Engineering of HIV gp120 by Yeast Surface Display for
Neutralizing Antibody Characterization and Immunogen Design
ARCHTV
by
Jordi Mata-Fink
B.S., M.S. Chemical Engineering
Stanford University, 2005
SUBMITTED TO THE DEPARTMENT OF CHEMICAL ENGINEERING IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
AT THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
FEBRUARY 2013
© Massachusetts Institute of Technology. All rights reserved.
Signature of Author:
Certified by:
Dep*ment of Chemical Engineering
December 18, 2012
K. Dane Wittrup
C.P. Dubbs Professor of Chemical Engineering and Biological Engineering
Thesis Supervisor
Accepted by:
*
1
1~
F~
1
ratrick :. uoyle
Professor of Chemical Engineering
Chairman, Committee for Graduate Students
2
Engineering of HIV gp120 by Yeast Surface Display for
Neutralizing Antibody Characterization and Immunogen Design
by
Jordi Mata-Fink
Submitted to the Department of Chemical Engineering on December 19, 2012
in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy in Chemical Engineering
ABSTRACT
The sequence diversity of glycoprotein gp120 of the envelope spike of Human
Immunodeficiency Virus (HIV) allows the virus to escape from antibody
selection pressure. Certain conserved epitopes, like the CD4 binding site, are
required for viral fitness and antibodies against these epitopes are able to
neutralize HIV from multiple clades. Passive immunization experiments suggest
that eliciting such broadly reactive antibodies by vaccination may provide
protection, but so far this has proven impossible.
In this thesis, we establish a yeast surface display system for the development of
gp120-based molecules for antibody characterization and immunogen design. A
stripped core gp120 is constructed that retains the correct presentation of the
CD4 binding site. Epitopes of several CD4 binding site-directed antibodies,
including the gold standard antibody VRC01, are mapped with yeast displayed
mutant libraries. A panel of immunogens that share the epitope defined by
VRC01 but are diverse elsewhere on their surfaces is designed. Mice immunized
sequentially with the diverse immunogens elicit an antibody response that is
focused entirely on the VRC01 epitope. The serum cross-reacts with gp120 from
multiple clades. Monoclonal antibodies from these mice are isolated and
characterized.
Thesis Supervisor: K. Dane Wittrup
Title: C.P. Dubbs Professor, Chemical Engineering and Biological Engineering
3
4
Acknowledgments
Stephen Jay Gould, in his book Wonderful Life, describes the history of Cambrian
fossils from the Burgess Shale in British Columbia, Canada. The soft-bodied
organisms were first classified into existing taxa as precursors on a neatly
progressing tree of life. Decades later, the fossils were re-examined and
determined to be organisms with no modem counterpart, evolutionary deadends that met their demise by chance rather than a fundamental unfitness.
This thesis began with a vague idea-apply protein engineering to vaccinesand grew to the specific work described in the following pages. The process was
neither linear nor predetermined. Along the way there were side branches that,
like the fossils of the Burgess Shale, petered out. Others, like the HIV project,
which appeared as an intellectual detour two years into the work, flourished.
The document that follows reflects not a finished product, but the current state of
an idea that has already evolved and will continue to do so.
I am grateful to many people who made this work possible. Foremost among
them, I am beholden to my advisor, Dane Wittrup, whose intellectual power,
creativity, patience, and guidance has shaped me both personally and
professionally. My thesis committee members, Darrell Irvine and Chris Love,
were generous with their labs, resources, and ideas. The Wittrup lab has been a
phenomenal place to call home for the past six years. I am indebted to lab
members past and present for advice, assistance, and friendship. In particular,
Shanshan Howland mentored me through the early learning curve; Ben Hackel
set a standard of productivity and effort that I will forever strive to meet; Jamie
Spangler and Chris Pirie shared with me many of the joys of grad school;
Seymour de Picciotto was a fountain of scientific curiosity; Cary Opel and Katie
Maass provided non-scientific intellectual stimulation. I was fortunate to work
with undergraduates-Barry Kriegsman, Hanna Zhu, Yu Huixin, Moses
Leavens, and Matt Glassman-who contributed significantly to the project. In the
Irvine lab, Melissa Hanson and Pete DeMuth were wonderful colleagues. In the
Love lab, Tim Politano and Brittany Thomas were generous with their time and
expertise. Dennis Burton, Laura Walker, and Emilia Falkowska helped us
establish our footing in a field in which we were complete newcomers.
Outside of lab, CJ Zopf, Joe Scott, and many classmates helped make daily life
interesting. Pam Grich and Lou Fink introduced me to the White Mountains and
provided warm meals and family. My sister, Ana, lent a patient ear to research
troubles. My parents, David and Marina, offered guidance when needed,
distance when wanted, and encouragement always. My wife, Rebecca, stuck
with me through it all. This thesis is for you. Thank you.
Jordi Mata-Fink
December 15, 2012
5
6
Table of Contents
Chapter 1. Introduction
Background......................................................................---Thesis Summary.........................................................................
9
11
Works Cited................................................................................12
Chapter 2. Yeast display of HIV gp120
Background..........................................................................
Results......................................................................................
Display of gp120 on yeast
Modifications to gp120 to make strippedcore
Binding of stripped core to anti-gp120 antibodies
Stability of stripped core
Discussion........................................................21
Materials & Methods...................................................................
...
Figures & Tables.....................................................................
Works Cited..............................................................................
Chapter 3. Epitope mapping of CD4 binding site-directed antibodies
Background..........................................................................
Results......................................................................................
Mapping the VRCO1 epitope with a random mutagenesis library
Making a defined mutant panel
Mapping the VRCO1 epitope with the mutant panel
Mapping other anti-gp120 antibodies
Multi-mutation constructs that bind differentially to VRCO1, b12, and b13
Discussion..............................................................................
Materials & Methods.......................................................................
..-Figures & Tables...................................................................
Works Cited................................................................................
17
18
22
25
31
37
39
43
46
49
62
Chapter 4. Eliciting VRC01-like antibodies by vaccination
Background................................................................................67
............- 70
Results....................................................................
Generationof diversified immunogens by semi-random mutagenesis
Characterizationof diversified immunogens
Immunization of mice with diversified immunogens
Characterizationof serum binding to individual immunogens
Serum binding to gp120 mutant D368R
Serum competition with VRCO1 by ternary complex assay
Serum binds to stripped core gp120from multiple clades
Isolationof monoclonal antibodies
7
Table of Contents (continued)
Discussion..................................................................................
Materials & Methods...................................................................
Figures & Tables...........................................................................83
Works Cited................................................................................
74
79
99
Chapter 5. Perspectives on immunizations
Background...................................................................................107
Results.........................................................................................
108
Immunization routes andformulations tested
Intranasaladministrationrequires both Fc and adjuvant
Pharmacokineticsof antigen administeredintranasally
Timing of boost mattersfor liposomalformulation
Conjugation is requiredfor liposomalformulations but not for other adjuvants
Immunization with diverse immunogens on liposomesfails to steer the immune
response to the CD4 binding site
Comparison of intradermalinjection to microneedle application
Differences in inflammation, retention, and biodistributionof antigen administered
intradermallyor by microneedle array
Steering the immune response to the VRC01 epitope using microneedles
Relationship between dose and KA across adjuvants,formulations, and routes
Discussion.....................................................................................114
Materials & Methods.......................................................................117
Figures & Tables.............................................................................119
Works Cited..................................................................................131
Chapter 6. Targeting DEC-205 to deliver antigen to the cross-presentation
pathway in dendritic cells
Background....................................................................................
133
Results..........................................................................................
134
Engineeringfibronectins to bind DEC-205
Conjugationof Fn3 clones to CpG DNA
Effect on bone marrow-deriveddendritic cells
Discussion....................................................................................
136
Materials & Methods......................................................................
137
Figures & Tables............................................................................
139
Works Cited..................................................................................144
8
Chapter 1. Introduction
Background
Broadly neutralizing antibodies are the best hope for an effective anti-HIV
vaccine (1, 2). Glycoprotein gp120 of the envelope spike of human
immunodeficiency virus (HIV) is the target of most antibodies generated upon
HIV infection (3, 4), but HIV's fast mutation rate and its tolerance of sequence
diversity allow the virus to escape from nearly all antibodies raised against
epitopes on the envelope spike (5-9). There are, however, regions of the viral
envelope spike that are essential for viral fitness and therefore are highly
conserved. Antibodies against these conserved epitopes that can neutralize HIV
across clades are known as broadly neutralizing antibodies (4). The CD4 binding
site is one such target for broadly neutralizing antibodies, conserved because
HIV from any strain must first dock onto CD4 in order to infect a T cell (10-14).
Antibodies exert selection pressure on autologous HIV almost immediately
following infection (8, 15, 16), and most patients eventually develop neutralizing
antibodies against conserved epitopes (6, 17-21) though these antibodies are
unable to control infection. Even individuals who generate the most potent of
broadly neutralizing antibodies over the course of infection do not demonstrate
any clinical benefit from having elicited them (22). The hypothesis is that by the
time broadly neutralizing antibodies are elicited-typically 2-3 years after
infection (3)-there is a sufficient reservoir of virus to find escape mutations
against any antibody. VRC01 is the most potent CD4 binding site-directed
neutralizing antibody yet discovered, able to neutralize more than 90% of
circulating isolates (13), but analysis of envelope variants from the patient who
generated VRC01 found that no autologous viruses were sensitive to it (7).
Once an infection is established, HIV is able to sample enough sequence space to
evade neutralization. During mucosal transmission, however, a single founder
virus often initiates the infection, creating an evolutionary bottleneck (23-25). A
9
broadly neutralizing antibody present before transmission may therefore be able
to block HIV with the breadth observed in vitro. In passive immunization
experiments performed in macaques, animals given neutralizing antibodies
prophylactically were protected from SHIV challenge, whereas those given nonneutralizing antibodies or to whom the antibodies were administered after
infection were not protected (26-29). The studies also suggest that a low
concentration of antibody as may be present following a vaccine could be
sufficient to neutralize the virus (2, 8, 30).
There is tremendous interest in the field to discover new neutralizing antibodies,
understand their mechanisms, and ultimately to elicit them by vaccination (3137). The tasks of antibody discovery, characterization, and immunogen design
are at heart protein engineering challenges, but the full arsenal of protein
engineering tools has not yet been deployed in service of finding a solution.
To discover neutralizing antibodies, patient samples are screened against wildtype gp120 and mutants that exhibit differential binding (13). Epitopes are
typically defined by crystallography or alanine scanning (12, 19, 32, 38).
Immunogens are engineered by inserting selective glycosylation sites to mask
immunodominant epitopes, or grafting defined epitopes onto new scaffolds (3942). Most of the previously-described gp120 variants are made by rational
design, sometimes guided by computation, then individually constructed,
secreted, and tested for binding by ELISA (35, 40, 43). Yeast surface display and
directed evolution provides an alternative, simple and flexible method for
engineering complex glycoproteins (44,45). Surface displayed proteins can be
easily modified by random or rational mutagenesis, and binding phenotypes
assayed by flow cytometry. We and others have engineered complex
glycoprotein receptor ectodomains (46, 47), and validated yeast surface display
for fine resolution mapping of conformational epitopes (48).
10
Thesis summary
This thesis adapts and adopts the tools of protein engineering by yeast surface
display to characterize CD4 binding site-directed neutralizing antibodies and
develop immunogens that elicit these vital antibodies by vaccination. More
broadly, the methods described can be used to design immunogens that focus
the immune response on a single epitope for any target antigen and any model
antibody.
* In chapter 2 we describe the display of a stripped core gp120 construct on
the surface of yeast, and show that the conformational CD4 binding site is
presented correctly.
* In chapter 3 we map the epitopes of CD4 binding site-directed antibodies
by two different yeast display methods, and determine important
energetic interactions between the antibodies and stripped core. These
interactions are mined to design mutants that more selectively identify
CD4 binding site-directed neutralizing antibodies than the canonical
gp120 mutant D368R.
" In chapter 4 we generate four stripped core-based immunogens that are
diverse on their surfaces but share the epitope defined by antibody
VRCO1. Mice immunized sequentially with these immunogens elicit a
strong gp120-specific antibody response that is directed entirely to the
VRC01 epitope. This is the first demonstration that an antibody response
can be focused upon the CD4 binding site by vaccination. Serum and
monoclonal antibodies from these mice are characterized for breadth and
neutralization potential.
* In chapter 5 we describe several robust mouse immunization protocols
with different formulations, adjuvants, and routes of administration. Key
parameters for each protocol are elucidated.
* In chapter 6 we deviate from HIV and describe early work done to design
10 type III fibronectin domains that bind to DEC-205, a dendritic cell
surface receptor involved in cross-presentation of extracellular protein
antigens onto MHC-I molecules.
11
Works Cited
1. D. R. Burton, R. L. Stanfield, I. A. Wilson, Antibody vs. HIV in a clash of
evolutionary titans, Proc Natl Acad Sci USA 102, 14943-14948 (2005).
2. M. J.McElrath, B. F. Haynes, Induction of immunity to human
immunodeficiency virus type-1 by vaccination, Immunity 33, 542-554 (2010).
3.
J. R. Mascola, D. C. Montefiori, The role of antibodies in HIV vaccines, Annu Rev
Immunol 28, 413-444 (2010).
4. R. Pantophlet, D. R. Burton, GP120: target for neutralizing HIV-1 antibodies,
Annu Rev Immunol 24, 739-769 (2006).
5. B. Korber et al., Evolutionary and immunological implications of contemporary
HIV-1 variation, Br Med Bull 58, 19-42 (2001).
6. S. D. W. Frost et al., Neutralizing antibody responses drive the evolution of
human immunodeficiency virus type 1 envelope during recent HIV infection,
Proc Natl Acad Sci USA 102, 18514-18519 (2005).
7. X. Wu et al., Selection Pressure on HIV-1 Envelope by Broadly Neutralizing
Antibodies to the Conserved CD4-Binding Site, J Virol 86, 5844-5856 (2012).
8. K. J.Bar et al., Early Low-Titer Neutralizing Antibodies Impede HIV-1
Replication and Select for Virus Escape, PLoS Pathog 8, e1002721 (2012).
9. P. L. Moore et al., Evolution of an HIV glycan-dependent broadly neutralizing
antibody epitope through immune escape, Nat Med (2012), doi:10.1038 /nm.2985.
10. D. R. Burton et al., A large array of human monoclonal antibodies to type 1
human immunodeficiency virus from combinatorial libraries of asymptomatic
seropositive individuals, Proc Natl Acad Sci USA 88, 10134-10137 (1991).
11. D. R. Burton et al., Efficient neutralization of primary isolates of HIV-1 by a
recombinant human monoclonal antibody, Science 266, 1024-1027 (1994).
12. T. Zhou et al., Structural definition of a conserved neutralization epitope on HIV1 gp120, Nature 445, 732-737 (2007).
13. X. Wu et al., Rational design of envelope identifies broadly neutralizing human
monoclonal antibodies to HIV-1, Science 329, 856-861 (2010).
14. E. Falkowska et al., PG V04, an HIV-1 gp120 CD4 binding site antibody, is broad
and potent in neutralization but does not induce conformational changes
characteristic of CD4, J Virol (2012), doi:10.1128/JVI.06973-11.
12
15. R. M. Lynch et al., The development of CD4 binding site antibodies during HIV-1
infection, J Virol 86, 7588-7595 (2012).
16. A. J. Mcmichael, P. Borrow, G. D. Tomaras, N. Goonetilleke, B. F. Haynes, The
immune response during acute HIV-1 infection: clues for vaccine development,
Nat Rev Immunol 10, 11-23 (2010).
17. A. Nandi et al., Epitopes for broad and potent neutralizing antibody responses
during chronic infection with human immunodeficiency virus type 1, Virology
(2009), doi:10.1016/j.virol.2009.10.044.
18. D. N. Sather, L. Stamatatos, Epitope specificities of broadly neutralizing plasmas
from HIV-1 infected subjects, Vaccine 28 Suppl 2, B8-12 (2010).
19. L. M. Walker et al., A limited number of antibody specificities mediate broad and
potent serum neutralization in selected HIV-1 infected individuals, PLoS Pathog6
(2010), doi:10.1371/journal.ppat.1001028.
20. J.F. Scheid et al., Broad diversity of neutralizing antibodies isolated from
memory B cells in HIV-infected individuals, Nature 458, 636-640 (2009).
21.
J.Pietzsch et al., Human anti-HIV-neutralizing antibodies frequently target a
conserved epitope essential for viral fitness, I Exp Med 207, 1995-2002 (2010).
22. Z. Euler et al., Cross-reactive neutralizing humoral immunity does not protect
from HIV type 1 disease progression, J Infect Dis 201, 1045-1053 (2010).
23. B. F. Keele et al., Identification and characterization of transmitted and early
founder virus envelopes in primary HIV-1 infection, Proc Natl Acad Sci USA 105,
7552-7557 (2008).
24. M.-R. Abrahams et al., Quantitating the Multiplicity of Infection with Human
Immunodeficiency Virus Type 1 Subtype C Reveals a Non-Poisson Distribution
of Transmitted Variants, J Virol 83, 3556-3567 (2009).
25.
J.F. Salazar-Gonzalez et al., Genetic identity, biological phenotype, and
evolutionary pathways of transmitted/founder viruses in acute and early HIVinfection, Journal of ExperimentalMedicine 206, 1273-1289 (2009).
26. D. R. Burton et al., Limited or no protection by weakly or nonneutralizing
antibodies against vaginal SHIV challenge of macaques compared with a
strongly neutralizing antibody, Proc Natl Acad Sci USA 108, 11181-11186 (2011).
27. A. J.Hessell et al., Broadly neutralizing human anti-HIV antibody 2G12 is
effective in protection against mucosal SHIV challenge even at low serum
neutralizing titers, PLoS Pathog 5, e1000433 (2009).
28. A. J. Hessell et al., Effective, low-titer antibody protection against low-dose
repeated mucosal SHIV challenge in macaques, Nat Med 15, 951-954 (2009).
13
29. J. R. Mascola, Passive transfer studies to elucidate the role of antibody-mediated
protection against HIV-1, Vaccine 20, 1922-1925 (2002).
30. D. C. Montefiori, J.R. Mascola, Neutralizing antibodies against HIV-1: can we
elicit them with vaccines and how much do we need? Curr Opin HIV AIDS 4,
347-351 (2009).
31. L. Chen et al., Structural basis of immune evasion at the site of CD4 attachment
on HIV-1 gp120, Science 326, 1123-1127 (2009).
32. T. Zhou et al., Structural basis for broad and potent neutralization of HIV-1 by
antibody VRC01, Science 329, 811-817 (2010).
33. J.F. Scheid et al., Sequence and structural convergence of broad and potent HIV
antibodies that mimic CD4 binding, Science 333, 1633-1637 (2011).
34. X. Wu et al., Focused evolution of HIV-1 neutralizing antibodies revealed by
structures and deep sequencing, Science 333, 1593-1602 (2011).
35. W. R. Schief, Y.-E. A. Ban, L. Stamatatos, Challenges for structure-based HIV
vaccine design, Curr Opin HIV AIDS 4, 431-440 (2009).
36. L. M. Walker, D. R. Burton, Rational antibody-based HIV-1 vaccine design:
current approaches and future directions, CurrentOpinion in Immunology (2010),
doi:10.1016/j.coi.2010.02.012.
37. B. D. Walker, D. R. Burton, Toward an AIDS vaccine, Science 320, 760-764 (2008).
38. M. Pancera et al., Crystal structure of PG16 and chimeric dissection with
somatically related PG9: structure-function analysis of two quaternary-specific
antibodies that effectively neutralize HIV-1, J Virol 84, 8098-8110 (2010).
39. B. Dey et al., Characterization of human immunodeficiency virus type 1
monomeric and trimeric gp120 glycoproteins stabilized in the CD4-bound state:
antigenicity, biophysics, and immunogenicity, J Virol 81, 5579-5593 (2007).
40. R. Pantophlet, I. A. Wilson, D. R. Burton, Hyperglycosylated mutants of human
immunodeficiency virus (HIV) type 1 monomeric gp120 as novel antigens for
HIV vaccine design, J Virol 77, 5889-5901 (2003).
41. R. Pantophlet et al., Fine mapping of the interaction of neutralizing and
nonneutralizing monoclonal antibodies with the CD4 binding site of human
immunodeficiency virus type 1 gp120, J Virol 77, 642-658 (2003).
42. J. Guenaga et al., Heterologous epitope-scaffold prime:boosting immuno-focuses
B cell responses to the HIV-1 gp4l 2F5 neutralization determinant, PLoS ONE 6,
e16074 (2011).
14
43. B. E. Correia et al., Computational protein design using flexible backbone
remodeling and resurfacing: case studies in structure-based antigen design,
Journal of Molecular Biology 405, 284-297 (2011).
44. E. T. Boder, K. D. Wittrup, Yeast surface display for screening combinatorial
polypeptide libraries, Nat Biotechnol 15, 553-557 (1997).
45. E. T. Boder, K. S. Midelfort, K. D. Wittrup, Directed evolution of antibody
fragments with monovalent femtomolar antigen-binding affinity, Proc Natl Acad
Sci USA 97, 10701-10705 (2000).
46. Y.-S. Kim, R. Bhandari, J.R. Cochran, J.Kuriyan, K. D. Wittrup, Directed
evolution of the epidermal growth factor receptor extracellular domain for
expression in yeast, Proteins 62, 1026-1035 (2006).
47. S. A. Gai, K. D. Wittrup, Yeast surface display for protein engineering and
characterization, Curr Opin Struct Biol 17, 467-473 (2007).
48. G. Chao, J.R. Cochran, K. D. Wittrup, Fine epitope mapping of anti-epidermal
growth factor receptor antibodies through random mutagenesis and yeast
surface display, Journalof Molecular Biology 342, 539-550 (2004).
15
16
Chapter 2. Yeast display of HIV gp120*
Background
Much of the focus of current HIV vaccine research involves the isolation of new
neutralizing antibodies, understanding their structure and function, and
designing immunogens intended to elicit such antibodies by vaccination. These
efforts often require that gp120, the predominant glycoprotein of the envelope
spike, be modified or engineered.
New neutralizing antibodies are identified and characterized by screening
patient samples against engineered gp120 probes with differential binding
profiles to known epitopes (1-3). For example, antibody VRC01, the gold
standard CD4-competitive antibody, was isolated from a patient sample using a
hybrid HIV/SIV gp120 that had been "resurfaced" to remove all HIV epitopes
except for the CD4 binding site (4).
Modifications are also made to gp120 when designing an immunogen to elicit
neutralizing antibodies by vaccination (5). Deliberate amino acid substitutions
can, for example, lock the flexible molecule into an "open" conformation and
overcome the considerable entropic penalty incurred upon binding (6-8).
Mutations can also disrupt antibody binding to irrelevant surfaces and encode
glycosylation sites to shield undesired epitopes from immune recognition (9-11).
Immunodominant loops that distract from neutralizing epitopes can be removed
entirely (12, 13).
Stable gp160 trimers have recently been described in the literature (14, 15).
Trimeric envelope is an attractive alternative to monomeric gp120 for many of
these applications. The immune system elicits non-neutralizing antibodies to
monomer-specific epitopes that are sterically inaccessible on the native trimer
(16). Additionally, some neutralizing antibodies recognize quaternary epitopes
*Portions of this chapter were adapted from Mata-Fink et al., Rapid conformational epitope mapping
of anti-gp120 antibodies with a designed mutant panel displayed on yeast, I Mol Biol, in press.
17
and thus cannot be faithfully captured by monomeric gp120 (17). Engineered
viruses are also used to present gp120 in its native conformation (18). The
drawbacks to these methods are that the trimer is less forgiving of mutations
(Kovacs et al. have only reported stable trimers for two strains of HIV (14)), and
that not every lab has the capability to work with pseudoviruses.
To date, engineering gp120 is a low-throughput process. Most of the previouslydescribed gp120 variants are made by rational design, sometimes guided by
computation, then individually constructed, secreted, and tested for binding by
ELISA (10, 19, 20).
Yeast surface display provides an alternative, simple and flexible method for
engineering complex glycoproteins (21, 22). Surface displayed proteins can be
easily modified by random or rational mutagenesis, and binding phenotypes
assayed by flow cytometry. We and others have engineered complex
glycoprotein receptor ectodomains (23, 24), and validated yeast surface display
for fine resolution mapping of conformational epitopes (25).
Despite its utility, yeast display has not yet been used to engineer gp120 for
immunogens or as bait for isolating neutralizing antibodies. In this chapter we
describe the display of an engineered stripped core gp120 on yeast and
characterize its binding to a panel of CD4 binding site-directed neutralizing
antibodies to demonstrate that the CD4 binding site is correctly presented.
Results
Display of gp120 on yeast. The gene for gp120 from HIV strain YU2 (26) was
subcloned into the yeast display vector pCHA with a C-terminal Aga2p fusion
partner (Figure 2.1A). The traditional pCT vector with an N-terminal Aga2p
fusion was tested and worked equally well (data not shown). The advantage of
pCHA over pCT is that, since the surface anchor Aga2p is at the distal end of the
transcript, truncated proteins are not captured and displayed.
18
Modifications to gp120 to make stripped core. Yeast displaying the full-length
gp120 protein did not bind to antibody b12 (data not shown). Extensive
modifications were made to strip the protein of flexible loops that might cause
the protein to misfold or sterically occlude the b12 epitope. The resultant
"stripped core" is similar, though not identical, to the original core gp120
crystalized in complex with CD4 (27). The flexible gp41-interactive regions at the
N- and C-termini (aa 1-89, 493-511) (28) were excised entirely. The long hypervariable loops (29, 30)-V1 /V2 (aa 124-198), V3 (aa 298-329), and the bridging
sheet p20 / 21 (aa 422-437)-were replaced with the short, glycosylated amino
acid linkers GNGS, GNGSG, and GGNGS respectively. An additional set of
mutations (Q114N, L116T) introduced an N-linked glycosylation sequence at a
site previously identified as important for masking an epitope on the V1 /V2
stem (10). The full sequence of stripped core is listed in Table 2.1, and a structural
model is shown in Figure 2.1B.
Binding of stripped core to anti-gp120 antibodies. Despite these significant
modifications, stripped core gp120 retains the structurally rigid outer domain
that has been extensively studied elsewhere (31-33). The protein is expressed
well on the yeast surface (1x10 5 copies per cell; standard for yeast surface
display) and conserves the structure of the CD4 binding site as measured by
binding to CD4 and to a panel of neutralizing antibodies (Figures 2.1C, 1D).
Broadly neutralizing anti-gp120 antibodies VRC01, PGV04 (isolated from the
same patient as VRC01 but by a different protocol) (4), and b12 (34, 35), as well as
the non-neutralizing CD4-directed antibodies b6 and b13 (36) were titrated on
yeast displaying stripped core gp120 and the binding measured by flow
cytometry. The equilibrium binding constants ranged from 24 pM for b6 to 0.9
nM for PGV04 and b13. Yeast-displayed stripped core gp120 also bound CD4-Fc
with a binding constant of 5.6 nM, despite the removal of the bridging sheet that
comprises a substantial portion of the CD4 binding site (27, 37). These binding
data are consistent with the single-digit nM affinities measured for CD4 binding
site-directed antibodies to several gp120 constructs by other methods (4), and
19
with binding of multivalent CD4 constructs to both "liganded" and
"unliganded" conformations of gp120 (38).
Other antibodies of interest in the field of HIV vaccine design were tested as
well, but did not bind to the stripped core construct (data not shown). Antibodies
17b and X5 that recognize the chemokine receptor epitope on gp120 (39-41)called the CD4-induced or CD4i epitope because it is only induced upon CD4
engagement-did not bind, which is expected given that much of the CD4i
epitope was removed when the bridging sheet was replaced with a short linker.
Non-neutralizing antibody F105 (42), whose crystal structure (PDB ID: 3HI1)
indicates that it too makes significant contacts with the bridging sheet (16), did
not bind. Antibodies PGT-121 and PGT-128, known to recognize an epitope on
variable loop V3 (43, 44), did not bind to the V3-less construct. Antibody 2G12,
which recognizes an array of high-mannose glycans on the surface of gp120 (4547) also did not bind the stripped core displayed on yeast. S. cerevisiae hypermannosylates N-linked glycans and produces glycoforms different from those
recognized by 2G12 (48). Others have engineered S. cerevisiae such that its
glycoproteins are recognized by 2G12 (49); display of stripped core on this
modified yeast strain may introduce 2G12 sensitivity.
To ensure that the binding to CD4 binding site-directed antibodies was not an
artifact of yeast display, we secreted stripped core gp120 from HEK 293 cells
both as a monomer and as an Fc-fusion, and measured its binding to the singlechain variable fragment (scFv) of VRC01 displayed on yeast. The equilibrium
dissociation constant was found to be 8.7 nM for monomeric stripped core and
1.1 nM for the dimeric Fc-fusion (Figure 2.1E), consistent with previous titrations
in which the arrangement of gp120 and antibody was inverted.
Stability of stripped core. The stripped core also proved to be remarkably stable,
perhaps due to its six inter-chain disulfide bonds. Yeast-displayed protein retains
its binding to CD4 binding site-directed antibodies for months at 4C. Soluble Fcstripped core can be lyophilized and rehydrated, heated for 30 minutes in the
presence of SDS, and stored at room temperature in liquid formulation for at
least four weeks with no loss of binding to VRC01 (Figure 2.1F). Mass
20
spectrometry indicated that the protein is not aggregated, in keeping with these
observations. For the Fc-fusion, a species of radius 6.1 nm (22.4 %Pd)
corresponding to an Fc-dimer with two gp120 molecules accounts for 99.9% of
the mass. For the non-Fc-fusion, the predominant species has radius 5.4 nm (23.9
%Pd) and accounts for 96.2% of the mass. The non-Fc-fusion is purified by metal
affinity chromatography and size exclusion, whereas the Fc-fusion is purified by
sequential metal affinity and protein A chromatography. The additional
heterogeneity observed in non-Fc soluble stripped core is likely from incomplete
purification rather than aggregation.
This translation of behavior on the surface of yeast to behavior in solution did
not hold for all gp120 constructs tested. An engineered outer domain construct
(aa 254-476) with V3 and bridging sheet replaced with glycosylated linkers, a
glycosylation site inserted at the C-terminus, and a necessary deglycosylation at
position 386 (mutation N386D) bound well to CD4 binding site-directed
antibodies when displayed on yeast but lost all binding when secreted from HEK
293 cells (Figure 2.2). Our hypothesis is that the hydrophobic surface at the
interface of the inner and outer domains destabilizes the protein, but was
shielded by the close proximity of Aga2p or the yeast cell wall when displayed
on the cell surface. In another instance, immunogens were engineered to contain
diversity on their surfaces but retain the CD4 binding site (Chapter 4). 16 of 20
immunogens that bound VRC01 when displayed on yeast lost most or all
binding when secreted as Fc-fusions.
Discussion
There is much effort in the HIV vaccine field to modify gp120 to make subunit
vaccine immunogens and bait for isolating and characterizing antibodies from
patient samples. Yeast surface display is a simple and powerful tool for
engineering complex glycoproteins (21, 22, 24). We report for the first time the
display of a stripped core gp120 on yeast, which opens the toolkit of protein
engineering-including directed evolution, high-throughput screening by flow
21
cytometry, and conformational epitope mapping-to the problem of HIV vaccine
design.
Stripped core is well displayed on yeast, and presents the CD4 binding site in the
correct conformation as measured by binding to CD4 and a panel of CD4 binding
site-directed antibodies. The platform thus enables rapid screening of candidate
immunogens and high-throughput analysis of serum samples.
The modifications to stripped core ablate binding to many anti-gp120 antibodies
that are specific for epitopes other than the CD4 binding site. A different starting
scaffold must be designed in order to screen these antibodies by yeast display.
Stripped core is highly stable both on the cell surface and when secreted from
mammalian cells, and binds similarly to VRC01 in both instances. This is
promising for the generation of candidate immunogens, which relies on the
accurate translation of selected characteristics on the cell surface to behavior in
solution and in vivo.
Materials & Methods
Display of gp120 on yeast. The gene for gp120 from HIV strain YU2 (26)
(Genbank Accession No. M93258) was subcloned into the yeast display vector
pCHA between NheI and BamHI restriction sites, immediately downstream of
an HA epitope tag (YPYDVPDYA) and upstream of a CMyc tag (EQKLISEEDL)
and the Aga-2 fusion partner. The protein was displayed on Saccharomyces
cerevisiae strain EBY100 using a standard surface display protocol (50). In brief,
EBY100 yeast were transformed with plasmid (Frozen-EZ Yeast Transformation
II Kit, Zymo Research, Irvine, CA), grown to mid-log phase in SD-CAA media at
30*C, and induced in galactose-containing SG-CAA media for 24 hr at 20"C.
Approximately 1x10 5 gp120 molecules are displayed per yeast, as measured by
quantitative flow cytometry (data not shown).
22
Modifications to gp120. Modifications to gp120 were made by site-directed
mutagenesis or the Geiser variant thereof (51) following the QuickChange
mutagenesis protocol (Agilent Technologies, Santa Clara, CA). The flexible gp4linteractive regions at the N- and C-termini were removed entirely (aa 1-89, 493511) (28). The hyper-variable loops V1 /V2 (aa 124-198), V3 (aa 298-329), and the
bridging sheet p20/21 (aa 422-437) were replaced with the short, glycosylated
amino acid linkers, GNGS, GNGSG, and GGNGS respectively. An additional
double mutation (Q114N, L116T) introduced an N-linked glycosylation sequence
at a site previously identified as important for masking an epitope on the V1 /V2
stem (10). All amino acid numbering is based on the HxB2 numbering
convention (52). Stripped core is 285 amino acids long. The full sequence with
these modifications listed is in Table 2.1.
Secretion of gp120 from HEK 293 cells. The stripped core gp120 was subcloned
into a mammalian expression vector based on gWiz (Genlantis, San Diego, CA),
either by itself or downstream of the Fc domain of mouse IgG2a (53). All
constructs also have a his6 tag. There are two variants of the Fc-fusion construct.
gWiz-Fc-X-hisC has a TEV protease site upstream of the gene of interest, a his6
tag immediately downstream, and a terminal cysteine. gWiz-Fc-hisT-X has a his6
tag and TEV protease site between the Fc and the gene of interest, and no
terminal modification. Protein secreted equally well from both Fc-fusion formats.
A suspension culture of HEK 293 cells grown in serum-free medium (Freestyle
293, Life Technologies, Carlsbad, CA) was transfected with DNA using PEI as
per manufacturers instructions. Supernatant was harvested after one week and
purified on TALON metal affinity chromatography resin (Clontech, Mountain
View, CA) followed by size-exclusion chromatography for the non-Fc-fusion,
and on sequential metal affinity chromatography and protein A affinity
chromatography (Genscript, Piscataway, NJ) for the Fc-fusion.
Binding assays. Anti-gp120 antibodies were either produced from HEK cells (the
plasmid for VRC01 was a generous gift from John Mascola, Vaccine Research
Center, NIAID, NIH) or a generous gift from Dennis Burton. CD4-Fc was
purchased from Sino Biological (Beijing, China). Yeast cells displaying protein
23
were incubated with anti-gp120 antibodies and the anti-epitope tag antibody
chicken anti-CMyc (Gallus Immunotech, Fergus, Ontario). Secondary labeling
was performed with goat anti-mouse, -human, or -chicken antibodies conjugated
to Alexa Fluor dyes (Life Technologies, Carlsbad, CA) at recommended
dilutions. Individual samples were analyzed on an Accuri C6 cytometer (BD
Accuri Cytometers, Ann Arbor, MI); 96-well plates were run on a FACSCalibur
HTS with a high-throughput plate sampler (Becton Dickinson, Franklin Lakes,
NJ).
24
A
C
B
10
Aga2p
7AgOa1p
CMye
00
HA
gplZO
CMycAga2
1.1
01
100
1
Display (anti-Cyc)
D
12-
-
-
V 13
141 -* 4Cslwap
12j
-. -~PG9
121 q-IM
--
ac4!
A
02~
-odCOMc(11"
a
Jou1
tR
00
0
0
1
gobooCoric (Wi)
10
100
1
10
100
Pc-CorsCmimetadon(Wi)
1000
Figure 2.1. Yeast display of gp120. (a) Schematic of gp120 displayed on the
surface of yeast. Beaneath it, the DNA map of vector pCHA showing gp120, the
C-terminal Aga2 fusion partner, and two epitope tags. (b) Homology model of
stripped core showing the inner domain (white), outer domain (blue), the short
glycosylated peptides that have replaced the V 1/V2 loop (orange), V3 loop (red),
and bridging sheet (purple), and the VRC01 contact residues (yellow). (c)
Representative flow cytometry dot-plot of yeast displaying stripped core gp120,
bound to both an anti-CMyc antibody (x axis) and the broadly neutralizing
antibody VRC01 (y axis). (d) Binding isotherms of yeast-displayed stripped core
gp120 to a panel of CD4 binding site-directed antibodies. Equilibrium KD values
for the antibodies are as follows: VRC01, 0.8 nM; PGV04, 0.9 nM; b12, 41 pM; b6,
24 pM; b13, 0.9 nM; CD4-Fc, 5.6 nM. (e) Binding of secreted stripped core gp120
to yeast displaying the scFv of antibody VRCO1. Equilibrium KD values for the
interactions are: Stripped Core, 8.7 nM; Fc Stripped Core, 1.1 nM. (f) Stability of
secreted Fc-stripped core. Binding of Fc-stripped core to yeast displaying the
scFv of antibody VRC01 after dissolution from microneedle array (black), stock
stored at 4C for several months (red), or stock stored at room temperature for 1
month (blue).
25
26
1.21
0.8
e0.4-
0.2
801
0.1
1
10
100
Protein (nM)
Figure 2.2. Differential binding of displayed and secreted outer domain-only
construct. Stripped outer domain (aa 254-476) has V3 and bridging sheet
replaced with glycosylated linkers, a glycosylation site inserted at the Cterminus, and deglycosylation mutation N386D (construct called 3G or stripped
OD). Fc-stripped core (solid) and Fc-stripped OD (dashed) secreted from HEK293 cells binding to VRC01 scFv displayed on yeast. Stripped OD displayed on
yeast (dash-dotted) binding to antibody VRC02, a somatic variant of VRCO1.
27
28
Table 2.1. Sequence of stripped core gp120 numbered both sequentially and by the HxB2
numbering convention (52). Extra glycosylation sites (aa 114/116) and modified linker
sequences (V1 /V2 aa124-199; V3 aa298-329; P20 /21 aa422-437) are highlighted in gray.
Sequential
position
HxB2
position
238
115
276
201
239
116
277
92
202
240
117
278
93
203
241
118
279
5
94
204
242
119
280
6
95
205
243
120
281
7
206
244
121
282
8
97
207
245
122
283
9
98
208
246
123
284
10
99
209
247
124
285
11
100
210
248
125
286
126
287
Amino
Acid
HxB2
position
Sequential
position
HxB2
position
1
90
200
2
91
3
4
Sequential
position
Amino
Acid
Sequential
position
HxB2
position
Amino
Acid
12
101
211
249
13
102
212
250
127
288
14
103
213
251
128
289
15
104
214
252
129
290
16
105
215
253
130
291
17
106
216
254
131
292
18
107
217
255
132
293
19
108
218
256
133
294
20
109
219
257
134
295
21
110
220
258
135
2%
22
111
221
259
136
297
23
112
222
260
Amino
Acid
24
113
223
261
25
114
N
224
262
26
115
S
225
263
27
116
T
226
264
28
117
K
227
265
142
330
H
29
118
P
228
266
143
331
C
30
119
C
229
267
144
332
N
31
120
V
230
268
145
333
L
32
121
K
231
269
146
334
S
33
122
L
232
270
147
335
K
34
123
T
233
271
148
336
T
234
272
149
337
235
273
150
338
Q
w
236
274
151
339
E
237
275
152
340
N
29
Sequential
position
HxB2
position
Amino
Acid
379
264
471
G
192
380
265
472
G
343
193
381
266
473
G
156
344
194
382
267
474
D
157
345
195
383
268
475
M
R
Sequential
position
HxB2
Amino
position
Acid
153
Sequential
position
HxB2
position
341
191
154
342
155
Amino
Acid
Sequential
position
HxB2
position
Amino
Acid
158
346
196
384
231
438
269
476
159
347
197
385
232
439
270
477
D
160
348
198
386
233
440
271
478
N
161
349
199
387
234
441
272
479
w
162
350
200
388
235
442
273
480
R
274
481
S
163
351
201
389
236
443
164
352
202
390
237
444
275
482
E
165
353
203
391
238
445
276
483
L
166
354
204
394
239
446
277
484
Y
K
167
355
205
395
240
447
278
485
168
356
206
396
241
448
279
486
Y
169
357
207
397
242
449
280
487
K
170
358
208
402
243
450
281
488
V
489
V
171
359
209
403
244
451
282
172
360
210
404
245
452
283
490
K
173
361
211
405
246
453
284
491
I
174
362
212
406
247
454
285
492
E
175
363
213
407
248
455
176
364
214
408
249
456
177
365
215
411
250
457
178
366
216
412
251
458
179
367
217
413
252
459
180
368
218
414
253
460
181
369
219
415
254
461
182
370
220
416
255
462
183
371
221
417
256
463
184
372
222
418
257
464
185
373
223
419
258
465
186
374
224
420
259
466
187
375
225
421
260
467
188
376
261
468
189
377
262
469
190
378
263
470
30
Works Cited
1. G. D. Tomaras et al., Polyclonal B Cell Responses to Conserved Neutralization
Epitopes in a Subset of HIV-1-infected Individuals, J Virol (2011),
doi:10.1128/JVI.05363-11.
2.
J. F. Scheid
et al., A method for identification of HIV gp140 binding memory B
cells in human blood, Journalof Immunological Methods 343, 65-67 (2009).
3. L. M. Walker, D. R. Burton, Rational antibody-based HIV-1 vaccine design:
current approaches and future directions, Current Opinion in Immunology (2010),
doi:10.1016 /j.coi.2010.02.012.
4. X. Wu et al., Rational design of envelope identifies broadly neutralizing human
monoclonal antibodies to HIV-1, Science 329, 856-861 (2010).
5. D. R. Burton et al., HIV vaccine design and the neutralizing antibody problem,
Nat Immunol 5, 233-236 (2004).
6. D. G. Myszka et al., Energetics of the HIV gp120-CD4 binding reaction, Proc Natl
Acad Sci USA 97, 9026-9031 (2000).
7. P. D. Kwong et al., HIV-1 evades antibody-mediated neutralization through
conformational masking of receptor-binding sites, Nature 420, 678-682 (2002).
8. B. Dey et al., Characterization of human immunodeficiency virus type 1
monomeric and trimeric gp120 glycoproteins stabilized in the CD4-bound state:
antigenicity, biophysics, and immunogenicity, I Virol 81, 5579-5593 (2007).
9. R. Pantophlet, D. R. Burton, Immunofocusing: antigen engineering to promote
the induction of HIV-neutralizing antibodies, Trends Mol Med 9, 468-473 (2003).
10. R. Pantophlet, I. A. Wilson, D. R. Burton, Hyperglycosylated mutants of human
immunodeficiency virus (HIV) type 1 monomeric gp120 as novel antigens for
HIV vaccine design, J Virol 77, 5889-5901 (2003).
11. R. Pantophlet, I. A. Wilson, D. R. Burton, Improved design of an antigen with
enhanced specificity for the broadly HIV-neutralizing antibody b12, ProteinEng
Des Sel 17, 749-758 (2004).
12. F. K. Ahmed, B. E. Clark, D. R. Burton, R. Pantophlet, An engineered mutant of
HIV-1 gp120 formulated with adjuvant Quil A promotes elicitation of antibody
responses overlapping the CD4-binding site, Vaccine (2011),
doi:10.1016/j.vaccine.2011.11.089.
31
13. S. Selvarajah et al., Comparing antigenicity and immunogenicity of engineered
gp120, J Virol 79, 12148-12163 (2005).
14. J. M. Kovacs et al., HIV-1 envelope trimer elicits more potent neutralizing
antibody responses than monomeric gp120, Proc Natl Acad Sci USA (2012),
doi:10.1073/pnas.1204533109.
15. J. P. Nkolola et al., Breadth of neutralizing antibodies elicited by stable,
homogeneous clade A and clade C HIV-1 gp140 envelope trimers in guinea pigs,
J Virol 84, 3270-3279 (2010).
16. L. Chen et al., Structural basis of immune evasion at the site of CD4 attachment
on HIV-1 gp120, Science 326, 1123-1127 (2009).
17. L. M. Walker et al., Broad and potent neutralizing antibodies from an African
donor reveal a new HIV-1 vaccine target, Science 326, 285-289 (2009).
18. L. M. Walker et al., A limited number of antibody specificities mediate broad and
potent serum neutralization in selected HIV-1 infected individuals, PLoS Pathog 6
(2010), doi:10.1371 /joumal.ppat.1001028.
19. W. R. Schief, Y.-E. A. Ban, L. Stamatatos, Challenges for structure-based HIV
vaccine design, Curr Opin HIV AIDS 4, 431-440 (2009).
20. B. E. Correia et al., Computational protein design using flexible backbone
remodeling and resurfacing: case studies in structure-based antigen design,
Journalof Molecular Biology 405, 284-297 (2011).
21. E. T. Boder, K. D. Wittrup, Yeast surface display for screening combinatorial
polypeptide libraries, Nat Biotechnol 15, 553-557 (1997).
22. E. T. Boder, K. S. Midelfort, K. D. Wittrup, Directed evolution of antibody
fragments with monovalent femtomolar antigen-binding affinity, ProcNatl Acad
Sci USA 97, 10701-10705 (2000).
23. Y.-S. Kim, R. Bhandari, J. R. Cochran, J. Kuriyan, K. D. Wittrup, Directed
evolution of the epidermal growth factor receptor extracellular domain for
expression in yeast, Proteins62, 1026-1035 (2006).
24. S. A. Gai, K. D. Wittrup, Yeast surface display for protein engineering and
characterization, Curr Opin Struct Biol 17, 467-473 (2007).
25. G. Chao, J. R. Cochran, K. D. Wittrup, Fine epitope mapping of anti-epidermal
growth factor receptor antibodies through random mutagenesis and yeast
surface display, Journalof Molecular Biology 342, 539-550 (2004).
32
26. Y. Li et al., Complete nucleotide sequence, genome organization, and biological
properties of human immunodeficiency virus type 1 in vivo: evidence for limited
defectiveness and complementation, J Virol 66, 6587-6600 (1992).
27. P. D. Kwong et al., Structure of an HIV gp120 envelope glycoprotein in complex
with the CD4 receptor and a neutralizing human antibody, Nature 393, 648-659
(1998).
28. M. Pancera et al., Structure of HIV-1 gp120 with gp41-interactive region reveals
layered envelope architecture and basis of conformational mobility, Proc Natl
Acad Sci USA (2009), doi:10.1073/pnas.0911004107.
29. B. Korber et al., Evolutionary and immunological implications of contemporary
HIV-1 variation, Br Med Bull 58, 19-42 (2001).
30. D. R. Burton, R. L. Stanfield, I. A. Wilson, Antibody vs. HIV in a clash of
evolutionary titans, Proc Natl Acad Sci USA 102, 14943-14948 (2005).
31. T. Zhou et al., Structural definition of a conserved neutralization epitope on HIV1 gp120, Nature 445, 732-737 (2007).
32. X. Yang et al., Characterization of the outer domain of the gp120 glycoprotein
from human immunodeficiency virus type 1, 1 Virol 78, 12975-12986 (2004).
33. L. Wu et al., Enhanced exposure of the CD4-binding site to neutralizing
antibodies by structural design of a membrane-anchored human
immunodeficiency virus type 1 gp120 domain, J Virol 83, 5077-5086 (2009).
34. D. R. Burton et al., A large array of human monoclonal antibodies to type 1
human immunodeficiency virus from combinatorial libraries of asymptomatic
seropositive individuals, Proc Natl Acad Sci USA 88, 10134-10137 (1991).
35. D. R. Burton et al., Efficient neutralization of primary isolates of HIV-1 by a
recombinant human monoclonal antibody, Science 266, 1024-1027 (1994).
36. P. Roben et al., Recognition properties of a panel of human recombinant Fab
fragments to the CD4 binding site of gp120 that show differing abilities to
neutralize human immunodeficiency virus type 1, J Virol 68, 4821-4828 (1994).
37. U. Olshevsky et al., Identification of individual human immunodeficiency virus
type 1 gp120 amino acids important for CD4 receptor binding, J Virol 64, 57015707 (1990).
38. T. Zhou et al., Structural basis for broad and potent neutralization of HIV-1 by
antibody VRCO1, Science 329, 811-817 (2010).
33
39. M. Thali et al., Characterization of conserved human immunodeficiency virus
type 1 gp120 neutralization epitopes exposed upon gp120-CD4 binding, J Virol
67, 3978-3988 (1993).
40. M. Moulard et al., Broadly cross-reactive HIV-1-neutralizing human monoclonal
Fab selected for binding to gpl20-CD4-CCR5 complexes, Proc Natl Acad Sci USA
99, 6913-6918 (2002).
41. A. F. Labrijn et al., Access of antibody molecules to the conserved coreceptor
binding site on glycoprotein gp120 is sterically restricted on primary human
immunodeficiency virus type 1, J Virol 77, 10557-10565 (2003).
42. M. R. Posner et al., An IgG human monoclonal antibody that reacts with HIV1/ GP120, inhibits virus binding to cells, and neutralizes infection, J Immunol 146,
4325-4332 (1991).
43. L. M. Walker et al., Broad neutralization coverage of HIV by multiple highly
potent antibodies, Nature 477, 466-470 (2011).
44. R. Pejchal et al., A Potent and Broad Neutralizing Antibody Recognizes and
Penetrates the HIV Glycan Shield, Science 334, 1097-1103 (2011).
45. C. N. Scanlan et al., The broadly neutralizing anti-human immunodeficiency
virus type 1 antibody 2G12 recognizes a cluster of alphal-->2 mannose residues
on the outer face of gp120, J Virol 76, 7306-7321 (2002).
46. R. W. Sanders et al., The mannose-dependent epitope for neutralizing antibody
2G12 on human immunodeficiency virus type 1 glycoprotein gp120, J Virol 76,
7293-7305 (2002).
47. K. J.Doores et al., Envelope glycans of immunodeficiency virions are almost
entirely oligomannose antigens, Proc Natl Acad Sci USA 107, 13800-13805 (2010).
48. E. Celik, P. Calik, Production of recombinant proteins by yeast cells, Biotechnol.
Adv. 30, 1108-1118 (2012).
49. R. J. Luallen et al., An engineered Saccharomyces cerevisiae strain binds the
broadly neutralizing human immunodeficiency virus type 1 antibody 2G12 and
elicits mannose-specific gp120-binding antibodies, J Virol 82, 6447-6457 (2008).
50. G. Chao et al., Isolating and engineering human antibodies using yeast surface
display, Nature Protocols 1, 755-768 (2006).
34
51. M. Geiser, R. Cebe, D. Drewello, R. Schmitz, Integration of PCR fragments at any
specific site within cloning vectors without the use of restriction enzymes and
DNA ligase, BioTechniques 31, 88-92 (2001).
52. Korber B, Foley BT, Kuiken C, Pillai SK, Sodroski JG (1998).Numbering Positions
in HIV Relative to HXB2CG. pp. 111-102-111 in Human Retroviruses and AIDS
1998. Edited by: Korber B, Kuiken CL, Foley B, Hahn B, McCutchan F, Mellors
JW, and Sodroski J. Published by: Theoretical Biology and Biophysics Group, Los
Alamos National Laboratory, Los Alamos, NM.
http://www.hiv.lanl.gov/content/sequence/HIV/REVIEWS/HXB2.html
53. K. M. Lo et al., High level expression and secretion of Fc-X fusion proteins in
mammalian cells, Protein Eng 11, 495-500 (1998).
35
36
Chapter 3. Epitope mapping of CD4 binding site-directed
antibodies*
Background
Advances in high-throughput screening techniques and new patient cohorts
have led to an explosion in the discovery of broadly neutralizing antibodies
against HIV. Whereas five years ago there were just a handful of known gp120specific neutralizing antibodies, there are now dozens of antibodies against a
well-defined set of conserved epitopes on the envelope spike. These epitopes are
the CD4 binding site (1-5), the co-receptor binding site (6-8), invariant linear
regions of the V 1/V2 and V3 loops (9, 10), a quaternary epitope formed by
adjacent gp120 monomers in the envelope trimer (11-14), a common glycan motif
(15), and the membrane-proximal external region of gp4l (16-18).
Determining the specificity of broadly neutralizing antibodies can provide
important insight into the mechanism of their action, and inform the design of
immunogens to elicit them by vaccination (19). For example, analysis of the
crystal structures of antibodies b12, b13, and F105 suggests that the angle at
which the antibodies approach the CD4 binding site explains why one is broadly
neutralizing and the other two are incapable of binding native trimer (20).
Alanine scanning of non-neutralizing CD4 binding site-directed antibodies
elicited by vaccination in rhesus macaques indicated that the mutation-sensitive
residues on gp120 for these antibodies match the epitope of non-neutralizing
antibody F105 more closely than they do the epitope of broadly neutralizing
antibody VRC01 (21). This data suggests that immunogens be designed expressly
to mask or otherwise distract from the non-neutralizing epitope recognized by
antibodies like b13 and F105.
*Portions of this chapter were adapted from Mata-Fink et al., Rapid conformational epitope mapping
of anti-gp120 antibodies with a designed mutant panel displayed on yeast, JMol Biol, in press.
37
Epitope mapping of serum samples from infected individuals has also shed
important light on virus-host dynamics during the course of natural infection.
Epitope mapping of serum with mutated pseudoviruses demonstrated that most
individuals develop antibodies against the same limited set of conserved
epitopes on the envelope spike described above, even if these antibodies do not
broadly neutralize heterologous virus (22). This important finding suggests that
the sites of vulnerability recognized by broadly neutralizing antibodies are not
privileged epitopes, and should be targetable by vaccination.
Several groups have used epitope mapping to elucidate pathways of viral escape
from immune pressure, which may also help inform immunogen design.
Longitudinal sequencing of envelope variants demonstrates that antibodies exert
selection pressure on autologous virus almost immediately following infection
(23-25). Analysis of envelope variants from the patient that generated antibody
VRC01 shows that the virus can escape from even the most broad and potent of
CD4 binding site-directed neutralizing antibodies (26). A separate study clearly
demonstrated that viral escape in two individuals was due to glycosylation at
position 332, a key neutralization determinant for V3-specific antibodies (27).
Yeast surface display is a simple and powerful tool for engineering complex
glycoproteins (28-30), and randomly mutated yeast-displayed libraries have been
used for conformational epitope mapping of EGFR (31), West Nile virus (32), and
influenza hemagglutinin (33).
In this chapter, we generate a fine resolution map of the conformational epitope
of antibody VRC01 using a random library of gp120 stripped core mutants.
Mutations can easily cause the protein to misfold and result in an allosteric loss
of binding. The random mutagenesis method requires a conformational control
antibody to guard against these false positive mutations (31), but one is not
readily available for stripped core gp120. To circumvent this problem, we create
a defined mutant panel using information from natural HIV sequence diversity
and a homology model of stripped core. With the panel we map the epitopes of
the CD4 binding site-directed antibodies VRC01, b12, and b13 and elucidate
residues that are energetically important for binding.
38
Antibodies VRCO1, b12, and b13 are competitive with one another and have
nearly overlapping epitopes but differ in their neutralization phenotypes. The
canonical mutation used to tease out CD4 binding site specificity, D368R,
disrupts binding to all three antibodies equally. The epitope maps made with the
mutant panel point to combinations of mutations that can discriminate between
these three antibodies. The new multi-mutation constructs may be a more
sensitive screening tool for binding specificity.
Results
Mapping the VRC01 epitope with a random mutagenesis library. The broadly
neutralizing antibody VRC01 was used as a test case for epitope mapping with
yeast-displayed stripped core gp120. The crystal structure of the antibody in
complex with gp120 from HIV strain 93TH057 has been solved (PDB ID: 3NGB)
(34) and can be used to validate mapping results. The antibody is one of the
broadest and most potent neutralizers of HIV yet discovered, and a deeper
understanding of its function many help inform immunogen design and guide
the search for new neutralizing antibodies.
We first mapped the VRC01 epitope using a library of randomly mutagenized
gp120 molecules, as described previously (31). The library was generated by PCR
with an error-prone polymerase under conditions of low mutagenesis. Its
diversity was 7x10 7, sufficient to sample all possible single amino acid
substitutions at each of 285 positions. Yeast displaying the gp120 library were
incubated with 50 nM VRC01 and sorted for loss of binding to the antibody. 0.6%
of the original library was collected in this first sort, and a second round at the
same antibody concentration further enriched the population (Figure 3.1A).
A third round of sorting was performed with a conformational control antibody
to exclude allosteric mutations that ablate VRC01 binding from afar without
making direct contact with the antibody. gp120-specific antiserum from a mouse
immunized with the stripped core construct was used as a conformational
control in this third sort. The antiserum bound more weakly to gp120 whose
39
structure had been disrupted by heating in the presence of 8 M urea (Figure
3.1D), suggesting that the antiserum recognized some conformational epitopes
on the protein. When the yeast library was incubated with antiserum, only 5.5%
of cells retained binding and were collected (Figure 3.1C). This suggests that the
majority of clones isolated in the VRC01 sorting rounds contained allosteric
mutations, and underscores the importance of a conformational control antibody
for this method.
Thirty-two clones from the library were sequenced after the third sort. Mutations
that introduced or replaced a cysteine or proline were removed from the
analysis, as were positions that were mutated only once. The remaining 21
mutations occurred at five positions: 280, 368, 456, 458, and 469 in standard HxB2
numbering (35). These positions fall within the VRC01 footprint as defined by
the antibody contact residues in crystal structure 3NGB (Figure 3.1E). Of the nine
single mutations that were discarded from the analysis, one (N278Y) resides in
the VRC01 epitope and seven came from clones that also contained either a
VRC01 epitope mutation or a change to/ from cysteine / proline.
The random mutagenesis epitope map is consistent with the structural epitope,
but the method requires access to a cell sorter and repeated rounds of sorting,
sequencing and analysis. To enable higher-throughput analysis of antibodies and
antisera, we constructed a small mutant library useful for rapid mapping of
gp120-specific antibodies simply by flow cytometry, without sorting, sequencing,
or the need for a validated conformational control antibody.
Making a defined mutant panel. We generated a homology model of stripped
core gp120 with the protein structure modeling program MODELLER (36), and
31 evenly-spaced surface residues were chosen for mutagenesis. Amino acid
substitutions were selected to be highly disruptive based on differences in
polarity, charge, and size. To guard against allosteric mutations, only amino
acids present with at least 0.1% frequency in the Los Alamos National
Laboratory HIV Sequence Database were considered at each position. Thus,
prevalence in nature was used as a proxy for properly-folded gp120. The only
exception to this rule is the mutation D368R-a known CD4 binding site
40
disruptor commonly used to screen for CD4-directed antibodies from patient
samples (37)-which is included in the library though absent from the database.
The location and identities of selected amino acid substitutions are shown in
Figure 3.2A and Table 3.1.
Mapping the VRC01 epitope with the mutant panel. The epitope of VRC01 was
mapped with the defined mutant panel. VRC01 was incubated with the library at
300 pM, and binding signal analyzed using a plate-reader cytometer. Six
mutations were found to disrupt binding to VRC01: the expected D368R, as well
as S365K, T455E, G459E, D461I, and G473R (Figure 3.2C). These residues are
consistent with the map from the random mutagenesis library, and nicely
overlay the antibody footprint as defined by crystal structure 3NGB (Figure
3.2B). The positions identified by the epitope map fall along the ridge comprised
of the CD4 binding loop, variable loop V5, and beta strand p24, all known to be
energetically important for VRC01 binding (38). Two mutations in loop D, T278K
and K282V, were found not to disrupt VRC01 binding to stripped core despite
the fact that the loop appears to make contact with VRC01 in the native trimer
(34).
An important consideration is that the quality of the map generated with this
panel is sensitive to the antibody concentration in the assay (Figure 3.2D). When
the library was incubated with VRC01 at 30 nM, only D368R showed any
appreciable disruption of binding. Incubations at 0.3 and 3 nM resulted in a clear
distinction between the six mutations listed above and all others. At lower
VRC01 concentrations, this difference diminished as the overall binding signal
decreased across the board. Fitting binding isotherms to the mapping data for
each mutant illustrates this phenomenon (Figure 3.3A, Table 3.2). Non-epitope
mutants have equilibrium binding constants between 0.1-1 nM, similar to that of
wild-type stripped core gp120. By contrast, the KD for D368R was greater than
100 nM, which explains why the mutant fails to bind at even the highest tested
concentrations of VRCO1. A cluster of weaker mutations-the same S365K,
T455E, G459E, D461I, and G473R identified above-have a KD an order of
41
magnitude higher than wild-type gp120. There is a narrower window of
antibody concentration for identifying these five mutations in our screen.
Mapping other anti-gp120 antibodies. Two other important CD4-directed antigp120 antibodies, b12 and b13, were then mapped with the mutant panel.
Antibody b12 is neither as broadly cross-reactive nor as potent as VRC01 but for
many years was the gold standard of HIV neutralization. Antibody b13 also
binds the CD4 binding site, but its offset footprint renders it incapable of binding
trimeric gp120 envelope on HIV (20) and thus incapable of neutralizing the virus.
Antibodies b12 and b13 were mapped over a range of concentrations near their
KD (Figure 3.3B, 3C; Table 3.2, and Table 3.3). Mutations D368R, R419G, and
T455E strongly disrupt binding to b12, while mutations K282V, S365K, and 1467K
have weaker effects. For b13, D368R is strongly disruptive and R419G is
marginally so. When these mutations are mapped onto the homology model,
they are consistent with the contact residues from the corresponding crystal
structures (2ND7 and 3IDY, respectively) (Figure 3.3E, 3F).
The residues identified by epitope mapping by yeast surface display are also
consistent with those identified by previous mapping efforts reported in the
literature. Amino acid substitutions at positions 368 and 457 were found to
disrupt binding of CD4 to gp120 (39). Alanine scanning identified residues 365,
455, 458, and 473 as important for b12 binding (40), and residues 474-476 for
binding of CD4 binding site-reactive antibodies cloned from infected individuals
(41). Computational analysis of envelope sequences found that conservation at
positions 364, 369 and 461 were indicative of b12 neutralization (42).
Multi-mutation constructs that bind differentially to VRC01, b12, and b13.
D368R ablates binding to all three CD4 binding site-directed antibodies tested,
despite variation in their epitopes and important phenotypic differences. The
fine resolution conformational epitope maps suggest that some of the weaker
mutations may be more specific disruptors than D368R, but they are very
sensitive to antibody concentration. We decided to combine the weaker
42
mutations to see if they can act cooperatively to knock out binding over a
broader antibody concentration range while maintaining specificity.
Six constructs with either two or three mutations in combination were made by
site-directed mutagenesis and displayed on the surface of yeast (Table 3.4). The
constructs were first titrated against gp120-specific mouse antiserum (Figure
3.4A). For all constructs, the binding thermodynamics were similar to that of
D368R, suggesting that the multiple mutations do not grossly disturb the overall
protein structure.
The constructs were then titrated against antibodies VRC01, b12, and b13 (Figure
3.4B-D). Multi-mutation construct MM3 (S365K, T455E, G473R) knocked out
binding to VRC01 and b12 to a greater than or equal extent as canonical mutation
D368R, but had no effect on binding of b13. This construct may be useful for
screening out b13-like antibodies that approach the CD4 epitope at an angle not
conducive to neutralization. Multi-mutation construct MM5 (T459E, G473R) is
even more specific for VRCO1. It abrogates binding to VRC01 more completely
than D368R, but has negligible effects on both b12 and b13 binding. This
construct may be useful for identifying precisely VRC01-like antibodies from
patient samples or immunization antisera.
Discussion
There is much effort in the HIV vaccine field to modify gp120 to make subunit
vaccine immunogens, and as bait to isolate and characterize antibodies from
patient samples. Yeast surface display is a simple and powerful tool for
engineering complex glycoproteins (28-30), and random mutagenesis yeastdisplayed libraries have been used for conformational epitope mapping of EGFR
(31), West Nile virus (32), and influenza hemagglutinin (33) previously.
In this work we map the epitope of the broadly neutralizing antibody VRC01
using a random mutagenesis library screened by FACS. This is a convenient and
established way to generate fine resolution maps of conformational epitopes (31,
43
43-45). In the method, a random library of gp120 variants with single point
mutations is displayed on the surface of yeast and sorted for loss of binding to
the antibody. Mutants that disrupt binding likely form part of the epitope. We
isolated mutations at five positions-280, 368, 456, 458, and 469-that are
consistent with the known contact area of VRC01 from crystallography.
Because it can be difficult to distinguish mutations in the epitope from allosteric
mutations that disrupt the binding site at a distance, the random library method
requires a conformational control antibody that binds non-competitively
elsewhere on the protein and can report whether the molecule is correctly folded.
In this work we relied upon gp120-specific mouse antiserum collected in-house
to eliminate mutations that allosterically disrupt VRC01 binding. The small
fraction of the library (5.5%) that retained binding to serum after having been
sorted for loss of binding to VRC01 underscores how critical the conformational
control is.
A small mutant panel such as we constructed in this work has both advantages
and disadvantages relative to the random library screening approach. The initial
effort to identify a set of ~30 mutants is balanced against the relative speed with
which a given antibody can then be tested. A defined library of rationally chosen
mutations, though requiring more work to make, can be screened and analyzed
without any DNA sequencing, giving results in just a few hours.
VRC01 was mapped with the mutant panel, and the results were consistent with
both the random mutagenesis mapping and crystallography. It is important to
note that the quality of the map from the defined mutant panel is sensitive to the
antibody concentration in the assay. This sensitivity is less evident with random
library mapping because only strongly disrupting mutations are isolated by that
method. By contrast, each mutant in the panel, although chosen to be disruptive,
will have a particular equilibrium binding constant. At too high a concentration
all the isotherms are saturated and will show similar binding signals. At too low
a concentration, noise will make it impossible to distinguish strong from weak
binders.
44
Two other anti-gp120 antibodies, b12 and b13, were mapped with the mutant
panel, and the results agreed with crystallography. Crystal structures give
accurate and detailed spatial data; epitope mapping can provide additional
information about epitope energetics. The three mapped antibodies all bind the
CD4 binding site but have significantly different neutralization phenotypes.
VRC01 is broadly neutralizing; b12 is weakly neutralizing; and b13 is nonneutralizing. Our epitope map highlights the different energetically important
residues in the epitopes of these three antibodies.
One shared mutation among the three epitope maps is D368R, the canonical CD4
knockout mutation. This mutation is widely used in the HIV vaccine field to test
the specificity of antibodies elicited by immunization and to screen patient
samples for new neutralizing antibodies (4, 22, 46-48). Our mapping titrations
indicate why this mutation is so favored. It abrogates binding to all CD4-directed
antibodies, and shifts the KD to such a degree that the effect is observed
regardless of the antibody concentration. This is particularly useful when
screening polyclonal antisera in which the precise concentration of individual
antibodies is unknown.
But the fact that D368R is common to all three epitope maps is also a detriment to
its use, because the mutation cannot distinguish between antibodies with similar
epitopes but significantly different phenotypes. Our mapping data suggests that
combinations of some of the weaker mutations may have equal potency but be
more specific disruptors than D368R. The multiple mutation construct MM3
(S365K, T455E, G473R) knocks out binding only to VRC01 and b12, but not b13.
The construct MM5 (T459E, G473R) abrogates binding to VRC01 alone. These
constructs may be useful for identifying precisely VRC01-like antibodies from
patient samples or immunization antisera. No binding assay can replace a
neutralization assay to determine antibody function, but improved specificity
can help improve the initial screen, particularly in cases where there may be
limiting quantities of serum (e.g. mouse immunization).
45
Materials & Methods
Making the random mutagenesis library. A random mutagenesis library was
generated by PCR with Mutazyme II error-prone polymerase (GeneMorph II Kit,
Agilent, Santa Clara, CA). To ensure a low mutagenesis rate, 400 ng target DNA
was amplified for 24 cycles with flanking primers. The re-amplified PCR product
was combined with digested pCHA vector and electroporated into EBY100 yeast,
where the full plasmid was reassembled by homologous recombination (49). The
final library size was estimated to be 7x10 7.
Homology modeling with MODELLER. The protein structure modeling
program MODELLER (36) (http:/ /salilab.org /modeller) was used to generate a
homology model of the yeast-displayed stripped core YU2 gp120. Though
several crystal structures of gp120 exist in the Protein Data Bank, none is an exact
analog for the yeast-displayed molecule. We used the crystal structure of gp120
bound by VRC01 (PDB ID: 3NGB, strain 93TH057) as a structural template,
though basing the model on alternate gp120 templates did not alter the model
significantly (Figure 3.5).
Algorithm for selecting surface residues and amino acid substitutions for the
mutant panel. The homology model of stripped core gp120 was rendered in
PyMOL (PyMOL Molecular Graphics System, Version 1.2r3pre, Schr6dinger,
LLC.) and 31 evenly-spaced sites with large solvent-exposed surface areas were
selected for mutagenesis. Amino acid substitutions were chosen to be highly
disruptive of binding, yet not so disruptive as to misfold the protein. We
evaluated each amino acid substitution on three biophysical axes: polarity
(hydophobic, neutral, or polar), charge (negative, neutral, or positive), and size
(greater than or less than 160 A2 surface area) (50-52). A value of +1 was assigned
for every unit step on each axis, and the total disruption score was calculated by
summing the net differences between two amino acids. Higher scores correspond
to more disruptive changes. For example, a mutation from glutamic acid to
tyrosine would be scored +3 total: +2 for the change from polar to hydrophobic,
+1 for the change from negative charge to neutral, and +0 for size because both
surface areas are greater than 160 A2.
46
To guard against allosteric mutations, we excluded all amino acid substitutions
that are present at less than 0.1% frequency in the Los Alamos National
Laboratory HIV Sequence Database (http:/ /hiv.lanl.gov). An extremely
disruptive mutation that that destroys the fold of gp120 would likely ablate viral
fitness too. The only exception to this rule is the mutation D368R, a frequently
used CD4 binding site disruptor, that is included in the library though absent
from the database.
Substitutions were made by site-directed mutagenesis. The mutant panel was
limited to 31 clones and one wild-type stripped core gp120 so that three
mappings could be performed on a single 96-well plate on a high-throughput
flow cytometer. A library with more members would provide a finer resolution
map at the cost of overall assay throughput.
Flow cytometry. Anti-gp120 antibodies were either produced from HEK cells
(the plasmid for VRC01 was a generous gift from John Mascola, Vaccine
Research Center, NIAID, NIH) or a generous gift from Dennis Burton. Yeast cells
displaying protein were incubated with anti-gp120 antibodies and the antiepitope tag antibody chicken anti-CMyc (Gallus Immunotech, Fergus, Ontario).
Secondary labeling was performed with goat anti-mouse, -human, or -chicken
antibodies conjugated to Alexa Fluor dyes (Life Technologies, Carlsbad, CA) at
recommended dilutions. Individual samples were analyzed on an Accuri C6
cytometer (BD Accuri Cytometers, Ann Arbor, MI); 96-well plates were run on a
FACSCalibur HTS with a high-throughput plate sampler (Becton Dickinson,
Franklin Lakes, NJ); sorting was performed on a MoFlo instrument (Beckman
Coulter, Brea, CA).
gp120-specific mouse serum. Stripped core gp120 was secreted from HEK 293
cells as a fusion to mouse IgG2a Fc, and purified by sequential TALON metal
affinity chromatography (Clontech, Mountain View, CA) and protein A affinity
chromatography (Genscript, Piscataway, NJ). The protein was mixed with
AddaVax (Invivogen, San Diego, CA) an oil-in-water emulsion adjuvant similar
to MF59 licensed for use in Europe. BALB / c mice (Charles River Laboratories,
Wilmington, MA) were immunized with 7.5 pmol gp120 subcutaneously at the
47
base of the tail, and boosted again six weeks later. Antisera were titrated on
gp120-displaying yeast (Figure 3.1D). A dilution of 1:200 saturated the yeastdisplayed gp120, and was used as the working dilution for antisera in all
experiments in this report. To demonstrate conformational specificity of the
antisera, titrations were performed on yeast displaying gp120 that had been
incubated with 8 M urea and heated for 30 minutes at 82C to denature the
protein. In this assay, binding MFU were scaled by display MFU (CMyc epitope
tag) to ensure that the loss of binding signal was not due to fewer total proteins
present on the yeast cell surface.
48
A'-
B
C '0
10
10 -
0.6%
e01'
L1Displa-y
E
-e-Core gp12
80
Core
'Denured
* 0
60'.
U.
40
200
Mutation
Freq.
N280D
D368G
D368E
R456G
G458D
R4691
5
1
1
8
3
3
n.-?
Yoo
1000
10000
Dflutio (fold)
100000
Figure 3.1. Epitope mapping of VRC01 with a random mutagenesis library. (a-c)
Sequential FACS sorting plots of the library incubated with (a) 50 nM VRC01, (b)
50 nM VRC01, and (c) a 1:200 dilution of anti-gp120 antiserum. The fraction of
total cells collected is shown on each plot. (d) Titration of serum from an
immunized mouse on yeast displaying gp120 (solid) and yeast displaying gp120
whose structure has been disrupted by incubation with 8 M urea (dashed). (e)
Mutated residues from 32 sorted clones highlighted on the homology model of
stripped core gp120. Residues in red are the sequenced mutations; residues in
yellow are the contact region of VRC01 as determined from crystal structure
3NGB. The accompanying table lists the mutations and their frequency.
49
50
0461
A
B
1so-t
D
C
100
FM12V
o
on
0A
0800
0
000
0A
T2WUL.
80o -AA8
T240K
30 nM
3 nM
0.3 nM
0 .1nM
0
120
A
20
o 8~
0
20
-
00
0
A
AA8
88
00
8 0 0 00
* 000
A AA
p
08
0RI~~
~jO
' '
- - - - - - - - - . .
03M0§800 1o 80.010!1
319
0808001800
a
0
0
-
'
AA AAA
00
00
einnowu
Figure 3.2. Epitope mapping of VRC01 with a defined mutant panel. (a)
Homology model of stripped core gp120 with mutant panel residues shown in
black. The contact surface of VRC01 from crystal structure 3NGB is shown in
yellow. (b) Close-up of the VRC01 epitope with select residues in the mutant
panel labeled and shown in red. (c) Binding of 300 pM VRC01 to the mutant
panel as measured by flow cytometry. Mutations are arranged in decreasing
order of binding MFU. (d) VRC01 binding to the mutant panel over a range of
VRC01 concentrations. At the highest concentration (30 nM) only D368R disrupts
binding. At the lowest concentration (100 pM) it is difficult to distinguish
differences in binding signal. At intermediate concentrations, an epitope is well
defined by loss of binding to: S365K, T455E, G473R, G459E, D461I, and D368R.
51
52
C
A
A
1014
100
40-4
20O1 01
D
10
100
D
01
A01
VRC.1 [nMAI
E
3NGB
1
10
S
b12{nM
1
10
b13 (pM)
10
F
2ND7
31DY
Figure 3.3. Epitope mapping of CD4 binding site-directed antibodies. (a-c)
Binding of antibodies VRC01 (a), b12 (b), and b13 (c) to the defined mutant panel
at a range of concentrations. Monovalent equilibrium binding isotherms are fit to
the data. Mutations that do not significantly change the binding thermodynamics
are shown in black. Those that weakly disrupt binding are shown in red with
dashed lines. Mutations that strongly disrupt binding are drawn in solid red
lines. Fitted KD values and classification of mutations are listed in Table 3.2 and
Table 3.3. (d-f) Mutations that affect binding to VRC01 (d), b12 (e), and b13 (f) are
shown on the homology model of stripped core gp120. Mutations that weakly
disrupt binding are colored pink; those that strongly disrupt binding are colored
red. The contact residues of each antibody as defined by crystallography are
shown in yellow (VRC01, PDB ID 3NGB), orange (b12, PDB ID 2ND7), and blue
(b13, PDB ID 3IDY) respectively.
53
54
Serum binding to CD4-KO gp120 mutants
b12 binding to CD4-KO
A
gp120 mutants
B
10
104
Serum (inverse dilution)
10b12 (nM)
VRCO1 binding to CD4-KO gpi20 mutants
CO
D
MU.
Y-
1
VOI (WM)
100
b13 (nM)
Figure 3.4 Binding of anti-gp120 antibodies to selective CD4 knockout mutants
displayed on yeast. Multi-mutation constructs MM3 (blue) and MM5 (red),
stripped core (black), and stripped core D368R (dashed black) are displayed on
yeast, and binding to various antibodies is assayed by flow cytometry: (a) gp120specific mouse antiserum; (b) VRC01; (c) b12; (d) b13. Mutations in multimutation constructs are listed in Table 3.3.
55
56
A
B
E
Exposed surface area
2
B
CS
Standard deviation
C
DD
0
50
I00
1
200
250
Residue (continuous numbering)
Figure 3.5. Homology model generation and evaluation using the protein
structure modeling program MODELLER (36). The stripped core sequence was
aligned to various subsets of known crystal structures of gp120 bound to CD4 or
neutralizing antibodies (tabulated by Zhou et. al.) (34). (a) Homology models of
the yeast-displayed stripped core YU2 gp120 generated with different structural
templates are overlaid in PyMOL: orange, gp120 of the same YU2 strain as the
stripped core (PDB ID: 1RZK, 2QAD); yellow, gp120 bound to VRC01 (PDB ID:
3NGB, strain 93TH057); blue, fourteen templates in the list (those mentioned
already and 3DNL, 3DNO, 3JWO, 3JWD, 2NXY, 2NXZ, 1RZJ, 1G9M, 1G9N,
1GC1, 2NY7, 2B4C). The homology models are self-consistent, and individual
ga341 model quality scores for all three models were 1.0 (53, 54). (b) Variability in
solvent-exposed surface area at each amino acid of the homology models is
plotted. In the top plot, the blue line is the mean surface area of the three models,
whereas the red curves are the mean ± standard deviation. The standard
deviation alone is plotted in the bottom plot. Amino acid residues in both plots
are numbered continuously, rather than the HxB2 convention. Areas of
disagreement between the homology models made from different structural
templates are lettered. As expected, the largest variability was observed in the
variable loops (A, C, E), the truncated bridging sheet (D), and in a highlyconserved, glycosylated, solvent-exposed loop (B).
57
58
Table 3.1. Mutations in defined mutant panel
N99Y
K121T
N197K
V200K
H216Y
K231T
T240K
L265R
Library point mutations
D397F
E269G
R403V
T278K
L405D
K282V
G327R
R412V
R419G
Q337V
G434K
Q344V
S365K
Q442V
S446K
D368R
T455E
G459E
D4611
1467K
G473R
R476V
K490S
59
Table 3.2 Binding affinities of gp120 variants in defined mutant panel to CD4
binding site-directed antibodies (Figure 3.3A-C). Affinities are reported as fitted
KD values ± one standard deviation. Mutations classified as strong disruptors of
binding are highlighted in black. Those classified as weak disruptors of binding
are highlighted in gray.
Mutation
N 99Y
K121T
N197K
V200K
H216Y
K231T
T240K
L265R
E269G
T278K
K282V
G327R
Q337V
0344V
VRC01
0.28±0.11
0.45 ± 0.19
0.40 ± 0.17
0.37 ± 0.18
0.28 ± 0.11
0.39 ± 0.16
0.47 ± 0.21
0.47 ± 0.21
1.12 ± 0.46
0.21 ± 0.08
0.95 ± 0.49
0.53 ± 0.25
0.37 ± 0.14
0.42 ± 0.17
b13
3.87±0.76
3.42 ± 1.24
2.36 ± 1.62
3.00 ± 0.94
3.51 ± 1.33
4.26 ± 1.53
4.64 ± 1.15
3.96 ± 1.22
3.50 ± 1.18
4.17 ± 1.13
3.26 ± 0.52
3.81 ± 1.56
4.02 ± 1.79
5.40 ± 2.16
b12
0.035±0.013
0.024 ± 0.012
0.043 ± 0.020
0.045 ± 0.018
0.034 ± 0.014
0.033 ± 0.012
0.045 ± 0.022
0.021 ± 0.009
0.031 ± 0.017
0.037 ± 0.017
0.028 ± 0.015
0.029 ± 0.009
0.024 ± 0.010
5.52 ± 1.41
S365K
14 .0
17 1
7845
10
D368R
.
25 7.
D397F
0.63 ± 0.32
0.40 ± 0.19
0.055 ± 0.034
4.36 ± 1.61
0.019 ± 0.009
3.47 ± 1.36
0.46 ± 0.22
0.39 ± 0.14
G434K
0.32 ± 0.14
0.37 ± 0.18
3.15 ± 1.28
0.036 ± 0.019
3.51 ± 1.04
0.037 ± 0.018
38.12 ±35.34J~'"
-3.64 ± 1.39
0.034 ± 0.012
Q4 42V
0.58 ± 0.25
0.031 ± 0.018
4.50 ± 1.56
S446K
0.33 ± 0.14
0.040 ± 0.011
4.12 ± 1.61
R40 3V
L405D
R412V
R419G
3.52 ± 0.83
T455E
W0-0400
D461I
1467K
3.30± 0.79
045 0.020
G459E
1.83
0
3.43
0 12
1.23
0.023 ± 0.007
4.65 1.18
6.74t 1.87
1.06
G473R
±
R476V
1.28 ± 0.75
0.032 ± 0.012
2.81 ± 1.02
K490S
0.51 ± 0.24
0.42 ± 0.21
0.063 ± 0.039
0.050 ± 0.019
4.25 ± 1.30
4.96 ± 1.38
WT
60
Table 3.3. Mutations that disrupt binding to CD4 binding site-directed
antibodies
Antibody
VRC01
b12
b13
Highly-disruptive mutations
D368R
D368R, R419G, T455E
D368R
Weak mutations
S365K, T455E, G459E, D461 1,G473R
K282V, S365K, 1467K
R419G
Table 3.4. Point mutations in multi-mutation (MM) constructs for selective
knockout of CD4 binding site-directed antibodies.
Construct
MMI
MM2
MM3
MM4
MM5
MM6
Mutations
S365K, D4611, G473R
S365K, G459E, G473R
S365K, T455E, G473R
T455E, G473R
G459E, G473R
D4611, G473R
61
Works Cited
1. D. R. Burton et al., A large array of human monoclonal antibodies to type 1
human immunodeficiency virus from combinatorial libraries of asymptomatic
seropositive individuals, Proc Natl Acad Sci USA 88, 10134-10137 (1991).
2. D. R. Burton et al., Efficient neutralization of primary isolates of HIV-1 by a
recombinant human monoclonal antibody, Science 266, 1024-1027 (1994).
3. T. Zhou et al., Structural definition of a conserved neutralization epitope on HIV1 gp120, Nature 445, 732-737 (2007).
4. X. Wu et al., Rational design of envelope identifies broadly neutralizing human
monoclonal antibodies to HIV-1, Science 329, 856-861 (2010).
5. E. Falkowska et al., PG V04, an HIV-1 gp120 CD4 binding site antibody, is broad
and potent in neutralization but does not induce conformational changes
characteristic of CD4, J Virol (2012), doi:10.1128/JVI.06973-11.
6. A. F. Labrijn et al., Access of antibody molecules to the conserved coreceptor
binding site on glycoprotein gp120 is sterically restricted on primary human
immunodeficiency virus type 1, J Virol 77, 10557-10565 (2003).
7. M. Moulard et al., Broadly cross-reactive HIV-1-neutralizing human monoclonal
Fab selected for binding to gpl20-CD4-CCR5 complexes, Proc Natl Acad Sci USA
99, 6913-6918 (2002).
8. M. Thali et al., Characterization of conserved human immunodeficiency virus
type 1 gp120 neutralization epitopes exposed upon gpl20-CD4 binding, J Virol
67, 3978-3988 (1993).
9. L. M. Walker et al., Broad neutralization coverage of HIV by multiple highly
potent antibodies, Nature 477, 466-470 (2011).
10. R. Pejchal et al., A Potent and Broad Neutralizing Antibody Recognizes and
Penetrates the HIV Glycan Shield, Science 334, 1097-1103 (2011).
11. L. M. Walker et al., Broad and potent neutralizing antibodies from an African
donor reveal a new HIV-1 vaccine target, Science 326, 285-289 (2009).
12. R. Pejchal et al., Structure and function of broadly reactive antibody PG16 reveal
an H3 subdomain that mediates potent neutralization of HIV-1, ProcNatl Acad
Sci USA 107, 11483-11488 (2010).
13. M. Pancera et al., Crystal structure of PG16 and chimeric dissection with
somatically related PG9: structure-function analysis of two quaternary-specific
antibodies that effectively neutralize HIV-1, J Virol 84, 8098-8110 (2010).
62
14. X. Wu et al., Immunotypes of a quaternary site of HIV-1 vulnerability and their
recognition by antibodies, J Virol 85, 4578-4585 (2011).
15. A. Trkola et al., Human monoclonal antibody 2G12 defines a distinctive
neutralization epitope on the gp120 glycoprotein of human immunodeficiency
virus type 1, J Virol 70, 1100-1108 (1996).
16. T. Muster et al., A conserved neutralizing epitope on gp4l of human
immunodeficiency virus type 1, J Virol 67, 6642-6647 (1993).
17. G. Stiegler et al., A potent cross-clade neutralizing human monoclonal antibody
against a novel epitope on gp4l of human immunodeficiency virus type 1, AIDS
Res. Hum. Retroviruses 17, 1757-1765 (2001).
18. G. Frey et al., A fusion-intermediate state of HIV-1 gp4l targeted by broadly
neutralizing antibodies, Proc Natl Acad Sci USA 105, 3739-3744 (2008).
19. W. R. Schief, Y.-E. A. Ban, L. Stamatatos, Challenges for structure-based HIV
vaccine design, Curr Opin HIV AIDS 4, 431-440 (2009).
20. L. Chen et al., Structural basis of immune evasion at the site of CD4 attachment
on HIV-1 gp120, Science 326, 1123-1127 (2009).
21. C. Sundling et al., High-Resolution Definition of Vaccine-Elicited B Cell
Responses Against the HIV Primary Receptor Binding Site, Sci Transl Med 4,
142ra96-142ra96 (2012).
22. L. M. Walker et al., A limited number of antibody specificities mediate broad and
potent serum neutralization in selected HIV-1 infected individuals, PLoS Pathog 6
(2010), doi:10.1371/journal.ppat.1001028.
23. R. M. Lynch et al., The development of CD4 binding site antibodies during HIV-1
infection, J Virol 86, 7588-7595 (2012).
24. K. J. Bar et al., Early Low-Titer Neutralizing Antibodies Impede HIV-1
Replication and Select for Virus Escape, PLoS Pathog8, e1002721 (2012).
25. A. J.Mcmichael, P. Borrow, G. D. Tomaras, N. Goonetilleke, B. F. Haynes, The
immune response during acute HIV-1 infection: clues for vaccine development,
Nat Rev Immunol 10, 11-23 (2010).
26. X. Wu et al., Selection Pressure on HIV-1 Envelope by Broadly Neutralizing
Antibodies to the Conserved CD4-Binding Site, J Virol 86, 5844-5856 (2012).
27. P. L. Moore et al., Evolution of an HIV glycan-dependent broadly neutralizing
antibody epitope through immune escape, Nat Med (2012), doi:10.1038/nm.2985.
28. E. T. Boder, K. D. Wittrup, Yeast surface display for screening combinatorial
polypeptide libraries, Nat Biotechnol 15, 553-557 (1997).
63
29. E. T. Boder, K. S. Midelfort, K. D. Wittrup, Directed evolution of antibody
fragments with monovalent femtomolar antigen-binding affinity, Proc Natl Acad
Sci USA 97, 10701-10705 (2000).
30. S. A. Gai, K. D. Wittrup, Yeast surface display for protein engineering and
characterization, Curr Opin Struct Biol 17, 467-473 (2007).
31. G. Chao, J. R. Cochran, K. D. Wittrup, Fine epitope mapping of anti-epidermal
growth factor receptor antibodies through random mutagenesis and yeast
surface display, Journalof Molecular Biology 342, 539-550 (2004).
32. T. Oliphant et al., Development of a humanized monoclonal antibody with
therapeutic potential against West Nile virus, Nat Med 11, 522-530 (2005).
33. T. Han et al., Fine epitope mapping of monoclonal antibodies against
hemagglutinin of a highly pathogenic H5N1 influenza virus using yeast surface
display, Biochem. Biophys. Res. Commun. 409, 253-259 (2011).
34. T. Zhou et al., Structural basis for broad and potent neutralization of HIV-1 by
antibody VRC01, Science 329, 811-817 (2010).
35. Korber B, Foley BT, Kuiken C, Pillai SK, Sodroski JG (1998).Numbering Positions
in HIV Relative to HXB2CG. pp. 111-102-111 in Human Retroviruses and AIDS
1998. Edited by: Korber B, Kuiken CL, Foley B, Hahn B, McCutchan F, Mellors
JW, and Sodroski J. Published by: Theoretical Biology and Biophysics Group, Los
Alamos National Laboratory, Los Alamos, NM.
http://www.hiv.lanl.gov/content/sequence/HIV/REVIEWS/HXB2.html
36. A. Sali, T. L. Blundell, Comparative protein modelling by satisfaction of spatial
restraints, Journal of Molecular Biology 234, 779-815 (1993).
37. M. Thali et al., Discontinuous, conserved neutralization epitopes overlapping the
CD4-binding region of human immunodeficiency virus type 1 gp120 envelope
glycoprotein, J Virol 66, 5635-5641 (1992).
38. Y. Li et al., Mechanism of neutralization by the broadly neutralizing HIV-1
monoclonal antibody VRC01, J Virol 85, 8954-8967 (2011).
39. U. Olshevsky et al., Identification of individual human immunodeficiency virus
type 1 gp120 amino acids important for CD4 receptor binding, J Virol 64, 57015707 (1990).
40. R. Pantophlet et al., Fine mapping of the interaction of neutralizing and
nonneutralizing monoclonal antibodies with the CD4 binding site of human
immunodeficiency virus type 1 gp120, J Virol 77, 642-658 (2003).
41.
J. Pietzsch et al., Human anti-HIV-neutralizing
antibodies frequently target a
conserved epitope essential for viral fitness, J Exp Med 207, 1995-2002 (2010).
64
42. S. Gnanakaran et al., Genetic signatures in the envelope glycoproteins of HIV-1
that associate with broadly neutralizing antibodies, PLoS Comput. Biol. 6,
e1000955 (2010).
43. J. B. Spangler et al., Combination antibody treatment down-regulates epidermal
growth factor receptor by inhibiting endosomal recycling, Proc Natl Acad Sci USA
(2010), doi:10.1073/pnas.0913476107.
44. R. Levy et al., Fine and domain-level epitope mapping of botulinum neurotoxin
type A neutralizing antibodies by yeast surface display, Journal of Molecular
Biology 365, 196-210 (2007).
45. T. Oliphant et al., Antibody recognition and neutralization determinants on
domains I and II of West Nile Virus envelope protein, J Virol 80, 12149-12159
(2006).
46. A. Nandi et al., Epitopes for broad and potent neutralizing antibody responses
during chronic infection with human immunodeficiency virus type 1, Virology
(2009), doi:10.1016/j.virol.2009.10.044.
47. G. D. Tomaras et al., Polyclonal B Cell Responses to Conserved Neutralization
Epitopes in a Subset of HIV-1-infected Individuals, J Virol (2011),
doi:10.1128/JVI.05363-11.
48. J.F. Scheid et al., A method for identification of HIV gp140 binding memory B
cells in human blood, Journal of Immunological Methods 343, 65-67 (2009).
49. G. Chao et al., Isolating and engineering human antibodies using yeast surface
display, Nature Protocols 1, 755-768 (2006).
50. R. Grantham, Amino acid difference formula to help explain protein evolution,
Science 185, 862-864 (1974).
51. D. Bordo, P. Argos, Suggestions for "safe" residue substitutions in site-directed
mutagenesis, Journalof Molecular Biology 217, 721-729 (1991).
52. C. Chothia, J. Janin, Principles of protein-protein recognition, Nature 256, 705-708
(1975).
53. F. Melo, R. Sinchez, A. Sali, Statistical potentials for fold assessment, Protein Sci
11,430-448 (2002).
54. B. John, A. Sali, Comparative protein structure modeling by iterative alignment,
model building and model assessment, Nucleic Acids Res 31, 3982-3992 (2003).
65
66
Chapter 4. Eliciting VRC01-like antibodies by vaccination
Background
Though the immune system has evolved to be redundant, most vaccine-induced
protection is mediated by antibodies (1). Glycoprotein gp120 of the envelope
spike of HIV is the target of most antibodies generated upon infection by the
virus (2, 3). HIV's fast mutation rate and its tolerance of sequence diversity allow
the virus to adapt in response to selection pressure from neutralizing antibodies
and evade recognition (4-8). There are, however, regions of the viral envelope
spike that are essential for viral fitness and are therefore highly conserved.
Antibodies against these conserved epitopes-the CD4 binding site (9-13), the coreceptor binding site (14-16), invariant linear regions of the V 1/V2 and V3 loops
(17, 18), a quaternary epitope formed by adjacent gp120 monomers in the
envelope trimer (19-22), a common glycan motif (23), and the membraneproximal external region (MPER) of gp4l (24-26)-can neutralize HIV across
clades. In recent years, our growing understanding of the mechanisms of action
of these broadly neutralizing raises the specter of someday eliciting them by
vaccination (27-33).
Individuals from whom broadly neutralizing antibodies are isolated do not
demonstrate any clinical benefit from having elicited these antibodies (34). The
natural response to infection sheds light on the dynamic interplay between virus
and neutralizing antibodies and helps to explain this phenomenon. Broadly
reactive neutralizing antibodies do not arise until several years after infection (2),
but antibodies exert selection pressure on autologous virus almost immediately
(7, 35, 36). These early antibodies have little heterologous neutralization potential
but tend to be specific for the same finite conserved epitopes that the broadly
neutralizing antibodies eventually recognize (37-41). Longitudinal tracking of
virus and serum samples from infected individuals shows that serum reliably
neutralizes past autologous virus but not present or future autologous strains (5).
This is true even in patients that develop broadly neutralizing antibodies. For
67
example, envelope variants from the individual who elicited antibody VRC01
were not sensitive to this antibody or its somatic variants, even though they are
the broadest and most potent CD4 binding site-directed neutralizing antibodies
known (6). These observations suggest that during persistent infection HIV is
able to sample enough sequence space to escape from any antibody.
During a typical sexual transmission event, however, a single founder virus
initiates the infection (42-44). A broadly neutralizing antibody present at this
evolutionary bottleneck may be able to block HIV with the breadth observed in
vitro. Indeed, in passive immunization experiments performed in macaques,
animals given neutralizing antibodies prophylactically were protected from
SHIV challenge, whereas those given non-neutralizing antibodies or to whom the
antibodies were administered after infection were not protected (45-48). These
and other studies suggest that low doses of antibody, as would be present
following a vaccine, may be sufficient to neutralize the virus in this setting (7,49,
50).
Eliciting neutralizing specificity by vaccination remains a challenge. Analysis of
immune correlates from the recently completed RV144 vaccination trial in
Thailand (51) found that the presence of IgG against the V1 /V2 loop on gp120
was inversely correlated with the rate of HIV infection in individuals who
received the vaccine (52). Some of these antibodies recognized a conserved
sequence in the otherwise variable loop, but it is does not appear that they were
broadly neutralizing (53, 54). Interestingly, plasma IgA reactivity to gp120
correlated with a higher risk of HIV infection, suggesting that antibodies to
irrelevant epitopes may interfere with protective functions of specific antibodies
(52). It may thus be important to focus the immune response only on the desired
neutralizing epitopes.
One way to focus the immune response on a particular epitope is to exclude
undesirable epitopes from the immunogen. Monomeric gp120 has been
extensively engineered by mutation and glycosylation to mask
immunodominant epitopes and focus the immune response on the CD4 binding
site (31, 55-57). However, most of these immunogens fail to elicit CD4 binding
68
site-directed antibodies upon vaccination (58, 59) or elicit non-neutralizing
antibodies (60). Immunization with stable trimers that more precisely mimic the
native envelope spike can elicit neutralizing antibodies in a guinea pig model,
though the specificity of the antibodies is not limited to a single epitope (61, 62).
A second way to steer the immune response is to administer multiple
immunogens that differ except for a shared epitope, such that only B cells
specific for the common epitope are boosted. In one elegant study, the MPER
peptide epitope of neutralizing antibody 2F5 was grafted onto protein scaffolds
selected to ensure that the peptide retained its proper conformation (63).
Sequential immunization of mice with the heterologous epitope scaffolds did
elicit 2F5-competitive antibodies, though the antibodies were non-neutralizing
because they did not engage the lipids flanking the MPER peptide on the virus
(64, 65). Similar work has been done with combinations of variable loop V3
immunogens grafted onto cholera toxin (66-68) or displayed on phage (69) to
focus the antibody response on conserved determinants within the variable loop
(70). In a non-vaccine context, a high-affinity cross-reactive antibody to multiple
chemokines was generated by sequential immunization with the chemokines
CCL3, CCL4, and CCL5 (71).
Antibodies against the CD4 binding site are among the broadest and most potent
neutralizers of HIV (12, 13, 17). Sequential immunization to focus the immune
response on the CD4 binding site has been tried with a stabilized gp120
monomer followed by a gp140 trimer from a different strain, but the observed
neutralization of Tier 1 viruses was traced predominantly to V3-directed
antibodies (72). The grafting method described above has proven difficult for this
epitope since the CD4 binding site is discontinuous and conformational. One
scaffold structure has been reported in the literature (73), and a hybrid HIV /SIV
gp120 that has been "resurfaced" to remove all non-CD4 binding site epitopes is
used to characterize antibodies (12), but no immunization results with either of
these molecules have been published.
The task of generating diverse immunogens that share a common epitope is well
suited to yeast surface display. The technology, coupled with random
69
mutagenesis and flow sorting, enables the rapid engineering large, complex
glycoproteins in the absence of detailed structural data (74-78). Yeast display is
sensitive enough to handle delicate epitopes, and has been used to map the
conformational epitopes of antibodies on a variety of target antigens (79-81).
In this chapter, we combine the yeast display system described in previous
chapters (Figure 4.1A) with semi-random mutagenesis to generate a panel of
immunogens based on stripped core gp120 that are diverse except for the shared
VRC01 epitope. We immunize mice with these immunogens, and develop yeastdisplay methods for analyzing the strength, specificity, and breadth of the serum.
Mice immunized sequentially with the diversified immunogens generate an
antibody response entirely focused on the CD4 binding site and that reacts with
stripped core from multiple clades of HIV. In ongoing work, hybridomas are
generated from these mice and monoclonal antibodies are isolated to assess their
neutralization potential and genetic lineage.
Results
Generation of diversified immunogens by semi-random mutagenesis.
Immunogen diversity was concentrated on surface residues because the
conformational CD4 binding site is sensitive to allosteric mutations that disrupt
the overall fold of the molecule. A homology model of stripped core gp120 was
generated with the protein structure modeling program MODELLER
(http:/ /salilab.org /modeller) (82), and 43 evenly-spaced, solvent-exposed
surface residues not in the CD4 binding site were selected for diversification
(Figure 4.1B and Table 4.1).
Diversity was introduced by polymerase chain reaction with degenerate
oligonucleotide primers. Segments of the gp120 gene were amplified with
primers containing random nucleotides (NNB) at chosen positions then pieced
together by overlap-extension PCR to reconstitute the full-length gene (Figure
4.1C). A library of 5x10 7 unique clones was transformed into yeast and displayed
by standard methods (83). The library was screened twice by FACS for binding
70
to 10 nM then 50 nM VRC01 to enrich for clones that retained correct
presentation of the CD4 binding site (Figure 4.1D). Based on the degree of library
oversampling (2.35x) and percentage of cells isolated at each sort (0.0067% and
7.1%, respectively), we estimate that 200-500 unique clones were collected (84,
85). Plasmid DNA from sorted cells was isolated and 288 variants were
sequenced. 20 mutants with high diversity were then secreted from HEK 293
cells as mouse Fc fusions, purified by sequential metal affinity and protein A
chromatography, and tested for binding to VRC01 scFv displayed on yeast.
Three gp120 variants that retained binding to VRC01 in this format were selected
as candidate immunogens (Figure 4.1E), henceforth called Clones A, B, and C.
Fitted KD values were 3.5 nM (stripped core), 6.2 nM (Clone A), 1.5 nM (Clone B),
and 12.2 nM (Clone C).
Characterization of diversified immunogens. The full amino acid sequences of
the four candidate immnogens were analyzed. 43 of 285 amino acids were
diversified, with a pairwise sequence identity of 13.6% at these positions (Table
4.1). The CD4 binding site-directed antibodies VRC01, b12, and b13 were titrated
against the individual immunogens displayed on yeast (Figure 4.2). Stripped
core binds with similar affinity to all three antibodies, whereas the mutants
exhibit varying degrees of weaker binding to b12 and b13 against which they
were not sorted. VRC01 neutralizes greater than 90% of circulating HIV isolates
(12). Antibody b12 neutralizes just 40% of circulating isolates (9, 10). Antibody
b13 binds the CD4 binding site on monomeric gp120 but not on trimeric
envelope and is thus incapable of neutralizing the virus (27).
Immunization of mice with diversified immunogens. BALB/c mice were
immunized intranasally with HEK 293-produced immunogens fused to mouse
Fc (Figure 4.3A) and CpG oligonucleotide adjuvant as described elsewhere (86,
87). Both the Fc and the adjuvant are necessary to generate a serum antibody
response to the antigen by this route (Figure 5.1A). Immunogens were
administered every two weeks at a total protein dose of 50 pmol (6 gg Fc-fusion,
equivalent to 3 yg stripped core gp120). Animals were divided into three groups
(Figure 4.3B). One group (Sequential, four mice) was immunized sequentially
71
with each of the four immunogens. A second group (Parallel, four mice) received
all four of the immunogens at each administration. The third group (Core-Only,
two mice) was given three sequential doses of the stripped core immunogen.
Serum was collected every week.
Characterization of serum binding to individual immunogens. Serum binding
to the various immunogens was assayed by cell-surface titration on yeast. Yeast
displaying each of the immunogens were incubated with serial dilutions of
serum. Binding was measured by flow cytometry. Representative curves for each
of the three immunization groups are shown in Figure 4.3C-E. Sequentially
immunized mice bind similarly to each of the four immunogens, a result
consistent with serum that recognizes a shared epitope presented on all clones
(Figure 4.3C). Mice immunized with all four immunogens in parallel show a
broader spread of binding affinities, a result consistent with dominant serum
specificities that are present on some immunogens but not others (Figure 4.3D).
In this example, it appears that the mouse generated a response to an epitope on
stripped core that is not present on Clones A or C. Serum from mice immunized
only with stripped core bind strongly to stripped core but not to the other
immunogens, a result consistent with the design objective that the four
immunogens not share many epitopes (Figure 4.3E).
Serum binding to gp120 mutant D368R. Serum specificity for the CD4 binding
site was determined in two ways. First, the binding of serum to the stripped core
mutant D368R, which is known to disrupt several CD4 binding site-directed
antibodies (88), was compared to binding to unmutated stripped core (Figure
4.4A). Sequentially immunized mice exhibit low fractional binding to D368R
(0.09-0.30 after four immunizations), similar to VRC01 at saturating
concentrations. Three of four mice immunized from the parallel group are very
sensitive to D368R after three immunizations (0.14-0.37 fractional binding) but
lose this specificity after the fourth immunization (0.40-0.54 fractional binding).
The two mice immunized solely with stripped core bind well to D368R (0.56 and
0.89 fractional binding), suggesting that some portion of the serum response is
specific for the disrupted epitope but much is directed elsewhere.
72
Serum competition with VRC01 by ternary complex assay. A second method
developed to measure specificity is a ternary complex assay. In this assay
(Figures 4.4B, 4C), yeast displaying the scFv of VRC01 is incubated with soluble
stripped core gp120 and mouse antiserum. The formation of a ternary complexa sandwich of VRC01, gp120, and mouse serum-can only occur if the serum and
VRC01 do not compete. A ternary complex is detected as the presence of mouse
antibodies on yeast by flow cytometry and reported as the percentage of positive
yeast cells. 40-50% of induced yeast lose their plasmid and do not display any
protein, so 50-60% positive cells is the maximum observable signal in the assay.
In the first iteration of the assay (Figure 4.4B), the gp120 is pre-incubated with
mouse serum before introducing the yeast to resolve whether the serum binds to
the VRC01 epitope on gp120. If so, the gp20-serum complex will not be captured
on VRC01-displaying yeast. When this experiment was performed, no binding
signal was observed for any of the serum samples (Figure 4.4D), suggesting that
all sera contained some fraction of VRC01-competitive antibodies. This result is
consistent with the D368R binding results discussed above.
In the second iteration of the assay (Figure 4.4C), the gp120 is loaded onto the
VRC01 yeast before the serum is introduced to determine whether the serum
contains any specificity other than to the CD4 binding site. Mice immunized with
just stripped core (mean = 54% ± 2%) or with all four immunogens in parallel
(mean = 32% ± 5%) develop antibodies that bind to other epitopes on the gp120
(Figure 4.4E). However, mice immunized sequentially do not (mean = 1.0% ±
0.3%). These experiments suggest that the entirety of the gp120-specific serum
response elicited by vaccination with four sequential immunogens is VRC01competitive.
Serum binds to stripped core gp120 from multiple clades. Genes for stripped
core from four other clades of HIV were synthesized and the proteins displayed
on yeast. Binding to the five strains-92UG037 (Clade A), 98ZADU156 (Clade C),
92UG021 (Clade D), 93BR037 (Clade F), and YU2 (Clade B, original stripped
core)-was used to simulate breadth of recognition of multiple virus subtypes
(Figure 4.5A, 5B, Table 4.2). The strains are all Tier 2 viruses that are commonly
73
used in representative neutralization panels. VRC01, b13, and serum from the
sequential group (day 55, after four immunizations) were tested for binding to
the panel (Figure 4.5C). VRC01 bound well to all except for the Clade D strain,
which has a mutation at the base of the V5 loop (Figure 4.5A) and is resistant to
multiple other neutralizing antibodies (12, 89). b13 bound poorly to all except for
stripped core YU2. Serum from all four sequential mice bound to all five gp120
strains, suggesting that they do in fact recognize a shared epitope. We do not yet
know how well this panel of stripped cores reflects natural diversity on the
envelope spike or predicts broad cross-reactivity of antibodies, but the above
results are encouraging.
Isolation of monoclonal antibodies. Serum was tested for neutralization by the
standard TZM-bl assay by Emilia Falkowska of Dennis Burton's lab. In this
assay, luciferase-expressing CD4* cells are exposed to virus that has been
incubated with antibody (90). The samples did not neutralize any of four viruses
tested, though quantification of gp120-specific antibodies in serum by sequential
immunoprecipitation suggested that the antibody concentration may have been
too low to observe any effect in the assay (Figure 4.6). To isolate monoclonal
antibodies, a mouse from the sequential immunization group were boosted with
150 pmol stripped core at day 237 and 150 pmol of Clone A at day 251, and
hybridomas made by splenic myeloma fusion. As of this writing, several
promising hybridoma clones that secrete antibodies that match the affinity,
specificity, and breadth of serum are being produced (Figure 4.7).
Discussion
Despite much effort over the past twenty-five years there is no successful HIV
vaccine. The primary culprit is the virus' propensity to mutate rapidly and the
ability of gp120 on the envelope spike to accept amino acid diversity, which
allows the virus to escape from nearly any antibody elicited against it (4-8). A
growing handful of antibodies, termed broadly neutralizing antibodies, have
been discovered that recognize conserved sites on the envelope spike such as the
74
CD4 binding site, which are required for viral fitness (9-21, 23-26, 30). These
antibodies potently neutralize HIV from multiple clades. Patients who elicit
broadly neutralizing antibodies over the course of infection are unable to control
the virus (34), but passive immunization studies suggest that broadly
neutralizing antibodies present at transmission, when only one or two virions
cross the mucosal barrier (42-44), may be sufficient to provide protection (45-48).
The recent explosion of structural information about neutralizing antibodies and
how they bind (2 7-30, 32, 33) has spawned much interest in structure-guided
vaccine design, in which detailed knowledge of the antigen's interaction with a
particular antibody is used to design immunogens to elicit this specificity (31).
This has taken the form of selective mutagenesis and glycosylation of the
immunogen (55, 56, 58-60, 91), as well as computational design and stability
engineering (61, 62, 92).
A compelling way to steer the immune response to a desired epitope is to
immunize with heterologous immunogens that are diverse except for a single
conserved epitope. In this strategy, only B cells specific for the conserved epitope
are boosted. This method has been used to elicit a cross-reactive antibody to
multiple chemokines (71), focus an immune response to conserved determinants
in the V2 loop of HIV (66-70), and generate antibodies specific for the 2F5 epitope
determinant on gp4l (63). In this last work, the short peptide epitope of interest
was grafted onto several heterologous scaffolds guided by protein structure
calculations such that the peptide retained its bound conformation. Mice
immunized with these diverse scaffolds elicited antibodies to the conserved
grafted peptide.
The CD4 binding site is more difficult to graft because it is a conformational
epitope comprised of four polypeptides (V5 loop, CD4 binding loop, loop D, and
strand p24) that must all be in the correct orientation relative to one another. To
date, just a single epitope scaffold of the CD4 binding site has been reported (73).
Another construct, resurfaced core RSC3, a hybrid HIV/SIV gp120 whose surface
residues have been replaced with their SIV analogues (12), is commonly used to
isolate CD4 binding site-directed antibodies from patient samples. These
75
molecules demonstrate that it is possible to engineer the surface milieu yet retain
the functionality of the CD4 binding site, but no immunization results with these
molecules have been reported.
In this work, a stripped core gp120 derived from strain YU2 is displayed on yeast
and used as a scaffold for immunogen design. Rather than grafting the native
epitope onto a new scaffold, we diversified the surrounding protein by semirandom mutagenesis to create an immunogen library. Previous experience with
stripped core suggested that the conformational integrity of the CD4 binding site
is sensitive to allosteric mutations, so a homology model of the core was used to
target the diversity to surface-exposed positions not in the CD4 binding site
(Table 4.1 and Figure 4.1B). This limited structural information is convenient but
not strictly necessary. For a protein with unknown structure, low mutagenesis
libraries can be sorted to ensure the target epitope is unaffected then shuffled
together to generate a high-diversity library.
Three immunogens were isolated from the library after sorting for binding to
VRC01, analyzing sequences for diversity, and confirming that the mammalianproduced protein was well behaved. The VRC01 concentrations used to sort the
library (10-50 nM) are a log higher than the equilibrium binding affinity for the
antibody to stripped core. We expressly selected for clones that retained binding
to VRC01, not those that bound better to VRCO1. The fitted KD values for the four
immunogens are 3.5 nM (stripped core), 6.2 nM (Clone A), 1.5 nM (Clone B), and
12.2 nM (Clone C).
The rationale is that the naive B cell repertoire of a vaccine recipient will not
contain an exact replica of VRCO1. VRC01 has a high degree of somatic
hypermutation and other features that contribute to its binding to gp120 (28). We
hypothesize that these attributes are artifacts of adaptation to a persistent
infection in vivo, not requirements for neutralization, and that the immune
system will be able to evolve the same specificity de novo if given the chance.
Immunogens with enhanced affinity for VRC01 lose fidelity to the native epitope
and may favor one B cell receptor with a particular germ line immunoglubulin
gene and specific VDJ recombination to the detriment of other possible solutions.
76
We characterized the immunogens in vitro by testing their binding to the CD4
binding site-directed antibodies VRC01, b12, and b13. VRC01 is the model
antibody whose specificity and breadth we aim to recapitulate in the serum.
Antibodies b12 and b13, though competitive with CD4, are less desirable.
Antibody b12 neutralizes only about 40% of circulating isolates (12), and
antibody b13 is unable to bind trimeric envelope on HIV and is thus incapable of
neutralizing the virus (27). We observe that stripped core binds well to all three
antibodies, whereas the three engineered immunogens retain VRC01 binding but
exhibit weaker or no binding to b12 and b13 (Figure 4.2). The fact that stripped
core retains the undesired epitopes is not cause for its dismissal. Since any
protein surface will function as an epitope for a B cell receptor, what matters is
that the immunogens be diverse such that any one epitope is presented only
once.
Our previous mapping work suggests that residue 419 may be the key mediator
of the differential binding of stripped core-based constructs to VRC01, b12, and
b13. In stripped core, the point mutation R419G was found to be a strong
disruptor of b12, a weak disruptor of b13, and to have no effect on VRC01
binding (Chapter 3). Clone B preserves the wild-type arginine at position 419 and
retains reasonable binding to b12 and b13. Clone A has a tyrosine at that position
and loses binding to both b12 and b13. The leucine in Clone C at position 419
disrupts b12 but not b13. If b13-like antibodies are elicited by vaccination, this
position is a good candidate for future focused randomization.
Another target for further engineering is the V5 loop (amino acids 460-464),
which makes intimate contacts with VRC01 (28). It is hypothesized that VRC01
tolerates diversity in the V5 loop because the polypeptide rests in the cleft
between the heavy and light chains of the antibody. Introducing amino acid
diversity at the tip of the V5 loop may ensure that elicited antibodies are not too
reliant on any particular sequence in this highly variable epitope. An analysis of
10,000 published HIV sequences in the Los Alamos National Labs HIV Sequence
Database (http: / / lanl.hiv.gov) indicates that the aspartate at position 461 in
stripped core is conserved in 7.9% of sequences. The similar residues asparagine
77
(26.8%), threonine (12.4%), and serine (8.2%) could be substituted to introduce
mild diversity. Their frequency in existing HIV sequences suggests that the risk
of allosteric effects on the VRC01 epitope is small. In our previous mapping
work, the hydrophilic-to-hydrophobic mutation D4611, which is observed in only
1.9% of published sequences, reduced binding to VRC01 by a factor of 10 (Table
3.2).
Existing sequence and structural data were useful in designing the immunogens
for this study and in proposing future variations, but the power of the yeastdisplay technique for immunogen design is that it requires neither complicated
protein structure calculations nor accurate crystallographic data. All that is
necessary to generate diverse immunogens with a shared epitope is an antigen
displayed on yeast and a model antibody with which to perform the sorting. This
simplicity makes the method translatable. With other immunogens it should be
possible to direct the immune response to other epitopes on gp120, other vaccine
targets, or multiple conserved epitopes on a single antigen.
BALB / c mice were immunized with the engineered immunogens either
sequentially, in parallel, or with just the stripped core (Figure 4.3B). Serum from
mice immunized sequentially with each of the four immunogens bound equally
well to all immunogens displayed on yeast, suggesting it recognized a shared
epitope (Figure 4.3C). Serum binding was significantly disrupted by the D368R
mutation known to knock out binding to CD4 binding site-directed antibodies
(Figure 4.4A), and the serum was able to bind stripped core from other clades
(Figure 4.5C). Furthermore, in a ternary complex assay with VRC01 scFv and
soluble stripped core gp120, it appears the sequential serum recognizes only the
VRC01 epitope and has no other specificities (Figure 4.4D, 4E). Mice that were
exposed to the same immunogen multiple times (parallel immunization or coreonly) generated some VRC01-competitve antibodies, but also antibodies to other
epitopes. Only sequential immunization elicited a response that is entirely
VRC01-competitive.
Serum did not neutralize a panel of four viruses by TZM-bl assay (Figure 4.6).
There are two competing explanations for this result: specificity and
78
stoichiometry. It is possible that the epitope presented by stripped core against
which the serum response is focused does not accurately mimic the CD4 binding
site as presented by the virus. If this were the case, a solution might be to design
immunogens based on native trimer that would be more faithful to the true
target. An alternative hypothesis is that the concentration of antibody is too low
to observe any effect in the neutralization assay. We quantified the amount of
gp120-specific antibody in the serum samples and found that it ranged from 10100 yg/mL. At working dilutions of these concentrations, even VRC01 would not
neutralize above the assay baseline. We have generated hybridomas from a
mouse in the sequential group to isolate the monoclonal antibodies that make up
the serum response. As of this writing, fusions that secrete monoclonal
antibodies that are specific for the VRC01 epitope and broadly cross-reactive
have been identified (Figure 4.7). In addition to answering the neutralization
question, the sequences of these antibodies will shed light on the requirement for
particular germ lines or CDR lengths, as well as demonstrate the robustness of
the immunization strategy.
Materials & Methods
Library construction and sorting. Degenerate oligonucleotide primers
(Integrated DNA Technologies, Coralville, IA) were designed with overlapping
homology. Eleven fragments were amplified with these degenerate primers by
PCR for 30 cycles with Taq polymerase (New England Biolabs, Ipswich, MA) and
purified by gel electrophoresis. The full-length gene was reconstituted by
pairwise overlap-extension PCR of these fragments. In each reaction, fragments
were allowed to self-prime for 8 cycles then outer primers were added for an
additional 30 cycles. Following gel purification, the new, longer fragments were
amplified for an additional 30 cycles and re-purified. The full-length library was
transformed into yeast by electroporation.
The induced library was incubated with VRC01 (expression plasmids a generous
gift from John Mascola, Vaccine Research Center, NIAID, NIH) at the stated
79
concentrations and a 10-fold molar excess of chicken anti-CMyc (Gallus
Immunotech, Fergus, Ontario). Secondary labeling was performed with Alexa
Fluor-labeled goat anti-human and goat anti-chicken antibodies (Life
Technologies, Carlsbad, CA) at recommended dilutions. The library was sorted
on a MoFlo instrument (Beckman Coulter, Brea, CA).
Secretion of Fc-gp120 from HEK 293 cells. gp120 constructs were subcloned into
a mammalian expression vector based on gWiz (Genlantis, San Diego, CA)
downstream of the Fc domain of mouse IgG2a and a his6 tag. A suspension
culture of HEK 293 cells grown in serum-free medium (Freestyle 293, Life
Technologies, Carlsbad, CA) was transfected with DNA using PEI as per
manufacturers instructions. Supernatant was harvested after one week and
purified by sequential metal affinity chromatography (TALON resin, Clontech,
Mountain View, CA) and protein A affinity chromatography (Genscript,
Piscataway, NJ).
Yeast cell surface binding assays. Cell surface binding titrations were performed
with various combinations of displayed and soluble proteins, e.g. displayed
immunogens binding to soluble anti-gp120 antibodies; displayed VRC01 scFv
binding to soluble immunogens; displayed gp120 from various clades binding to
anti-gp120 antibodies and serum from immunized mice. In all instances, display
was detected by chicken anti-CMyc (Gallus Immunotech, Fergus, Ontario).
Secondary labeling was performed with Alexa Fluor-labeled goat anti-human or
anti-mouse, and goat anti-chicken antibodies (Life Technologies, Carlsbad, CA)
at a 1:400 dilution, and analyzed on a FACSCalibur HTS with a high-throughput
plate sampler (Becton Dickinson, Franklin Lakes, NJ).
Immunizations. 6-8 week-old female BALB /c mice (Taconic, Hudson, NY) were
immunized with 50 pmol Fc-gp120 (6 yg total protein, or 3 yg gp120 equivalents)
intranasally with 3 nmol CpG-1826 adjuvant (Integrated DNA Technologies,
Coralville, IA) in PBS in the method of (86, 87). The maximum administered
volume was 15 yL. Mice were boosted at two-week intervals, and serum was
collected weekly. In a separate experiment, 6-8 week-old female C57BL / 6 mice
(Taconic, Hudson, NY) were immunized with 150 pmol Fc-stripped core, 150
80
pmol Fc-Clone A, and 1 nmol Fc-Clone B at two-week intervals, each time with 3
nmol CpG-1826 adjuvant.
Analysis of serum binding. Antisera were titrated on gp120-displaying yeast. In
each well of a 96-well plate, 1.5x10' induced yeast were incubated with serial
dilutions of serum and 40 ng chicken anti-CMyc (Gallus Immunotech, Fergus,
Ontario) in 50 yL for several hours. Secondary labeling was performed with
Alexa Fluor-labeled goat anti-human and goat anti-chicken antibodies (Life
Technologies, Carlsbad, CA) at a 1:400 dilution, and analyzed on a FACSCalibur
HTS with a high-throughput plate sampler (Becton Dickinson, Franklin Lakes,
NJ).
Median binding signals were fit using MATLAB (MathWorks, Natick, MA) to a
monovalent binding isotherm of the form y =
Ymax di
'
where "dil" is the serum
dilution and y is the binding signal in MFU. The two fitted parameters are the
maximal binding signal, y. and the equilibrium association constant, KA' max
depends on the display level of a particular gp120 variant and the number of
antibodies that can bind to a single variant simultaneously. KA, the dilution at
which the binding signal is one-half the maximum value, is a combination
measure of the affinity and number of antigen-specific antibodies in the serum.
Ternary complex assay. This assay has two iterations. In the first, an excess of
serum at a saturating dilution (as calculated by cell-surface titration) is incubated
with soluble stripped core (not an Fc-fusion) at 4*C overnight. This solution (20
nM gp120) is then added to 5x10' yeast displaying the scFv of VRC01 in a 96-well
plate and incubated for 1 hr at 4"C. In the second iteration, yeast displaying the
scFv of VRC01 are incubated with a large molar excess of 20 nM soluble stripped
core at 4*C overnight. The yeast are washed and distributed into a 96-well plate
at 5x104 cells per well. Serum is added at a saturating dilution and the plate is
incubated for 1 hr at 4". The formation of a ternary complex-a sandwich of
VRC01, gp120, and mouse serum-is detected by secondary labeling with goat
anti-mouse 647 (Life Technologies, Carlsbad, CA) at a 1:400 dilution, and
81
analyzed on a FACSCalibur HTS with a high-throughput plate sampler (Becton
Dickinson, Franklin Lakes, NJ).
Neutralization assay. Standard TZM-bl assay described by Li et al. (90). TZM-bl
cells are HeLa cells expressing CD4, CCR5 and CXCR4, and luciferase under a
Tat-inducible promoter. HIV-infected cells will express luciferase. In brief, virus
(supernatant from 293T-transfected cells) is incubated with antibodies for 1 hr at
37C then added to TZM-bl cells at 2x10 5 mL1 and incubated for 48 hr at 37C.
Luciferase substrate is added and lysed cells are visualized.
Hybridoma production. Two mice from the sequential immunization group
were boosted with 150 pmol stripped core at day 237 and 150 pmol of Clone A at
day 251, both times with 3 nmol CpG-1826 adjuvant (Integrated DNA
Technologies, Coralville, IA) intranasally. A third C57BL /6 mouse was
immunized with three immunogens as described previously. Hybridomas were
produced by Green Mountain Antibodies, Burlington, VT, by standard splenic
fusion methods. Fusions were screened for binding to stripped core by ELISA.
Positive supernatants were titrated on yeast-displayed stripped core and
specificity determined by the ternary complex assay described above. Promising
wells were subcloned by limiting dilution to isolate single cells.
82
HA
gp120
B
Aga2
gp120
Agaip
Front
HA
CMyc
Yeast
NN-
C
D
10
30 -.-OoA
NNO
Q
Cltone8
?#48
Introduce diverst
25 -cinc
by PCIR
~l0~
50'
10;S
QCombne mutations
Ill
M
H!
iff-
K
81
1
10
Prowen(nm)
100
10,00
CMyc Display
Figure 4.1. Generation of diversified immunogens. (a) Map of yeast display
vector with gp120 and two epitope tags (HA and CMyc) fused to Aga2, and
schematic of the construct displayed on the surface of yeast. (b) Homology model
of stripped core gp120 showing the VRC01 contact residues (yellow, from PDB
ID: 3NGB(28)) and diversified surface residues (red). (c) Schematic of PCR-based
method for introducing random amino acids at defined positions. Degenerate
oligonucleotides were used as amplification primers to introduce diversity at
chosen positions. The amplified fragments were pieced together by overlapextension PCR to form the final full-length library. The library was displayed on
yeast and sorted twice for binding to VRCO1. (d) FACS plot of the second library
sort against 50 nM VRCO1. Clones in the red gate were collected and sequenced.
Selected clones were secreted from HEK 293 cells as fusions to mouse Fc. (e)
Titration of stripped core and candidate immunogens (Clones A-C) on yeast
displaying VRC01 scFv. Fitted KD values are 3.5 nM (stripped core), 6.2 nM
(Clone A), 1.5 nM (Clone B), and 12.2 nM (Clone C).
83
84
C
*
-vRcO1
0.8
*-b12
+-b13
10.8
D
-- VRC01
-*-b12
0.8 -+b13
*1
I
0.6
..
0.4
0.4
10
100
1
Antibody Conc ()
1000
-
.1
10
100
Antibody Conc (nM)
1
--
1000
Figure 4.2. Binding of immunogens to CD4 binding site-directed antibodies. CD4
binding site-directed antibodies VRC01 (solid line), b12 (dashed line), and b13
(dash-dotted line) were titrated on yeast displaying (a) stripped core, (b) Clone
A, (c) Clone B, or (d) Clone C. An anti-CMyc antibody was used to confirm
display, and secondary detection was performed with fluorescently labeled antichicken and anti-mouse antibodies.
85
86
A
B
2 Wk
Group
Sequential
Core-only
C
**D
L
0
2Wk
2wk
I
Dose 1
Dose 2
Dose 3
Core
Clone A
Clone
B
Clone C
core
Clone A
Clone B
Clone C
core
Clone A
clone B
Clone C
Core
Clone A
Clone B
Clone C
core
Clone A
Clone B
Cone C
Core
core
core
Doee 4
**E*
o
i.
L?40!
w.g60
0F
F
20
90oom0
Serumn Dilution (fold)
00
100
00
20-,
100
i00
Serum Dilution (fold)
90-
100
100
10000
Serum Dilution (fold)
Figure 4.3. Immunization of mice with diversified immunogens. (a) Schematic of
an Fc-gpl2O immunogen. Mouse IgG2a Fc with a gpl2O variant fused at the C
terminus was secreted from HEK 293 cells and purified by sequential metal
affinity and protein A chromatography. (b) BALB/c mice were divided into three
groups. The sequential group (4 mice) was immunized sequentially with each
immunogen. The parallel group (4 mice) was immunized with all four
immunogens each time. The core-only group (2 mice) was immunized just with
stripped core. Immunizations were given every two weeks. Each immunization
was administered intranasally with 50 pmol total protein (equivalent to 6 yg Fcgp120 or 3 sg gp120 protein weight) and 3 nmol CpG oligonucleotide adjuvant
(CpG 1826(93)). (c-e) Titration of serum on yeast displaying gp120 variants.
Serum at day 48 post prime from a representative mouse from each group-(c)
sequential, (d) parallel, (e) core-only-is titrated on yeast displaying stripped
core (black), Clone A (blue), Clone B (red), or Clone C (green). Each plot also
includes serum from an unimmunized mouse binding to yeast displaying Clone
C (black dashed). Binding data is fit to a monovalent binding isotherm of the
form y =Ymax
di-' where "dil" is the serum dilution and y is the binding signal in
form
=
"+Kj
MFU. For the sequential mouse (c) the KA values are 1833 (core), 1242 (Clone A),
1631 (Clone B), 1735 (Clone C). For the parallel mouse (d) the KA values are 2600
(core), 238 (Clone A), 210 (Clone B), 68 (Clone C). For the core-only mouse (e) the
KA values are 604 (core), 65 (Clone A), 50 (Clone B), 37 (Clone C).
87
88
A
B
1
C
0Fu
.4.
Serum
i~a
Agpi2o
U?
J
04
d r (f
e
o.~gpI2O
m)
ias
par
Seq
c
yastt
D 10-
60
E~ eo
D --- Par
sp core M
a 1t
ore
- ------
--
r---Core
40
~30j
i
2
0
01
2
at
3
3 42
ous
4
doses
cNo.
Figure 4.4. Serum specificity for CD4 binding site. (a) Binding of serum to
stripped core D368R mutant. Serum from each group after three (open markers)
and four (filled markers) immunizations was assayed at 1:100 dilution on yeast
displaying stripped core and stripped core D368R. The ratio of D368R MFU to
stripped core MFU for each mouse is plotted. The fractional binding to D368R for
antibody VRC01 at various concentrations is shown as lines: 0.21 at 93 nM
(dashed line), 0.10 at 75 nM (dash-dotted line), and 0.06 at 8.2 nM (dotted line).
(b-c) Schematic of ternary complex assay with yeast displaying the VRC01 scFv,
soluble stripped core gpo2n with the CD4 binding site shown in gray, and mouse
serum. The presence of mouse serum on yeast--only possible if a ternary
complex is formed between the three assay components-is assayed by flow
cytometry. In the first iteration (b), serum is pre-incubated with gpl2O before
introducing the VRC01 yeast to determine whether any component of serum
competes with VRC01 for its binding site. In the second iteration (c), gp120 is
loaded onto VRC01 yeast before the serum is introduced to determine whether
the serum binds any other epitopes on stripped core. (d) Percentage of yeast cells
that are positive for mouse serum in the ternary complex assay (version (b)) after
incubation with serum from each immunization group-sequential (solid line),
parallel (dashed line), core-only (dash-dotted line). (e) Percentage of yeast cells
that are positive for mouse serum in the ternary complex assay (version (c)) after
incubation with serum from each immunization group-sequential (solid line),
parallel (dashed line), core-only (dash-dotted line).
89
90
B
A
BYU2
C_98ZADU156
* Clade A92UG037
Clade B YU2
* CladeC98ZADU156
* Clade D92UG021
* Clade F 93BR029
o
""
F_93BR029
""
D_92UG021
a
A_92UG037
C
80
15W
A
0
1
40
A Clade C
v Clade D
0 Clade F
U5W
.
A
20[
Clade A
0 Clade B
V
0
LL
CMt
0
a
V
a
V
VRC01
I
b13
V
Sequential Serum
Figure 4.5. Binding to stripped core from multiple HIV clades. (a) Homology
model overlay of stripped core from five HIV strains: 92UG037 (Clade A), YU2
(Clade B, original stripped core), 98ZADU156 (Clade C), 92UG021 (Clade D), and
93BR037 (Clade F). VRC01 contact surface is highlighted in yellow. Side chains
for positions within this surface that differ among the clades are shown in the
color indicated by the legend. (b) Phylogenetic tree of the five strains. (c) Binding
of antibodies and serum to stripped core from multiple strains displayed on
yeast. VRC01 and b13 are at 100 nM and labeled with goat anti-human 647.
Serum from sequential mice is at 1:100 dilution and labeled with goat anti-mouse
647.
91
92
A
8
Neutralization of JR-CSF
Neutralization of 94UG
40-
40
20-
c2 20
-20
10
10
10 -2
10
102
10
10-2
102
Ab conc. (ug/mL)
Ab conc. (ug/mL)
Figure 4.6. Neutralzation potential of serum by TZM-bl assay. Neutralization of
(a) JR-CSF and (b) 94UG pseudoviruses by VRC01 (positive control, black solid),
DEN3 (negative control, black dashed), sera from two sequential mice (red and
blue), and sera from three unimmunized mice (yellow). Antibody concentration
for the serum samples was measured by sequential immunoprecipitation on
gp120 yeast and quantitative flow cytometry. The starting concentrations for
serum samples (1:100 dilution) was 0.1 and 1 yg/mL.
93
94
50
on-target
Off-target
Non-binders
A
S40!+
10
8
0
e
-10-
0
20
40
04
60
80
Maximal binding (MFU)
B
o000
100
120
120
10 0
0
h
80
0
0
M 8o
U-
.
0
a
Clade A
Clade B
A Clade C
v Clade D
* Clade F
0
0
S60!
0
iD 401
A
20V
0
01A4
3D4
---4C9
0
587
6B9
#
a__,L__
---~
ANW
-_ -7D9 7E10 8C8 lOA6 13D10 1401 14F7
Figure 4.7. Screening of hybridomas. Hybridomas made by splenic fusion from a
sequentially immunized mouse were screened for binding to stripped core by
yeast surface display, and for specificity by ternary complex assay. (a) Scatter
plot of 32 ELISA-positive fusions showing three distinct populations of clones:
on-target binders (circles), false-positive non-binders (squares), and off-target
binders (pluses). Off-target binders are likely interacting with the his6 tag or
other artifact that is different between the yeast-displayed and mammalianproduced protein used in the immunization, ELISA, and ternary complex assay.
(b) Hybridoma supernatant (26.7x dilution) binding to yeast displaying stripped
core from five clades. Cross-reactive clones 3D4 and 10A6, and core-only clones
6B7 and 13D10, were subcloned for further analysis.
95
96
Table 4.1. Surface positions modified in engineered immunogens. Amino acid
position in sequential and HxB2 numbering (94), solvent-accessible surface area
on the homology model, and residue for stripped core and three immunogen
clones.
Position
Position
Solvent-accessible
Amino Acid
Amino Acid
Amino Acid
Amino Acid
(Sequential)
(HxB2)
surface area (AA2)
(Core)
(Clone A)
(Clone B)
(Clone C)
92
97
92.1
131.6
N
K
D
D
R
T
T
D
10
99
93.4
N
V
P
G
18
24
32
107
113
121
65.8
114.8
98.6
D
D
K
E
R
A
R
G
A
K
A
1
33
122
138.6
L
R
P
R
34
123
49.4
T
T
M
Q
35
196
64.7
G
F
L
W
36
39
40
41
197
200
201
202
164.1
102.1
63.4
106.1
N
V
I
T
P
S
G
Y
F
A
K
P
G
N
D
S
42
203
95.0
Q
R
L
V
43
46
52
55
70
79
85
204
207
213
216
231
240
246
45.8
91.0
103.7
108.4
116.5
113.2
137.5
A
K
I
H
K
T
Q
N
N
T
M
A
S
V
L
F
L
S
R
S
E
P
T
K
P
S
C
L
108
139
269
327
66.0
70.1
E
G
E
G
E
G
G
T
149
337
75.9
Q
Q
Q
Y
156
164
170
344
352
358
79.0
72.8
76.1
Q
Q
T
Q
R
T
Q
Q
T
A
Q
A
193
381
3.5
E
E
E
K
201
389
85.7
Q
Q
N
A
207
397
127.7
D
A
P
P
208
402
48.2
T
T
L
R
209
210
403
404
147.6
165.7
R
K
K
R
D
L
A
R
211
212
405
406
64.5
136.0
L
N
T
A
R
R
E
R
213
214
216
407
408
412
112.7
3.6
168.1
N
T
R
L
R
H
T
S
T
E
A
G
223
419
154.6
R
Y
R
L
235
442
120.9
Q
H
P
N
239
446
62.3
S
A
G
V
268
475
82.5
M
V
M
M
283
490
138.7
K
R
P
A
3
8
97
Table 4.2. Sequence alignment of stripped core gp120 from multiple clades.
aln. pos
SC4Gv
A_92UG037
B_YU2
C_98ZADU156
D_92UG021
F_93BR029
_consrvd
10
20
30
40
50
60
TENFNMWKNNMVEQMHEDIISLWDNSTKPCVKLTGNGSVITQACPKVSFEPIPIHYCAPAGFAIL
TEEFNMWKNNMVEQMHTDIISLWDNSTKPCVQLTGNGSALTQACPKVTFEPIPIHYCAPAGYAIL
TENFNMWKNNMVEQMHEDIISLWDNSTKPCVKLTGNGSVITQACPKVSFEPIPIHYCAPAGFAIL
TENFNMWKNDMVDQMHEDIISLWDNSTKPCVKLTGNGSTITQACPKVSFDPIPIHYCAPAGYAIL
TENFNIWKNNMVEQMHDDIISLWDNSTKPCVKLTGNGSTITQACPKISFEPIPIHYCAPAGFAIL
TENFDMWKNNMVEQMHTDIISLWDNSTKPCVKLTGNGSTLTQACPKVSWDPIPIHYCAPAGYAIL
_aln.pos
SC4Gv
A_92UG037
B_YU2
C_98ZADU156
D_92UG021
F_93BR029
70
80
90
100
110
120
130
KCNDKKFNGTGPCTNVSTVQCTHGIRPVVSTQLLLNGSLAEEEIVIRSENFTNNAKTIIVQLNES
KCNDKEFNGTGLCKNVSTVQYTHGIRPVVSTQLLLNGSLAEGKVMIRSENITNNVKNIIVQLNES
KCNDKKFNGTGPCTNVSTVQCTHGIRPVVSTQLLLNGSLAEEEIVIRSENFTNNAKTIIVQLNES
KCTDKKFNGTGSCNNVSTVQCTHGIKPVVSTQLLLNGSLAEEEIIIKSENLTDNIKTIIVQLNQS
KCNDKKFNGTGPCKNVSTVQCTHGIKPVVSTQLLLNGSLAEEEIIIRSKNLTNNAKIIIVHLNES
KCNDKKFNGTGPCTNVSTVQCTHGIKPVVSTQLLLNGSLAEKDIIIRSQNISDNAKTIIVQLNVS
_consrvd
**
**
*
**
***
*****
**
***
*
**************
******
******
****
******
***********
***************
*
*
*
*
***
*
***
**
*
_aln.pos
SC4Gv
A_92UG037
B_YU2
C_98ZADU156
D_92UG021
F_93BR029
_consrvd
140
150
160
170
180
190
VVINCTGNGSGHCNLSKTQWENTLEQIAIKLKEQFGNNKTIIFNPSSGGDPEIVTHSFNCGGEFF
VTINCTGNGSGHCNVSGSQWNKTLHQVVEQLRK-YWNNNTIIFNSSSGGDLEITTHSFNCAGEFF
VVINCTGNGSGHCNLSKTQWENTLEQIAIKLKEQFGNNKTIIFNPSSGGDPEIVTHSFNCGGEFF
IGINCTGNGSGHCNISRNQWNETLEQVKKKLGEHFHNQTKIKFEPPSGGDLEITTHSFNCRGEFF
VPINCTGNGSGHCNISGEKWNKTLQQVAVKLRD-LLNQTAIIFKPSSGGDPEITTHSFNCGGEFF
VPINCTGNGSGHCNVSGTQWNQTLERVRAKLKSHFPN-ATIKFNSSSGGDLEITMHSFNCRGEFF
_aln.pos
SC4Gv
A_92UG037
B_YU2
C_98ZADU156
D_92UG021
F_93BR029
_consrvd
200
210
220
230
240
250
260
YCNSTQLF--TWNDTRKLNNTGRNITLPCRIKGGNGSPIRGQIRCSSNITGLLLTRDGG---KDT
YCNTSGLFNSTWVNGTTSSMSNGTITLPCRIKGGNGSPIQGVIKCESNITGLILTRDGG--V-NS
YCNSTQLF--TWNDTRKLNNTGRNITLPCRIKGGNGSPIRGQIRCSSNITGLLLTRDGG---KDT
YCNTTDLFTNATKLVNDTE-NKAVITIPCRIKGGNGSPIEGNITCNSNITGLLLTRDGGGNVTEI
YCNTSGLFNNSVWTSNSTIGANGTITLPCRIKGGNGSPIEGQINCSSTITGLLLTRDGG--VKNN
YCNTSGLFNDTV--------DNNTITLPCRIKGGNGSPIAGNITCSSNITGLLLTRDGG---QNN
_aln.pos
SC4Gv
A_92UG037
B_YU2
C_98ZADU156
D_92UG021
F_93BR029
_consrvd
270
280
290
NGTEIFRPGGGDMRDNWRSELYKYKVVKIE
SDSETFRPGGGDMRDNWRSELYKYKVVKIE
NGTEIFRPGGGDMRDNWRSELYKYKVVKIE
NRTEIFRPGGGNMKDNWRNELYKYKVVEIK
SQNETFRPGGGDMRDNWRNELYKYKVVRIE
QTEETFRPGGGNMKDNWRSELYKYKVVEIE
************
***
*
*
**
*
******
**
*
**
*
****
************
********
*
98
*
*
*
*
*
****
*
*
**
****
*****
******
****
Works Cited
1. S. A. Plotkin, Correlates of protection induced by vaccination, Clin. Vaccine
Immunol. 17, 1055-1065 (2010).
2.
J.R. Mascola, D. C. Montefiori, The role of antibodies in HIV vaccines, Annu Rev
Immunol 28, 413-444 (2010).
3. R. Pantophlet, D. R. Burton, GP120: target for neutralizing HIV-1 antibodies,
Annu Rev Immunol 24, 739-769 (2006).
4. B. Korber et al., Evolutionary and immunological implications of contemporary
HIV-1 variation, Br Med Bull 58, 19-42 (2001).
5. S. D. W. Frost et al., Neutralizing antibody responses drive the evolution of
human immunodeficiency virus type 1 envelope during recent HIV infection,
Proc Natl Acad Sci USA 102, 18514-18519 (2005).
6. X. Wu et al., Selection Pressure on HIV-1 Envelope by Broadly Neutralizing
Antibodies to the Conserved CD4-Binding Site, J Virol 86, 5844-5856 (2012).
7. K. J. Bar et al., Early Low-Titer Neutralizing Antibodies Impede HIV-1
Replication and Select for Virus Escape, PLoS Pathog 8, e1002721 (2012).
8. P. L. Moore et al., Evolution of an HIV glycan-dependent broadly neutralizing
antibody epitope through immune escape, Nat Med (2012), doi:10.1038/nm.2985.
9. D. R. Burton et al., A large array of human monoclonal antibodies to type 1
human immunodeficiency virus from combinatorial libraries of asymptomatic
seropositive individuals, Proc Natl Acad Sci USA 88, 10134-10137 (1991).
10. D. R. Burton et al., Efficient neutralization of primary isolates of HIV-1 by a
recombinant human monoclonal antibody, Science 266, 1024-1027 (1994).
11. T. Zhou et al., Structural definition of a conserved neutralization epitope on HIV1 gp120, Nature 445, 732-737 (2007).
12. X. Wu et al., Rational design of envelope identifies broadly neutralizing human
monoclonal antibodies to HIV-1, Science 329, 856-861 (2010).
13. E. Falkowska et al., PG V04, an HIV-1 gp120 CD4 binding site antibody, is broad
and potent in neutralization but does not induce conformational changes
characteristic of CD4, J Virol (2012), doi:10.1128/JVI.06973-11.
99
14. A. F. Labrijn et al., Access of antibody molecules to the conserved coreceptor
binding site on glycoprotein gp120 is sterically restricted on primary human
immunodeficiency virus type 1,J Virol 77, 10557-10565 (2003).
15. M. Moulard et al., Broadly cross-reactive HIV-1-neutralizing human monoclonal
Fab selected for binding to gpl20-CD4-CCR5 complexes, Proc Natl Acad Sci USA
99, 6913-6918 (2002).
16. M. Thali et al., Characterization of conserved human immunodeficiency virus
type 1 gp120 neutralization epitopes exposed upon gpl20-CD4 binding, J Virol
67, 3978-3988 (1993).
17. L. M. Walker et al., Broad neutralization coverage of HIV by multiple highly
potent antibodies, Nature 477, 466-470 (2011).
18. R. Pejchal et al., A Potent and Broad Neutralizing Antibody Recognizes and
Penetrates the HIV Glycan Shield, Science 334, 1097-1103 (2011).
19. L. M. Walker et al., Broad and potent neutralizing antibodies from an African
donor reveal a new HIV-1 vaccine target, Science 326, 285-289 (2009).
20. R. Pejchal et al., Structure and function of broadly reactive antibody PG16 reveal
an H3 subdomain that mediates potent neutralization of HIV-1, Proc Natl Acad
Sci USA 107, 11483-11488 (2010).
21. M. Pancera et al., Crystal structure of PG16 and chimeric dissection with
somatically related PG9: structure-function analysis of two quaternary-specific
antibodies that effectively neutralize HIV-1, J Virol 84, 8098-8110 (2010).
22. X. Wu et al., Immunotypes of a quaternary site of HIV-1 vulnerability and their
recognition by antibodies, J Virol 85, 4578-4585 (2011).
23. A. Trkola et al., Human monoclonal antibody 2G12 defines a distinctive
neutralization epitope on the gp120 glycoprotein of human immunodeficiency
virus type 1, J Virol 70, 1100-1108 (1996).
24. T. Muster et al., A conserved neutralizing epitope on gp4l of human
immunodeficiency virus type 1, J Virol 67, 6642-6647 (1993).
25. G. Stiegler et al., A potent cross-clade neutralizing human monoclonal antibody
against a novel epitope on gp4l of human immunodeficiency virus type 1, AIDS
Res. Hum. Retroviruses 17, 1757-1765 (2001).
26. G. Frey et al., A fusion-intermediate state of HIV-1 gp4l targeted by broadly
neutralizing antibodies, Proc Natl Acad Sci USA 105, 3739-3744 (2008).
100
27. L. Chen et al., Structural basis of immune evasion at the site of CD4 attachment
on HIV-1 gp120, Science 326, 1123-1127 (2009).
28. T. Zhou et al., Structural basis for broad and potent neutralization of HIV-1 by
antibody VRC01, Science 329, 811-817 (2010).
29. J. F. Scheid et al., Sequence and structural convergence of broad and potent HIV
antibodies that mimic CD4 binding, Science 333, 1633-1637 (2011).
30. X. Wu et al., Focused evolution of HIV-1 neutralizing antibodies revealed by
structures and deep sequencing, Science 333, 1593-1602 (2011).
31. W. R. Schief, Y.-E. A. Ban, L. Stamatatos, Challenges for structure-based HIV
vaccine design, Curr Opin HIV AIDS 4, 431-440 (2009).
32. L. M. Walker, D. R. Burton, Rational antibody-based HIV-1 vaccine design:
current approaches and future directions, Current Opinion in Immunology (2010),
doi:10.1016/j.coi.2010.02.012.
33. B. D. Walker, D. R. Burton, Toward an AIDS vaccine, Science 320, 760-764 (2008).
34. Z. Euler et al., Cross-reactive neutralizing humoral immunity does not protect
from HIV type 1 disease progression, J Infect Dis 201, 1045-1053 (2010).
35. R. M. Lynch et al., The development of CD4 binding site antibodies during HIV-1
infection, J Virol 86, 7588-7595 (2012).
36. A. J. Mcmichael, P. Borrow, G. D. Tomaras, N. Goonetilleke, B. F. Haynes, The
immune response during acute HIV-1 infection: clues for vaccine development,
Nat Rev Immunol 10, 11-23 (2010).
37. A. Nandi et al., Epitopes for broad and potent neutralizing antibody responses
during chronic infection with human immunodeficiency virus type 1, Virology
(2009), doi:10.1016/j.virol.2009.10.044.
38. D. N. Sather, L. Stamatatos, Epitope specificities of broadly neutralizing plasmas
from HIV-1 infected subjects, Vaccine 28 Suppl 2, B8-12 (2010).
39. L. M. Walker et al., A limited number of antibody specificities mediate broad and
potent serum neutralization in selected HIV-1 infected individuals, PLoS Pathog 6
(2010), doi:10.1371/journal.ppat.1001028.
40. J.F. Scheid et al., Broad diversity of neutralizing antibodies isolated from
memory B cells in HIV-infected individuals, Nature 458, 636-640 (2009).
101
41.
J. Pietzsch et al., Human anti-HIV-neutralizing antibodies frequently target a
conserved epitope essential for viral fitness, J Exp Med 207, 1995-2002 (2010).
42. B. F. Keele et al., Identification and characterization of transmitted and early
founder virus envelopes in primary HIV-1 infection, Proc Natl Acad Sci USA 105,
7552-7557 (2008).
43. M.-R. Abrahams et al., Quantitating the Multiplicity of Infection with Human
Immunodeficiency Virus Type 1 Subtype C Reveals a Non-Poisson Distribution
of Transmitted Variants, J Virol 83, 3556-3567 (2009).
44. J. F. Salazar-Gonzalez et al., Genetic identity, biological phenotype, and
evolutionary pathways of transmitted! founder viruses in acute and early HIV-1
infection, Journalof Experimental Medicine 206, 1273-1289 (2009).
45. D. R. Burton et al., Limited or no protection by weakly or nonneutralizing
antibodies against vaginal SHIV challenge of macaques compared with a
strongly neutralizing antibody, ProcNatl Acad Sci USA 108, 11181-11186 (2011).
46. A. J. Hessell et al., Broadly neutralizing human anti-HIV antibody 2G12 is
effective in protection against mucosal SHIV challenge even at low serum
neutralizing titers, PLoS Pathog5, e1000433 (2009).
47. A. J. Hessell et al., Effective, low-titer antibody protection against low-dose
repeated mucosal SHIV challenge in macaques, Nat Med 15, 951-954 (2009).
48. J. R. Mascola, Passive transfer studies to elucidate the role of antibody-mediated
protection against HIV-1, Vaccine 20, 1922-1925 (2002).
49. D. C. Montefiori, J. R. Mascola, Neutralizing antibodies against HIV-1: can we
elicit them with vaccines and how much do we need? Curr Opin HIV AIDS 4,
347-351 (2009).
50. M. J. McElrath, B. F. Haynes, Induction of immunity to human
immunodeficiency virus type-1 by vaccination, Immunity 33, 542-554 (2010).
51. S. Rerks-Ngarm et al., Vaccination with ALVAC and AIDSVAX to prevent HIV-1
infection in Thailand, N. Engl. J.Med. 361, 2209-2220 (2009).
52. B. F. Haynes et al., Immune-correlates analysis of an HIV-1 vaccine efficacy trial,
N. Engl. J.Med. 366, 1275-1286 (2012).
53. N. Karasavvas et al., The Thai Phase III HIV Type 1 Vaccine Trial (RV144)
Regimen Induces Antibodies That Target Conserved Regions Within the V2
Loop of gp120, AIDS Res. Hum. Retroviruses (2012), doi:10.1089/aid.2012.0103.
102
54. D. C. Montefiori et al., Magnitude and breadth of the neutralizing antibody
response in the RV144 and Vax003 HIV-1 vaccine efficacy trials, J Infect Dis 206,
431-441 (2012).
55. B. Dey et al., Characterization of human immunodeficiency virus type 1
monomeric and trimeric gp120 glycoproteins stabilized in the CD4-bound state:
antigenicity, biophysics, and immunogenicity, J Virol 81, 5579-5593 (2007).
56. R. Pantophlet, I. A. Wilson, D. R. Burton, Hyperglycosylated mutants of human
immunodeficiency virus (HIV) type 1 monomeric gp120 as novel antigens for
HIV vaccine design, J Virol 77, 5889-5901 (2003).
57. R. Pantophlet et al., Fine mapping of the interaction of neutralizing and
nonneutralizing monoclonal antibodies with the CD4 binding site of human
immunodeficiency virus type 1 gp120, J Virol 77, 642-658 (2003).
58. R. Pantophlet, I. A. Wilson, D. R. Burton, Improved design of an antigen with
enhanced specificity for the broadly HIV-neutralizing antibody b12, Protein Eng
Des Sel 17, 749-758 (2004).
59. S. Selvarajah et al., Comparing antigenicity and immunogenicity of engineered
gp120, J Virol 79, 12148-12163 (2005).
60. F. K. Ahmed, B. E. Clark, D. R. Burton, R. Pantophlet, An engineered mutant of
HIV-1 gp120 formulated with adjuvant Quil A promotes elicitation of antibody
responses overlapping the CD4-binding site, Vaccine (2011),
doi:10.1016/j.vaccine.2011.11.089.
61.
J.P. Nkolola et al., Breadth of neutralizing antibodies elicited by stable,
homogeneous clade A and clade C HIV-1 gp140 envelope trimers in guinea pigs,
J Virol 84, 3270-3279 (2010).
62. J.M. Kovacs et al., HIV-1 envelope trimer elicits more potent neutralizing
antibody responses than monomeric gp120, Proc Natl Acad Sci USA (2012),
doi:10.1073/pnas.1204533109.
63.
J. Guenaga et al., Heterologous epitope-scaffold prime:boosting immuno-focuses
B cell responses to the HIV-1 gp4l 2F5 neutralization determinant, PLoS ONE 6,
e16074 (2011).
64. G. R. Matyas, Z. Beck, N. Karasavvas, C. R. Alving, Lipid binding properties of
4E10, 2F5, and WR304 monoclonal antibodies that neutralize HIV-1, Biochim
Biophys Acta 1788, 660-665 (2009).
103
65. G. R. Matyas et al., Neutralizing antibodies induced by liposomal HIV-1
glycoprotein 41 peptide simultaneously bind to both the 2F5 or 4E10 epitope and
lipid epitopes, AIDS (London, England) 23, 2069-2077 (2009).
66. S. Zolla-Pazner et al., Cross-clade HIV-1 neutralizing antibodies induced with V3scaffold protein immunogens following priming with gp120 DNA, J Virol 85,
9887-9898 (2011).
67. S. Zolla-Pazner et al., Cross-clade neutralizing antibodies against HIV-1 induced
in rabbits by focusing the immune response on a neutralizing epitope, Virology
392, 82-93 (2009).
68. M. Totrov et al., Structure-guided design and immunological characterization of
immunogens presenting the HIV-1 gp120 V3 loop on a CTB scaffold, Virology
405, 513-523 (2010).
69. C. Charles-Nif5o, C. Pedroza-Roldan, M. Viveros, G. Gevorkian, K.
Manoutcharian, Variable epitope libraries: New vaccine immunogens capable of
inducing broad human immunodeficiency virus type 1-neutralizing antibody
response, Vaccine (2011), doi:10.1016/j.vaccine.2011.05.007.
70. S. Zolla-Pazner, T. Cardozo, Structure-function relationships of HIV-1 envelope
sequence-variable regions refocus vaccine design, Nat Rev Immunol 10, 527-535
(2010).
71. M. L. Scalley-Kim et al., B. Ryffel, Ed. A Novel Highly Potent Therapeutic
Antibody Neutralizes Multiple Human Chemokines and Mimics Viral Immune
Modulation, PLoS ONE 7, e43332 (2012).
72. Y. Feng et al., Biochemically defined HIV-1 envelope glycoprotein variant
immunogens display differential binding and neutralizing specificities to the
CD4-binding site, Journal of Biological Chemistry 287, 5673-5686 (2012).
73. M. L. Azoitei et al., Computation-guided backbone grafting of a discontinuous
motif onto a protein scaffold, Science 334, 373-376 (2011).
74. E. T. Boder, K. D. Wittrup, Yeast surface display for screening combinatorial
polypeptide libraries, Nat Biotechnol 15, 553-557 (1997).
75. E. T. Boder, K. S. Midelfort, K. D. Wittrup, Directed evolution of antibody
fragments with monovalent femtomolar antigen-binding affinity, Proc Natl Acad
Sci USA 97, 10701-10705 (2000).
104
76. Y.-S. Kim, R. Bhandari, J. R. Cochran, J. Kuriyan, K. D. Wittrup, Directed
evolution of the epidermal growth factor receptor extracellular domain for
expression in yeast, Proteins 62, 1026-1035 (2006).
77. E. V. Shusta, P. D. Holler, M. C. Kieke, D. M. Kranz, K. D. Wittrup, Directed
evolution of a stable scaffold for T-cell receptor engineering, Nat Biotechnol 18,
754-759 (2000).
78. S. A. Gai, K. D. Wittrup, Yeast surface display for protein engineering and
characterization, Curr Opin Struct Biol 17, 467-473 (2007).
79. G. Chao, J. R. Cochran, K. D. Wittrup, Fine epitope mapping of anti-epidermal
growth factor receptor antibodies through random mutagenesis and yeast
surface display, Journal of Molecular Biology 342, 539-550 (2004).
80. T. Oliphant et al., Development of a humanized monoclonal antibody with
therapeutic potential against West Nile virus, Nat Med 11, 522-530 (2005).
81. T. Han et al., Fine epitope mapping of monoclonal antibodies against
hemagglutinin of a highly pathogenic H5N1 influenza virus using yeast surface
display, Biochem. Biophys. Res. Commun. 409, 253-259 (2011).
82. A. Sali, T. L. Blundell, Comparative protein modelling by satisfaction of spatial
restraints, Journalof Molecular Biology 234, 779-815 (1993).
83. G. Chao et al., Isolating and engineering human antibodies using yeast surface
display, Nature Protocols 1, 755-768 (2006).
84. A. E. Firth, W. M. Patrick, Statistics of protein library construction, Bioinformatics
21, 3314-3315 (2005).
85. W. M. Patrick, A. E. Firth, J. M. Blackburn, User-friendly algorithms for
estimating completeness and diversity in randomized protein-encoding libraries,
Protein Eng 16, 451-457 (2003).
86. L. Ye, R. Zeng, Y. Bai, D. C. Roopenian, X. Zhu, Efficient mucosal vaccination
mediated by the neonatal Fc receptor, Nat Biotechnol 29, 158-163 (2011).
87. L. Lu et al., An FcRn-targeted mucosal vaccine strategy effectively induces HIV-1
antigen-specific immunity to genital infection, J Virol (2011),
doi:10.1128/JVI.05441-11.
88. M. Thali et al., Discontinuous, conserved neutralization epitopes overlapping the
CD4-binding region of human immunodeficiency virus type 1 gp120 envelope
glycoprotein, J Virol 66, 5635-5641 (1992).
105
89.
J.M. Binley et al., Comprehensive cross-clade neutralization analysis of a panel of
anti-human immunodeficiency virus type 1 monoclonal antibodies, J Virol 78,
13232-13252 (2004).
90. M. Li et al., Human immunodeficiency virus type 1 env clones from acute and
early subtype B infections for standardized assessments of vaccine-elicited
neutralizing antibodies, J Virol 79, 10108-10125 (2005).
91. R. Pantophlet, D. R. Burton, Immunofocusing: antigen engineering to promote
the induction of HIV-neutralizing antibodies, Trends Mol Med 9, 468-473 (2003).
92. J.S. McLellan et al., Design and characterization of epitope-scaffold immunogens
that present the motavizumab epitope from respiratory syncytial virus, Journalof
Molecular Biology 409, 853-866 (2011).
93. A. M. Krieg, CpG motifs in bacterial DNA and their immune effects, Annu Rev
Immunol 20, 709-760 (2002).
94. Korber B, Foley BT, Kuiken C, Pillai SK, Sodroski JG (1998).Numbering Positions
in HIV Relative to HXB2CG. pp. 111-102-111 in Human Retroviruses and AIDS
1998. Edited by: Korber B, Kuiken CL, Foley B, Hahn B, McCutchan F, Mellors
JW, and Sodroski J. Published by: Theoretical Biology and Biophysics Group, Los
Alamos National Laboratory, Los Alamos, NM.
httD://www.hiv.lanl.ov/content/seauence/HIV/REVIEWS/HXB2.html
106
Chapter 5. Perspectives on immunizations
Background
The formulation and route of administration of a vaccine can have a profound
impact on its in vivo efficacy. In the HIV vaccine field, clinical trials have been
performed with soluble adjuvanted proteins, plasmid DNA, viral vectors, and
heterologous prime-boost regimens that combine the above formulations (1).
These trials have demonstrated beyond doubt that the choice of formulation and
adjuvant can determine the strength and type of elicited response, which is
important because several analyses suggest that the number of antibodies
required for neutralization or clearance in vivo may be achievable by vaccination
(1-3).
Traditional adjuvants like aluminum phosphate or hydroxide (alum) or
oil/water emulsions (Incomplete Freund's Adjuvant and derivatives) have been
used for many years in commercial antibody production and laboratory scale
clinical trials. Their mechanism of action is largely unknown, but they have been
empirically and clinically validated (4). More recently, molecularly defined
adjuvants like TLR ligands and enterotoxins, as well as particulate vaccines that
present antigen in multivalent, pathogen-like arrays, have been developed and
studied (5-9).
Since sexual transmission of HIV occurs at mucosal surfaces, a successful vaccine
will likely have to elicit protective mucosal immunity beyond simply circulating
serum antibodies. Much work has recently been done on mucosal vaccination
strategies that may promote the induction of immunity at these sites (10, 11).
The choice of animal model is an important consideration. Immunologists have
traditionally favored rabbits because they generate high antibody titers and large
volumes of serum can be collected, but analysis of their V gene usage, VDJ
recombination, and antibody maturation pathways suggest that mice are more
similar to humans than rabbits are (12-14). A recent immunization study in
107
rhesus macaques notes that they have similar numbers of heavy chain V, D, and J
segments to humans, as well as much homology in the VH sequences (15).
Macaques have the added benefit that challenge experiments can be performed
on them and the attendant drawback that they are significantly more expensive
than mice. Advances in the development of humanized mice may one day make
a cheap, user-friendly animal challenge model available to researchers.
In this chapter we describe our work with several immunization routes and
formulations, and establish important optimization parameters to consider for
each. For intranasal administration of antigen, we demonstrate that both the Fc
fusion partner and GpG adjuvant are necessary to elicit a serum response, and
characterize the pharmacokinetics of the immunogen. For subcutaneous
liposomal vaccines, we show that conjugation of the immunogen to the particle is
required and that the timing of the boost can affect the post-boost serum titers.
We also show that immunization with heterologous immunogens loaded on the
current generation of protein-coated liposomes is unable to steer the immune
response to the CD4 binding site. For intradermal vaccines, we demonstrate that
micronedle application increases local inflammation over direct intradermal
injection and that the biodistribution of antigen is different between these two
routes. We also show that only a fraction of the antigen loaded onto a
microneedle array is delivered. In a broad comparison across routes,
formulations, and adjuvants, we find that antigen dose is the primary parameter
driving serum titer.
Results
Immunization routes and formulations tested. Several different immunization
routes and formulations were tested over the course of this work. The steering
experiments described in the previous chapter were performed by intranasal
administration of soluble protein with CpG DNA adjuvant in the method of (16,
17). In collaboration with Melissa Hanson of the Irvine lab, stripped core was
conjugated to the surface of liposomes by maleimide-thiol linkage, and then
108
injected subcutaneously at the base of the tail with monophosphoryl lipid A
(MPLA) adjuvant. In collaboration with Peter Demuth of the Irvine lab, stripped
core was coated onto poly-L lactic acid (PLLA) microneedle arrays with poly I:C
adjuvant, and administered intradermally in the ear. Liposome work was
performed using BALB / c mice; microneedle and intranasal immunizations were
done with both BALB/c and C57BL /6 strains.
Intranasal administration requires both Fc and adjuvant. BALB / c mice were
immunized with stripped core or Fc-stripped core (300 pmol gp120 equivalents)
and CpG DNA adjuvant (3 nmol). Two mice were given Fc-stripped core with
CpG, two were given stripped core with CpG, and one was given Fc-stripped
core without CpG. Serum titers were measured by yeast cell surface titration as
described in previous chapters. A strong gp120-specific serum response was only
observed when mice were immunized with the Fc-fusion construct in the
presence of CpG DNA adjuvant (Figure 5.1A). This result is consistent with the
hypothesis put forth by Xiaoping Zhu and colleagues that the antigen is actively
transported across the mucosal barrier by the neonatal Fc receptor (FcRn) (16,
17).
Pharmacokinetics of antigen administered intranasally. A C57BL /6 mouse was
immunized intranasally with 300 pmol fluorescently labeled Fc-stripped core
and CpG adjuvant. Blood samples were collected at various time points, and
blood fluorescence was measured to estimate the serum pharmacokinetics of the
antigen (Figure 5,1B). The signal did not peak until 0.5-1 hr post administration,
in contrast to the immediate bolus observed with intravenous delivery. The peak
fluorescent signal measured was equivalent to 1000x lower concentration than
expected if the full dose were delivered intraveinously, suggesting that only a
small fraction of antigen reached the bloodstream. A bi-exponential curve was fit
to the data with an a-phase time constant of 0.966 min' and a p-phase time
constant of 0.010 min'. The long serum half-life is likely due to FcRn-mediated
recycling of the Fc-containing antigen. After four days, cervical lymph nodes
were harvested from the immunized mouse and an unimmunized control
mouse. The lymph node from the immunized mouse was more fluorescent than
109
that of the control (Figure 5.1C), though the absolute amount of antigen present
was not quantified.
Timing of the boost matters for liposomal formulation. BALB / c mice were
immunized with stripped core (7.5 pmol, approximately 250 ng protein)
conjugated to liposomes by maleimide-thiol linkage, with 1.3 sg MPLA adjuvant,
injected subcutaneously at the base of the tail. Mice were boosted two, four, or
six weeks after the prime. Serum titers were measured by yeast cell surface
titration. A spike in serum titer was observed for all mice that were boosted
(Figure 5.2A). There appeared to be two tiers of responses: a high response KA ~
4000 fold dilution, and a low response KA~ 1000 fold dilution. Three of four mice
in the six-week group had a high response, two of which sustained this response
for at least 100 days after the boost. Only two of five mice in the four-week group
elicited a high response, but both of these responses waned by 50 days post
boost. One mouse of five in the two-week group reached a KA of 3500, but
regressed to the low level response by 20 days after the boost, joining the other
four mice of its cohort. Our conclusion from these experiments is that allowing
six weeks to pass between prime and boost may improve the magnitude and
duration of the serum response, perhaps by allowing more time for germinal
centers to form and somatic hypermutation to occur.
Conjugation is required for liposomal formulations, but not for other
adjuvants. BALB / c mice were immunized with stripped core or Fc-stripped core
(3.75 pmol protein) that was either conjugated to liposomes or simply mixed
with liposomes prior to injection, with 1.3 pg MPLA as adjuvant, injected
subcutaneously at the base of the tail. Mice were boosted six weeks after the
prime. Serum titers were measured by yeast cell surface titration. Regardless of
the antigen, conjugated liposomes elicited an anti-gp120 response whereas
unconjugated antigen and liposomes did not (Figure 5.2B). The presence of the
Fc in the conjugated liposome formulation may generate a stronger initial peak
titer following the boost, but the effect disappears after two weeks.
The serum titer elicited by MPLA liposomes was compared to the response to
other common adjuvants. Alum, AddaVax, and Montanide are all depot-forming
110
adjuvants that have been tested in clinical settings (18). Alum is the only FDAapproved adjuvant. It promotes an antibody and Th2-type immune response,
though its mechanism of action is still under investigation (19). AddaVax is a
squalene-in-water emulsion analogous to MF59 recently approved for use with
influenza vaccines in Europe. Montanide, a water-in-oil emulsion analogous to
Incomplete Freund's Adjuvant, is used routinely in animals and occasionally in
human trials, though severe local inflammation limits its widespread use (4).
Mice were immunized with 3.75 pmol Fc-stripped core mixed with alum,
AddaVax, or Montanide and 1.3 pg MPLA subcutaneously at the base of the tail,
and boosted six weeks later. Serum titers were measured by yeast cell surface
titration. Immunization with alum elicited titers significantly lower than
conjugated liposomes (Figure 5.2C). Three of four mice immunized with each
AddaVax and Montanide elicited significantly stronger responses than
conjugated liposomes (KA~ 5000-15000). One AddaVax mouse elicited a
liposome-like response, and one Montanide mouse elicited a significantly higher
titer (peak KA ~ 30000 before reverting to ~10000 40 days after the boost).
Immunization with diverse immunogens on liposomes fails to steer the
immune response to the CD4 binding site. BALB/c mice were immunized with
5 pmol Fc-gp120 conjugated to liposomes with 1.3 gg MPLA injected
subcutaneously at the base of the tail. As with the steering experiment described
in the previous chapter, mice were immunized with each immunogen
sequentially (Core, Clone A, Clone B, Clone C), all four immunogens in parallel,
or just stripped core every four weeks for four administrations. Serum titer was
analyzed by yeast cell surface titration on yeast displaying each of the
immunogens. Serum titers for all groups after three immunizations are shown in
Figure 5.3A. Serum from core-only mice predominantly binds to stripped core,
whereas serum from parallel mice binds to all four immunogens. Serum from
sequential mice binds to all immunogens as well, though with lower titers than
parallel mice. This is consistent with the serum binding to a single epitope shared
by all immunogens, rather than multiple epitopes on all immunogens.
111
To confirm this initial impression, sera from these mice were analyzed by the
ternary complex assay described in a previous chapter. In the iteration for
competition with the VRC01 binding site, it appears that two mice in each the
parallel and sequential groups begin to lose all specificity for the VRC01 epitope
by the third immunization (Figure 5.3B). In the iteration for non-VRC01
specificities, there is no difference between the groups (Figure 5.3C). All mice,
regardless of the order of immunogens given, raise antibodies to non-VRC01
epitopes on stripped core.
Comparison of intradermal injection to microneedle application. In close
collaboration with Pete DeMuth of the Irvine lab, microneedles were investigated
as a means of intradermal delivery of protein antigen. Fc-stripped core was
coated onto PLLA microneedle arrays in a sucrose shell (Figure 5.4A). Protein
stability on microneedles was tested by dissolving the microneedle arrays at
various points in time and measuring protein binding to VRC01 scFv displayed
on yeast. Stripped core that had been stored on microneedle arrays at room
temperature for several months retained its binding to VRC01 (Figure 2.1F).
Intradermal delivery by microneedle application was compared to direct
intradermal injection in BALB / c and C57BL /6 mice. 1 nmol Fc-stripped core was
coated onto microneedle arrays with poly I:C adjuvant and pressed into ear skin,
or injected as soluble protein with poly I:C adjuvant into the ear. Mice were
boosted 4 weeks after the prime. Serum titers were measured by yeast cell
surface titration (Figure 5.4B) All mice immunized with microneedles (three from
each strain) elicited steady serum responses of KA~ 5000. All three BALB / c mice
and two of three C57BL /6 mice injected intradermally elicited substantially
stronger responses of KA ~10000-20000. We attribute this difference in serum
titers to difference in delivered dose. IVIS imaging of microneedle arrays coated
with fluorescently labeled protein before and after application, as well as total
fluorescence in the ear, suggested that only ~10% of the coated dose enters the
skin from microneedle array during the five minute application time (data not
shown). In contrast, intradermal injection delivers the entire dose.
112
Differences in inflammation, retention, and biodistribution of antigen
administered intradermally or by microneedle array. Despite the difference in
delivered dose and resultant difference in serum titers, administration with
microneedles resulted in much stronger local inflammation as measured by in
vivo myeloperoxidase activity, which is due to phagocytic activity by neutrophils
(20). Bioluminescence was measured each day for several days after
administration of the prime, immediately after intraperitoneal injection of 200
mg/kg luminol (protocol identical to that described in (20)). A spike of
myeloperoxidase activity was observed one and two days post application in
mice administered the antigen via microneedles but not in mice that received the
intradermal injection (Figure 5.4C).
To determine whether the retention time and biodistribution of antigen differed
for the two delivery methods, fluorescently labeled Fc-stripped core was
administered and necropsies were performed four days later. In mice that were
injected intradermally, fluorescence was detected in the axillary lymph node. In
mice that received microneedle application, on the other hand, fluorescent signal
was retained in the skin of the cheek and neck (the ear was removed), and no
signal was observed in the draining lymph node (Figure 5.4D).
Steering the immune response to the VRC01 epitope using microneedles.
C57BL / 6 mice were immunized with 1 nmol Fc-stripped core coated on
microneedles with poly I:C adjuvant, then boosted four weeks later with either
the same immunogen or clone A. Serum titer was analyzed by yeast cell surface
titration. All three mice in the homologous immunogen group and two of three
in the heterologous immunogen group generated strong anti-core responses KA
1000-2000 (Figure 5.5A). One mouse in the heterologous immunogen group had
a weak response to core. When these sera were analyzed by the ternary complex
assay, it appears that one of three mice from the homologous group and one of
three mice from the heterologous group have specificity for the VRC01 epitope
(Figure 5.5B). The titer from another mouse in the heterologous group is too low
to be confident in the specificity measure. Further investigation of these mice is
113
warranted, but these results suggest that epitope focusing may be possible using
microneedles with only two immunogens.
Relationship between dose and KA across adjuvants, formulations, and routes.
The post-boost peak KA was plotted versus the protein dose per immunization
for all BALB / c mice immunized with Fc-stripped core (Figure 5.6). The data
includes intradermal injection of soluble protein with poly I:C, intradermal
administration of protein with poly I:C on microneedles, intranasal
administration of soluble protein with CpG DNA, subcutaneous injection of
protein conjugated to liposomes with MPLA, and subcutaneous injection of
protein with MPLA formulated with alum, AddaVax, or Montanide. A few
points merit comment. There appears to be a log-linear relationship between
protein dose and peak KA, particularly at concentrations above 50 pmol protein,
independent of route of administration or adjuvant. Below 50 pmol protein, the
KA is pretty consistently 1000-2000. Only AddaVax and Montanide buck this
trend with consistently high titers at low protein dose.
Discussion
Despite many efforts in the field, an HIV vaccine that elicits broad antibodybased protection has not yet been developed (1). Much of our work involves the
design of engineered antigens, but the successful translation of these proteins
into immunogens requires a functioning vaccination model. In the course of the
thesis, we have explored various formulations and routes of delivery in mice.
The species was chosen for its ease of use and the similarities in V gene usage,
VDJ recombination, and germinal center formation between it and humans (1214). In this chapter, we report the successful development of four different
vaccination modes-intranasal administration of soluble protein, subcutaneous
injection of protein-loaded liposomes, intradermal injection of soluble protein,
and intradermal administration of protein-coated microneedles-that elicit
reliable titers to stripped core and the engineered immunogens.
114
The different routes have different requirements for success. Intranasal
administration requires that the antigen be fused to an Fc in order to cross the
mucosal barrier, and also co-administered with GpG DNA adjuvant in order to
elicit anti-gp120 titers (Figure 5.1A). Subcutaneously injected liposomes require
that the antigen be conjugated to the particle surface and not simply mixed
(Figure 5.2B). The timing of the boost matters as well for the liposome
formulation, with a six-week gap between prime and boost eliciting stronger and
more consistent titers than four-week or two-week gaps (Figure 5.2A).
Intradermal administration of antigen-coated microneedles efficiently elicits
antibodies, but only -10% of the loaded dose is delivered to the animal per
application. Direct intradermal injection delivers the entire dose, but elicits less
local neutrophil activation than microneedle application (Figure 5.4). The
biological significance of the enhanced activation for this system has yet to be
determined.
We attempted to focus the immune response on the CD4 binding site by
sequential immunization with liposomal formulations of the same four
immunogens described in the previous chapter. We were surprised that the
series of immunization did not steer the immune response as measured by the
ternary complex assay (Figure 5.3B, 3C). There are two plausible hypotheses for
why the liposomes fail to steer: persistence and avidity. The persistence
hypothesis argues that the liposomes form a subcutaneous antigen depot and
persist longer than the four weeks between boosts. When the second immunogen
is administered with immunostimulatory adjuvant, there is enough of the first
immunogen present to boost all primed B cells. Experimental validation would
involve boosting at longer intervals after the prime. The avidity hypothesis
argues instead that the multivalent presentation of antigen on liposomes
counteracts the loss of affinity due to surface mutations on the diversified clones.
That is, a point mutation that causes a significant loss of binding to a particular
antibody in isolation may not disrupt binding enough to prevent B cell activation
when the antigen is presented multivalently on a liposome. Experimental
validation for this hypothesis would require adjusting the degree of protein
115
loading per liposome by, for example, varying the molar fraction of maleimide
lipid in the liposome.
Regardless of whether the depot effect is an important parameter in steering, it is
important to understand the pharmacokinetics and biodistribution of antigen
after administration by these different routes, and the cells and tissues involved
in mounting the immune response. The serum pharmacokinetics of antigen
administered intranasally was measured. The concentration of Fc-antigen peaks
in the bloodstream after 0.5-1 hour, then undergoes rapid a-phase clearance and
slower P-phase clearance due to FcRn recycling, as is observed with antibody
therapeutics and other Fc-containing molecules. Some fluorescence was observed
in the draining cervical lymph node four days after protein administration. The
fluorescence signals in serum and tissue were orders of magnitude lower than
what would be expected from the delivered dose, however, suggesting that only
a small fraction of the administered antigen is being captured in this assay. The
antigen may have drained to more distal nodes, or perhaps much did not cross
the mucosal barrier and was instead swallowed or expelled. Further
investigation is warranted.
Biodistribution of the two intradermal formulations was studied as well. Antigen
that was injected directly into the ear drained primarily to the axillary lymph
node and was present there four days after injection. In contrast, antigen that
was delivered via microneedle array appeared to be retained in the skin below
the affected ear (Figure 5.4D). Whether this retention is due to the Fc on the
antigen interacting with Fc-receptors on skin-resident macrophages or somehow
related to the enhanced neutrophil activation described earlier is not yet known.
Further investigation is required to establish the tissues and cells in which the
antigen-specific immune response is elicited via this route. Intriguingly, there is
evidence that one mouse may have generated a steered immune response after
the administration of just two antigens-stripped core and clone A-on
microneedles (figure 5.5B). This response should be explored further.
As the experiments described in this chapter suggest, the parameter space for
optimizing immunization route and formulation is enormous. There is no a priori
116
reason to prefer a particulate vaccine to a soluble one, or to suppose one recipient
tissue will be superior to another in eliciting the desired epitope-specific
antibody response. In an attempt to bring order to this landscape, we compared
post-boost peak serum titers across routes, formulations, and adjuvants.
Surprisingly, with the exception of depot-forming emulsion adjuvants AddaVax
and Montanide that give strong responses at a very low dose, we observe that
antigen dose is the primary parameter that determines serum titer (Figure 5.6).
Depot formation may hinder the sequential immunization strategy, which relies
upon clearance of one antigen before administration of a second, so these potent
adjuvants may not be suitable for repeated use. Perhaps, however, they could be
used as the final adjuvant, to strongly boost a weak but focused humoral
response.
Materials & Methods
Intranasal immunization. Mice were anesthetized with isoflurane then held by
the scruff with their necks along the edge of a cage. Gentle pressure was applied
to close the oropharynx, and liquid (maximum 20 yL) was held at the nostril
using a pipette. After the liquid was aspirated, mice were laid on their backs to
encourage draining into the nasopharynx. Studies of intranasal administration of
technetium-labeled colloid particles have shown that the particles are dispersed
throughout both the upper and lower respiratory tracts (21). Beyond the
pharmacokinetic study and lymph node biopsy described above, we did not
investigate the in vivo fate of intranasally-delivered antigen.
Liposome conjugation and purification. Liposomes were made by sonication
(70 nm diameter) or extrusion (200 nm diameter) of DOPC and DOPG lipids in a
4:1 molar ratio with 5-10% PEG-Maleimide-DSPE (Avanti Polar Lipids,
Alabaster, AL). Surface amines on the antigen were converted to thiols with the
bifunctional molecule SAT(PEG)4 (Thermo Fisher Scientific, Rockford, IL), an
amine-reactive NHS ester group linked to an acetylated thiol by a short PEG
chain. Deacetylated thiol-modified proteins were conjugated to the maleimide
117
liposomes overnight. Liposomes were purified from unreacted protein by size
exclusion chromatography on a Sepharose CL-4B column (GE Healthcare
Biosciences, Pittsburgh, PA). Purified fractions were concentrated using a 30 kDa
centrifugal filter (EMD Millipore, Billerica, MA) and sterilized with a 0.45 Pm
filter (note that a 0.22 pm filter will retain most of the 200 nm particles) (Pall, Port
Washington, NY). Injections were performed subcutaneously at the base of the
tail. Injection volumes were 50-100 yL.
Microneedle fabrication. Antigen was dissolved at 200 yM in PBS, and mixed
1:1 with a 5% sucrose and 13.9% Tween PBS solution. Poly I:C (Invivogen, San
Diego, CA) was added to 0.24 mg / mL. 75 jyL antigen solution was added to the
top surface of a PLLA microneedle array and carefully spread to cover the base,
before removing 35 yL to leave 40 y L solution. The coated arrays were incubated
under vacuum overnight. Mice were anesthetized by isoflurane inhalation and
ears were immobilized with tape. Microneedle arrays were applied to the ear and
held in place for five minutes.
118
A
B"~
Fc-gpl20 + CpG
gp120 + CpG
-a- Fc-gp120 no CpG
-49000
-o.01o6n'
bs
14000
soc-o20001
0
20
40
W0
8
100
TO*e (hrs)
C
4000
8000
700m
3000k
6000
2000
s500_____
4000-
Days post prime
0
Treated
Untreated
Figure 5.1. Intranasal immunization. (a) Serum titer KA for BALB/c mice
immunized intranasally with 150 pmol Fc-stripped core and 3 nmol CpG (blue),
300 pmol stripped core and 3 nmol CpG (green), or 150 pmol Fc-stripped core
without CpG (red). Mice were primed at day 0 and boosted at day 14. (b) Serum
fluorescence from a C57BL6 mouse immunized with 300 pmol Alexa 647-labeled
Fc-stripped core and 3 nmol CpG. (c) Fluorescence from a cervical lymph node
from the same immunized mouse ("Treated") and an unimmunized control
("Untreated").
119
120
B
-
rp
r o (sis
Boostweek 2
Boost wek 4
p
-0\
i
D"
sh
(
a
Al
~POWP&*M
pre20
ps2
asic8
13
C
y
Co(ttid
qp120
FC-PI2
-0- Unconjugatcor
edgp12O
-- Loncorjted Fc-gp2
-
..
lue
-a- C~fnUgd
/c
(l-a
ouBoostwed
w
-
o~
8
6
Boost
3-
S
- Uposomes
-0- Akum
41- Adit69
-0 Monterlde
Figure 5.2. Subcutaneous liposome immunizations of BALBc mice. (a) Serum
titer KA for mice immunized with 7.5 pmol stripped core conjugated to liposomes
with 1.3 pg MPLA adjuvant. Mice were boosted at week 2 (blue), week 4 (green),
or week 6 (red) after prime, or unboosted (black). (b) Serum titer KA for mice
immunized with 7.5 pmol stripped core (solid blue) or 3.75 pmol Fc-stripped,
core (solid green) conjugated to liposomes, or 7.5 pmol stripped core (dashed
blue) or 3.75 pmol Fc-stripped core (dashed green) mixed with liposomes. All
preps also included 1.3 pg MPLA. (c) Serum titer KA for mice immunized with
3.75 pmol Fc-stripped core conjugated to liposomes (solid green) or mixed with
alum (dashed red), AddaVax also called Adju,59 (solid black), or Montanide
(dashed black). All preps also included 1.3 jug MPLA.
121
122
A
Mouse 1
Seq
Par
Core
B
Serum pre-incubated withcore gp120
2
i-
-cr
_Q"
I to
~S
0
Mouse 2
Day
VRC01
yeawt
pre-incubated withcore p120
-
Mouse 3
C-50
7;0
-eq
-
Per
-COW
40
20
Mouse 4
10
VO
30
410
Da'*
so
so
710
Figure 5.3. Heterologous immunogen immunization with liposomes to focus the
immune response on the VRC01 epitope. (a) Serum titer KA for mice immunized
with immunogens sequentially (Seq), in parallel (Par), or with stripped core only
(Core). Titers are measured against yeast displaying stripped core (black), Clone
A (blue), Clone B (red), or Clone C (green). Mice have been immunized with Fcstripped core (day 0), Fc-Clone A (day 28), and Fc-Clone B (day 56). (b) Ternary
complex assay in which serum from day 63 (after three immunizations) is preincubated with stripped core gp120 before introducing VRC01 yeast. Data points
are mean ± standard deviation for four mice. Y-axis is the percent of cells that
stain positive for mouse serum. (c) Ternary complex assay in which yeast is preloaded with stripped core gp120 before introducing serum from day 63 (after
three immunizations). Data points are mean ± standard deviation for four mice.
Y-axis is the percent of cells that stain positive for mouse serum.
123
124
BBo :
A
LumMoIAssayofMPO
.C
Black6 ID
MBaoM
MN
1000
.
ka ..
maw
)
c0000
W
teooo
aAt ID
D
Aftw
BALB/MN
awauum
I
4-MN W~b
I
200M.0
a
0
0
s Post ""
Figure 5.4. Intradermal microneedle delivery. (a) Array of Fc-stripped corecoated poly-L lactic acid microneedles. (b) Serum titrations on gp120 yeast for
mice immunized with 1 nmol Fc-stripped core by direct injection-C57BL6
(black), BALB/c (red)-or by microneedle array-C57BL6 (blue), BALB/c
(green). Mice were primed on day 0 and boosted on day 28. Serum was collected
on day 35. (c) Luminol assay of myeloperoxidase activity. Three mice from either
strain C57BL6 (Black6) or BALB/c were administered 1 nmol Fc-stripped core by
direct intradermal injection (ID) or by microneedle array (MN). Signal was
measured daily for five days, and is quantified in the lower plot. (d) IVIS
imaging of BALB / c mice administered 1 nmol Alexa 647-labeled Fc-stripped core
by direct injection (ID) or microneedle array (MN). The lower images shown the
axillary lymph node and the large patch of skin to which the signal localized
after three days.
125
126
A
Cell surface titration of MN steering expt
90
-m-only
MN.i Day351
-4-
Ce-o
-o-
SueWMN.i Da 35!
B
Ternary complex assay
50
MN.1 Day 351
40 4-
60
50
aSer-gp120+Y
20
* Ser +gp120-Y
10O 420
40
iCore 1 Core 2 Core 3
Het 1
Hot 2
Het 3
-10
Figure 5.5. Heterologous immunogen immunization with microneedle arrays.
C57BL6 mice were primed with 1 nmol Fc-stripped core then boosted 4 weeks
later with either 1 nmol Fc-stripped core or 1 nmol Fc-Clone A on microneedle
arrays. (a) Cell surface titrations for serum collected on day 35. (b) Ternary
complex assay with serum from day 35 from homologous mice (Core) or
heterologous mice (Het). The blue bars are the percent serum-positive cells when
serum is pre-incubated with stripped core gp120 before introducing VRC01
yeast. The red bars are the percent serum-positive cells when yeast is pre-loaded
with stripped core gp120 before introducing serum.
127
128
Serum titer (Ka) vs. Dose
18 0 0 0
16000
4
-
14000
---
-
-
-
-
+Intradermal
12000
N Microneedle
10000 I ---
28000
A Intranasal
----
-
-Uposomes
6000
--
-
-
4000
gAlum
-
2000
1
0 Adju59
* Montanide
-
-
10
100
1000
gplZO Immunization dose (pmol)
Figure 5.6. Serum titer dependence on immunization dose. Serum titer KA for
BALB /c mice immunized with various doses of Fc-stripped core by intradermal
injection with poly I:C (blue), intradermal microneedle array with poly I:C (red),
intranasal administration with CpG DNA (green), liposomes with MPLA
(purple), AddaVax also called Adju59 (orange), or Montanide (black).
129
130
Works Cited
1. M. J. McElrath, B. F. Haynes, Induction of immunity to human
immunodeficiency virus type-1 by vaccination, Immunity 33, 542-554 (2010).
2. D. C. Montefiori, J. R. Mascola, Neutralizing antibodies against HIV-1: can we
elicit them with vaccines and how much do we need? Curr Opin HIV AIDS 4,
347-351 (2009).
3. K. J.Bar et al., Early Low-Titer Neutralizing Antibodies Impede HIV-1
Replication and Select for Virus Escape, PLoS Pathog 8, e1002721 (2012).
4. H. F. Stills, Adjuvants and antibody production: dispelling the myths associated
with Freund's complete and other adjuvants, ILAR J 46, 280-293 (2005).
5. N. Lycke, Recent progress in mucosal vaccine development: potential and
limitations, Nat Rev Immunol 12, 592-605 (2012).
6. R. Rappuoli, C. W. Mandl, S. Black, E. De Gregorio, Vaccines for the twenty-first
century society, Nat Rev Immunol 11, 865-872 (2011).
7. C. Ludwig, R. Wagner, Virus-like particles-universal molecular toolboxes,
Current Opinion in Biotechnology 18, 537-545 (2007).
8. E. V. L. Grgacic, D. A. Anderson, Virus-like particles: passport to immune
recognition, Methods 40, 60-65 (2006).
9. C. Ruedl, T. Storni, F. Lechner, T. Bachi, M. F. Bachmann, Cross-presentation of
virus-like particles by skin-derived CD8(-) dendritic cells: a dispensable role for
TAP, Eur.J. Immunol. 32, 818-825 (2002).
10.
J. Holmgren, C. Czerkinsky, Mucosal immunity and vaccines, Nat Med 11, S45-53
(2005).
11. M. F. Pasetti, J.K. Simon, M. B. Sztein, M. M. Levine, Immunology of gut
mucosal vaccines, Immunol Rev 239, 125-148 (2011).
12. D. K. Dunn-Walters, A. Belelovsky, H. Edelman, M. Banerjee, R. Mehr, The
dynamics of germinal centre selection as measured by graph-theoretical analysis
of mutational lineage trees, Dev. Immunol. 9, 233-243 (2002).
13. R. Mehr, H. Edelman, D. Sehgal, R. Mage, Analysis of mutational lineage trees
from sites of primary and secondary Ig gene diversification in rabbits and
chickens, J Immunol 172, 4790-4796 (2004).
131
14. H. W. Schroeder, Similarity and divergence in the development and expression
of the mouse and human antibody repertoires, Dev Comp Immunol 30, 119-135
(2006).
15. C. Sundling et al., High-Resolution Definition of Vaccine-Elicited B Cell
Responses Against the HIV Primary Receptor Binding Site, Sci Transl Med 4,
142ra96-142ra96 (2012).
16. L. Ye, R. Zeng, Y. Bai, D. C. Roopenian, X. Zhu, Efficient mucosal vaccination
mediated by the neonatal Fc receptor, Nat Biotechnol 29, 158-163 (2011).
17. L. Lu et al., An FcRn-targeted mucosal vaccine strategy effectively induces HIV-1
antigen-specific immunity to genital infection, J Virol (2011),
doi:10.1128/JVI.05441-11.
18. R. L. Coffman, A. Sher, R. A. Seder, Vaccine adjuvants: putting innate immunity
to work, Immunity 33, 492-503 (2010).
19. R. Spreafico, P. Ricciardi-Castagnoli, A. Mortellaro, The controversial
relationship between NLRP3, alum, danger signals and the next-generation
adjuvants, Eur. J. Immunol. 40, 638-642 (2010).
20. S. Gross et al., Bioluminescence imaging of myeloperoxidase activity in vivo, Nat
Med 15, 455-461 (2009).
21. D. S. Southam, M. Dolovich, P. M. O'Byrne, M. D. Inman, Distribution of
intranasal instillations in mice: effects of volume, time, body position, and
anesthesia, Am J Physiol Lung Cell Mol Physiol 282, L833-9 (2002).
132
Chapter 6. Targeting DEC-205 to deliver antigen to the crosspresentation pathway in dendritic cells
Background
Neutralizing antibodies are the correlate of protection for successful vaccines
against smallpox, poliovirus, rubella, tetanus, and bacterial toxins. Some diseases
for which effective vaccines do not yet exist, such as tuberculosis, parasitic
infections, and hepatitis C, may require CD8* T cell activation in addition to
antibodies in order to be cleared (1). Eliciting a robust CD8* T cell response by
vaccination remains a challenge.
Dendritic cells (DC) are the key innate mediators of adaptive immunity (2). Like
all nucleated cells, DCs present fragments of intracellular antigens on MHC-I
molecules for immune surveillance. Like other antigen presenting cells, DCs can
display peptides from extracellular antigens on MHC-II molecules to elicit CD4*
T cell responses. Unlike most other cell types, however, DCs can also crosspresent antigen-that is, process peptides from extracellular antigens onto MHCI molecules to prime CD8* T cells (3). This is the pathway we seek to exploit by
vaccination.
DEC-205 is a scavenger receptor on the surface of DCs that recognizes necrotic
self antigen (4). It actively transports its cargo into the cross-presentation
pathway, most likely in order to induce tolerance to self antigens. Delivery of
antigen conjugated to an antibody specific for DEC-205 induces peripheral
tolerance, whereas antigen delivered in an immunostimulatory context can
activate CD8* lymphocytes (5-8). Other receptors in the family including DCSIGN and the mannose receptor have also been targeted for their crosspresentation potential (9-12).
CpG DNA is a promising class of immunostimulatory molecules that engage
TLR9 and promote Th1-type inflammation. CG-rich sequences have been
implicated in responses to bacterial infection as well as autoimmune disorders
133
(13-16), and their mechanism of engagement of TLR9 has been investigated (1721). CpG DNA has been explored in a number of vaccine and
immunostimulatory contexts, both in the lab and in the clinic (22-28). Studies
have shown, however, that conjugating antigen to CpG is insufficient to elicit
CD8* T cell responses in the absence of robust cross-presentation (29).
The objective of this portion of the thesis is to combine antigen cross-presentation
with CpG immunostimulation to generate a dendritic cell-targeted vaccine that
elicits a strong CD8* T cell response. We engineer fibronectin domains (Figure
6.1A) to bind to DEC-205, express these molecules as fusions to a model
ovalbumin antigen, and conjugate CpG DNA to them. This all-in-one package
may be small enough to drain to follicular dendritic cells in the lymph node (30,
31). An original design-targeted antigen fused to zinc finger domains bound to
double stranded immunostimulatory and/or structured DNA-was discarded in
favor of direct conjugation, and is not discussed herein.
Results
Engineering fibronectins to bind DEC-205. The two outermost extracellular
domains of DEC-205-the N-terminal cysteine-rich and type II fibronectin
domains-were secreted from HEK-293 cells (Figure 6.1B). This extracellular
subunit was the immunogen used to elicit the rat anti-mouse DEC-205 studied by
the Steinman laboratory (5, 32, 33), and so was chosen as the antigen target for
library sorting.
The G4 library of human 10* type III fibronectin domains (Fn3) was displayed on
yeast (34, 35). This library of 1x10' unique clones has diversity in antigen binding
loop length and amino acid usage that mimics the human antibody CDR
repertoire. The library was sorted against antigen-loaded magnetic beads for four
rounds (36) before switching to flow cytometry-based sorting for three additional
rounds. In between rounds of selection, the library was diversified by errorprone PCR in the manner of Hackel et al. (37). By the fifth round of sorting the
library was able to bind well to 20 nM soluble DEC-205, and this affinity
134
increased in the sixth round before reaching a plateau in the seventh round
(Figure 6.1C).
After seven total rounds of sorting, 16 individual clones from the library were
isolated and sequenced. Four of these were displayed on yeast as single clones
and binding to soluble DEC-205 was measured. The fitted KD values ranged from
0.5-1.25 nM (Figure 6.1D). Ten clones were then subcloned into a pET vector for
expression in bacteria, immediately upstream of the ovalbumin MHC-I peptide,
SIINFEKL, and a terminal cysteine. SDS-PAGE analysis of bacterial lysates
indicated that four clones were expressed at high levels, three at medium levels,
and three hardly at all (data not shown). Sequences of select Fn3 are shown in
Table 6.1.
Conjugation of Fn3 clones to CpG DNA. Expressed proteins were purified on
Talon resin and conjugated to CpG DNA by means of a heterobifunctional crosslinker SM(PEG)8 . Amine-terminated CpG DNA was reacted with the NHS ester
of the cross-linker, desalted to remove excess linker, then conjugated to the
terminal thiol on the reduced Fn3. Gel electrophoresis indicated that the
conjugation had occurred and that the single conjugate was the dominant
product (Figure 6.2A), but further purification by ion exchange or size-exclusion
chromatography did not produce a pure product.
Effect on bone marrow-derived dendritic cells. We attempted to measure the
binding of select clones to DEC-205 on primary bone marrow-derived dendritic
cells (BMDC-hematopoetic stem cells cultured with GM-CSF). DEC-205 is
expressed at low levels on BMDCs (2000-5000 copies per cell) and BMDCs are
sticky, which made generating clean binding isotherms difficult. Equilibrium
binding constants measured both with two-step labeling and competition
titrations were calculated to be single-digit nanomolar (data not shown).
Endotoxin removal from protein preparations was an enduring challenge.
BMDCs were treated with Fn3-OVAp and Fn3-OVAp-CpG at various
concentrations for 24 hours and activation was assayed by IL12-p70 secretion
(Figure 6.2B). The unconjugated fibronectins generated higher cytokine secretion
135
than their conjugated counterparts, which suggests that the endotoxin present in
the preparation outweighed any immunostimulation provided by the CpG.
A preliminary attempt to measure cross-presentation suggested that BMDCs
might be able to cross present OVA peptide following treatment with Fn3OVAp. BMDCs were treated with Fn3-OVAp for 11 hours and the number of
peptide-MHC complexes was detected by flow cytometry by labeling with
antibody 25D1-16 specific for SIINFEKL-Kb (38). All four Fn3-OVAp clones tested
had peptide-MHC levels similar to a 1 nM pulse of SIINFEKL peptide (Figure
6.2C, 2D). There was no clear dose response, perhaps because the few DEC-205
molecules per BMDC were saturated.
Discussion
This section of the thesis remains largely preliminary. We developed fibronectin
binders to the dendritic cell surface receptor DEC-205 by standard directed
evolution and yeast surface display. The molecules bound soluble DEC-205 as
well as primary bone marrow-derived dendritic cells. Proteins were expressed in
bacterial cells, purified, and conjugated to immunostimulatory CpG DNA, but
the presence of CpG did not enhance the secretion of the inflammatory cytokine
IL12-p70, perhaps because of residual endotoxin.
Many fundamental experiments were shelved when the immunogen engineering
project took off. For example, conjugation of CpG to the validated rat anti-mouse
DEC-205 antibody would demonstrate whether the DEC-205 internalization
pathway crosses paths with TLR9, indicating whether this combination of
receptor and immunostimulant have synergistic potential.
There are other potential directions for this project that do not involve direct
cross-presentation and immunostimulation. For example, the DEC-205-binding
Fn3 could be conjugated to IgGs, like the multi-epitopic EGFR-targeting
molecules developed by Jamie Spangler and now in wide use in the lab (39), to
engage dendritic cells in immunotherapy or other applications.
136
Materials & Methods
Fibronectin engineering. The library was diversified by error-prone PCR in the
manner of Hackel et al. (37). See Ben Hackel's thesis for a detailed protocol.
Briefly, loops were subjected to a high mutagenesis rate then shuffled together,
while the whole gene was subject to a low mutagenesis rate to search for
beneficial, stabilizing framework mutations. The "gene" and "loop" libraries
were then transformed into yeast and sorted on beads (36) or by flow cytometry
(40) using standard methods.
Protein expression, purification, and conjugation to CpG. Proteins were
expressed in E. coli and purified on Talon resin. 5'-amino CpG DNA 1826
(phosphorothioate backbone 5'- TCCATGACGTTCCTGACGTT - 3') was reacted
with the NHS ester of SM(PEG),, desalted to remove excess linker, and
conjugated in excess to the reduced terminal thiol on Fn3.
Culturing dendritic cells. BMDCs were derived by culturing hematopoetic stem
cells in GM-CSF, as described by Inaba et al. (41). Briefly, C57BL / 6 mice were
euthanized, and tibias and femurs collected and cleaned. Hematopoetic stem
cells were extracted by flushing out the bone marrow with PBS and lysing the
red blood cells in hypotonic solution. Cells were resuspended in DC medium
(RPMI 1640 + 5% FBS + 1% Pen-Strep + L-Glutamine + GM-CSF 1:30 dilution J5
supernatant) at 2x105 mL-' and distributed in 24-well plates (1 mL per well). On
days 2 and 4, old media was replaced with 1 mL fresh media. On day 6, loosely
adherent cells were collected by gentle pipetting and resuspended at 1x10 6 mL-1
in fresh media. These cells were promptly used for flow cytometry, or plated for
treatment assays.
137
138
A
FG -
BC
B
----1
2
tni
*= 0 -M4
C
4
4.2
so
10
0
125
0
Ubrwy
*3.2
go-
7
0
5
-r"**WPf
D
100
3
1
A
+
721
-
Ka'.050W4
729
4 ' 80WW
72 11
-
72,12
K . 254M
- r4 0-714A
-4.
0.5
5.2
40
6.2
*7.2
20
0 10
0.25
00.01
10E
C
101
0.1
1
10
AIGX&488DEC-205 $nho
Alexa488-DEC-205
Figure 6.1. Engineered fibronectin binders to DEC-205. (a) Fibronectin
domain with antigen binding loops shown in red, blue, and orange. (b)
Schematic of DEC-205, with outermost two ectodomains marked with dashed
line. SDS-PAGE of soluble DEC-205 secreted from HEK 293 cells. (c) Libraries
from sequential rounds of directed evolution binding to 20 nM soluble DEC205. Sorting was performed by FACS beginning in round 5. (d) Titration of
soluble DEC-205 on yeast displaying select fibronectin clones.
139
140
IL-12
p70EUSA
A
1 23
4
5
B
6
+-Dimer
q#--- Conjugate
I
C
D
SINFEKL peptide pulse
20
Is
10
hkAu loriI
pMHC complexes- I hr Incubation
1
1hU
2.
0.01
01
1
10
too
~11
1000
+ Owcl-C
Dsc2-C
D~O4-C
XDso"
A Dm00-C
100
Conoma*01
SWFM 0'-
P10005 00n.1mn (11t"
Figure 6.2. Effect on bone marrow-derived dendritic cells (BMDC). (a) Select
fibronectin clones with terminal cysteine (1) and conjugated to CpG DNA by
maleimide-thiol chemistry (2-6). (b) Activation of BMDCs as measured by IL-12
p70 secretion for various concentrations of select fibronectins with and without
conjugation to CpG DNA. (c) Detection of OVA peptide-MHC complexes on
BMDCs pulsed with SIINFEKL using antibody 25D1-16. (d) Detection of OVA
peptide-MHC complexes on BMDCs incubated with DEC-205-targeted
fibronectins fused to OVAp using antibody 25D1-16.
141
142
Table 6.1. Sequences of select fibronectin clones
Clone
7.2.1
7.2.2
7.2.4
7.2.5
7.2.9
7.2.11
7.2.12
WT Fn3
BC
RSTPYAAY
RNTPYAAH
RSTPYAAY
RNTPYAAH
RSTPYAAH
HSSAYAYS
HPSAYAYS
DAPAVTVR
DE
RSASS
RSASS
RSNSS
RRFSV
RSRSK
WRRTN
WRRTN
GSKST
FG
WTGYPL
WTGYPL
WTGYSL
WTGYPL
WTGYPL
WTNYSL
WTGYPL
GRGDSPASSK
143
Framework
V11A, V72A
L8Q, S24N, A71T, V751, Y92H
L19P, P85S, 190T
L8Q, S24N
V1 A, P44S, T49A, D67G, T94A
V10A, V45A
V10A, T39A, N42S, K63R, D67G
Works Cited
1. R. M. Zinkernagel, On natural and artificial vaccinations, Annu Rev Immunol 21,
515-546 (2003).
2. P. Guermonprez, J. Valladeau, L. Zitvogel, C. Thdry, S. Amigorena, Antigen
presentation and T cell stimulation by dendritic cells, Annu Rev Immunol 20, 621667 (2002).
3. P. Guermonprez et al., ER-phagosome fusion defines an MHC class I crosspresentation compartment in dendritic cells, Nature 425, 397-402 (2003).
4. R. E. Shrimpton et al., CD205 (DEC-205): a recognition receptor for apoptotic and
necrotic self, Molecular Immunology 46, 1229-1239 (2009).
5. W. Jiang et al., The receptor DEC-205 expressed by dendritic cells and thymic
epithelial cells is involved in antigen processing, Nature 375, 151-155 (1995).
6. K. Inaba et al., Efficient presentation of phagocytosed cellular fragments on the
major histocompatibility complex class II products of dendritic cells, J Exp Med
188, 2163-2173 (1998).
7. D. Hawiger et al., Dendritic cells induce peripheral T cell unresponsiveness
under steady state conditions in vivo, J Exp Med 194, 769-779 (2001).
8. L. C. Bonifaz et al., In vivo targeting of antigens to maturing dendritic cells via
the DEC-205 receptor improves T cell vaccination, J Exp Med 199, 815-824 (2004).
9.
J. S.
Lam, M. K. Mansour, C. A. Specht, S. M. Levitz, A model vaccine exploiting
fungal mannosylation to increase antigen immunogenicity, J Immunol 175, 74967503 (2005).
10. T. Keler, L. He, V. Ramakrishna, B. Champion, Antibody-targeted vaccines,
Oncogene 26, 3758-3767 (2007).
11. E. J. Mckenzie et al., Mannose receptor expression and function define a new
population of murine dendritic cells, J Immunol 178, 4975-4983 (2007).
12. L.-Z. He et al., Antigenic targeting of the human mannose receptor induces tumor
immunity, J Immunol 178, 6259-6267 (2007).
13. A. M. Krieg, CpG motifs in bacterial DNA and their immune effects, Annu Rev
Immunol 20, 709-760 (2002).
144
14. A. M. Krieg et al., CpG motifs in bacterial DNA trigger direct B-cell activation,
Nature 374, 546-549 (1995).
15. G. A. Viglianti et al., Activation of autoreactive B cells by CpG dsDNA, Immunity
19, 837-847 (2003).
16. M. B. Uccellini et al., Autoreactive B cells discriminate CpG-rich and CpG-poor
DNA and this response is modulated by IFN-alpha, I Immunol 181, 5875-5884
(2008).
17. K. Yasuda et al., CpG motif-independent activation of TLR9 upon endosomal
translocation of "natural" phosphodiester DNA, Eur. J. Immunol. 36, 431-436
(2006).
18. A. M. Krieg et al., Sequence motifs in adenoviral DNA block immune activation
by stimulatory CpG motifs, Proc Natl Acad Sci USA 95, 12631-12636 (1998).
19. M. Rutz et al., Toll-like receptor 9 binds single-stranded CpG-DNA in a sequenceand pH-dependent manner, Eur.J. Immunol. 34, 2541-2550 (2004).
20. E. Latz et al., TLR9 signals after translocating from the ER to CpG DNA in the
lysosome, Nat Immunol 5, 190-198 (2004).
21. H. Wagner, The sweetness of the DNA backbone drives Toll-like receptor 9,
CurrentOpinion in Immunology 20, 396-400 (2008).
22. Y. Krishnamachari, A. K. Salem, Innovative strategies for co-delivering antigens
and CpG oligonucleotides, Advanced Drug Delivery Reviews 61, 205-217 (2009).
23. S. Jain, W. T. Yap, D. J. Irvine, Synthesis of protein-loaded hydrogel particles in
an aqueous two-phase system for coincident antigen and CpG oligonucleotide
delivery to antigen-presenting cells, Biomacromolecules 6, 2590-2600 (2005).
24. D. M. Klinman, D. Currie, I. Gursel, D. Verthelyi, Use of CpG
oligodeoxynucleotides as immune adjuvants, Immunol Rev 199, 201-216 (2004).
25. S. M. Standley et al., Incorporation of CpG oligonucleotide ligand into proteinloaded particle vaccines promotes antigen-specific CD8 T-cell immunity,
Bioconjug Chem 18, 77-83 (2007).
26. D. E. Speiser et al., Rapid and strong human CD8+ T cell responses to vaccination
with peptide, IFA, and CpG oligodeoxynucleotide 7909, Journalof Clinical
Investigation 115, 739-746 (2005).
145
27. C. Bourquin et al., Targeting CpG oligonucleotides to the lymph node by
nanoparticles elicits efficient antitumoral immunity, JImmunol 181, 2990-2998
(2008).
28. H. Wagner, The immunogenicity of CpG-antigen conjugates, Advanced Drug
Delivery Reviews (2009), doi:10.1016/j.addr.2008.12.010.
29. S. K. Datta, H. J. Cho, K. Takabayashi, A. A. Horner, E. Raz, Antigenimmunostimulatory oligonucleotide conjugates: mechanisms and applications,
Immunol Rev 199, 217-226 (2004).
30. S. T. Reddy et al., Exploiting lymphatic transport and complement activation in
nanoparticle vaccines, Nat Biotechnol 25, 1159-1164 (2007).
31. S. F. Gonzalez, L. A. Pitcher, T. Mempel, F. Schuerpf, M. C. Carroll, B cell
acquisition of antigen in vivo, Current Opinion in Immunology 21, 251-257 (2009).
32. W. J.Swiggard, A. Mirza, M. C. Nussenzweig, R. M. Steinman, DEC-205, a 205kDa protein abundant on mouse dendritic cells and thymic epithelium that is
detected by the monoclonal antibody NLDC-145: purification, characterization,
and N-terminal amino acid sequence, Cell Immunol 165, 302-311 (1995).
33. M. Guo et al., A monoclonal antibody to the DEC-205 endocytosis receptor on
human dendritic cells, Human Immunology 61, 729-738 (2000).
34. B. J. Hackel, M. Ackerman, S. W. Howland, K. D. Wittrup, Stability and CDR
Composition Biases Enrich Binder Functionality Landscapes, Journal of Molecular
Biology (2010), doi:10.1016/j.jmb.2010.06.004.
35. B. J. Hackel, K. D. Wittrup, The full amino acid repertoire is superior to
serine / tyrosine for selection of high affinity immunoglobulin G binders from the
fibronectin scaffold, Protein Eng Des Sel 23, 211-219 (2010).
36. M. Ackerman et al., Highly avid magnetic bead capture: an efficient selection
method for de novo protein engineering utilizing yeast surface display, Biotechnol
Prog25, 774-783 (2009).
37. B. J.Hackel, A. Kapila, K. D. Wittrup, Picomolar affinity fibronectin domains
engineered utilizing loop length diversity, recursive mutagenesis, and loop
shuffling, Journalof Molecular Biology 381, 1238-1252 (2008).
38. A. Porgador, J. W. Yewdell, Y. Deng, J.R. Bennink, R. N. Germain, Localization,
quantitation, and in situ detection of specific peptide-MHC class I complexes
using a monoclonal antibody, Immunity 6, 715-726 (1997).
146
39.
J. B.
Spangler, M. T. Manzari, E. K. Rosalia, T. F. Chen, K. D. Wittrup, Triepitopic
antibody fusions inhibit cetuximab-resistant BRAF- and KRAS-mutant tumors
via EGFR signal repression, Journalof MolecularBiology (2012),
doi:10.1016/j.jmb.2012.06.014.
40. G. Chao et al., Isolating and engineering human antibodies using yeast surface
display, Nature Protocols1, 755-768 (2006).
41. K. Inaba et al., Generation of large numbers of dendritic cells from mouse bone
marrow cultures supplemented with granulocyte/macrophage colonystimulating factor, J Exp Med 176, 1693-1702 (1992).
147