading in immune-mediated diseases:
implications for immunotherapy. Nat Rev Immunol 2, 85-95.
Verdeil, G., Marquardt, K., Surh, C.D., and Sherman, L.A. (2008). Adjuvants targeting innate
and adaptive immunity synergize to enhance tumor immunotherapy. Proc Natl Acad Sci U S A
105, 16683-16688.
Willimsky, G., Czeh, M., Loddenkemper, C., Gellermann, J., Schmidt, K., Wust, P., Stein, H.,
and Blankenstein, T. (2008). Immunogenicity of premalignant lesions is the primary cause of
general cytotoxic T lymphocyte unresponsiveness. J Exp Med 205, 1687-1700.
Zou, W. (2005). Immunosuppressive networks in the tumour environment and their therapeutic
relevance. Nat Rev Cancer 5, 263-274.
209
CHAPTER 5
DISCUSSION AND FUTURE DIRECTIONS
210
In this thesis research, I used tractable and reproducible models of autochthonous
mouse cancers that accurately recapitulate the stepwise progression of human cancer, both
histopathologically and genetically, to study cancer immune surveillance. This led to several key
observations that were not possible with previous technologies. I show that TSA expression,
along with the origin of the disease and the type of model used to study immune responses to
tumors, can all influence the dynamics of immune-tumor interactions, driving T cells to become
tolerant or forcing tumors to escape. Thus, this research emphasizes the considerable
contextual diversity in immune responses to cancer that should be further explored. Lessons
learned from our attempts to improve immune responses against cancer utilizing therapeutic
vaccines have also revealed important insights into immune-tumor interactions. Finally, I will
end by proposing that the regulated timing of adaptive immunity may provide a unifying theme
to rationalize the range of immune responses and tumor phenotypes we and others have
observed while investigating the processes of cancer immune surveillance.
I. Tumor antigens in immunotherapy: TAA versus TSA
The presence of antigens expressed in tumors that are recognizable by T cells is
essential for the detection and the potential regulation of cancer. However, the type of antigens
targeted by an immune response may be antigens specific to cancer cells or antigens also
present in normal host tissues. Whether or not the type of antigen targeted by T cells alters the
effectiveness of the immune response against cancer has long been debated, and a clear
answer is still elusive.
Transgenic mouse models of cancer that express tumor-associated antigens (TAAs)
indicate that T cells responding against these types of tumor antigens routinely become tolerant
(Pardoll, 2003). Several groups have shown that immune tolerance toward tumors occurs due to
211
immune ignorance – T cells are not stimulated to respond to the TAAs expressed by tumors
(Ochsenbein et al., 1999; Speiser et al., 1997). Ignorance likely occurs in these scenarios as a
direct consequence of the actions of central and peripheral tolerance mechanisms that lead to
the deletion of T cells capable of recognizing these types of antigens (Cheung et al., 2008;
Ohashi et al., 1991). Therapeutic vaccinations can often bolster endogenous T cell responses
against the autochthonous tumors, but the responses are not sustained. This has important
implications for human cancer, as T cells directed against TAAs may require special conditions,
not only to become activated, but also to persist in a host environment that naturally removes
these cells. If ignorance or deletional tolerance can be overcome in these models, ultimately
anergy or the loss of the functional activity of TAA-specific T cells seems to universally occur
(Anderson et al., 2007; Cheung et al., 2008; Drake et al., 2005; Lyman et al., 2004).
Transplantable models have provided the only means to express true tumor-specific
antigens (TSAs) in cancer models, and interestingly, these tumors can be routinely rejected
after transplantation (Schreiber, 2003). This has led many investigators to speculate that T cells
targeting TSAs are critical for productive T cell responses against cancer (Parmiani et al., 2007;
Schietinger et al., 2008). However, it is unclear whether transplantation faithfully recapitulates
the human disease, leading many to believe that the importance of TSAs is exaggerated due to
the artificial nature of these models (Khong and Restifo, 2002; Scott, 1991). Several attempts
have been made to create endogenous mouse models of cancer that express TSAs (Cheung et
al., 2008; Soudja et al., 2010; Willimsky and Blankenstein, 2005). Interestingly, systemic T cell
tolerance toward the tumor antigens was found to inhibit productive anti-tumor immune
responses in all of these systems. Leaky expression of the putative TSAs in normal tissues or
prior to tumor formation may be responsible for the systemic tolerance in these models (Cheung
et al., 2008; Getnet et al., 2009; Willimsky et al., 2008). Therefore, TSA expression from
autochthonous mouse models of cancer may be difficult to achieve.
212
In an effort to investigate the effect of T cell responses against TSAs expressed from
endogenously arising tumors, I developed a novel system to introduce exogenous antigens into
spatiotemporally controlled autochthonous mouse cancers (Chapters 2 and 3). By mimicking
TSA expression in lung tumors and sarcomas, I found that endogenous T cell responses to
tumors are generated that can persist for long periods of time. Therefore, tumor antigens are not
ignored in the context of TSA expression, and T cells targeting TSAs, as opposed to TAAs, are
not rapidly deleted. In the setting of TSA expression in lung cancer, T cells ultimately became
tolerant. This could be overcome by vaccination with dendritic cells or therapeutic influenza
infection, indicating that priming by autochthonous tumors in vivo was suboptimal even in
response to TSAs. In addition, in autochthonous lung-tumor-bearing mice, T cell responses
against transplanted tumors expressing the same TSAs were unaltered. Therefore, T cell
tolerance against TSAs expressed from autochthonous tumors was specific to the endogenous
tumors, and unlike TAAs, TSA expression from endogenous tumors did not induce a systemic
state of T cell tolerance.
These results indicate that TSAs may serve as better targets for human cancer
immunotherapy because T cells directed against these antigens do not need to overcome host
tolerance networks that can remove or suppress T cells specific for TAAs. Furthermore, T cells
targeting TSAs will only target tumors, reducing the risk of inducing autoimmune reactions.
Indeed, many cancer immunotherapies in animal models and human clinical trials result in
autoimmune side-effects (Caspi, 2008). These side-effects may be the result of the generalized
blockade of immune tolerance networks, such as with CTLA-4 antibodies, or the
immunotherapeutic targeting of TAAs. These risks could potentially be minimized if the
therapies were tailored against TSAs. Lastly, TSAs may originate from mutations in protooncogenes or tumor suppressors that are necessary for driving the disease, whereas TAA
expression may not be essential to the cancer’s etiology (Schietinger et al., 2008). Therefore,
213
selection for tumor cells with reduced expression of TAA may provide an easy mechanism of
immune escape that is not viable for tumors expressing oncogenic TSAs.
In our model, the TSAs were not oncogenic in nature and thus mimicked so called
“passenger mutations.” Not surprisingly, lung tumors that grew in mice vaccinated against the
TSAs or sarcomas that developed in immune-competent mice, lost expression of the model
TSAs. Nevertheless, the expression of these TSAs still provided significant protection from
tumor initiation and progression in both types of cancer. Models that link TSAs to the driving
oncogenic event need to be developed to investigate whether T cell responses to oncogenic
antigens can be even more effective. We are developing new lentiviral vectors that will express
model TSAs fused directly to an oncogenic K-ras allele to study the T cell response to such
oncoantigens. In addition, our lab is screening multiple oncogenic mutations in K-ras to
determine whether novel peptides are generated that can be presented on MHC class I
molecules from mice with different H-2 backgrounds.
Models in which tumor initiation and expression of TSAs are temporally separable are
also needed to discern whether T cell responses are different if the induction of antigen
expression occurs in established tumors, where mutated proteins likely accumulate in human
cancers. We are developing lentiviral systems that utilize a tetracycline operon to control TSA
expression, allowing TSAs to be turned on after tumors are initiated (see Chapter 2). The
combined use of ligand-regulatable Cre and Flp recombinases may also provide a means to
separate the genetic induction of cancer from the expression of model TSAs (see Introduction,
Figure 4).
Despite the advantages of TSAs, these antigens may be more difficult to target with
current immunotherapeutic techniques because they are different in each cancer. However, if
targeting TSAs provides the best therapeutic option, greater focus needs to be placed on
methods to tailor therapies to individual patients. The post-genomic era provides us with a new
214
means to rapidly identify mutations in individual tumors and computationally predict which
peptides will prompt the best T cell responses in these patients (Segal et al., 2008). Streamlined
methods to harvest tumor-infiltrating T cells from patients and selectively expand T cells
recognizing TSAs for reinfusion back into patients has already proven to be quite successful
against some forms of cancer (Rosenberg et al., 2008). Alternatively, utilizing patients’ own
resected tumors as the basis for tailored vaccinations may also provide a means to target TSAs
without prior identification of the targeted antigens.
II. Different models of cancer have altered T cell responses
The models of cancer used to draw conclusions about immune responses to tumors are
seldom emphasized in cancer immunology investigations. This has led to an acceptance in the
field that discoveries made using a particular cancer model actually reveal important
mechanisms underlying all immune-tumor interactions. However, there may be tremendous
diversity in how T cells respond to tumors in different settings – TAA or TSA expression,
autochthonous or transplanted, carcinogen-induced or arising spontaneously, or originating in
different tissues (see next section). In this thesis research, I provide evidence to suggest that
the cancer model significantly influences immune-tumor interactions.
In Chapter 2, I compared homologous forms of lung cancer in autochthonous and
transplantable settings to determine whether immune-tumor interactions differed. We
discovered that while autochthonous lung tumors expressing TSAs had delayed malignant
progression but always retained antigen expression, transplanted tumors expressing TSAs were
either rejected or tumors that did grow always lost expression of TSAs. Orthotopic
transplantation of cell lines back into the lung by intravenous injection also led to the outgrowth
of tumors that lost expression of TSAs. Tumor protection and antigen loss only occurred in the
autochthonous tumor setting after the transfer of in vitro activated TSA-specific T cells or
215
vaccination against the TSAs. This indicates that priming in response to endogenous lung
tumors is less efficient than priming against transplanted tumors. Using mouse models of
pancreatic cancer, another study also found that autochthonous tumors grew progressively
while transplanted cell lines derived from these tumors were rejected (Garbe et al., 2006). In
contrast to our results, all the cell lines generated from the autochthonous pancreatic tumors
were immunogenic upon transplantation, whereas the immunogenicity of lung tumors from our
model required the introduction of model TSAs. The investigators were hindered by a lack of
knowledge about the tumor antigens targeted by T cells, making it impossible to compare T cell
priming in both contexts. Instead they found that CD8+ T cells could not infiltrate the
endogenous tumors as efficiently as transplanted tumors. Additional mechanisms by which
transplanted tumors may artificially promote anti-tumor T cell responses are detailed in section
five of the Introduction. These results indicate that immune responses and their effect on tumor
development may be the direct result of the cancer models used for the investigation.
To understand cancer immunoediting, investigators have relied heavily on the use of
mouse models of cancer in which carcinogens, primarily methylcholanthrene (MCA), are used to
induce the development of tumors. However, the immunogenicity of carcinogen-induced
sarcomas has been questioned as artificial because spontaneously arising tumors are rarely
immunogenic (Hewitt et al., 1976). The mutagenic activity of carcinogens may generate tumorspecific neoantigens in experimentally induced tumors that grossly exceed the number of
neoantigens harbored in human cancers (Khong and Restifo, 2002). By utilizing genetically
engineered mouse models of sarcomagenesis that could be conditionally induced to express or
lack TSAs, we provide evidence to suggest that TSA expression is required for the process of
immunoediting against sarcomas (Chapter 3). Only sarcomas derived in lymphocyte-deficient
mice that express model TSA antigens are rejected upon transplantation into immunecompetent mice. Thus tumor immunogenicity is not a necessary byproduct of the tumorigenic
216
process. The sarcomas induced to express TSAs in immune-competent mice lost expression of
the antigens during their development, indicating that immunoediting during tumor development
acts by selecting for tumor cells that have lost the expression of potent antigens. Both of these
results could potentially explain the reduced immunogenicity of spontaneous tumors derived in
immune-competent mice. Therefore, these data provide a missing link between the discordant
results found using either carcinogen-induced or spontaneously arising tumors.
III. Tumor origin affects the T cell response and a tumor’s immunogenicity
Over four decades ago, R.T. Prehn speculated that the tissue in which a cancer arises
could influence how the immune system responds to cancer (Prehn, 1968). However, this issue
has scarcely been addressed experimentally. By comparison of autochthonous models of lung
cancer and sarcomagenesis described in Chapters 2 and 3, it is clear that the tissue in which a
tumor originates can significantly modulate the ensuing T cell response. Importantly, both
models contained the same genetic alterations to initiate the cancers, by using K-rasLSLG12D/+
;p53fl/fl mice, and expressed the same model TSAs, by using Lenti-LucOS to induce tumor
formation. Only the site of tumor formation and the type of cancers generated were distinct.
In both scenarios, T cell responses directed against the model TSAs could be detected,
however, only T cells generated in the sarcoma model were fully functional, providing protection
from sarcoma formation or driving TSA loss in palpable sarcomas. T cells responding to TSAs in
lung tumors led to a delay in the malignant progression of the cancers, but were not completely
functional and ultimately dissipated from tumors despite maintained expression of the TSAs in
tumors. It is not clear what underlies these divergent immune responses to the same TSAs in
the context of these two forms of cancer, but it is enticing to speculate that different immune
environments surrounding the tumors may dictate the anti-tumor immune response.
217
Disruption or damage to tissues underlying the surface of the skin, such as during an
abrasion, can lead to the introduction of pathogens that require rapid and effective immune
responses for host protection. This has perhaps led to the evolution of fewer regulatory
requirements for activating cells of the adaptive immune system in these tissues, making most
disturbances in this environment immune stimulatory (Strid et al., 2009). Indeed, this has made
skin the tissue of choice for vaccination regimens and may also explain why primary sarcomas,
induced either by carcinogens or genetically, and sub-cutaneously transplanted tumors are so
susceptible to anti-tumor immune responses. The relatively rare encounter with external
elements in tissues underneath the skin contrasts with the epithelial cells of the lung, which are
perturbed constantly by the external environment (Strid et al., 2009). This has led to the
evolution of an immune-tolerant environment in the lung, which prevents inappropriate (and
potentially detrimental) immune responses to innocuous encounters (Holt et al., 2008). Thus the
immune environment of the lung maintains a more stringent threshold for adaptive immune
activation. In this context, T cell responses to autochthonous lung tumors may be driven to
inactivity and tolerance because of unique immune-regulatory networks that have specifically
evolved in the lung. In fact, early investigations found that carcinogen-induced lung adenomas
were less immunogenic than carcinogen-induced sarcomas. However, Prehn predicted at the
time (perhaps incorrectly), that stronger immune responses were actually induced in lung
tissues, which led to the outgrowth of lung tumors with reduced immunogenicity compared to
sarcomas. An important next step for cancer immunologists should be to embrace the diversity
of immune environments that likely shape immune responses to cancers that arise in different
tissues by broadening our investigations to encompass a greater diversity of cancer models.
218
IV. Cancer vaccination
The weak and incomplete endogenous T cell response generated in our autochthonous
model of lung cancer recapitulates many of the features of T cell responses against human
cancers, making it an ideal system to test immunotherapeutic strategies. Successful therapeutic
vaccination has proven to be very difficult in patients with cancer as well as in many
experimental mouse cancer models (Drake, 2010; Goldman and DeFrancesco, 2009). Similar to
other investigations in mouse models of cancer that express TAAs, work from our lab has also
found that it is impossible to sustain T cell responses against model TAAs expressed in our
model of lung cancer after vaccination with a strain of influenza expressing the same TAA
(Cheung et al., 2008). However, in Chapter 4, we showed that by targeting TSAs in lung tumors
with the same strains of influenza, we could generate fully functional T cell responses against
the tumor antigens that were sustained, again providing support for the use of TSAs in vaccine
strategies rather than TAAs.
Nevertheless, the tumors in these mice were not eradicated. Potential explanations for
tumor escape after therapeutic vaccination include lost or reduced expression of TSAs and a
decreased magnitude of the T cell response after vaccination in mice with established tumors.
However, it is still not clear why vaccination did not more dramatically reduce the tumor burden
since antigen expression was clearly maintained in some tumors and functionally competent T
cells did persist. Understanding the ineffectiveness of the vaccine in this context may provide
insights into the failures of therapeutic vaccines in cancer patients.
A kinetic argument has often been used to describe the failure of immune responses to
eradicate cancers (Hanson et al., 2000). It is possible that immune responses can only control a
limited tumor burden due to a limited cytotoxic capacity of T cells and the rapid growth of
tumors. This could explain why adoptive cell transfer therapy (ACT) has proven to be more
effective than vaccination therapies. The tremendous number of anti-tumor T cells generated
219
and transferred during ACT (on the order of 1011 cells) may exceed the productive capacities of
vaccine-induced immune responses. Yet the generation of a sustained memory T cell response
in our vaccination scheme would seem to eventually overcome this problem by providing a
limitless supply of T cells to attack the relatively slow-growing lung tumors. Although employing
the immunoediting hypothesis provides limitless potential explanations for tumor evasion,
antigen loss would seem to provide the easiest mode of escape in the context of functional antitumor T cells. One potential explanation is that antigen-expressing tumor cells within the core of
the tumors are sheltered or protected from T cell recognition by tumor cells that have lost
antigen expression. Other investigations have shown that impeded access into the tumor can
allow tumors to escape (Buckanovich et al., 2008). Alternatively, the T cell response against
tumors may end with the clearance of the influenza infection because T cell activity may require
reencounter with tissue-resident, antigen-presenting cells (APCs) that are only active during
infections (Woodland and Kohlmeier, 2009). If this were the case, repeated stimulation of APCs
at the site of tumors might prolong T cell responses against tumors. In section VI, I will discuss
the implications of a temporally regulated immune response in productive immune responses
against cancer.
V. Immune tolerance or immunoediting?
Whether cancers progress because of the induction of anti-tumor immune tolerance or
the evolution of tumor-escape mechanisms has been a topic of great debate and has
unfortunately forged the development of two opposing ideological camps in the field of cancer
immunology (Qin and Blankenstein, 2004). However, by examination of different forms and
models of cancer in Chapters 2 and 3, we find support for both mechanisms of tumor evasion.
As described in sections I, II and III of this Discussion, the immune response and its outcome on
220
tumor development are affected by the types of tumor antigens expressed, the models used,
and the types of cancers studied.
Interestingly, evidence for both immunoediting and immune tolerance occurring within
the progression of a single cancer has even been documented. In this particular study,
investigators found that T cell tolerance to a tumor antigen was preceded by reduced
expression of the antigen in tumors (Zhou et al., 2004). It is interesting to speculate that both
mechanisms may complement, rather than exclude, one another. For example, if strong
immune responses arise in response to tumors naturally, or by induction with vaccines,
immunoediting may be required for tumors to escape. Indeed, we found evidence of decreased
TSA expression in sarcomas and also lung tumors in the context of vaccination. Evidence of
immunoediting can also be found in the clinic, particularly after therapeutic vaccination against
specific tumor antigens (Khong and Restifo, 2002). However, prolonged interactions between
immune cells and tumors would seem to naturally drive immune tolerance (Kim and Ahmed,
2010). This might occur in the context of relatively weak immune responses that are unable to
rapidly eradicate the tumors. In this way, autoregulatory tolerance mechanisms may provide an
alternative route for “unedited” tumors (or tumor cells) to progress by outlasting the functional
immune response.
VI. Timing in adaptive immune responses: a unifying theme?
In closing, it is interesting to speculate about the role of the immune system in regulating
cancer from the perspective of its clear evolutionary role in fighting pathogenic infections.
Foremost, it is important to consider that perhaps the most critical aspect of the evolution of the
immune system was the generation of appropriate checks and balances to regulate this
powerful weapon, rather than the creation of a more effective killer of infectious agents. Just by
considering that 25% of deaths worldwide each year are due to infection (from W.H.O.), while
221
the prevalence of autoimmune disease is only perhaps 4% (from Cureresearch.com), one may
infer that the immune system has evolved to err on the side of tolerance rather than immunity. A
key mechanism for preventing autoimmune reactions may be the evolution of systems that
restrict adaptive immune responses to a defined period of time. This is not a new idea, as it has
long been understood that both intrinsic and extrinsic factors contribute to the natural
contraction of responding adaptive immune cells after infections (Badovinac et al., 2004;
Jameson and Masopust, 2009). However, the natural cessation of immune responses after an
acute infection is rarely considered in the context of immune responses to cancer.
Transplantable tumors are the only cancers that have been cured by an adaptive
immune response. More importantly, these tumors have always been rejected within a short
window of time between 10 and 14 days after transplantation or following various immune
priming protocols. When tumors grow for extended periods of time, immune responses may
delay the growth of tumors, but never reject them (Hanson et al., 2000). In a similar manner to
transplanted tumors, acute viral infections are always cleared within the first two weeks after
infection (Wherry and Ahmed, 2004). Viral infections that are not cleared ultimately go on to
become chronic infections that are essentially tolerated by the immune system. In contrast to
transplanted tumors, autochthonous tumors have never been eradicated in any setting, but
therapeutic treatments have always begun very late during cancer progression. Immunity
against autochthonous tumors has only come in the form of protection after prophylactic
vaccination or very early therapeutic vaccination as evidenced in Chapters 2 and 4 of this
thesis. All of these results seem to emphasize that there is an inherent clock that controls the
timing of active T cell responses, whether it’s against viruses or cancers. What is responsible for
maintaining this “leash” on the immune response is unclear. As pointed out in the previous
section, the presence of activating signals at the sites of T cell function (potentially provided by
activated APCs) may be necessary to license T cell activity. Alternatively, persistent encounters
222
with antigens may drive T cells to lose function (Bucks et al., 2009). Perhaps activated T cells
are only permitted to retain effector function for limited periods of time after initial priming, and
paradoxically, antigen withdrawal is required for T cells to ultimately retain function. Antigen
withdrawal would normally occur with the clearance of pathogens. However, antigen withdrawal
would not occur if T cell responses were directed against self antigens. Thus, T cells may be
hard-wired to lose function in the context of persistent antigen exposure as a means to prevent
inappropriate responses to self.
In any case, for the immune system to effectively treat cancer, it may be important to
consider that the immune system works best in an acute fashion. Indeed, considering the acute
timing of the immune response can provide a unifying explanation for the diversity of results
from many investigations using different cancer models. From the perspective of the immune
system, however, cancer may more accurately be viewed as a chronic disease. Therefore,
immune therapy against cancer must not only act to boost T cell responses to tumors, but also
attempt to counteract the many, as yet unidentified, regulatory networks that likely restrict the
timing of active immune responses. Autoimmune diseases represent rare breakdowns in these
regulatory systems that enable immune responses to continue indefinitely and may provide
clues as to how to better treat cancer immunotherapeutically.
223
REFERENCES
Anderson, M.J., Shafer-Weaver, K., Greenberg, N.M., and Hurwitz, A.A. (2007). Tolerization of
tumor-specific T cells despite efficient initial priming in a primary murine model of prostate
cancer. J Immunol 178, 1268-1276.
Badovinac, V.P., Porter, B.B., and Harty, J.T. (2004). CD8+ T cell contraction is controlled by
early inflammation. Nat Immunol 5, 809-817.
Buckanovich, R.J., Facciabene, A., Kim, S., Benencia, F., Sasaroli, D., Balint, K., Katsaros, D.,
O'Brien-Jenkins, A., Gimotty, P.A., and Coukos, G. (2008). Endothelin B receptor mediates the
endothelial barrier to T cell homing to tumors and disables immune therapy. Nat Med 14, 28-36.
Bucks, C.M., Norton, J.A., Boesteanu, A.C., Mueller, Y.M., and Katsikis, P.D. (2009). Chronic
antigen stimulation alone is sufficient to drive CD8+ T cell exhaustion. J Immunol 182, 66976708.
Caspi, R.R. (2008). Immunotherapy of autoimmunity and cancer: the penalty for success. Nat
Rev Immunol 8, 970-976.
Cheung, A.F., Dupage, M.J., Dong, H.K., Chen, J., and Jacks, T. (2008). Regulated expression
of a tumor-associated antigen reveals multiple levels of T-cell tolerance in a mouse model of
lung cancer. Cancer Res 68, 9459-9468.
Drake, C.G. (2010). Prostate cancer as a model for tumour immunotherapy. Nat Rev Immunol
10, 580-593.
Drake, C.G., Doody, A.D., Mihalyo, M.A., Huang, C.T., Kelleher, E., Ravi, S., Hipkiss, E.L.,
Flies, D.B., Kennedy, E.P., Long, M., et al. (2005). Androgen ablation mitigates tolerance to a
prostate/prostate cancer-restricted antigen. Cancer Cell 7, 239-249.
Garbe, A.I., Vermeer, B., Gamrekelashvili, J., von Wasielewski, R., Greten, F.R., Westendorf,
A.M., Buer, J., Schmid, R.M., Manns, M.P., Korangy, F., and Greten, T.F. (2006). Genetically
induced pancreatic adenocarcinoma is highly immunogenic and causes spontaneous tumorspecific immune responses. Cancer Res 66, 508-516.
Getnet, D., Maris, C.H., Hipkiss, E.L., Grosso, J.F., Harris, T.J., Yen, H.R., Bruno, T.C., Wada,
S., Adler, A., Georgantas, R.W., et al. (2009). Tumor recognition and self-recognition induce
distinct transcriptional profiles in antigen-specific CD4 T cells. J Immunol 182, 4675-4685.
Goldman, B., and DeFrancesco, L. (2009). The cancer vaccine roller coaster. Nat Biotechnol
27, 129-139.
Hanson, H.L., Donermeyer, D.L., Ikeda, H., White, J.M., Shankaran, V., Old, L.J., Shiku, H.,
Schreiber, R.D., and Allen, P.M. (2000). Eradication of established tumors by CD8+ T cell
adoptive immunotherapy. Immunity 13, 265-276.
Hewitt, H.B., Blake, E.R., and Walder, A.S. (1976). A critique of the evidence for active host
defence against cancer, based on personal studies of 27 murine tumours of spontaneous origin.
Br J Cancer 33, 241-259.
224
Holt, P.G., Strickland, D.H., Wikstrom, M.E., and Jahnsen, F.L. (2008). Regulation of
immunological homeostasis in the respiratory tract. Nat Rev Immunol 8, 142-152.
Jameson, S.C., and Masopust, D. (2009). Diversity in T cell memory: an embarrassment of
riches. Immunity 31, 859-871.
Khong, H.T., and Restifo, N.P. (2002). Natural selection of tumor variants in the generation of
"tumor escape" phenotypes. Nat Immunol 3, 999-1005.
Kim, P.S., and Ahmed, R. (2010). Features of responding T cells in cancer and chronic
infection. Curr Opin Immunol 22, 223-230.
Lyman, M.A., Aung, S., Biggs, J.A., and Sherman, L.A. (2004). A spontaneously arising
pancreatic tumor does not promote the differentiation of naive CD8+ T lymphocytes into effector
CTL. J Immunol 172, 6558-6567.
Ochsenbein, A.F., Klenerman, P., Karrer, U., Ludewig, B., Pericin, M., Hengartner, H., and
Zinkernagel, R.M. (1999). Immune surveillance against a solid tumor fails because of
immunological ignorance. Proc Natl Acad Sci U S A 96, 2233-2238.
Ohashi, P.S., Oehen, S., Buerki, K., Pircher, H., Ohashi, C.T., Odermatt, B., Malissen, B.,
Zinkernagel, R.M., and Hengartner, H. (1991). Ablation of "tolerance" and induction of diabetes
by virus infection in viral antigen transgenic mice. Cell 65, 305-317.
Pardoll, D. (2003). Does the immune system see tumors as foreign or self? Annu Rev Immunol
21, 807-839.
Parmiani, G., De Filippo, A., Novellino, L., and Castelli, C. (2007). Unique human tumor
antigens: immunobiology and use in clinical trials. J Immunol 178, 1975-1979.
Prehn, R.T. (1968). Tumor-specific antigens of putatively nonviral tumors. Cancer Res 28, 13261330.
Qin, Z., and Blankenstein, T. (2004). A cancer immunosurveillance controversy. Nat Immunol 5,
3-4; author reply 4-5.
Rosenberg, S.A., Restifo, N.P., Yang, J.C., Morgan, R.A., and Dudley, M.E. (2008). Adoptive
cell transfer: a clinical path to effective cancer immunotherapy. Nat Rev Cancer 8, 299-308.
Schietinger, A., Philip, M., and Schreiber, H. (2008). Specificity in cancer immunotherapy.
Semin Immunol 20, 276-285.
Schreiber, H. (2003). Tumor Immunology. In Fundamental Immunology, W.E. Paul, ed.
(Philadelphia: Lippincott Williams & Wilkins), pp. 1557-1592.
Scott, O.C. (1991). Tumor transplantation and tumor immunity: a personal view. Cancer Res 51,
757-763.
225
Segal, N.H., Parsons, D.W., Peggs, K.S., Velculescu, V., Kinzler, K.W., Vogelstein, B., and
Allison, J.P. (2008). Epitope landscape in breast and colorectal cancer. Cancer Res 68, 889892.
Soudja, S.M., Wehbe, M., Mas, A., Chasson, L., de Tenbossche, C.P., Huijbers, I., Van den
Eynde, B., and Schmitt-Verhulst, A.M. (2010). Tumor-initiated inflammation overrides protective
adaptive immunity in an induced melanoma model in mice. Cancer Res 70, 3515-3525.
Speiser, D.E., Miranda, R., Zakarian, A., Bachmann, M.F., McKall-Faienza, K., Odermatt, B.,
Hanahan, D., Zinkernagel, R.M., and Ohashi, P.S. (1997). Self antigens expressed by solid
tumors Do not efficiently stimulate naive or activated T cells: implications for immunotherapy. J
Exp Med 186, 645-653.
Strid, J., Tigelaar, R.E., and Hayday, A.C. (2009). Skin immune surveillance by T cells--a new
order? Semin Immunol 21, 110-120.
Wherry, E.J., and Ahmed, R. (2004). Memory CD8 T-cell differentiation during viral infection. J
Virol 78, 5535-5545.
Willimsky, G., and Blankenstein, T. (2005). Sporadic immunogenic tumours avoid destruction by
inducing T-cell tolerance. Nature 437, 141-146.
Willimsky, G., Czeh, M., Loddenkemper, C., Gellermann, J., Schmidt, K., Wust, P., Stein, H.,
and Blankenstein, T. (2008). Immunogenicity of premalignant lesions is the primary cause of
general cytotoxic T lymphocyte unresponsiveness. J Exp Med 205, 1687-1700.
Woodland, D.L., and Kohlmeier, J.E. (2009). Migration, maintenance and recall of memory T
cells in peripheral tissues. Nat Rev Immunol 9, 153-161.
Zhou, G., Lu, Z., McCadden, J.D., Levitsky, H.I., and Marson, A.L. (2004). Reciprocal changes
in tumor antigenicity and antigen-specific T cell function during tumor progression. J Exp Med
200, 1581-1592.
226
APPENDIX 1:
Conditional mouse lung cancer models using adenoviral or lentiviral delivery of
Cre recombinase
Michel DuPage1,3, Alison L. Dooley1,3, Tyler Jacks1,2
1
Koch Institute for Integrative Cancer Research and Department of Biology, Massachusetts
Institute of Technology, Cambridge, Massachusetts, USA.
2
Howard Hughes Medical Institute, Massachusetts Institute of Technology, Cambridge,
Massachusetts, USA.
3
These authors contributed equally to this work.
This appendix was taken from the work published in:
Nature Protocols, 4(7):1064-1072, 2009.
The author wrote the protocol describing the infection of conditionally regulated mouse models
together with Alison Dooley. Experiments were performed in the laboratory of Tyler Jacks.
227
ABSTRACT
The development of animal models of lung cancer is critical to our understanding and treatment
of the human disease. Conditional mouse models provide new opportunities for testing novel
chemopreventatives, therapeutics and screening methods that are not possible with cultured
cell lines or xenograft models. This protocol describes how to initiate tumor formation in two
conditional genetic models of human non-small cell lung cancer (NSCLC) utilizing the activation
of oncogenic K-ras and the loss of function of p53. We discuss methods for sporadic expression
of Cre in the lungs via engineered adenovirus or lentivirus and provide a detailed protocol for
the administration of the virus by intranasal inhalation or intratracheal intubation. The protocol
requires 1-5 minutes per mouse with an additional 30-45 minutes to set-up and allow for the
recovery of mice from anesthesia. Mice may be analyzed for tumor formation and progression
starting 2-3 weeks after infection.
228
INTRODUCTION
Lung cancer is the leading cause of cancer deaths worldwide with non-small cell lung
cancer (NSCLC) being the most prevalent form of lung cancer (Sun et al. 2007; Jemal et al.
2008; Herbst et al. 2009). Progress within the last decade has led to the sophisticated
engineering and application of advanced preclinical models of human cancer in the mouse (Van
Dyke and Jacks 2002; Frese and Tuveson 2007). These models are critical to our
understanding of the human disease because they shed light on events and processes that
cannot be easily studied using transplantable or chemically-induced cancer models (Meuwissen
and Berns 2005; Shaw et al. 2005; Frese and Tuveson 2007). Several laboratories have
constructed genetically engineered mouse models of NSCLC that mimic the genetic and
histopathological features of the human disease (Fisher et al. 2001; Jackson et al. 2001;
Johnson et al. 2001; Meuwissen et al. 2001; Meuwissen and Berns 2005; Ji et al. 2006; Dankort
et al. 2007). The development of Cre recombinase-controlled (Cre/LoxP) tumor models has
allowed for the generation of autochthonous tumors derived from a limited number of somatic
cells that become transformed in their natural location, surrounded by a normal tissue
microenvironment (Frese and Tuveson 2007). By engineering LoxP DNA elements into the
mouse genome that either surround (‘flox’) exons critical to a tumor suppressor gene’s function
or surround a synthetic ‘stop’ element (‘LSL’) inserted in front of an oncogene, investigators can
‘turn-off’ tumor suppressors or ‘turn-on’ oncogenes with delivery of Cre recombinase to the
appropriate cell types (Supp. Fig. 1 and see reference 5 for a review of Cre/LoxP-controlled
genetically engineered mouse models of cancer). With this technology, investigators can not
only recapitulate the genetic alterations found in the human disease, but also the timing of onset
and potentially the cellular origin of the disease.
Common mutations in human NSCLC are activating mutations in K-RAS (10-30%) and loss
of function point mutations in p53 (50-70%) (Herbst et al. 2009). Our laboratory has modeled an
229
oncogenic mutation in K-ras by changing a glycine to aspartic acid at codon 12 in the gene’s
endogenous locus. To control the expression of K-rasG12D, a lox-stop-lox (LSL) cassette was
engineered into the first intron of the K-ras gene. The LSL cassette consists of transcriptional
and translational stop elements flanked by LoxP sites that prevents the expression of the mutant
allele until the stop elements are removed by the activity of Cre recombinase (Jackson et al.
2001; Tuveson et al. 2004). The LSL cassette thus creates a null version of the gene. It is
important to note that K-ras null mice are embryonic lethal (Johnson et al. 1997); therefore, mice
can only be heterozygous for the K-rasLSL-G12D allele. To mimic the loss of p53 function in our KrasLSL-G12D-driven tumor model, we have utilized a conditional p53 allele from the laboratory of
Anton Berns. This ‘floxed’ p53 allele (p53fl) has LoxP sites flanking exons two through ten of p53
that are deleted after Cre-mediated recombination, abolishing p53 function (Jonkers et al. 2001)
(Supp. Fig. 1). Prior to Cre-mediated recombination, the p53 locus is maintained in its wildtype
state and p53 activity is normal. To more accurately recapitulate the p53 loss of function
mutations commonly observed in human NSCLC, our laboratory has generated two conditional
point mutant (mt) versions of p53 (R172H, R270H) that are engineered into the endogenous
p53 locus, but silenced by a LSL cassette in the absence of Cre (Olive et al. 2004). We use
mice that are p53LSL-mt/fl or p53LSL-mt/+ to specifically express mutant p53 alone or with wildtype
p53 (respectively) in tumors upon Cre-mediated recombination (Olive et al. 2004; Jackson et al.
2005). Importantly, as with the K-rasLSL-G12D allele, the LSL cassette in p53LSL-mt alleles blocks
p53 expression in the absence of Cre, effectively creating a p53 null allele. Therefore, p53LSL-mt/fl
or p53LSL-mt/+ mice are heterozygous for wildtype p53 and phenocopy p53+/- mice. To specifically
express either of these mutant p53 alleles sporadically in cells of the lung, mice inhale viruses
engineered to express Cre either by intranasal instillation or intratracheal intubation.
230
Additional Cre recombinase controlled animal models of lung cancer
In addition to the models of NSCLC utilized in this protocol, other conditional lung cancer
models have been described which may be initiated with inhalation of viruses expressing Cre.
The activation of oncogenic K-ras along with loss of p16Ink4a or Ink4a/Arf tumor suppressors,
which are mutated in 20-50% of human cases, have been described (Fisher et al. 2001; Ji et al.
2007). In another mouse model, two other subtypes of NSCLC, squamous cell and large cell
carcinoma, develop following the combined activation of oncogenic K-ras and loss of the LKB1
tumor suppressor, which is mutant in 10-30% of human cases (Ji et al. 2007). NSCLC models
driven by conditionally activated mutations in Braf or EGFR, mutated in 3% or 10-40%
(respectively) of human lung cancers, have also been generated (Ji et al. 2006; Dankort et al.
2007) (and unpublished, K. Lane). A model of small cell lung cancer (SCLC) has been created
in which tumors arise following the loss of both the p53 and Rb tumor suppressors (Meuwissen
et al. 2003).
Limitations of viral delivery of Cre recombinase
Although these Cre/LoxP models provide the most sophisticated means to sporadically
control genetic events at their endogenous loci, they are also limited by the difficult requirement
to introduce Cre into an initiating cancer cell. In this protocol, investigators cannot restrict the
genetic events exclusively to initiating cells of the disease because of the inherently non-specific
nature of the viruses used to deliver Cre to cells of the lung. However, this has in fact been
beneficial in our laboratory because it does not require that investigators identify and target Cre
specifically to the cell of origin of the disease. Perhaps the best system to deliver Cre
specifically to the cells that give rise to the disease are next generation regulatable Cre alleles,
such as the tamoxifen-inducible Cre-estrogen receptor (CreERT) fusion protein (Frese and
Tuveson 2007). By expressing a CreERT transgene under the control of a cell type-specific
231
promoter, investigators can induce Cre activity specifically in the cells of a given tissue by
injecting animals with tamoxifen. With proper dosing of tamoxifen, investigators can in theory
control the penetrance of Cre activity and therefore the multiplicity of the disease. However,
various promoter-driven CreERT alleles are not yet available and CreERT models succumb to
problems encountered with many transgenic models in that the expression and activity of these
alleles are often leaky and do not provide tight control of tumor initiation in the organ of interest
(Frese and Tuveson 2007). Therefore, despite the drawback of using potentially hazardous
viruses, this technique is the most effective means to sporadically deliver Cre to cells of the lung
to initiate lung tumor formation.
Alternative animal models of lung cancer
There are several alternative lung cancer models to those described here that do not require
viral delivery of Cre to the lungs to generate tumors (see reference 6 for a comprehensive
review of mouse models of NSCLC). Transgenic models have been created that utilize lung
tissue specific promoters to drive expression of viral oncoproteins, such as E6/E7 and large T
antigen, or cellular oncogenes, such as c-myc, c-Raf-1 and v-H-ras (Meuwissen and Berns
2005). However, these models are limited in recapitulating the human disease because the
oncogenic transgene is expressed in all of the cells of a targeted organ beginning early in the
organ’s development. To overcome this limitation, our laboratory has engineered a latent
oncogenic allele of K-ras that is expressed spontaneously in only a limited number of cells in the
lung to initiate lung tumor formation (Johnson et al. 2001). However, the stochastic nature of this
allele does not allow investigators to control tumor onset or multiplicity. Double transgenic
models that rely on doxycycline-regulated transcriptional transactivators have allowed for the
induction of oncogene expression in adult animals, making it possible to control tumor initiation
(Meuwissen and Berns 2005). However, these models still fail to provide sporadic oncogene
232
expression at physiological levels due to the use of doxycyline-regulated transcription factors.
Furthermore, these transgenic systems make it difficult to control tumor multiplicity and require
continuous doxycycline dosing to maintain oncogene expression.
Applications of intranasal and intratracheal delivery methods
The intranasal and intratracheal infection techniques described in this protocol are not
restricted to tumor-initiating studies. Investigators may theoretically probe the role of any
Cre/LoxP-controlled genes in various cells of the lung using this protocol to deliver Cre.
Alternatively, the intranasal or intratracheal delivery methods can be used in applications other
than the viral delivery of Cre recombinase. The viral delivery systems may be adapted to
express cDNAs or shRNAs in cells of the lung by infection with lentiviruses (Marumoto et al.
2009). Viruses, such as the influenza virus, can be delivered for pathological studies (Shen et
al. 2008). siRNAs can be delivered as therapeutic agents for disease (Ge et al. 2004). While
either the intranasal or intratracheal techniques can be utilized for the aforementioned studies,
only intratracheal intubation is recommended for the orthotopic transplantation of lung tumor
cells or lung tumor cell lines to the alveolar space of the lungs (unpublished, C. Kim and T.
Jacks). Certain chemical injurants, such as bleomycin, can be delivered intratracheally to study
repair mechanisms in the lung and the effect of injury on tumor development (Daly et al. 1997).
Therefore, there are a number of uses of this technique for a variety of studies involving the
mouse lung. Here we describe the use of the intranasal and intratracheal delivery methods to
introduce viruses expressing Cre to initiate the K-rasLSL-G12D/+ (K) and the K-rasLSL-G12D/+;p53fl/fl
(KP) conditional mouse models of NSCLC (Supp. Fig. 1).
233
EXPERIMENTAL DESIGN
After deciding to utilize autochthonous tumor models, investigators must consider the
genetic tumor model, the viral system to express Cre recombinase, and the viral delivery
method when initiating an experiment. Each option has certain advantages and limitations that
are highlighted in the text and the Tables.
Genetic Model
Although both the K-rasLSL-G12D/+ (K) and the K-rasLSL-G12D/+;p53fl/fl (KP) NSCLC models induce
tumors that resemble the human disease histopathologically, they have distinct features that
make them more suitable for different applications (Table 1). In our laboratory’s mouse models
of NSCLC, the activation of an oncogenic allele of K-ras is sufficient to initiate the tumorigenesis
process, while the additional deletion or point mutation of p53 significantly enhances tumor
progression, leading to a more rapid development of adenocarcinomas that have features of a
more advanced disease. K-ras and p53 mutant tumors exhibit a greater incidence of cellular
and nuclear pleomorphism, desmoplasia, and a high frequency of metastases to the mediastinal
lymph nodes and the pleural spaces of the thoracic cavity, and less frequently to the liver and
kidneys.
Table 1: Choosing a tumor model
K-rasLSL-G12D/+ (K)
K-rasLSL-G12D/+;p53fl/fl (KP)
Limited tumor progression Rapid tumor progression
(modifier screens)
No metastases
Metastatic disease
Local: draining lymph node, pleural cavity
Distant: liver, kidney
Long survival
Short survival
Single allele
Multiple alleles
(easy to breed)
Small tumors
Large tumors
(limited tumor tissue)
(more tumor tissue for DNA, RNA, protein)
(easier to track with live imaging, microCT)
234
Cre System
To generate tumors sporadically in the lungs of K-rasLSL-G12D/+ mice, our laboratory initially
used replication-deficient adenoviruses expressing Cre (Ad-Cre) to deliver transient Cre
expression to infected cells of the lung (Jackson et al. 2001; Meuwissen et al. 2001). Prior to
administration, Ad-Cre is precipitated with calcium phosphate (see procedure) to improve the
delivery of Cre by increasing the efficiency of viral infection of the lung epithelium (Fasbender et
al. 1998). Recently, we have used lentiviruses to deliver Cre to the lung (Lenti-Cre) (Kumar et
al. 2008). Lentiviruses are beneficial because they integrate into the genome of infected cells
(Naldini et al. 1996), allowing for further modification of the tumors by simultaneously
introducing Cre recombinase and stable expression of cDNAs to overexpress, or short-hairpin
RNAs to silence, genes of interest (Table 2).
In order to control for the number of tumors generated, viruses are titered prior to use in
experiments. Ad-Cre is titered at the University of Iowa, while Lenti-Cre is titered in our
laboratory by assaying for Cre activity after infection of the 3TZ reporter cell line, a mouse
fibroblast cell line modified to express β-galactosidase after Cre-mediated recombination
(Psarras et al. 2004). We infect mice with 2.5x107 infectious particles of Ad-Cre (titered at Univ.
Iowa) or approximately 104-105 infectious particles of Lenti-Cre (3TZ titered in-house). It is
important to note that Ad-Cre and Lenti-Cre are titered differently, and as a result, the titers
cannot be directly compared.
Table 2: Choosing the Cre system
Adenovirus Cre (Ad-Cre)
High, reproducible titer
(typically produces a greater number of
tumors, shorter survival)
Titered virus is commercially available
Cannot introduce cDNAs or shRNAs with
Cre recombinase
Lentivirus Cre (Lenti-Cre)
Variable titer
In-house production may be required
Potential to modify tumors by introduction of
cDNAs or shRNAs in lentivirus
(gain-of-function or loss-of-function experiments)
235
Viral Administration Method
While we initially delivered Cre to the lungs of anesthetized mice using intranasal (IN)
instillation of the virus, we now prefer to deliver the virus to the lungs by intratracheal (IT)
intubation. IT delivery provides the most direct and consistent method for the virus to reach the
lungs. Reproducible delivery of the virus is critical because it directly affects the number of
tumors generated in the mice. However, intratracheal intubation requires additional equipment
and practice to perform it correctly and in a timely manner (Table 3). Therefore, it may be easier
to begin with the IN delivery method to assess the tumor model and practice the IT method
while continuing to breed animals for future experiments.
Table 3: Choosing the viral administration method
Intranasal (IN)
No training required
(can implement rapidly)
No additional equipment necessary
Indirect and variable delivery of virus to lungs
• virus displaced in nasal passage, mouth or
esophagus/stomach
• low penetrance sinonasal adenocarcinomas in KP
infected mice (Jackson et al. 2005)
Intratracheal (IT)
Technique requires practice
for proficiency
Requires catheters, light
source, platform
Direct delivery of virus to
lungs
Anesthesia
We recommend using avertin (2-2-2 Tribromoethanol) to anesthetize the mice. The amount of
avertin administered to the mice is crucial to the success of the procedure. Mice administered
too much avertin are more likely to stop breathing during the infection procedure, and recover
poorly from the anesthesia. Conversely, mice administered too little avertin may struggle to
inhale the virus and should be given more avertin before continuing. Therefore, we recommend
using the smallest volume of avertin required to keep mice anesthetized during the procedure.
236
Following the procedure, mice will recover better if they are kept warm to maintain their normal
body temperature after anesthesia.
Age of Mice
Our laboratory has utilized mice between 6 and 12 weeks of age for tumor initiation by IN or IT
delivery of viruses expressing Cre. Mice of this age are old enough to recover from the
anesthetic, the volume of virus administered to the lung, and the intubation of the trachea with
the catheter.
Volume of virus
Mice can be infected with a volume ranging from 50-125 µl per mouse, but we recommend
using a total volume of 75 µl per mouse. Although a volume of 125 µl can be administered, it is
not recommended for very young mice (6 weeks of age or younger). If 125 µl is administered,
then the mice should receive two doses of 62.5 µl each, with a 1-5 minute break in between the
doses, to allow the mice to recover a normal breathing pattern before receiving the second
dose.
MATERIALS
REAGENTS
CRITICAL The following list of reagents represents our laboratory’s preference. All reagents,
with the exception of Ad-Cre, can be modified according to investigator preference.
•
Mice: LSL-KrasG12D (Mouse Models of Human Cancers Consortium (MMHCC) Strain
01XJ6, Jackson Laboratory #008179 (B6), #008180 (129)); p53fl (MMHCC Strain 01XC2,
Jackson Laboratory #008462 (B6)); p53 LSL-R270H (MMHCC Strain 01XM3, Jackson
Laboratory #008651 (129svj)); p53LSL-R172H (MMHCC Strain 01XM2, Jackson Laboratory
237
# 008652 (129svj)). Researchers in industry can obtain these mice by contacting the MIT
Technology Licensing Office (http://web.mit.edu/tlo/www/industry/).
CAUTION All experiments should be done in accordance with protocols approved by the
Institutional Animal Care and Use Committee.
•
K-rasLSL-G12D (LSL K-ras G12D) and p53 mutant genomic targeting vectors are available
from Addgene (http://www.addgene.org/pgvec1?f=c&cmd=showcol&colid=171). Koch
Institute Core Facilities can perform chemotherapeutic testing and small animal imaging
(http://web.mit.edu/ki/facilities/core.html)
CAUTION Mice should be kept under specific pathogen-free conditions during
experiments.
Protocols to genotype the mouse strains listed above can be found on the Jacks Lab
website (http://web.mit.edu/jacks-lab/protocols_table.html)
•
Protocols to determine the recombination efficiency of the KrasLSL-G12D and p53fl alleles
can be found on the Jacks Lab website (http://web.mit.edu/jackslab/protocols_table.html)
•
2-2-2 Tribromoethanol (Avertin, Sigma Aldrich T48402)
•
2-methyl-2-butanol (Tert-amyl alcohol, Sigma Aldrich 152463)
•
Phosphate Buffered Saline (PBS)
•
Minimal Essential Media (MEM, Sigma, M4655)
•
CaCl2 (Mallinckrodt, Catalog #4160)
•
Bleach
•
Adenovirus-Cre (University of Iowa, Gene Transfer Vector Core,
http://www.uiowa.edu/~gene/; Jacks Lab website http://web.mit.edu/jackslab/protocols/AdenovirusCre.html)
238
CAUTION Preparation and administration of viruses should occur in a biosafety hood
and follow all guidelines for biosafety level 2 research.
•
Lentivirus-Cre (Three plasmid transfection system of CMV-VSV-G (Addgene plasmid
8454), Δ8.2 (gag/pol) (Addgene plasmid 8455), and transfer vector expressing Cre
(modified from Addgene plasmid 17408)) (Naldini et al. 1996; Pfeifer et al. 2001; Lois et
al. 2002; Stewart et al. 2003; Tiscornia et al. 2006; Kumar et al. 2008)
CAUTION Preparation and administration of viruses should occur in a biosafety hood
and follow all guidelines for biosafety level 2 research. Research with lentiviruses
pseudotyped with VSV-G containing known or putative oncogenes or short hairpin RNAs
to silence tumor suppressor genes should abide with all institute safety practices.
EQUIPMENT
•
Bottle top vacuum filter (.22 µM, 150 ml, Corning, cat. no. 430624)
•
Scale (Ohaus, cat. no. HH120D)
•
Needles (30 gauge, ½ inch, Becton Dickinson, cat. no. 305106)
•
Syringes (1 ml, Becton Dickinson, cat. no. 309602)
•
Flat forceps (Roboz, cat. no. RS-8260)
•
Exel Safelet IV catheters (22 gauge, 1 inch, Fisher, cat. no. 14-841-20)
•
Intubation platform (Steve Boukedes, labinventions@gmail.com)
•
Fiber-Lite Illuminator (Dolan-Jenner Industries, Inc., Model 3100-1)
•
Heat lamp (or latex gloves filled with warm water)
REAGENT SETUP
Avertin Stock (1.6 g ml-1) Add 15.5 ml tert-amyl alcohol to 25 g of avertin. Stir overnight to
dissolve. Stable at room temperature (18-25 °C) for approximately 1 year.
239
CRITICAL Seal bottle tightly and protect from light. Discard the solution if it yellows.
Avertin Working Solution (20 mg ml-1) Dilute avertin stock in PBS, stir overnight, and protect
from light. Sterilize by passing solution through a 0.22 µm filter. Aliquots may be safely stored at
4 °C in the dark for approximately 4 months.
2 M CaCl2 Dissolve in distilled water and store at 4 °C for up to 5 years.
Virus Ad-Cre should be prepared fresh each time and used within one hour of preparation. For
extended storage times, Lenti-Cre should be stored at -80 °C or alternatively, at 4 °C for periods
of a few days. Keep viruses on ice prior to infection.
EQUIPMENT SET-UP
Biosafety hood To set-up for either IN or IT administrations, arrange the virus, a heat lamp (or
gloves filled with warm water), and a beaker filled with 50% bleach (to disinfect catheters and
pipette tips that have contacted the virus) in a biosafety hood.
Intubation platform and light source For IT administration, set up the platform and light
source on a flat surface near the biosafety hood (Fig. 1a). Insert the catheter into the trachea
outside of the hood, and then move the mouse into the biosafety hood to inhale the virus. A
sharps waste container is also required for the proper disposal of the needles from the Exel
Safelet IV Catheter.
PROCEDURE
Virus Preparation TIMING 30 min
1 For infection with Lenti-Cre, proceed to step 5. For Ad-Cre, prepare the adenovirus-precipitate
mixture. Determine the number of mice that you will infect and pipette the appropriate volume of
MEM into a tube. Refer to the Experimental Design Section to determine the volume of virus to
administer per mouse.
240
CAUTION Preparation and administration of viruses should occur in a biosafety hood and follow
all guidelines for biosafety level 2 research. Research with lentiviruses pseudotyped with VSV-G
containing known or putative oncogenes or short hairpin RNAs to silence tumor suppressor
genes should abide with all institute safety practices.
CAUTION All experiments should be done in accordance with protocols approved by the
Institutional Animal Care and Use Committee.
CRITICAL STEP The adenovirus-precipitate mixture should be used within one hour of
preparation. If many mice will be infected, precipitates should be prepared sequentially.
2 After the adenovirus thaws on ice, pipette the adenovirus directly into the MEM to obtain a
final titer of 2.5x107 PFU per mouse. Flick the tube to mix.
CRITICAL STEP Do not freeze/thaw the adenovirus as this reduces the titer of the virus 10 fold
with each thaw.
3 Add 2M CaCl2 into the MEM mix to obtain a final CaCl2 concentration of 10 mM. Flick the tube
to mix.
4 Incubate at room temperature for 20 min to allow for the formation of calcium phosphate
precipitates.
Delivery of Ad-Cre or Lenti-Cre using the intratracheal or intranasal infection method
5 Avertin Administration TIMING 5 min per mouse
Anesthetize mice via intra-peritoneal injection of room temperature 20 mg ml-1 avertin (use 0.4
mg g-1 body weight for females and 0.45 mg
-1
body weight for males). Confirm the mice are fully
anesthetized by ensuring that they lack a toe reflex.
CRITICAL STEP Administering the correct amount of avertin is crucial to successfully delivering
the virus.
TROUBLESHOOTING
241
6 Delivery of Ad-Cre or Lenti-Cre can be carried out using Option A the intratracheal infection
method or Option B the intranasal infection method
A. Intratracheal infection method TIMING 1-5 min per mouse
i.
Place mouse on the platform so that it is hanging from its top front teeth on
the bar (Fig. 1b, c).
ii. Push the mouse towards the bar so that the chest is vertical underneath the
bar (perpendicular to the platform) (Fig. 1b).
iii. Direct the Fiber-Lite Illuminator to shine on the mouse’s chest, in between the
front legs (Fig. 1c).
iv. Prepare the Exel Safelet IV catheter for the infection procedure. To ensure
that the needle does not become exposed and impale the mouse, hold the
square part of the needle with one’s thumb and index finger, and using one’s
middle finger, push the catheter over the end of the needle completely and
continue to hold the catheter in place during the infection protocol (Supp. Fig.
2 a, b).
TROUBLESHOOTING
v. Using the Exel Safelet IV catheter, open the mouth and gently pull out the
tongue with the flat forceps (Fig. 1d).
vi. Locate the opening of the trachea by peering into the mouth and looking for
the white light emitted from the trachea (Fig. 1e).
TROUBLESHOOTING
vii. While holding the Exel Safelet IV catheter vertically, position the catheter over
the white light emitted from the opening of the trachea, and allow the catheter
to slide into the trachea until the top of the catheter reaches the mouse’s front
242
teeth (Fig. 1f). There should be no resistance while inserting the catheter into
the trachea.
viii. While stabilizing the Exel Safelet IV catheter with one hand, remove the
needle from the mouth (Fig. 1g).
CRITICAL STEP Prior to removing the needle, the mouse cannot breathe
through the catheter. Once the Exel Safelet IV catheter has been inserted
into the trachea, promptly remove the needle to allow the mouse to breathe
through the catheter.
ix. The proper placement of the catheter in the trachea can be confirmed by
visualizing the white light shining through the opening of the catheter in the
mouth (Fig. 1h).
x. Move the platform, mouse, and catheter into the biosafety hood.
xi. Pipette the virus directly into the opening of the catheter to ensure the entire
volume is inhaled (Fig. 1i).
xii. If the catheter is correctly inserted into the trachea, the mouse will begin
inhaling the virus immediately. Once the virus is no longer visible in the
opening of the catheter, wait a few seconds for the entire volume to travel
down the catheter before removing the catheter from the trachea and
disposing of it in 50% bleach.
TROUBLESHOOTING
B. Delivery of Ad-Cre or Lenti-Cre using the intranasal delivery method TIMING 2-5 min
per mouse
i.
In the biosafety hood, lay the mouse in your hand, ventral side up. Tilt the
mouse so that the head is positioned above its feet (Fig. 2a).
243
CRITICAL STEP Do not grasp the mouse tightly as this will inhibit the
mouse’s breathing
ii. Hold the end of a pipet tip over the opening of one nostril and dispense the
virus dropwise until the entire volume of virus has been inhaled (Fig. 2b, c).
CRITICAL STEP Do not attempt to insert the pipet tip into the nostril.
TROUBLESHOOTING
Animal Recovery TIMING 10-15 min
7 Place the mouse under a heat lamp (Supp. Fig. 3a) or on a latex glove filled with warm water
(Supp. Fig. 3b) to recover in the biosafety hood.
TROUBLESHOOTING
TIMING
Steps 1-4 Virus Preparation (Ad-Cre only): 30 min
Step 5 Avertin Administration: 5 min per mouse
Step 6A Intratracheal Infection: 1-5 min per mouse after an initial training phase
Step 6B Intranasal Delivery: 2-5 min per mouse
Step 7 Animal Recovery: 10-15 min
TROUBLESHOOTING
Troubleshooting advice can be found in Table 4.
244
Table 4: Troubleshooting advice
Step
Problem
Number
Mouse is not falling
asleep
5
6Aiv
6Avi
6Axii
6B
7
Mouse is overanesthetized and
breathing very slowly
Blood appears in
mouth
Cannot visualize the
white light
Mouse does not
inhale the virus (virus
stays in the catheter)
Mouse is coughing
and sputtering virus
Mouse does not
recover well following
anesthesia
Possible Reason
Not enough avertin
administered
Too much avertin
administered
Catheter has slipped
and exposed the
needle
Light is not in the
correct position
Not looking at the
ventral surface of
the throat
Saliva is covering
the opening of the
trachea
Catheter inserted
into esophagus
Too much virus
placed on nostril
before being inhaled
Virus dispensed into
both nostrils
Mouse is too cold
Mouse is too hot
under heat lamp
Solution
Administer more avertin, 50-100 µl at
a time
Wait until breathing becomes more
regular but ensure the mouse still
lacks a reflex response before
attempting to infect the mouse
Make sure that the catheter is held in
place to cover the needle correctly
(Supp. Fig. 2a, b)
Re-direct the light on the upper chest
Lean further over the mouth and
push the tongue with the catheter
towards the ventral surface of the
throat
Gently probe at the back of the throat
with the Exel Safelet IV catheter to
expose the trachea
Pipette the virus out of the catheter
for reuse, dispose of the catheter in
50% bleach, and begin the procedure
again at step 6Aiv with a new
catheter
Let coughing subside before
continuing. Pipette the virus resting
on nostril for reuse
Only dispense virus into one nostril
Mouse should be kept warm until it
starts to wake up
Turn off heat lamp temporarily or let
mouse recover on a glove filled with
warm water
ANTICIPATED RESULTS
We have reported that recombination of K-rasLSL-G12D and p53fl alleles in tumors can be
examined by PCR for the presence of a “1 lox”, or recombined, product (see Reagents in
Materials section for the protocols) (Jackson et al. 2001). Although it is possible to perform PCR
on DNA isolated from whole lung after infection to assess infection efficiency, this is not
recommended. Typically very few cells in the lung have undergone recombination of these
245
alleles, making it difficult to detect recombination. Instead, it is more informative to examine Cre
expression after infection by using conditional reporter strains such as Rosa26LSL-LacZ or
Rosa26LSL-eGFP and examining reporter expression by immunohistochemistry,
immunofluorescence, or fluorescent activated cell sorting (Soriano 1999). Polyclonal antibodies
that can specifically detect the oncogenic K-rasG12D protein are no longer available; however,
increased Ras-GTP levels in tumors or cells can be assessed using a Raf-GST pulldown assay
(Tuveson et al. 2004).
The time to tumor development and progression will vary depending upon the model chosen
(K v. KP), while survival time will also depend on the amount and type of virus administered to
the mice. The survival of mice is reduced approximately twofold in KP mice compared to K mice
(median survival with Ad-Cre: K, 185 days; KP, 76 days, with Lenti-Cre: K, 266 days; KP, 170
days) (Fig. 3a, b). Decreased survival is due to a greater growth rate of tumors lacking p53,
leading to the more rapid development of a tumor burden that disrupts the normal function of the
lungs. Reduced survival after p53 loss is not due to an increased number of tumors or
metastatic disease. Survival of mice is also reduced after infection with Ad-Cre as compared to
Lenti-Cre (Fig. 3a, b). This is due to the higher titer of virus typically administered to mice with
Ad-Cre but may also reflect the viral tropism or the efficiency of the virus to infect the cell of
origin of the disease. Mice infected with 2.5x107 infectious particles of Ad-Cre (Univ. Iowa, Gene
Transfer Vector Core) can generate greater than 200 tumors per mouse, whereas mice infected
with roughly 104-105 infectious particles of Lenti-Cre (3TZ titered) can generate 10-100 tumors
per mouse. Our laboratory has had success titrating the viruses to lower levels (Ad-Cre: 5x106
infectious particles/mouse, Lenti-Cre: 5x103 infectious particles/mouse) which reduce the
number of primary tumors and increase the survival time of mice, allowing for a greater
frequency of metastatic disease in the KP model.
246
Monitoring tumor development and progression histologically is an important, though
difficult, way to follow the disease, and it is especially useful in experiments where genetic
events or modifications in addition to K-ras activation and/or p53 loss may be expected to
impact tumor progression. We stress that the NSCLC model generates a multi-focal disease
and therefore, investigators should expect some tumor heterogeneity; tumors do not always
progress in exactly the same way or at the same time. To follow these changes objectively, we
have employed a 4-stage grading system for tumor progression in our NSCLC models (adapted
from ref. 18). The earliest lesions, designated as grade 1, are atypical adenomatous
hyperpasias (AAH) or small adenomas that feature uniform nuclei and can appear as early as 23 weeks post-infection (Fig. 4a). These earliest lesions can only be identified through careful
histological analysis. To see visible lesions on the surface of the lung, investigators should wait
until 6-8 weeks after tumor initiation or later. Grade 2 tumors are larger adenomas that have
slightly enlarged nuclei with prominent nucleoli and are observed 6-8 weeks post-infection (Fig.
4b, c). Adenocarcinomas are classified as grade 3; they have a great degree of cellular
pleomorphism and nuclear atypia and can develop as early as 16 weeks post-infection (Fig. 4d,
e). Grade 4 tumors are invasive adenocarcinomas (Fig. 4f) that harbor all the cellular
characteristics of Grade 3 tumors but with a higher mitotic index - including irregular mitoses, a
distinctive highly invasive stromal reaction (desmoplasia) (Fig. 4g), and invasive edges
bordering lymphatic vessels, blood vessels, or the pleura (Fig. 4h). Grade 4 tumors may
develop as early as 18 weeks post-infection in KP animals, but are not observed in K animals.
Finally, locally metastatic disease to the mediastinal lymph nodes (Fig. 4i) or the pleural cavity
develops in approximately 50% of KP mice as early as 18-20 weeks post-infection. In some
mice, distant metastases can be found seeding the liver or the kidneys as early as 20 weeks
post-infection (Fig. 4j). Although the time to progression in K and KP mice is described here to
be similar up to Grade 3 lesions, mice with p53 deficient tumors often harbor cells that exhibit
247
nuclear atypia at very early time points after tumor initiation. In addition, a greater proportion of
tumors that lack p53 progress to higher grades during tumor development (Jackson et al. 2005).
For example, at 6 weeks after infection the distribution of tumor grades in the K model is ~90%
grade 1 and ~10 grade 2, whereas in the KP model it is ~40% grade 1, ~40% grade 2 and
~20% grade 3 (Jackson et al. 2005). At 19 weeks after infection in the KP model, the tumor
distribution is ~5% grade 1, ~20% grade 2, ~70% grade 3 and ~5% grade 4 (Jackson et al.
2005). At 26 weeks after infection in the K model, ~30% of tumors are grade 1, ~40% are grade
2 and ~30% are grade 3 (Jackson et al. 2005). However, as with most autochthonous mouse
tumor models, there is some variation in the results, such as tumor number and approximate
time to progression, depending on the strain/ background of mice as well as other factors that
vary between institutions.
248
Figure 1. Intratracheal infection technique. Anesthetized mice are placed on the platform by
their front teeth so that their chest hangs vertically beneath them (a, b). The light is directed on
the mouse’s upper chest (a, c), on the spot marked by the ‘X’ (c). The mouth is opened using
the Exel Safelet IV catheter (d), and the tongue is gently pulled out using the flat forceps. After
locating the white light emitted from the trachea (e), the Exel Safelet IV catheter is slid into the
trachea (f), and the needle is removed (g). The mouse with the inserted catheter (h) on the
platform is moved into a biosafety hood, where the virus is dispensed into the opening of the
catheter (i).
CAUTION All experiments should be done in accordance with protocols approved by the
Institutional Animal Care and Use Committee.
249
Figure 2. Intranasal infection technique. The anesthetized mouse lies gently in the hand of
the investigator (a), and the virus is administered dropwise (b) into one nostril until the virus is
completely inhaled (b, c).
CAUTION All experiments should be done in accordance with protocols approved by the
Institutional Animal Care and Use Committee.
250
Figure 3. Survival is reduced in KP model compared to K model. Kaplan-Meier plot of KP
mice (median survival 76 days) and K mice (median survival 185 days) infected with 2.5x107
PFU of Ad-Cre per mouse (provided by T. Oliver) (a). Kaplan-Meier plot of KP mice (median
survival 170 days) and K mice (median survival 266 days) infected with 105 Lenti-Cre viruses
per mouse (provided by P. Sandy and M. DuPage) (b).
CAUTION All experiments should be done in accordance with protocols approved by the
Institutional Animal Care and Use Committee.
251
Figure 4. Tumor progression and histopathological phenotype in KP model. Haematoxylin
and eosin stained (H&E) tumors from KP mice at various times after infection with Lenti-Cre: (a)
Grade 1 lesion of an AAH progressing to a small adenoma. (b) Grade 2 adenoma. (c) Uniform
nuclei in a grade 2 adenoma. (d) Pleomorphic nuclei in a Grade 3 adenocarcinoma. (e) Grade 3
adenocarcinoma displaying mixed cellular phenotypes. (f) Grade 4 invasive adenocarcinoma.
(g) Grade 4 adenocarcinoma with glandular/acinar architecture and desmoplasia. (h)
Adenocarcinoma from the lung (L) invading across the mesothelium into the pleural cavity (P).
(i) Local lung tumor metastasis (T) to mediastinal lymph node (LN). (j) Distant lung tumor
metastasis (T) to the kidney (K). Panels a, b, e, f, i and j are at 100X magnification. Panels c, d,
g and h are at 400X magnification. Scale bar = 100 µm.
CAUTION All experiments should be done in accordance with protocols approved by the
Institutional Animal Care and Use Committee.
252
!"#$%
!"#$%
!"
"#$
$%
!%
-+./*010+!"#$%
-+./*23%
!%
!
9:&%
-+./*8('+!"#$%
-+./*23%
4567%
456!"
4567%
456!"
&'()*%#+",%
&'()*
# ",
Supplementary Figure 1. Genetically controlled events in a mouse model of lung
cancer. Engineering of LoxP elements and a stop element allow for the controlled
expression of oncogenic K-ras and the loss of p53 function after Cre expression.
253
Supplementary Figure 2. Preparation of the Exel Safelet IV catheter for intratracheal
infection. Upon opening the Exel Safelet IV catheter, the needle is exposed (a). Slide the
catheter over the end of the needle to completely cover the tip (b) and the Exel Safelet IV
catheter is now ready to use.
254
Supplementary Figure 3. Recovery following intranasal or intratracheal infection. Mice
can be placed under a heat lamp (a) or on a glove filled with warm water (b) to recover following
anesthesia in the biosafety hood.
CAUTION All experiments should be done in accordance with protocols approved by the
Institutional Animal Care and Use Committee.
255
ACKNOWLEDGEMENTS
We would like to thank Carla Kim and Amber Woolfenden for originally training the authors to
perform the intratracheal intubation technique. The technique was implemented in our lab with
help from Kwok-Kin Wong and Samanthi Perera. We would like to thank Trudy Oliver and Peter
Sandy for providing data to compile the Kaplan-Meier survival curves as well as Denise Crowley
and Roderick Bronson for providing key histological and pathological advice. We also thank
Keara Lane, Etienne Meylan, Eric Snyder, and Anne Deconinck for reviewing this protocol. This
work was supported by funding from the Howard Hughes Medical Institute, the NCI (including a
Cancer Center Support grant), and the Ludwig Center for Molecular Oncology at MIT. T.J. is the
David H. Koch Professor of Biology and a Daniel K. Ludwig Scholar. Research was conducted
in compliance with the Animal Welfare Act Regulations and other Federal statutes relating to
animals and experiments involving animals and adheres to the principles set forth in the Guide
for Care and Use of Laboratory Animals, National Research Council, 1996.
256
REFERENCES
Daly, H. E., Baecher-Allan, C. M., Barth, R. K., D'Angio, C. T. and Finkelstein, J. N. (1997).
Bleomycin Induces Strain-Dependent Alterations in the Pattern of Epithelial Cell-Specific Marker
Expression in Mouse Lung. Toxicology and Applied Pharmacology 142(2): 303-310.
Dankort, D., Filenova, E., Collado, M., Serrano, M., Jones, K. and McMahon, M. (2007). A new
mouse model to explore the initiation, progression, and therapy of BRAFV600E-induced lung
tumors. Genes & Development 21(4): 379-384.
Fasbender, A., Lee, J. H., Walters, R. W., Moninger, T. O., Zabner, J. and Welsh, M. J. (1998).
Incorporation of adenovirus in calcium phosphate precipitates enhances gene transfer to airway
epithelia in vitro and in vivo. The Journal of Clinical Investigation 102(1): 184-193.
Fisher, G. H., Wellen, S. L., Klimstra, D., Lenczowski, J. M., Tichelaar, J. W., Lizak, M. J.,
Whitsett, J. A., Koretsky, A. and Varmus, H. E. (2001). Induction and apoptotic regression of
lung adenocarcinomas by regulation of a K-Ras transgene in the presence and absence of
tumor suppressor genes. Genes & Development 15(24): 3249-3262.
Frese, K. K. and Tuveson, D. A. (2007). Maximizing mouse cancer models. Nat Rev Cancer
7(9): 654-658.
Ge, Q., Filip, L., Bai, A., Nguyen, T., Eisen, H. N. and Chen, J. (2004). Inhibition of influenza
virus production in virus-infected mice by RNA interference. Proceedings of the National
Academy of Sciences of the United States of America 101(23): 8676-8681.
Herbst, R. S., Heymach, J. V. and Lippman, S. M. (2009). Lung Cancer. New England Journal
of Medicine 359(13): 1367-1380.
Jackson, E. L., Olive, K. P., Tuveson, D. A., Bronson, R., Crowley, D., Brown, M. and Jacks, T.
(2005). The differential effects of mutant p53 alleles on advanced murine lung cancer. Cancer
Res 65(22): 10280-8.
Jackson, E. L., Olive, K. P., Tuveson, D. A., Bronson, R., Crowley, D., Brown, M. and Jacks, T.
(2005). The Differential Effects of Mutant p53 Alleles on Advanced Murine Lung Cancer. Cancer
Res 65(22): 10280-10288.
Jackson, E. L., Willis, N., Mercer, K., Bronson, R. T., Crowley, D., Montoya, R., Jacks, T. and
Tuveson, D. A. (2001). Analysis of lung tumor initiation and progression using conditional
expression of oncogenic K-ras. Genes Dev. 15(24): 3243-3248.
257
Jemal, A., Siegel, R., Ward, E., Hao, Y., Xu, J., Murray, T. and Thun, M. J. (2008). Cancer
Statistics, 2008. CA: A Cancer Journal for Clinicians 58(2): 71-96.
Ji, H., Li, D., Chen, L., Shimamura, T., Kobayashi, S., McNamara, K., Mahmood, U., Mitchell,
A., Sun, Y., Al-Hashem, R., et al. (2006). The impact of human EGFR kinase domain mutations
on lung tumorigenesis and in vivo sensitivity to EGFR-targeted therapies. Cancer Cell 9(6): 485495.
Ji, H., Ramsey, M. R., Hayes, D. N., Fan, C., McNamara, K., Kozlowski, P., Torrice, C., Wu, M.
C., Shimamura, T., Perera, S. A., et al. (2007). LKB1 modulates lung cancer differentiation and
metastasis. Nature 448(7155): 807-810.
Johnson, L., Greenbaum, D., Cichowski, K., Mercer, K., Murphy, E., Schmitt, E., Bronson, R. T.,
Umanoff, H., Edelmann, W., Kucherlapati, R., et al. (1997). K-ras is an essential gene in the
mouse with partial functional overlap with N-ras. Genes & Development 11(19): 2468-2481.
Johnson, L., Mercer, K., Greenbaum, D., Bronson, R. T., Crowley, D., Tuveson, D. A. and
Jacks, T. (2001). Somatic activation of the K-ras oncogene causes early onset lung cancer in
mice. Nature 410(6832): 1111-1116.
Jonkers, J., Meuwissen, R., van der Gulden, H., Peterse, H., van der Valk, M. and Berns, A.
(2001). Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model
for breast cancer. Nat Genet 29(4): 418-425.
Kumar, M. S., Erkeland, S. J., Pester, R. E., Chen, C. Y., Ebert, M. S., Sharp, P. A. and Jacks,
T. (2008). Suppression of non-small cell lung tumor development by the let-7 microRNA family.
Proceedings of the National Academy of Sciences 105(10): 3903-3908.
Lois, C., Hong, E. J., Pease, S., Brown, E. J. and Baltimore, D. (2002). Germline transmission
and tissue-specific expression of transgenes delivered by lentiviral vectors. Science 295(5556):
868-872.
Marumoto, T., Tashiro, A., Friedmann-Morvinski, D., Scadeng, M., Soda, Y., Gage, F. H. and
Verma, I. M. (2009). Development of a novel mouse glioma model using lentiviral vectors. Nat
Med 15(1): 110-116.
Meuwissen, R. and Berns, A. (2005). Mouse models for human lung cancer. Genes Dev. 19(6):
643-664.
258
Meuwissen, R., Linn, S. C., Linnoila, R. I., Zevenhoven, J., Mooi, W. J. and Berns, A. (2003).
Induction of small cell lung cancer by somatic inactivation of both Trp53 and Rb1 in a
conditional mouse model. Cancer Cell 4(3): 181-189.
Meuwissen, R., Linn, S. C., van der Valk, M., Mooi, W. J. and Berns, A. (2001). Mouse model
for lung tumorigenesis through Cre/lox controlled sporadic activation of the K-Ras oncogene.
Oncogene 20(45): 6551-6558.
Naldini, L., Blomer, U., Gallay, P., Ory, D., Mulligan, R., Gage, F. H., Verma, I. M. and Trono, D.
(1996). In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector.
Science 272(5259): 263-267.
Olive, K. P., Tuveson, D. A., Ruhe, Z., Yin, B., Willis, N., Bronson, R., Crowley, D. and Jacks, T.
(2004). Mutant p53 Gain of Function in Two Mouse Models of Li-Fraumeni Syndrome. Cell
119(6): 847-860.
Pfeifer, A., Brandon, E. P., Kootstra, N., Gage, F. H. and Verma, I. M. (2001). Delivery of the
Cre recombinase by a self-deleting lentiviral vector: Efficient gene targeting in vivo. Proceedings
of the National Academy of Sciences of the United States of America 98(20): 11450-11455.
Psarras, S., Karagianni, N., Kellendonk, C., Tronche, F., Cosset, F. L., Stocking, C.,
Schirrmacher, V., Boehmer, H. v. and Khazaie, K. (2004). Gene transfer and genetic
modification of embryonic stem cells by Cre- and Cre-PR-expressing MESV-based retroviral
vectors. The Journal of Gene Medicine 6(1): 32-42.
Shaw, A. T., Kirsch, D. G. and Jacks, T. (2005). Future of Early Detection of Lung Cancer: The
Role of Mouse Models. Clinical Cancer Research 11(13): 4999s-5003s.
Shen, C.-H., Ge, Q., Talay, O., Eisen, H. N., Garcia-Sastre, A. and Chen, J. (2008). Loss of IL7R and IL-15R Expression Is Associated with Disappearance of Memory T Cells in Respiratory
Tract following Influenza Infection. The Journal of Immunology 180(1): 171-178.
Soriano, P. (1999). Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat
Genet 21(1): 70-71.
Stewart, S. A., Dykxhoorn, D. M., Palliser, D., Mizuno, H., Yu, E. Y., An, D. S., Sabatini, D. M.,
Chen, I. S. Y., Hahn, W. C., Sharp, P. A., et al. (2003). Lentivirus-delivered stable gene
silencing by RNAi in primary cells. RNA 9(4): 493-501.
Sun, S., Schiller, J. H. and Gazdar, A. F. (2007). Lung cancer in never smokers - a different
disease. Nat Rev Cancer 7(10): 778-790.
259
Tiscornia, G., Singer, O. and Verma, I. M. (2006). Production and purification of lentiviral
vectors. Nat. Protocols 1(1): 241-245.
Tuveson, D. A., Shaw, A. T., Willis, N., Silver, D., Jackson, E. L., Chang, S., Mercer, K.,
Grochow, R., Hock, H., Crowley, D., et al. (2004). Endogenous oncogenic K-rasG12D stimulates
proliferation and widespread neoplastic and developmental defects. Cancer Cell 5(4): 375-387.
Van Dyke, T. and Jacks, T. (2002). Cancer Modeling in the Modern Era. Cell 108(2): 135-144.
260