PROTEIN CAGE ARCHITECTURES FOR
TARGETED THERAPEUTIC AND
IMAGING AGENT DELIVERY
by
Michelle Lynne Flenniken
A dissertation submitted in partial fulfillment
of the requirements for the degree
of
Doctor of Philosophy
In
Microbiology
MONTANA STATE UNIVERSITY
Bozeman, Montana
July 2006
© COPYRIGHT
by
Michelle Lynne Flenniken
2006
All Rights Reserved
ii
APPROVAL
of a dissertation submitted by
Michelle Lynne Flenniken
This dissertation has been read by each member of the dissertation
committee and has been found to be satisfactory regarding content, English
usage, format, citations, bibliographic style, and consistency, and is ready for
submission to the Division of Graduate Education.
Dr. Trevor Douglas and Dr. Mark Young
Approved for the Department of Microbiology
Dr. Tim Ford
Approved for the Division of Graduate Education
Dr. Joseph J. Fedock
iii
STATEMENT OF PERMISSION TO USE
In presenting this dissertation in partial fulfillment of the requirements for a
doctoral degree at Montana State University – Bozeman, I agree that the Library
shall make it available to borrowers under rules of the Library. I further agree
that copying of this dissertation is allowable only for scholarly purposes,
consistent with “fair use” as prescribed in the U.S. Copyright Law. Requests for
extensive copying or reproduction of this dissertation should be referred to
ProQuest Information and Learning, 300 North Zeeb Road, Ann Arbor, Michigan
48106, to whom I have granted “the exclusive right to reproduce and distribute
my dissertation in and from microfilm along with the non-exclusive right to
reproduce and distribute my abstract in any format in whole or in part.”
Michelle Lynne Flenniken
July, 2006
iv
ACKNOWLEDGEMENTS
I would like to thank Dr. Mark Young and Dr. Trevor Douglas for being
excellent scientific mentors, providing state of the art facilities, helping to develop
my presentation and writing skills, and for fostering an exciting and enjoyable
research environment.
This dissertation work was funded by grants from NIH (RO1 GM61340, RO1
EB00432 and the Ruth L. Kirschstein National Research Service Award - F31
EB005093-01) and the National Aeronautics and Space Administration (NAG58807).
v
TABLE OF CONTENTS
1. A BIOMIMETIC APPROACH: BIOLOGICALLY PRODUCED, NANOMETER
SIZED PROTEIN CAGE ARCHITECTURES FOR BIOMEDICAL
APPLICATIONS .............................................................................................. 1
Abstract ........................................................................................................... 1
Inspired by Nature .......................................................................................... 2
Protein Cage Architectures ............................................................................. 4
Cowpea chlorotic mottle virus (CCMV)....................................................... 6
CCMV Structure................................................................................... 8
Production of CCMV Capsids .............................................................. 9
CCMV as a Template for Nanotechnology............................................... 10
Genetic Incorporation of Peptides within the CCMV Capsid .............. 14
CCMV as a Platform for Magnetic Resonance
Contrast Agent Delivery .............................................................. 15
CCMV’s Endogenous Metal Binding Site ........................................... 15
Genetic Incorporation of Metal Binding Peptide in CCMV .................. 16
Chemical Conjugation of Metal Chelators to CCMV ........................... 16
Small Heat Shock Protein (Hsp) from M. jannaschii................................. 17
Hsp Structure ................................................................................ 18
Production of Hsp Cages .............................................................. 20
Hsp as a Platform for Nanotechnology .................................................... 20
Biomineralization of Hsp Cages..................................................... 21
Hsp Architectures for Catalysis...................................................... 21
Hsp Cages as Magnetic Resonance Imaging Agents.................... 22
Human H-chain Ferritin (HFn) .................................................................. 22
Ferritin Structure ................................................................................. 22
Human H-chain Ferritin Production ..................................................... 24
Human H-Chain Ferritin as a Platform for Nanotechnology ..................... 24
Protein Cages for Targeted Therapeutic and Imaging Agent Delivery .......... 25
Attributes of Protein Cage Architectures
Beneficial for Nanotech Applications ........................................................ 26
Additional Protein Cage Architectures Being
Explored for Nanotechnology ..................................................................... 27
Cowpea Mosaic Virus (CPMV)................................................................. 28
CPMV as a Vaccine Vector ........................................................... 29
Chemical Derivatization of CPMV ................................................. 31
Other Nanoscale Therapeutic Delivery Systems........................................... 31
vi
TABLE OF CONTENTS - CONTINUED
2. THE SMALL HEAT SHOCK PROTEIN CAGE FROM M. jannaschii
IS A VERSATILE NANOSCALE PLATFORM FOR GENETIC
AND CHEMICAL MODIFICATION ................................................................ 38
Abstract ......................................................................................................... 38
Introduction ................................................................................................... 39
Results and Discussion ................................................................................. 40
Small Heat Shock Protein (Hsp) from Methanococcus jannaschii;
Wild Type and Genetic Variants of Hsp (HspG41C and HspS121C) . 40
Reactivity of Engineered Thiols and Endogenous Amines of Hsp ........... 42
Inorganic Modification of Hsp .................................................................. 47
Summary....................................................................................................... 49
3. SELECTIVE ATTACHMENT AND RELEASE OF A CHEMOTHERAPEUTIC
AGENT FROM THE INTERIOR OF A PROTEIN CAGE ARCHITECTURE.. 50
Abstract ......................................................................................................... 50
Introduction ................................................................................................... 51
Results and Discussion ................................................................................. 53
Derivatization of HspG41C with Doxorubicin............................................ 53
Release of Doxorubicin from HspG41C in Acidic Conditions ................... 57
Summary....................................................................................................... 58
4. MELANOMA AND LYMPHOCYTE CELL SPECIFIC TARGETING
INCORPORATED INTO A HEAT SHOCK PROTEIN CAGE
ARCHITECTURE .......................................................................................... 59
Abstract ......................................................................................................... 59
Introduction ................................................................................................... 60
Cell Specific Targeting Peptides Identified by the Phage Display
Technique........................................................................................... 60
Results and Discussion ................................................................................. 63
Tumor Targeted HspG41C-RGD4C ......................................................... 63
Chemical Derivatization of HspG41C-RGD4C ......................................... 66
HspG41C-RGD4C Cages Bind C32 Melanoma Cells .............................. 69
Chemical Introduction of Cell-Specific Targeting to Hsp
via Antibody Conjugation ......................................................... 75
Cell Specific Binding of Ab-Hsp Conjugates............................................. 76
Summary....................................................................................................... 78
vii
TABLE OF CONTENTS – CONTINUED
5. INVESTIGATION OF THE INTERACTION BETWEEN MAMMALIAN
CELLSAND PROTEIN CAGE ARCHITECUTRES AND EFFICACY OF
THERAPEUCTIC CONTAINING Hsp CAGES .............................................. 80
Abstract ......................................................................................................... 80
Introduction ................................................................................................... 81
Results and Discussion ................................................................................. 84
Derivatization of HspG41C-RGD4C with Doxorubicin ............................. 85
Evaluation of the Stability of HspG41C-RGD4C ...................................... 89
Studies of Protein Cage-Mammalian Cell Interactions in vitro ................. 90
Subcellular Localization of Protein Cage Architectures ........................... 92
Cell Proliferation Assay............................................................................ 94
Efficacy of Non-Targeted HspG41C Cages Containing
Doxorubicin in vitro ............................................................................. 97
Tumor Targeted HspG41CRGD4C-Dox .................................................. 98
Subcellular Localization of Doxorubicin in vitro...................................... 102
Summary..................................................................................................... 105
6. CHARACTERIZATION OF THE BIOCOMPATIBILITY OF PROTEIN CAGE
ARCHITECTURES IN MURINE MODELS: INITIAL TUMOR LOCALIZATION,
BIODISTRIBUTION AND IMMUNE RESPONSE DATA ............................. 107
Abstract ....................................................................................................... 107
Introduction ................................................................................................. 108
Background Information of the Biocompatibility and/or Biodistribution of
Other Nanometer-Scale Delivery Systems ....................................... 109
Background Relevant to the Characterization of the Immune Response
to Protein Cage Architectures........................................................... 112
Biodistribution of Other Types of Nanoparticles ..................................... 114
Results and Discussion ............................................................................... 116
Localization of Hsp Cages within Tumor Vasculature............................ 116
Biodistribution of Protein Cage Architectures (HspG41C, HspG41C-GFE,
CCMV K42R) in Naïve and Immunized Mice.................................... 119
Iodination of Protein Cages
(HspG41C, HspG41C-GFE, CCMV K42R) ................................. 120
Specific Activity of Radiolabeled Protein Cage Architectures ................ 124
Percent 125I Incorporation Studies.......................................................... 126
Biodistribution of 125I Labeled Protein Cages (HspG41C, HspG41C-GFE,
CCMV K42R) in Naïve and Immunized Mice .............................. 131
viii
TABLE OF CONTENTS – CONTINUED
Pilot Study Results – Biodistribution of Lung Targeted
HspG41C-GFE vs. Non-Targeted HspG41C in Naïve Mice........ 131
Biodistribution of Different Protein Cage Architectures
(HspG41C and CCMV K42R) in Naïve and Immunized Mice ..... 133
Biodistribution of Free 125I in Mice .............................................. 139
State of Excreted Cages – Intact or Not?.................................... 141
Initial Evaluation of the Immune Response to Protein
Cage Architectures ............................................................................ 143
Characterization of the Immune Response to Protein Cage
Architectures........................................................................... 143
Antibody Response to CCMV and Hsp Protein
Cage Architectures ................................................................. 143
Hypersensitivity to CCMV and Hsp Evaluated in Mice ................ 145
Summary..................................................................................................... 146
7. ADDITIONAL GENETIC VARIANTS OF TWO PROTEIN CAGE PLATFORMS
(Hsp and HFn) WITH CELL-SPECIFIC TARGETING POTENTIAL ............ 148
Abstract ....................................................................................................... 148
Introduction ................................................................................................. 149
Results and Discussion ............................................................................... 153
HspG41C-GFE Engineered to Target Membrane Dipeptidase
Expressed On Lung Endothelium ....................................................... 153
Tumor Targeting HspG41C-RGD2C and HspG41C-RGDWXLE ........... 155
Tumor Targeted Human H-Chain Ferritin (RGD-HFn)............................ 156
Cell Targeting Assay of Mineralized RGD4C-Fn ......................... 157
Cell Targeting of Fluorescein Labeled RGD4C-Fn...................... 157
Summary..................................................................................................... 159
8. CHEMICAL DERIVATIZATION OF GENETIC VARIANTS OF COWPEA
CHLOROTIC MOTTLE VIRAL (CCMV) CAPSIDS...................................... 160
Introduction and Abstract ............................................................................ 160
Results and Discussion ............................................................................... 161
CCMV Virus-Like Particle (VLP) Production in Pichia pastoris .............. 161
Genetic Engineering of CCMV Variants with Internally
Exposed Cysteines ............................................................................. 162
Characterization of CCMV K42R T48C.................................................. 165
Derivatization of CCMV K42R T48C ...................................................... 169
Summary ............................................................................................... 173
ix
TABLE OF CONTENTS – CONTINUED
9. SUMMARY AND CONCLUSIONS .............................................................. 172
Comparison of Hsp, HFn, and CCMV Protein Cage Platforms for
Their Development as Targeted Therapeutic and Imaging Agent
Delivery Systems ........................................................................................ 176
Future Studies............................................................................................. 180
REFERENCES CITED ..................................................................................... 181
APPENDIX A – MATERIALS AND METHODS ................................................ 205
x
LIST OF TABLES
Table
Page
2.1 Extent of Hsp Labeling with Fluorophores................................................... 46
6.1 Summary of 125I Labeling Reaction Results............................................... 125
6.2 TCA Precipitation Data from 125I Labeled Hsp and CCMV......................... 129
6.3 Percent of Injected Dose per Gram of Tissue; T-test Results.................... 136
6.4 Percentage of Intact Protein Cage (CCMV and Hsp) Excreted By Mice... 143
7.1 Hsp Genetic Variants Summarized......................................................... 150-1
8.1 CCMV Mutagenesis Primers .................................................................. 165-6
8.2 Degree of CCMV Labeling with Fluorescein-5-Maleimide ......................... 170
9.1 Summary of the Utility of Different Protein Cage Architectures
(Hsp, HFn, CCMV) .................................................................................... 178
xi
LIST OF FIGURES
Figure
Page
1.1. Bio-Inspired: Viral Capsids as Protein Cage Architectures ............................ 2
1.2 Protein Cage Architectures for Therapeutic and Imaging Agent Delivery ...... 3
1.3 Douglas and Young Lab Protein Cage Library .............................................. 6
1.4 Structure of the Cowpea chlorotic mottle viral Capsid ................................... 9
1.5 Image of Methanococcus jannaschii............................................................ 17
1.6 Structure of the Small Heat Shock Protein from M. jannaschii .................... 19
1.7 Structure of Human Ferritin ......................................................................... 23
2.1 Space Filling Model of the 24 Subunit MjHsp Cages................................... 41
2.2 MjHsp Cage Characterization...................................................................... 42
2.3 Fluorescence Emission of the Fluorescein Labeled Hsp Cages.................. 44
2.4 Characterization of the Fluorescently Labeled HspS121C .......................... 44
2.5 Characterization of Fluorescein Labeled HspG41C..................................... 45
2.6 SDS-Polyacrylamide Gel Electrophoresis of Hsp ........................................ 46
2.7 Transmission Electron Microscope (TEM) Images ...................................... 48
2.8 Dynamic Light Scattering (DLS) Data.......................................................... 49
3.1 Reaction Schematic..................................................................................... 52
3.2 Characterization of Doxorubicin Labeled HspG41C .................................... 54
3.3 SDS-PAGE of Doxorubicin Linked to HspG41C .......................................... 56
3.4 Mass Spectrometry of HspG41C-Doxorubicin (Dox) ................................... 56
xii
LIST OF FIGURES – CONTINUED
Figure
Page
3.5 Acid Triggered Release of Doxorubicin from HspG41C Cages ................... 57
4.1 Phage Display Technique for Discovering Peptides That Interact with
Particular Targets.......................................................................................... 62
4.2 Characterization of the αvβ3 Integrin Targeted HspG41C-RGD4C Protein
Cage ............................................................................................................. 65
4.3 Characterization of Fluorescein Labeled HspG41C-RGD4C Cages............ 67
4.4 Deconvoluted Mass Spectra of HspG41CRGD4C-Fluorescein ................... 69
4.5 Epifluorescence Microscopy of C32 Melanoma Cells with Hsp CageFluorescein Conjugates ................................................................................ 70
4.6 Specific Binding of HspG41CRGD4C-Fluorescein Labeled Cages to C32
Melanoma Cells ............................................................................................ 72
4.7 Binding of HspG41CRGD4C-Fluorescein Labeled Cages to C32 Melanoma
Cells .............................................................................................................. 74
4.8 HspG41CRGD4C-Fluorescein Cage Interaction with C32 Melanoma Cells is
Blocked by Unlabeled HspG41C-RGD4C but not by HspG41C or Albumin.. 74
4.9 Dynamic Light Scattering (DLS) Analysis of Anti-CD4 mAb HspG41C Cage
Conjugates .................................................................................................... 76
4.10 Epifluorescence Microscope Images Illustrating the Specific Binding of AntiCD4 mAb Conjugated HspG41C-Fluorescein Cages (Ab-HspG41C-Fl) to
CD4+ Lymphocytes....................................................................................... 77
4.11 Specific Binding of Anti-CD4 mAb Conjugated HspG41C-Fluorescein
Cages (Ab-HspG41C-Fl) to CD4+ Lymphocytes........................................... 78
5.1 Characterization of HspG41C-RGD4C Cages
Derivatized with Doxorubicin ....................................................................... 86
xiii
LIST OF FIGURES – CONTINUED
Figure
Page
5.2 SDS-PAGE of HspG41C-RGD4C Subunits Demonstrates Covalent
Linkage of Doxorubicin.................................................................................. 87
5.3 Deconvoluted Mass Spectrum of HspG41CRGD4C-Doxorubicin................ 88
5.4 The Stability of HspG41C-RGD4C (37˚C, DPBS pH 7.4) ............................ 90
5.5 Epifluorescence Microscope Images of HeLa Cells Transfected with
HspG41C-FlMal ............................................................................................ 92
5.6 Epifluorescence Microscope Images of HeLa Cells Transfected with CCMVFlMal ............................................................................................................. 96
5.7 MTS Assay of HeLa Cell Proliferation ......................................................... 95
5.8 C32 Melanoma MTS Cell Proliferation Assays ............................................ 96
5.9 Efficacy of HspG41C-Dox, Free Dox, and Mal-Dox Evaluated
In HeLa Cells................................................................................................. 97
5.10 Preferential Adherence of C32 Melanoma Cells to HspG41C-RGD4C
Protein Spots................................................................................................. 99
5.11 C32 Melanoma Cell Proliferation Assay .................................................. 100
5.12 C32 Melanoma Cell Proliferation Assay .................................................. 101
5.13 Nuclear Localization of Doxorubicin within C32 Melanoma Cells ............ 102
5.14 Subcellular Localization of Doxorubicin within C32 Melanoma Cells....... 103
5.15 Subcellular Localization of Doxorubicin within C32 Melanoma Cells....... 104
6.1 Tumor Vasculature Localization of Hsp ..................................................... 117
6.2 HspG41C-RGD4C is Take Up by RES System: Not Other Organs........... 118
6.3 Iodination Reaction.................................................................................... 121
xiv
LIST OF FIGURES – CONTINUED
Figure
Page
6.4 Characterization of HspG41C, HspG41C-GFE, and CCMV K42R Before
(A,C,E) and After (B, D, F) Iodination....................................................... 122-3
6.5
125
I Standard Curve ................................................................................... 124
6.6 Differential Presentation of Tyrosines on the Hsp and CCMV Protein Cage
Architectures ............................................................................................... 126
6.7 SDS-PAGE Analysis of TCA Precipitation of CCMV and Hsp ................... 130
6.8 Lung Localization Data of HspG41C-GFE ................................................. 132
6.9 Selected Organ Distribution of 125I-Labeled Cages.................................... 135
6.10 Percent of Dose per Organ in Selected Tissues...................................... 137
6.11 CCMV and Hsp Specific Antibody Titer Reduced by PEGylation ............ 144
7.1 Characterization of HspG41C-GFE ........................................................... 154
7.2 Lung Targeted HspG41C-GFE Binds ex vivo Lung Tissue ....................... 154
7.3 TEM Images of HspG41C-RGD2C and HspG41C-RGDWLE ................... 155
7.4 C32 Melanoma Cell Binding Capability of Derivatized RGD4C-Fn Protein
Cage Architectures...................................................................................... 158
8.1 Space Filling Model of CCMV K42R T48C ................................................ 168
8.2 Characterization of CCMV K42R T48C ..................................................... 168
8.3 TEM Characterization of CCMV K42R T48C............................................. 169
8.4 Size Exclusion Chromatography Elution Profiles of CCMV Cages ........... 172
8.5 Texas-red Derivatized CCMV K42R T48C ................................................ 173
xv
ABSTRACT
Protein cage architectures such as viral capsids, heat shock proteins, and
ferritins are naturally occurring spherical structures that are potentially useful
nanoscale platforms for biomedical applications. This dissertation work
demonstrates the utility of protein cages including their use as therapeutic and
imaging agent delivery systems. Protein cage architectures have clearly
demarcated exterior, interior, and interface surfaces and their structures are
known to atomic level resolution. This information is essential for the engineering
of functionalized nanoparticles via both chemical and genetic modification. In the
process of tailoring protein cage architectures for particular applications,
fundamental information about the architectures themselves is gained.
The present work describes endeavors toward the use of three different
protein cage architectures, the Cowpea chlorotic mottle viral capsid (CCMV), a
small heat shock protein (Hsp) architecture originally isolated from the
hyperthermophilic archaeon Methanococcus jannaschii, and human H-chain
ferritin, as cell-specific therapeutic and imaging agent delivery systems. Each
protein cage is roughly spherical, but their sizes differ; CCMV is 28 nm in
diameter, whereas Hsp and HFn are 12 nm in diameter. The advantages and
disadvantages of all three architectures are described.
Wild type and genetic variants of the Hsp, HFn, and CCMV cages were
reacted for the site specific attachment of organic molecules such as therapeutic
agents, imaging agents, and targeting ligands. Inorganic chemical modification
of the cages was employed for the formation iron oxide nanoparticles which are
potentially useful as magnetic resonance imaging (MRI) contrast agents. Toward
the development of the Hsp platform for therapeutic delivery, the antitumor agent
doxorubicin was covalently bound on the interior of the cage and selectively
released via a pH dependant trigger. In addition, mammalian cell-specific
targeting was imparted to the Hsp and HFn cages by both genetic and chemical
strategies. Biodistribution studies of Hsp and CCMV were performed in naïve
and pre-immunized mice and Hsp cages localized to human tumor xenografts in
mice. Together these results demonstrate the utility of the protein cages as
robust nanoscale platforms for the synthesis of both soft (organic) and hard
(inorganic) materials.
1
CHAPTER 1
A BIOMIMETIC APPROACH: BIOLOGICALLY PRODUCED,
NANOMETER SIZED PROTEIN CAGE ARCHITECTURES
FOR BIOMEDICAL APPLICATIONS
Abstract
In the next twenty years it is anticipated that the medical field will be
transformed by the use of nanotechnologies to diagnose, image, treat, and
prevent diseases and disorders [18-20]. The generation of these nanotechnologies requires basic research at the interfaces of biology, chemistry,
physics, and engineering. The work presented here is an investigation into the
use of biologically produced nanoparticles for biomedical applications. Nature
has provided nanometer sized templates including viral capsids, ferritins, and
heat shock protein architectures. The fundamental properties of these protein
cage architectures, such as their structure and biological function, have been
investigated. Based on our current understanding of these architectures, genetic
and chemical modifications were employed to alter their properties and endow
them with biomedical functionality. This work is iterative in nature and as we
learn more about the protein cage architectures through our efforts to develop
them for specific applications we gain insight about the architectures themselves.
2
Inspired by Nature
Viruses are biological entities that have co-evolved with the host organisms in
which they thrive. They are both simple (encoding relatively few proteins since
many key enzymes are ‘borrowed’ from their hosts) and complex (their ability to
precisely assemble their capsids within which their nucleic acid genomes are
specifically packaged). Viruses are obligate intracellular pathogens which are
highly evolved to deliver their genomes into host cells (Figure 1.1). They are cellspecific delivery experts. It is this property that we have tried to mimic, in our
efforts to understand and engineer protein cage architectures with biomedical
purposes such as cell-specific delivery of therapeutic and imaging agents.
EXTERIOR
Cell Specific Targeting
SUBUNIT INTERFACE –
Precise Molecular Assembly with Symmetry
INTERIOR
Package Nucleic Acid ‘Cargo’
Figure 1.1. Bio-Inspired: Viral Capsids as Protein Cage Architectures
The cryo-reconstruction of the Sulfolubos turreted icosahedral virus (STIV) [3, 6]. STIV is
used to highlight the structural features of viral capsids (and other protein cage
architectures) that can be chemically and genitically modified to transform them into nanovessels for targeted delivery of ‘cargo molecules’ including therapeutic and imaging
agents.
3
Nature has provided us with many ‘protein cage’ platforms, including viral
capsids, ferritins, and heat shock proteins, which are comprised of multiple
subunits assembled into precisely defined structures with three distinct surfaces
(exterior, subunit interface, and interior) (Figure 1.1). Protein cages can be
viewed as molecular ‘nano-vessels’ which are inherently multifunctional. Thus,
therapeutic molecules can be housed within their interior, imaging agents can be
incorporated at subunit interfaces, and the exterior surfaces are available for the
presentation of cell or tissue specific targeting ligands (Figures 1.1 & 1.2). The
multivalent nature and nanometer size range (5 -100 nm) of protein cages allows
for the delivery of multiple therapeutic molecules per ‘delivery event’. Housing
therapeutics within protein cage architectures may restrict their bioavailability
until their release is triggered by a particular cellular or tissue specific signal. In
addition, some protein cages, including CCMV, have structural features that
allow the for the controlled encapsulation and release of material [10, 21, 22].
RG
D
RG
D
Gd
Gd3+
3+
3+ Gd
Gd
Gd3+3+
Incorporation of
3+
Gd
RGD
D
Cell Specific Targeting RG
Gd3+
Gd3+
Gd3+
Gd3+
RG D
Gd3+
Gd3+
Gd3+
Gd3+
D
RG
R GD
Gd3+
Gd3+
Gd3+R
Gd3+ GD
D
RG D
D
R
G
D
RG
(Hsp, viral capsid)
R
RG
GD
Gd3+
Gd3+
Gd3+
Gd3+
Protein Cage
Architecture
Gd3+
RGD
Therapeutic
Encapsulation
Gd3+3+
Gd
D
RG
Functionalized
Protein Cage
Figure 1.2 Protein Cage Architecture for Therapeutic and Imaging Agent Delivery.
This schematic depicts the step-wise modification of two different protein cage
architectures resulting a nanometer sized therapeutic and imaging agent delivery vehicle.
RG
D
4
The overall goal of my dissertation work was to further our understanding of
protein cage architectures by exploring their potential utility for biomedical
applications. This approach combined biology (protein cage architectures,
peptides, antibodies, cells) with synthetic and materials chemistry. The
similarities and differences in biological packaging of cargo were investigated in
three distinct protein cage architectures: the Cowpea chlorotic mottle viral
(CCMV) capsid, an archaeal small heat shock protein (Hsp) cage, and the
human H-chain ferritin (HFn) architecture. These protein cages were utilized to
encapsulate therapeutic and imaging agents on their interiors and present
mammalian cell-specific targeting functionalities on their exteriors [23, 24]. The
distinct advantages and disadvantages associated with individual architectures,
as a consequence of their unique structural and biochemical features, are
discussed. In addition, initial research on the efficacy of these constructs in cell
culture, their biocompatibility, and in vivo tumor localization and bio-distribution is
reported.
The hypothesis tested was that nanometer scale protein cage architectures,
including CCMV, Hsp and HFn, can serve as targeted therapeutic and imaging
agent delivery systems.
Protein Cage Architectures
Protein cage architectures are naturally occurring, self-assembled, hollow
spheres that are 5 -100 nm in diameter. They include viral capsids and cellular
5
proteins such as ferritins, heat shock proteins (Hsp), and DNA binding proteins
from starved cells (Dps). These architectures play critical biological roles. For
example, viral capsids serve to house, protect, and deliver nucleic acid genomes
to specific host cells. Heat shock protein assemblies act as chaperones that
prevent protein denaturation, and ferritins are known to store iron (which is both
essential and toxic) as a nanoparticle of iron oxide [25, 26]. While each of these
structures has evolved to perform a unique biological function, they are similar in
that they are all essentially proteinaceous containers with three distinct surfaces
(interior, exterior, and subunit interface) to which one can potentially impart
function by design (Figure 1.1). Protein cage architectures have demonstrated
utility in nanotechnology with applications including inorganic nanoparticle
synthesis and the development of targeted therapeutic and imaging delivery
agents [27-47].
Protein cage architectures are naturally diverse. Each has unique attributes
including size, structure, solvent accessibility, chemical and temperature stability,
structural plasticity, assembly and disassembly parameters, and electrostatics
which are useful for particular applications. Importantly, one can capitalize on
these features or alter them via genetic or chemical modification. Atomic level
structural information identifies the precise location of amino acids within protein
cage architectures and in turn allows for the rational inclusion, exclusion, and
substitution of amino acid(s) (at the genetic level) resulting in protein cages with
novel functional properties.
6
The laboratories of Mark Young and Trevor Douglas have developed a ‘library
of protein cages’ (including CCMV, ferritin, Hsp, and Dps) as size constrained
reaction vessels and as platforms for genetic and chemical modification (Figure
1.3) [22, 27, 30, 32, 36, 38, 39, 48-56]. My work has focused on the use of three
protein cage architectures: a small heat shock protein (Hsp) from Methanococcus
jannaschii, the Cowpea chlorotic mottle virus (CCMV) capsid, and human Hchain ferritin (HFn) (boxed in Figure 1.3).
12 nm HFn
28 nm CCMV
12 nm Hsp
38 nm NV
70 nm STIV
9nm SsDps
9nm LiDps
18 x 300 nm TMV
Figure 1.3. Douglas and Young Lab Protein Cage Library
The structures of these protein cage architectures illustrate their similarities and differences.
From left to right: Sulfolobus turreted icosahedral virus (STIV) [3]; Norwalk virus [11, 12];
Cowpea chlorotic mottle virus (CCMV) [10, 12]; human ferritin [13]; the small heat shock
protein (Hsp) from Methanococcus jannaschii [4]; Dps-like protein from Sulfolobus
sofaltaricus [14, 15]; Dps protein from Listeria innocua [16]; and the tobacco mosaic virus
[17]. Red boxes indicate those protein cages described herein.
Cowpea chlorotic mottle virus (CCMV)
The Cowpea chlorotic mottle virus (CCMV) was the first protein cage
architecture utilized by the Douglas and Young Lab [22]. CCMV is a member of
7
the Bromoviridae family [57]. This virus naturally infects the cowpea plant (Vigna
unguiculata) which is found in the Southern United States (reviewed in [58]). Its
genome consists of three positive sense, single stranded RNA molecules
packaged separately [10, 57, 58]. Di-cistronic RNA3 encodes the 190 amino
acid coat protein (20 kDa) (Figure 1.4). As described below, the crystal structure
of CCMV has been solved to near atomic resolution (3.2 Å) [59, 60]. The first 26
or 42 amino acids of each subunit, at the 6-fold and 5-fold capsid symmetry axes,
respectively, were disordered in the crystal structure and therefore their structure
is unknown [59, 60]. The first 26 amino acids contain a lot of positively charged
amino acids (Arg (R), Lys (K)) which interact with the negatively charged nucleic
acid within the interior of the capsid. Recent data indicate that the N-terminus of
the subunit protein is sometimes exposed to the exterior of the capsid as it is
susceptible to protease cleavage [61]. One of the initial motivations for exploring
the use of CCMV as a platform for genetic and chemical modification (described
in detail below) was the wealth of structural information available. Further
exploration of CCMV capsids by screening for naturally occurring mutants led to
the discovery of new properties, such as the ‘salt stable’ mutant, which did not
disassemble under conditions that normally result in wild type (wt) capsid
disassembly (ionic strength > 1.0M, pH 7.5) [62-66].
Later work determined that
a single amino acid change in the capsid subunit protein, a lysine to arginine
change at position 42 (K42R), was responsible for the increase in capsid stability
under conditions of high ionic strength [62-66]. This example illustrates the
8
importance of each amino acid to capsid assembly. By subjecting the CCMV
structure to genetic and chemical modification, we continue to learn about its
inherent properties.
CCMV Structure. 180 copies of the coat protein assemble into a capsid with
icosahedral (T=3) symmetry. The exterior diameter of the CCMV capsid is 28 nm
and the interior cavity in which the viral RNA genome is housed is 25 nm in
diameter [10, 67, 68]. The structure of wild type CCMV has been solved to a 3.2
Å resolution [10] (Figure 1.4). Naturally, Ca2+ metal ions are bound at the
pseudo 3-fold axis of the CCMV capsid. Specifically, amino acids E81, E143,
and D153 comprise the metal binding site (discussed further in the next section,
in the context of binding other metals) [29, 58, 69]. The CCMV viral capsid is
capable of undergoing a reversible structural transition that is pH and metal ion
dependent [10, 70]. This ‘reversible gating’ or ‘swelling’ results in a 10%
increase in viral dimension at pH > 6.5 in the absence of metal cations; at pH <
6.5 the virion is in it’s ‘closed’ confirmation [10]. In addition, this structural
transition results in the creation of sixty ~20 Å pores that allow access between
the interior and exterior of the cage [10, 70]. This property will be discussed
further in the context of entrapping metallic nanoparticles within the capsid
interior.
9
A.
Surface exposed
loops
N-terminal extension
β-barrel
core
exterior
B.
interior
C-terminal extension
Figure 1.4. Structure of the Cowpea chlorotic mottle viral capsid
A. Ribbon diagram of the CCMV coat protein subunit (Speir et al., 1995).
The N (amino acids 1– 51) and C (amino acids 176–190) terminal arms extend away from
the central, eight-stranded, anti-parallel β-barrel core (amino acids 52–176) [9, 10].
B. Cryo-electron microscopy image reconstruction of the assembled 28 nm diameter
CCMV capsid composed of 180 subunits [10, 12].
Production of CCMV Capsids. In addition to harvesting CCMV capsids from
cowpea plants, two heterologous expression systems (Escherichia coli and
Pichia pastoris) have been developed [55, 67]. Both RNA containing and ‘empty’
viral capsids are produced. Capsids containing viral RNA are produced in plants.
Importantly, the RNA can be removed from capsids resulting in hollow
proteinaceous nano-containers [71]. The first heterologous expression system
developed was in E. coli (bacteria) [9, 67]. The DNA encoding the CCMV coat
protein was cloned into a bacterial plasmid, from which it was expressed [9, 67].
CCMV coat protein subunits were purified and assembled into empty capsids by
dialysis, but the capsid assembly efficiency of coat proteins expressed in E. coli
was low [9, 67]. This system often produced subunit proteins that did not selfassemble or were insoluble [9, 67]. Therefore, another heterologous expression
10
system, utilizing Pichia pastoris (yeast), was developed and is more prevalently
utilized for CCMV capsid production [55]. CCMV virus-like particle (VLP)
production from the yeast heterologous expression system is discussed further in
Chapter 8 and in the Materials and Methods (Appendix A) [55, 72]. Since only
the DNA encoding the CCMV coat protein was cloned into the expression vector,
the VLPs produced in yeast are inherently devoid of viral nucleic acid, whereas
CCMV purified from plants are infectious viral particles [55]. The amount of viral
particles purified from both cowpea plants and yeast varies depending on the
particular genetic variant, but is approximately 1-2 mg per gram of infected plant
tissue and 0.1 – 1 mg per gram of fresh yeast cell weight [55, 73].
CCMV as a Template for Nanotechnology
The natural role of the viral capsid is to house and protect the viral nucleic acid
and ‘deliver’ it to a host cell. Likewise, we would like to house ‘cargo’ molecules,
such as therapeutic and imaging agents, within the capsid and re-direct it to
deliver this cargo to other cells of interest. Toward this goal, CCMV capsids have
been employed as scaffolds for chemical conjugation of biologically important
molecules including imaging agents (fluorescein, gadolinium chelators) and as
size-constrained reaction vessels for mineralization of metallic and metal oxide
nanoparticles [22, 39, 41, 74].
The inherent properties of the CCMV protein cage are sufficient for many
types of chemical modification, but these properties can be expanded by genetic
11
modification. As discussed below and in Chapter 8, the endogenous properties
of CCMV can be dramatically altered by design and implemented via polymerase
chain reaction (PCR) mediated mutagenesis of the gene encoding the capsid
protein. This technique enables genetic alteration of the DNA encoding the viral
capsid protein and in turn the assembled cage architecture. Based on the atomic
level crystal structure of CCMV, it is possible to genetically introduce single
amino acid substitutions at precisely defined locations (i.e., interior, subunit
interface, exterior) (Chapter 8). In addition, genetic modifications allow for the
introduction of novel peptide sequences as extensions from the N- and C- termini
or within the ‘loop’ regions of the subunit structure (Figure 1.4) [41]. Importantly,
not all regions of the capsid tolerate change. There are limitations to the length
and site of incorporation of peptide sequences as well as the amount of
endogenous sequence that may be deleted, before the modifications prove too
drastic and the protein subunits are no longer able to assemble into capsid
architectures.
In initial work utilizing CCMV as a template for nanotechnology, wild type
capsids were employed for the mineralization of polyoxometalate species
(paratungstate and decavanadate) [22]. The virion’s interior cavity constrained
mineral growth, resulting in a spherical nanoparticle with a maximum diameter of
~24 nm [22]. This work took advantage of the distinct chemical environments on
the interior and exterior of the CCMV capsid. The interior surface is more
positively charged than the exterior, thus allowing it to serve as a nucleation site
12
for aggregation of anionic precursors and crystal growth. After nucleation the
size and shape of the mineral were defined by the interior of the virion. During
the polyoxometalate mineralization reaction the pH was lowered entrapping the
mineral within the viral capsid [22]. In addition to the endogenous properties of
size, shape, and delineation of charge on the interior / exterior surfaces, the
‘gating’ property of CCMV cage was utilized for the entrapment of an organic
polyanion, polyanetholesulphonic acid (PAS) [22].
Inspired by the iron storage protein cage, ferritin, and its ability to mineralize
iron oxide nanoparticles in a size constrained fashion, the SubE mutant of CCMV
was designed. SubE was made by replacing 8 of the positively charged amino
acid residues (5-Arg, 3-Lys) on the N-terminus of the protein with negatively
charged glutamic acids (Glu). This created a cage with a negatively charged
interior, illustrating the plasticity of the CCMV cage towards genetic manipulation
[55, 75]. The electrostatically altered viral protein cages did not differ in overall
structure from wild type (wt) CCMV, but they exhibited different mineralization
capabilities [55]. SubE’s negatively charged interior promoted interaction with
cationic Fe(II) ions and subsequent oxidative hydrolysis led to the formation of
~24 nm particles of lepidocrocite (γ-FeOOH) constrained within the interior cavity
of the protein cage [75].
Similar to the alteration of charge distribution in CCMV SubE described above,
genetic introduction of cysteines at the quasi-three fold axis resulted in capsids
with novel properties [21]. In this genetic variant, CCMV A14C R82C, the
13
structural transition of CCMV is under redox control. Specifically, the capsid is in
a compact, closed conformation under oxidizing conditions due to intersubunit
disulfide bond formation and undergoes a transition to an open conformation
under reducing conditions [21].
Wild type and genetic variants of CCMV have been chemically derivatized with
activated dye molecules, a small peptide, biotin, and digoxigenin [39, 76]. For
example, exteriorly exposed carboxylate groups (glutamic and aspartic acids) of
CCMV were reacted with 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide
hydrochloride (DEC) and N-hydroxysuccinimide (NHS) to generate a succinimidyl
ester which was subsequently reacted with an amine containing fluorophore, (5((5-aminopentyl)thioureidyl)eosin [39]. Derivatization of 560 of the 1980 available
carboxylate groups was achieved by utilizing 1000-fold molar excess dye per
CCMV subunit concentration [39]. Two variants of CCMV were engineered to
present thiol (cysteine) reactive groups on their exterior (CCMV R82C & CCMV
A141C) [39]. These variants were labeled with fluorescein-5-maleimide. A
maximum labeling of 100 thiol groups (out of 180 total) was achieved [39].
Genetic variants of CCMV with introduced cysteines also proved useful for
conjugation of a 24 amino acid peptide to the exterior surface [39]. In addition,
the surface exposed cysteines of another genetic variant, CCMV A163C, bound
to gold surfaces. This enabled the formation of an patterned 2-dimensional (2-D)
array of CCMV capsids on the gold surface [74]. This work involved pre-blocking
14
most of the exposed cysteines except those on one particular CCMV surface
which then bound the gold surface [74].
Differentially labeled CCMV particles were created utilizing the disassembly
and reassembly of the capsid in vitro [76]. CCMV capsids disassemble in the
presence of 400 mM CaCl2 at pH 7.6 and reassemble in 50 mM sodium citrate
and 1M NaCl at pH 5.25 [67, 76-78]. In order to produce dual labeled capsids
with varying label ratios, two populations of CCMV VLPs were chemically
derivatized with either biotin or digoxigenin, disassembled into subunits,
combined in desired ratios, and reassembled [76].
Genetic Incorporation of Peptides within the CCMV Capsid. Toward the
ultimate goal of vaccine generation, an effort to produce CCMV capsids that
displayed human immunodeficiency viral (HIV) antigens on their surface was
made in collaboration with Dr. M. Teintze’s lab [79]. Their results indicated that
CCMV was not a superior platform, as compared to keyhole limpet hemocyanin
(KLH), for antigen display. In addition, there was difficulty in the generation of
genetic variants of CCMV with peptide insertions within the ‘loop’ regions of the
capsid protein (Figure 1.4) [79]. Introduction of peptide sequences as N-terminal
fusions, as described in the next section, proved more successful.
One CCMV genetic variant, CPPep11, tolerated a peptide insertion within a
surface exposed loop region [55]. An 11 amino acid component (pep11) of
laminin was genetically incorporated into the CCMV coat protein, resulting in
CPPep11. Although low yields were obtained, a sufficient amount of CPPep11
15
VLPs were produced to test this genetic variant’s ability to bind laminin binding
protein (LBP), a cell surface protein involved in cell attachment to basement
membranes. The goal of this work was to use CPPep11 to block cancer
metastasis. Tumor cells require attachment to the basement membrane by
binding laminin, one protein component of the extracellular matrix, via cell
surface expressed LBP. Pep11 is the component of laminin that which binds
cellular LBP, therefore if it is presented on CCMV cages (CPPep11)
metastasizing tumor cells in circulation could bind the CCMV-Pep11 genetic
variant (CPPep11) instead of binding laminin within the basement membrane
thus inhibiting metastasis [80].
CCMV as a Platform for Magnetic
Resonance Contrast Agent Delivery.
CCMV’s Endogenous Metal Binding Site. The road to developing the CCMV
viral capsid as a magnetic resonance imaging (MRI) contrast agent has involved
utilization of the endogenous properties of the cage, as well as both genetic and
chemical modification. In its natural environment CCMV coordinates divalent
calcium (Ca2+) ions at its quasi-3-fold axis, but in vitro other metal ions including
paramagnetic Gd3+ can bind these sites [29, 69]. Each CCMV capsid can bind
up to 180 Gd3+ ions [29]. The T1 and T2 relaxivities of water protons, as
measured at 61 MHz Larmor frequency were, 202 and 376 mM-1s-1 respectively.
These are the highest values reported for a molecular paramagnetic material to
date [29]. The ‘large’ molecular size of CCMV and its slow tumbling rate in
16
solution, relative to small molecule contrast agents such as DOTA and DTPA
(see below), is in part responsible for the high relaxivity observed for CCMVGd3+. Although Gd3+-CCMV exhibited high relaxivity values, the dissociation
constant (Kd) for Gd3+ (31 μM) indicated that the Gd binding was too weak for in
vivo applications [29].
Genetic Incorporation of Metal Binding Peptide in CCMV. In order to increase
the binding affinity the metal binding motif from calmodulin (DKDGDGWLEFEE),
referred to as the EF hand motif, was genetically incorporated the into the CCMV
architecture as an N-terminal fusion [41, 81]. Although incorporation of this
peptide did increase Gd3+ binding affinity, without decreasing the relaxivity values
(R1 = 213, R2 = 407 mM-1 sec-1 at 61 MHz), the Kd was still too weak (100 nM)
for clinical applications [41].
Chemical Conjugation of Metal Chelators to CCMV. In order to achieve high
Gd3+ binding affinity, Gd3+ chelators (diethylenetriaminepentaacetic acid (DTPA),
tetraazacyclododecanetetraacetic acid (DOTA) were chemically linked to the
CCMV capsid and subsequently loaded with Gd3+ [41]. The resulting protein
cage architecture exhibits both high relaxivity values (R1 = 46, R2 = 142 mM-1
sec-1 at 61 MHz) and high metal binding affinity [41].
Additionally, the iron oxide
containing SubE CCMV (described above) has potential utility as an MRI contrast
agent as iron oxide nanoparticles have proven useful as dark field MRI agents
[30-32, 53].
17
Small Heat Shock Protein (Hsp)
from Methanococcus jannaschii
Proteins cage architectures, both viral and non-viral, originating from extreme
(high temperature, high pressure, acidic) environments, including the hot springs
within Yellowstone National Park and oceanic hydrothermal vents, are being
explored for applications in biomedicine and nano-materials synthesis.
Methanococcus jannaschii is a hyperthermophilic archaeon originally isolated in
1983 from a 2,600 meter deep hydrothermal vent in the Pacific Ocean [82]. It is
an organism that thrives at high temperature and pressure. The optimal growth
temperature for M. jannaschii is 85˚C and it is grown in culture under pressure
(30 psi) (Figure 1.5). The M. jannaschii genome consists of a 1,664,970 base
pair (bp) circular chromosome, a large (58,407 bp) extra-chromosomal element,
and a small (16,550 bp) extra-chromosomal element [83]. It was the first
archaeal genome sequenced (completed in 1996) [83].
Figure 1.5. Image of Methanococcus jannaschii
M. jannaschii grown at 78˚C and 30 psi; Image from [1].
18
Hsp Structure. The small heat shock protein (Hsp) from Methanococcus
jannaschii assembles into an empty 24 subunit cage with octahedral (4:3:2)
symmetry [4, 84, 85] (Figure 1.6). Each of the identical protein subunits is
composed of 147 amino acids (MW 16.5 kDa) [4]. The N-terminal 32 residues of
the heat shock protein subunit were not resolved crystallographically, but density
observed by cryo-electron microscopy within the cage interior suggest these
residues are internally exposed [4, 85, 86].
The structure of the Hsp protein cage architecture was resolved to 3.2 Å
resolution by X-ray crystallography [4, 84]. This near atomic level structural
information provides the foundation for precisely defining the location of reactive
groups on both the exterior and interior of the cage [84]. The Hsp cage
architecture has an exterior diameter of 12 nm and an interior diameter of 6.5
nm. Large, 3 nm diameter, triangular pores at the 3-fold axes (8 per cage) allow
free exchange between the interior of the cage and the bulk solution (Fig.1.6)
[84, 85]. In addition, there are smaller square windows each 1.7 nm in diameter
(6 per cage). Therefore the overall structure of Hsp resembles a multi-windowed
hollow sphere (Figure 1.6B). The Hsp cage assembly is stable up to ~70oC [87]
and in a pH range of 5-11 [38].
MjHsp16.5 (referred to in this work as simply Hsp) belongs to the small heat
shock protein family which encompasses proteins from all three domains of life
[88, 89]. The unifying characteristics of proteins within this family include a
19
A.
B.
B.
Figure 1.6. Structure of the Small Heat Shock Protein from Methanococcus jannaschii
A. Topology of the interaction between 2 subunits; MjHSP 16.5 dimer.
B. Space filling model of the 12 nm diameter assembled Hsp architecture; an interior
view along the four-fold axis. Images from [4].
conserved α-crystallin domain (~100 residues, in Hsp residues 46-135), a
subunit size ranging from 12 to 42 kDa, and increased expression in response to
elevated temperatures and/or other stress conditions [4, 85]. The function of the
conserved α-crystallin domain in the prototype lenticular α-crystallin protein is to
prevent protein precipitation and cataract formation within the vertebrate eye lens
[26, 90]. In addition, α-crystallins, which are also found in other organs (heart,
liver, brain, kidney) and have homologues in nearly every organism have been
shown to serve as molecular chaperones that aid in the proper folding of proteins
[26, 91]. The 20% identity between MjHsp16.5 and mammalian small heat
20
shock proteins is found within the α-crystallin domain [87]. MjHsp16.5 has been
shown to protect E. coli protein extracts and single chain monellin from complete
denaturation in vitro, but its in vivo function remains unknown [87, 92, 93].
Interestingly, the in vitro data suggest that MjHsp16.5 is a more efficient
chaperone under physiological conditions of elevated temperature [87], 80˚C
versus 37˚C, and pressure [94]. Fluorescence resonance energy transfer
(FRET) data from differentially labeled Hsp populations reveal that the multimeric
‘cage’ structures undergo subunit exchange at temperatures >65˚C (incubation
time > 20 min.) [87]. This property could prove useful as a strategy for obtaining
multi-labeled Hsp constructs, similar to the disassembly and reassembly strategy
used to obtained variably labeled CCMV architectures (described above) [76].
Production of Hsp Cages. Wild type and genetic variants of Hsp are
expressed in an E. coli expression system in which they self assemble into cagelike architectures (see Materials and Methods). The amount of Hsp protein cage
purified per gram of E. coli varies, but is approximately 5 mg per gram of fresh E.
coli cell weight [38].
Hsp as a Platform for Nanotechnology.
In addition to my work exploring the potential of Hsp cages to serve in targeted
delivery of therapeutics (described in the following chapters), parallel work in our
lab has focused on the use of this architecture for biomineralization, catalysis,
and magnetic resonance imaging [95-97].
21
Biomineralization of Hsp Cages. One genetic variant of Hsp (CP_Hsp) was
designed to serve as a size constrained vessel for the growth of metallic CoPt
nanoparticles (6.5 nm) [95]. The CP_Hsp genetic variant was engineered to
direct the formation of a particular CoPt crystal phase (L10) that exhibits
ferromagnetism. Ferromagnetism is defined as a property of materials (i.e., iron,
nickel or cobalt) that become magnetized in a magnetic field and retain their
magnetism when the field is removed at room temperature (300 K). The CoPt
binding peptide (KTHEIHSPLLHK), discovered utilizing a phage display library
screen (phage display described in Chapter 4) [98], was introduced as an Nterminal addition to HspG41C [95]. The resulting CP_Hsp protein cage with
internal CoPt binding peptides was used to produce CoPt nanoparticles that
exhibited room temperature ferromagnetism [95]. Magnetic alloys including the
L10 phase of CoPt have potential to serve as addressable bits in future magnetic
storage applications [95]. In addition, another genetic variant, HspG41C, served
as a size constrained reaction vessel for iron oxide mineralization. This work is
described further in Chapter 2.
Hsp Cage Architectures for Catalysis. Hsp cage architectures were utilized in
the synthesis and encapsulation of platinum (Pt) nanoparticles (1 or 2 nm
diameter) [97]. The Pt-Hsp protein cage composites served as catalysts for the
reduction of H+ to form H2, which is an important alternative fuel source [97]. In
addition, the ability of Hsp-Ru(bpy)3+2 composites to produce singlet oxygen (1O2)
has been demonstrated. Future studies will investigate the efficacy of this
22
construct to serve as a photodynamic therapy (PDT) agent. PDT is a treatment
that involves the cellular uptake of photoactivatable molecules, which can destroy
cells upon exposure to a specific light source. These results are an exciting
demonstration of the potential role of protein cage architectures for catalysis.
Hsp Cages as Magnetic Resonance Imaging Agents. Genetic variants of Hsp
with either N- or C- terminal peptide extensions were designed to incorporate the
metal binding motif from calmodulin (GDDGDGWLEFEE), similar to the genetic
variant of CCMV described previously [99]. Unfortunately, neither genetic variant
was very successful for a number of reasons. The metal binding peptide of the
N-terminal variant was found to proteolytically cleave from the subunit and the Cterminal variant was not soluble at high concentrations (>0.1 mg/ml) after the
addition of gadolinium (Gd3+) [41]. Therefore, Gd3+ chelators (DOTA & DTPA)
were chemically conjugated to lysines of Hsp cage architectures. However,
chemical conjugation resulted in less than 5 labels per cage, therefore relaxivity
data were not collected with these constructs [41].
Human H-chain Ferritin
Ferritin Structure. Ferritins are 12 nm diameter spherical protein cage
architectures present in nearly all organisms including bacteria, archaea, and
eukaryotes [100-102]. They are important proteins that serve to sequester and
store iron, an essential but toxic element [100]. The ferritin cage is composed of
24 subunits which assemble with octahedral (4:3:2) symmetry resulting into a
23
N terminal
8 nm
C terminal
12 nm
Figure 1.7. Structure of Human Ferritin
A. Ribbon diagram of a single ferritin H-chain subunit.
B. Space filling model of the 24 subunit containing ferritin architecture; exterior
diameter 12 nm, interior diameter 8 nm.
hollow shell with an interior diameter of 8 nm (Figure 1.7) [100]. It is within this 8
nm cavity that iron (Fe) is stored as a small nanoparticle of iron oxide mineral
known as ferrihydrite. The structure of ferritins from different organisms is highly
conserved, but the primary amino acid sequence of the subunits from which they
are assembled can share as little at 14% sequence identity [100]. Endogenous
human ferritin is composed of two different subunit types, the H-chain and the Lchain [100]. The H-chain contains the catalytic site, which catalyzes the
oxidation of two Fe2+ atoms, and the L-chain subunit plays a role in iron oxide
core formation [100, 101]. Interestingly, the percentage of H- and L-chains
composing ferritin cages varies depending on the tissue specific expression level
of each of these chains [100, 101]. This is a mechanism by which vertebrates
regulate the function of ferritins. For example, in organs that contain and require
a lot of iron, such as the liver, a greater percentage of the ferritin is composed of
L-chain subunits, which in turn increases the iron storage capacity of ferritin [25,
24
100]. The work described herein utilizes heterologously produced human ferritin
solely comprised of the H-subunit type.
Human H-chain Ferritin Production. Human H-chain ferritin (HFn) and genetic
variants of HFn are expressed in an E. coli heterologous expression system (see
Materials and Methods). The amount of HFn protein cage purified per gram of E.
coli varies depending on the particular genetic variant, but is approximately 5 mg
per gram of fresh E. coli cell weight [38].
Human H-chain Ferritin as a Platform for Nanotechnology.
The most prevalent use of ferritin and ferritin-like architectures as platforms for
nanotechnology is their use as constrained reaction vessels for the production of
nano-sized metal and metal oxide particles (reviewed in [103]) [48, 103-109].
Nanoparticle production employs empty ferritin cages (appoferritins) for
constrained synthesis of iron oxide and other metallic species and by controlling
the mineralization conditions different crystallographic forms are obtained. For
example, magnetite (Fe3O4) is formed at elevated temperatures (85˚C) and pH
(8.5) (Materials and Methods). The synthesis of metal containing particles within
these cages is not restricted to iron oxides. Cobalt, manganese, nickel,
chromium, and indium oxide nanoparticles have all been formed within apoferritin
cages (reviewed in [103]). The same approach has been used with a newly
reclassified DNA binding protein (Dps) from the bacteria Listeria innocua
(formerly known as a ferritin-like protein cage architecture), and within
25
recombinant horse L-ferritin [36, 48, 103, 108-110]. In addition, ferrihydrite
containing horse ferritin was used to photocatalytically synthesize copper colloids
[107]. In the work presented in this dissertation, the human H-chain ferritin and
genetic variants were utilized as size constrained reaction vessels for magnetite
synthesis and as templates for conjugation of organic molecules (see Chapter 7).
Protein Cages for Targeted Therapeutic and Imaging Agent Delivery
One of the goals of this research is to develop protein cage architectures that
serve as cell-specific therapeutic and imaging agent delivery systems while
simultaneously gaining understanding of protein cage assembly and dynamics.
Targeted therapeutic delivery systems can enhance the effective dose at the site,
such as a tumor, while decreasing general exposure to the drug and its
associated side effects [19]. Protein cage architectures have three surfaces
(interior, subunit interface, and exterior) amenable to both genetic and chemical
modification. Figure 1.2 depicts a schematic representation of how each surface
can play a distinct role in the development of new targeted therapeutic and
imaging agent delivery systems. The cage interior can house therapeutics, the
subunit interface can incorporate gadolinium ions (an MRI contrast agent), and
the exterior can present cell-specific targeting ligands (such as peptides and
antibodies) (Figure 1.2).
26
Attributes of Protein Cage Architectures Beneficial for Nanotech Applications
The inherent attributes of protein cage architectures make them attractive
platforms for the design of targeted therapeutic and imaging delivery systems.
Protein cages have a variety of endogenous properties including: size, structure,
solvent accessibility, cellular tropisms, amino acid composition, and release
mechanisms that might prove useful for biomedical applications. Structural
integrity and particle uniformity are additional inherent properties of protein cage
assemblies. These features enable the size-constrained synthesis of nanoparticles whereas monodispersity is sometimes difficult to achieve solely by
chemical synthesis. High resolution structural information [4, 60, 111] allow the
incorporation of molecules, peptides, or particles at precise locations [38, 39, 4446]. Protein cages have been chemically derivatized and utilized as spatially
constrained reaction vessels for materials, thus illustrating that they can serve as
chemical building blocks [22, 27, 38, 39, 44-46, 104, 112-114]. The ability to
genetically modify protein cages has allowed incorporation of novel functions,
including cell-specific targeting.
Their size falls into the nano-meter range shown to localize in tumors due to
the enhanced permeability and retention effect (EPR; described below) [19, 115,
116]. Their multivalent nature enables the incorporation of multiple functionalities
(including targeting peptides and imaging agents) on a single protein cage. Both
size and multivalency enable a relatively large amount of therapeutic to be
encapsidated or covalently linked to the cages which in turn can result in a high
27
therapeutic payload per delivery event. The bioavailability of therapeutics may
also be sequestered within the cage interior until release is triggered at a desired
time or by a given condition.
Heterologous expression of H-chain ferritin and Hsp cages in E. coli and
CCMV capsids in Pichia pastoris facilitates the production of these cages in
sufficient quantities [27, 38, 55]. This dissertation focuses on the use of three
protein cage systems; the 12 nm diameter small heat shock protein (MjHsp), the
12 nm diameter human H-chain ferritin (HFn) and the 28 nm diameter Cowpea
chlorotic mottle viral capsid (CCMV).
Additional Protein Cage Architectures Being Explored for Nanotechnology
In addition to the CCM, Hsp, and HFn protein cage platforms described above,
there are many other viruses and virus-like particles (VLPs) being developed as
platforms for applications in nanotechnology. The most similar to CCMV is
another plant virus, the Cowpea mosaic virus (CPMV) (described in detail below).
Other virus and VLP nanotech platforms include the tobacco mosaic virus [113,
117, 118], the flock house virus [119] , polyoma virus [120, 121], the MS2
bacteriophage [40], canine parvo virus [122], MS2 bacteriophage [123] and
adenovirus [124-128]. Non-viral protein cage architectures being explored as
platforms for nanotechnology include a ring shaped heat shock protein [129, 130]
and ferritin and ferritin-like architectures (described previously).
28
The use of viruses themselves for cancer treatment is an exciting area of
research [131, 132]. The fields of viral mediated gene therapy [133-137], viral
immunotherapy [138] and oncolytic viral therapy (which utilizes replication
competent viruses to kill cancer cells) [139-150], could all be considered under
the umbrella of biomedical nanotechnology. However, these fields are beyond
the scope of this dissertation, which focuses on the use of ‘empty’ viral capsids
and other protein cage architectures for biomedical applications [139, 151-153].
One of the most similar platforms to the CCMV platform (described above and in
Chapter 8) is the CPMV platform which has been developed as a vaccine
construct; this work is described below.
Cowpea mosaic virus (CPMV).
CPMV has been used as a platform for vaccine generation, imaging agent
delivery, and is currently being explored as an anti-cancer therapeutic delivery
system. CPMV is a 31 nm icosahedral virus in the Comoviridae viral family
within the same superfamily as picornaviruses, including polio virus [59]. These
viruses share structural similarities [59]. The CPMV capsid is composed of 60
copies each of a large (L), 41 kDa, and a small (S), 24 kDa, subunit protein [59].
The L subunit comprises two domains and the S subunit one domain, of the
asymmetric unit. Sixty asymmetric units self assemble into an icosahedral
capsid with pseudo T=3 symmetry [59, 60, 154]. Although, CCMV and CPMV
are both small (28 nm and 31nm respectively) icosahedral cowpea plant infecting
29
viruses, their primary structures are unrelated. Like CCMV, CPMV is amenable
to genetic and chemical modification and has been utilized as a chemical building
block for vaccine and imaging agent generation [33-35, 44-46, 155-158].
CPMV as a Vaccine Vector. One of the first potential biomedical applications
for which the use of CPMV was investigated was vaccine generation [159].
Based on crystallographic information, an exteriorly exposed ‘loop’ region of the
small capsid protein was genetically engineered to contain additional amino acid
(a.a.) sequences ranging in length from 15-34 a.a. [160]. These short peptide
sequences mimic antigenic epitopes from a variety of pathogens including
viruses and bacteria [154, 155, 160-163]. Specifically, CPMV chimeras were
engineered to express epitopes from human rhinovirus (HRV-14) [154], human
immunodeficiency virus (HIV-1 ) [154], mink enteritis virus (MEV; a parvovirus)
[160], canine parvovirus (CPV)[163], hepatitis B [164], merozoite surface antigen1 of the malaria parasite Plasmodium falciparum [158], outer membrane protein
F of Pseudomonas aeruginosa [155], and the fibronectin-binding protein of S.
aureus [162]. The list presented above is not inclusive and over fifty CPMV
chimeras have been generated (see [154, 165, 166]). While it is important to
note that the structural integrity of these CPMV chimeras has been demonstrated
it has also been reported that some CPMV chimeras do not properly assemble
and in turn lose their ability to infect and be propagated in cowpea plants [160,
167]. It was determined that the site of peptide insertion is important; most
inserts have been introduced within the βB-βC loop of the S protein [168]. In
30
addition, the tolerance of the CPMV capsid to the insertion of various peptide
sequences was found to be dependent on the size and isoelectric point of the
peptide insert [167].
Investigation of the potential efficacy of the above CPMV vaccine constructs in
animals typically involves their injection in combination with adjuvants (adjuvants
are substances such as alum and saponin that nonspecifically enhance the
immune response to an antigen [169]) [160]. These studies have demonstrated
that antigen specific antibodies are produced in experimental animals and some
are neutralizing (meaning that they bind the virus and prevent it from infecting a
cell).
Excitingly, some CPMV chimeras, such as the canine parvovirus chimera
(CPMV-PARVO1), generate a protective response [163]. Langeveld et al.
demonstrated that administration of CPMV-PARVO1 in the context of adjuvant
protected dogs from a lethal challenge with canine parvovirus [163]. Another
CPMV chimera displaying an epitope from the outer membrane protein F of
Pseudomonas aeruginosa also proved efficacious in the protection mice from P.
aeruginosa bacterial infection [155]. These data are very exciting, and they
illustrate the potential for plant derived vaccines, but it is also important to note
that in this study similar levels of anti-VP2 specific antibody were generated in
dogs that were given another type of antigen-protein chimera (VP2-keyhole
limpet hemocyanin, KLH). This is noteworthy because this was also shown with
the CCMV- antigen chimeras described above. The CCMV constructs were
31
made in yeast and some chimeras were only produced in low yields. Therefore
production of sufficient quantities was difficult. In contrast, these CPMV vaccine
constructs were produced in plants from which they could be produced easily
and cheaply, and maybe further developed as “edible vaccines” [170-172].
Chemical Derivatization of CPMV. Wild type and genetic variants of CPMV
with exteriorly exposed lysines (amines), cysteines (thiols), aspartic and glutamic
acids (carboxylates) have been chemically derivatized with fluorescent dyes,
quantum dots, oligonucleotides, polyethylene glycol (PEG), nanometer sized gold
particles, and redox active moieties [35, 42, 44, 45, 47, 173-175]. These types of
chemical derivatizations are described further herein in the context of CCMV,
Hsp and HFn cages. The in vivo data obtained from PEGylated and fluorescently
labeled CPMV will be described in detail in Chapter 6. In addition, CPMV was
used as a platform for the development of “click chemistry”, which utilizes nonbiological molecules including azides and alkynes. Click chemistry has
expanded the types of chemical reactions available for conjugation of proteins
(peptides, antibodies) and synthetic molecules to protein cage architectures [112,
176-179].
Other Nano-scale Therapeutic Delivery Systems
Drug carriers are often used to alleviate non-ideal properties of therapeutics
(poor solubility, non-specific tissue damage, rapid breakdown, too rapid
clearance, poor distribution) by positively altering circulation time, solubility, drug
32
release, and via site-specific targeting [19, 180]. In addition to the properties
listed above, nanoparticles may preferentially accumulate in tumor tissues due
to the enhanced permeability and retention (EPR) effect [19, 181]. It has been
shown that in certain disease states, including tumor growth, the permeability of
the vasculature increases, and particulate carriers are able to pass through and
accumulate [19]. This phenomenon is especially relevant for drug delivery
systems which target tumors. Tumor growth requires formation of new blood
vessels (angiogenesis). These newly formed vessels have large spaces (600 –
800 nm) between endothelial cells in contrast to the tight packing of endothelial
cells that make up normal mature blood vessels [19]. Passive targeting of drug
delivery systems can result in a ten-fold increased accumulation as compared to
free drug. This is, in part, due to poor drainage of tumor lymphatic systems [19].
Many nanoparticle (< 200 nm) drug delivery systems are composed of lipids
(liposomes, micelles, lipid emulsions, lipid-drug complexes) or polymers
(polymer-drug conjugates, polymer microspheres). Additionally, in some cases
conjugation of therapeutic molecules to either polyethylene glycol (PEG) or
antibodies has proven advantageous [19]. PEG is a hydrophilic flexible polymer
that usually increases the circulation time of conjugated therapeutics and
nanoparticle platforms [132]. Other nano-scale therapeutic delivery systems also
being explored include: silica nanoparticles, polysaccharide colloids, PEGylated
liposomes, polyamidoamine (PAMAM) dendrimer clusters, hydrogel dextran
nanoparticles, biodegradable poly(DL-lacticide-co-glycolide) or PLG
33
microspheres, latex and polystyrene beads, polyalkyl-cyanoacrylate
nanocapsules, and nanoparticles made up of fatty acid polymers and titanium
dioxide particles [19, 24, 50, 170, 180, 182-207]. Additionally, antibody mediated
therapeutic delivery has proven successful in both lab and clinical settings and
antibodies themselves may serve as therapeutic agents [5, 208-223]. PAMAM
dendrimers are hyper-branched synthetic macromolecules presenting amine
groups, and are currently being investigated for use as target-specific MRI
contrast agents [224-230]. Recently, some new drug delivery systems have been
approved for clinical use (reviewed in [19]). One of the most relevant examples
to the work described herein is Doxil/Caelyx, which is a doxorubicin (an anticancer agent described further in Chapter 3) containing PEGylated liposome
[231]. Another example is N-2-hydroxypropyl methacrylamide copolymer-linked
doxorubicin which can be administered in higher doses than free doxorubicin,
because the bioavailability of the drug is restricted when linked to the polymer
and the release rate is slow [232]. The polymer doxorubicin linkage is a peptide
sequence that is cleaved by lysosomal enzymes after cellular uptake via
pinocytosis [232]. In another example, PAMAM dendrimers were conjugated
with methotrexate (MTX), an anti-cancer drug, and folic acid, which served as a
cell targeting agent [180]. In vitro cytotoxicity data demonstrated that
dendrimers, each loaded with 5 MTX molecules, were efficacious in cell lines
known to be MTX-resistant. This result demonstrates that nanoparticle delivery
of currently used anti-cancer agents can result in beneficial properties, including
34
the evasion of therapeutic resistance [180]. Additionally, PEGylated liposomes
containing doxorubicin and cell-targeting folic acid were 45 times more
efficacious than non-targeted liposomes [233-235]. Some general properties of
drug delivery systems in development are described below.
Properties to consider when comparing the advantages and disadvantages
of drug delivery systems include the ability to synthesize them in sufficient
quantities, their therapeutic carrying capacity (or payload per delivery event),
their biodistribution and biocompatibility, and the ability to incorporate cellspecific targeting capability. For example, antibodies are excellent cell specific
targeting agents whose biocompatibility has been significantly improved by the
use of humanized antibodies and antibody fragments (mini-bodies, di-bodies,
etc.), but their therapeutic carrying capacity is limited to a few drug molecules per
antibody [19, 214-217, 236]. Liposomes have a high therapeutic molecule
carrying capacity, on the order of tens of thousands, but the types of compounds
that can be entrapped are somewhat limited [19]. Hydrophilic molecules are
readily entrapped within liposomes and a pH or chemical gradient can be used to
load hydrophobic weak bases (such a doxorubicin), but neutral hydrophobic
molecules or those with intermediate solubility tend to be rapidly released in the
presence of cell membranes [3]. One advantage of PEG coupled therapeutics is
that the clearance rate via the kidneys is usually decreased [3]. Polymeric
micelles prove useful for delivering water insoluble therapeutics due to their
hydrophobic cores [225]. Polymer based delivery systems can be conjugated to
35
a wide array of molecules and this linkage restricts the bioavailability of the drug
until it is dissociated from the carrier, which is advantageous for most drug
delivery scenarios [3]. One advantage of dendrimers is that they provide surface
exposed functional groups whereas the functional groups of glycopolymers are
folded inside [229]. On the other hand, as the size of PAMAM dendrimers is
increased, so does their in vitro cytotoxicity, but these ill effects can be alleviated
by PEGylation which has been shown to reduce immunogenicity and increase
circulation time [224-230]. Dendrimers have been used to encapsulate anticancer agents (doxorubicin and methotrexate), but the release rates are slow
with only 15 % of the drug released after 24 hours.
Several different strategies for cell-specific targeting of drug delivery systems
are being explored including, (1) utilization of the natural targeting ligands of viral
systems (e.g., the canine parvo virus (CPV) naturally binds the transferrin
receptor upregulated on cancer cells [122]), (2) antibody mediated targeting, (3)
peptide mediated targeting, involving ligands that bind cellular receptors and are
internalized by receptor mediated endocytosis (described further in Chapter 4),
(4) vitamin mediated targeting (e.g., B12 / folate), (5) small synthetic molecules
for targeting [237], and (6) passive targeting (based on size and charges of
particulate carrier) (e.g., the EPR effect described above).
This introduction has narrowly focused on the delivery of therapeutic
molecules, but nanometer scale platforms including liposomes and viruses are
also being utilized in the field of gene therapy. The lessons learned on that front
36
are applicable to the drug delivery field and vice versa. Efforts toward cellspecific delivery of nucleic acid have had a recent boom due to the discovery of
small interfering RNAs, which decrease expression of their corresponding cellular
target mRNAs (RNA interference, RNAi). RNAi mediated therapy has potential in
the treatment of cancer and other diseases.
In summary, there are many drug delivery systems, each having its own
inherent advantages and disadvantages, many of which can be altered by
chemical engineering. The area of targeted therapeutic and imaging agent
delivery is a rapidly growing field. Cancer treatment is a very broad topic
encompassing numerous types of cancer, many of which require specialized
drug delivery strategies. Therefore, although a lot of research is directed toward
the development of new drug delivery systems, there is room for new systems
which might prove advantageous for particular applications.
The long term goal of my dissertation work was to further our understanding of
protein cage architectures by exploring their potential utility for biomedical
applications. This approach has combined biology (protein cage architectures,
peptides, antibodies, cells) with synthetic and materials chemistry. As depicted in
Figure 1.2, protein cage architectures (CCMV, Hsp, HFn) can be chemically and
genetically derivatized in order to encapsulate therapeutic and / or imaging
agents on their interiors, and present mammalian cell-specific targeting
functionalities on their exteriors. These modifications result in a functionalized
protein cage architecture that has potential to specifically deliver therapeutic
37
agents to diseased tissue (such as tumors) while simultaneously enabling
diagnosis and/or detection by techniques such a MRI.
38
CHAPTER 2
THE SMALL HEAT SHOCK PROTEIN CAGE FROM Methanococcus jannaschii
IS A VERSATILE NANOSCALE PLATFORM FOR
GENETIC AND CHEMICAL MODIFICATION
Reproduced in part with permission from Nano Letters 3:1573-1576.
Copyright 2003 American Chemical Society.
Abstract
Nature has provided us with a range of reactive nanoscale platforms, in the
form of protein cage architectures such as viral capsids and the cages of ferritinlike proteins. Protein cage architectures have clearly demarcated exterior,
interior, and interface surfaces consisting of precisely located chemical
functionalities. In this chapter, we demonstrate that the small heat shock protein
(MjHsp) cage from Methanococcus jannaschii is a new and versatile nanoscale
platform with exterior and interior surfaces amenable to both genetic and
chemical modification. Wild type and genetic mutants of the Hsp cage are shown
to react with activated fluorescein molecules in a site specific manner. In
addition, the 12 nm Hsp cage served as a size constrained reaction vessel for
the oxidative mineralization of iron, resulting in the formation of monodispersed
9 nm iron oxide nanoparticles. These results demonstrate the utility of the Hsp
cage to serve as a nanoscale platform for the synthesis of both soft (organic) and
hard (inorganic) materials.
39
Introduction
A major goal of nanomaterials is the design and synthesis of multifunctional
materials with precise molecular level control. One promising approach is the
biomimetic synthesis of nanomaterials using organic assemblies as templates for
directed fabrication [98, 104, 238]. Organic assemblies such as viral based
protein cages and ferritin-like protein cages have been utilized as templates for
constrained nanomaterials synthesis [22, 27, 39, 44-46, 75, 113, 130]. Protein
cage architectures provide exterior, interior, and subunit interface surfaces on a
nano-scale platform that can be genetically and chemically modified to impart
function by design [22, 27, 39, 44-46, 75, 113, 130]. In this chapter, the small
heat shock protein (Hsp) from the hyperthermophilic archaeon, Methanococcus
jannaschii, is shown to be a highly versatile cage-like structure whose exterior
and interior surfaces are amenable to both genetic and chemical modification.
Organic ligands were attached to the Hsp cages in a site specific manner to the
exterior and interior surfaces. In addition, the Hsp cage was employed as a size
constrained reaction vessel for the synthesis of inorganic nanomaterials. The
demonstration of molecular level control in the Hsp cage system illustrates its
versatility as a multifunctional nanoscale platform. The ability to perform both
inorganic and organic functionalization of this template indicates its potential
applicability in nano-materials [239] and biomedical [31] sciences.
40
Results and Discussion
Small Heat Shock Protein (Hsp) from Methanococcus jannaschii;
Wild Type and Genetic Variants (HspG41C and HspS121C)
As described in Chapter 1, the small heat shock protein (Hsp) assembles into
an empty 24 subunit cage with octahedral symmetry [4, 84, 85]. The assembled
protein cage has an exterior diameter of 12 nm. Large, 3 nm diameter, pores at
the 3-fold axes allow free exchange between the interior and bulk solution
(Figure 2.1A) [4, 84, 85]. Each 16.5 kDa Hsp monomer is composed of 147
amino acids, but contains no endogenous cysteine residues.
The 3.2 Å resolution X-ray crystal structure provides the foundation for
precisely defining the location of reactive groups on both the exterior and interior
of the cage [4, 84, 85]. The endogenous cage provides spatially defined reactive
groups and through genetic engineering we are able to introduce additional
reactive functional groups in a site specific manner. Naturally occurring lysines,
bearing primary amino groups, are found both on the exterior surface (amino acid
positions 55, 65, 82, 110, 116, 123, 141, 142) and the interior surface (position
40) of the assembled MjHsp cage (Figure 2.1). In order to engineer spatial
selectivity for chemical modification, unique thiol functional groups were
introduced into via polymerase chain reaction (PCR) mediated site directed
mutagenesis (Materials and Methods). Glycine at position 41 on the interior
surface was replaced with cysteine to generate the HspG41C mutant (Figure
2.1C). Similarly, serine at position 121 on the exterior surface was replaced by
41
cysteine to generate HspS121C (Figure 2.1D). In brief, the gene encoding the
small heat shock protein MjHsp16.5 was amplified by PCR, cloned into the pET30a(+) plasmid (Novagen), and expressed in E. coli. The nucleic acid sequences
encoding wild type, G41C, and S121C Hsp clones were confirmed by DNA
sequencing (Applied BioSystems). The Hsp proteins were purified from E. coli
where they self assembled into the 24 subunit cages. These cages were purified
to homogeneity as described in the Materials and Methods (Appendix A).
Recombinant protein cages were characterized by size exclusion
chromatography (Superose 6, Amersham Pharmacia), dynamic light scattering
(Brookhaven 90Plus), transmission electron microscopy (Leo 912 AB) (Figure
2.2), mass spectroscopy (Esquire3000, Bruker), and SDS polyacrylamide gel
electrophoresis (SDS-PAGE).
A.
B.
C.
D.
Figure 2.1. Space Filling Models of the 24 Subunit MjHsp Cages
A. The exterior of the Hsp cage viewed along the 3-fold axis; lysines are colored blue.
B. The interior of the cage viewed along the 4-fold axis; lysines are colored purple.
C. The interior of the Hsp cage viewed along the 4- fold axis; the interior exposed amino acids,
at subunit position 41, are colored red. In wtHsp position 41 is glycine whereas in the
HspG41C genetic mutant position 41 is a cysteine.
D. The exterior of the cage viewed along the 3-fold axis; the externally exposed amino acid at
subunit position 121 is colored green. In wtHsp position 121 is serine whereas in the
HspS121C genetic mutant it is a cysteine.
42
A.
A b s o r b a n c e (a .u .)
12x10
B.
-3
A280
A500
10
8
6
4
2
0
50 nm
10
15
20
25
30
Time (minutes)
35
40
45
100
C.
Intensity
80
60
40
20
0
0
10
20
30
40
50
60
Diameter (nm)
70
80
90
100
Figure 2.2. MjHsp Cage Characterization
A. Size exclusion chromatography absorbance profile. Hsp elutes after 30 minutes (280 nm).
B. Transmission electron microscopy of the 12 nm Hsp cages; 50 nm scale bar.
C. Dynamic light scattering shows a mean diameter of 12.6 nm. The insert shows the
Correlation function of the raw data collected (red) and the fit of these data using a
Non-Negatively Constrained Least Squares (NNLS) analysis (blue).
.
Reactivity of Engineered Thiols and Endogenous Amines of Hsp
To assess the reactivity of the introduced thiol and endogenous amine groups,
the purified Hsp protein cages were reacted with activated fluorescein.
Specifically, fluorescein-5-maleimide (Fl-Mal) was used for reactions with thiols
and 5-(and-6) carboxy-fluorescein-succinimidyl ester (5(6)-FAM) for reactions
with amines [240]. For exposed thiol reactions, purified HspG41C cages (0.9
43
mg/ml, 54 μM) in HEPES buffer (100 mM, pH 6.5) were treated with tris(2carboxyethyl)phosphine (TCEP) for one hour at room temperature to ensure
reduction of the thiol groups prior to labeling. Later, this step was shown to be
unnecessary for routine reactions. Reactions to covalently bind 5(6)-FAM to the
endogenous amine groups, on both the exterior and interior surfaces, were
performed in 100 mM HEPES, pH 8.0. A range of fluorescein concentrations,
from 0.06 to 28 molar equivalents per subunit, were reacted with the protein
cages. The reactions proceeded at room temperature for 2 hours with stirring,
followed by incubation at 4°C overnight. After the reactions, modified Hsp was
separated from unreacted fluorescein by size exclusion chromatography. Since
fluorescein emission is pH sensitive [240], 100 μL of Tris buffer (0.5 M, pH 8.5)
was combined with column eluant (50 µL) prior to fluorescence analysis. The
results showed selective labeling in HspG41C and HspS121C reactions with
fluorescein-5-maleimide as compared to the wild type Hsp control (Figure 2.3).
HspG41C and HspS121C pre-reacted with iodoacetamide (HspG41C /
HspS121C + IAA), in order to passivate reactive thiols, served as additional
control reactions. As expected, subsequent exposure to activated fluorescein did
not result in covalent attachment of fluorescein, therefore no fluorescence signal
was observed in these reactions (Figures 2.3 & 2.4). The number of fluorophore
molecules covalently linked per Hsp cage was determined using absorbance
spectroscopy and ranged from 1 to 24 per protein cage depending on reaction
conditions (Table 2.1).
44
Hsp wt : 0.6x FlMal
HspG41C + IAA : 0.6x FlMal
HspG41C : 0.06x FlMal
HspG41C : 0.6x FlMal
HspG41C : 3x FlMal
HspG41C : 6x FlMal
HspG41C : 28x FlMal
Negative Control
3
Fluorescence (a.u.)
40x10
30
20
10
0
520
540
560
580
600
Wavelength (nm)
620
Figure 2.3 Fluorescence emmision of the Hsp cages (either wild type (wt) or
HspG41C labeled with fluorescein-5-maleimide (FlMal).
The fluorophore quantities listed are in molar equivalents of fluorophore per subunit.
Protein free controls were performed at all fluorophore concentrations, one example is
shown above (negative control = 3x fluorophore only). In each case, no free
fluorophore was detected by fluorescence.
B.
A.
6000
Hsp wt : 1x FlMal
HspS121C + IAA : 1x FlMal
HspS121C : 6x FlMal
HspS121C : 5x FlMal
HspS121C : 3x FlMal
HspS121C : 1x FlMal
HspS121C : 0.6x FlMal
HspS121C : 0.1x FlMal
HspS121C : 0.06x
Negative Control
F lu o resce nc e (a .u .)
5000
4000
3000
2000
1000
A bsorbance (a.u.)
2.0x10
-3
A280
A495
1.5
1.0
0.5
0.0
0
520
540
560
580
Wavelength (nm)
600
620
10
15
20
25
30
Time (minutes)
35
40
Figure 2.4. Characterization of Fluorescently Labeled HspS121C
A. Fluorescence emmision of the HspS121C cages (either wild type (wt) or S121C)
labeled with fluorescein-5-maleimide (FlMal). The fluorophore quantities listed are in
molar equivalents of fluorophore per subunit. Protein free controls were performed at
all fluorophore concentrations, one example is shown above (negative control = 10x
fluorophore only). In each case, no free fluorophore was detected by fluorescence.
B. Size exclusion chromatography elution profile of HspS121C reacted with a 28 fold
(28x) molar excess of fluorescein-5-maleimide per subunit. The profile illustrates the
co-elution of fluorescein (495 nm) and Hsp protein cage (280 nm).
45
45
Size exclusion chromatography, SDS PAGE, dynamic light scattering, and
transmission electron microscopy were used to characterize the Hsp protein
cages after fluorescein labeling. Size exclusion chromatography showed coelution of the protein cage (280 nm) and fluorescein (495 nm) demonstrating the
covalent attachment of fluorescein to the protein cage (Figures 2.4 & 2.5). The
elution profile of fluorescein labeled Hsp cages was identical to the unreacted
protein cages. These results were confirmed by SDS-PAGE, in which the band
corresponding to the fluorescein labeled HspG41C co-migrated with the 16.5 kDa
Hsp protein subunit (Figure 2.6). In addition, dynamic light scattering and
transmission electron microscopy both revealed 12 nm diameter protein cage
particles before and after the fluorescein labeling reactions (Figures 2.5 & 2.8).
-3
-3
3.0x10
A280
2.5
A500
6
2.0
1.5
4
1.0
2
0.5
0.0
0
10
15
20
25
30
35
40
B.
A bsorbance 280nm (a.u.)
Absorbance 500nm (a.u.)
A. 8x10
45
Time (minutes)
Figure 2.5. Characterization of Fluorescein Labeled HspG41C
A. Size exclusion chromatography elution profile of HspG41C reacted with a 10 fold (10x)
molar excess of fluorescein-5-maleimide per subunit. The profile illustrates the co-elution
of fluorescein (500 nm) and Hsp protein cage (280 nm).
B. TEM of fluorescein labeled HspG41C cages stained with uranyl acetate.
46
A.
1
2
3
4
5
6
7
B.
20
14.2
Figure 2.6. SDS-Polyacrylamide Gel Electrophoresis of Hsp
The same gel is visualized first by (A.) fluorescence (inverted image) (Molecular Imager
FX, BioRad) and subsequently by (B.) Coomassie Brilliant Blue staining. The lanes
corresponding to the following samples are numbered below the gels: Lane 1 =
Molecular weight ladder (kDa), Lane 2 = wtHsp only, Lanes 3&4 = wtHsp Fl-Mal
reaction, Lane 5 = HspG41C only, Lanes 6&7 = HspG41C labeled with Fl-Mal. The
gels show the covalent association of fluorescein-5-maleimide (Fl-Mal) with HspG41C;
there is little association of fluorophore with the wtHsp cage. The lower band of the
doublet (2 bands very close to each other) visualized by fluorescence imaging
(A.) was confirmed to be an Hsp degradation product by mass spectroscopy.
Hsp Cage
Fluorophore
Conc.**
# Fl-Mal/cage
# FAM/cage
WtHsp
0.6x
0.3
1
HspG41C + IAA;
HspS121C +IAA
0.6x
0.6;0
N/A
HspG41C; S12C
0.06x
1;0
0
HspG41C; S121C
0.6x
3;0
1
HspG41C; S121C
3x
7;2
4
HspG41C; S121C
6x
24;5
8
HspG41C; S121C
28x
11;5
17
Table 2.1 Extent of Hsp Labeling with Fluorophores (Fl-Mal = fluorescein-5-maleimide,
FAM = 5-(and-6) carboxy-fluorescein-succinimidyl ester)
The extent of labeling is shown by the number of fluorophore molecules covalently linked
per cage. **Fluorophore concentration is reported in molar equivalents per subunit in the
labeling reaction. The Fl-Mal data indicate 100% labeling of the HspG41C reactive groups
with 6x molar equivalents of fluorophore; compared to only 21% labeling of HspS121C. The
decreased labeling of HspS121C might be due to reduced accessibility. Importantly, wtHsp
which lacks cysteines does not show labeling with Fl-Mal. Control reactions, using
HspG41C / HspS121C treated with iodoacetamide (IAA) to passivate reactive thiols prior to
addition of Fl-Mal, showed low reactivity.
47
Inorganic Chemical Modification of Hsp
The Hsp cage was utilized as a template for inorganic chemical modification of
its interior. The Hsp cage served as a size constrained reaction vessel for iron
oxide mineralization similar to the iron-storage protein ferritin (described in
Chapter 1) which catalyses the oxidation of Fe(II) to Fe(III) resulting in the
spatially constrained mineralization of the iron oxyhydroxide mineral ferrihydrite
(Fe(O)OH) [241]. We followed Fe(II) oxidation activity in the presence of Hsp
cages by monitoring the increase in the ligand to metal charge transfer (LMCT)
absorbance in the visible spectrum at 400 nm. Addition of Fe(II) to G41C and
subsequent air oxidation resulted in the formation of a homogeneous rust colored
solution. When imaged by TEM, small iron oxide particles were seen with an
average diameter of 9 + 1.2 nm (Figure 2.7A). Both the wt Hsp and HspG41C
cages showed similar mineralization capabilities. The presence of iron in the
particle was confirmed by electron energy loss spectroscopy (EELS). In contrast,
Fe(II) oxidation in the absence of Hsp protein cages resulted in rapid bulk
precipitation of a ferric oxide (Fe2O3)due to unconstrained particle growth (Figure
2.7B). Electron diffraction of the bulk precipitate showed d-spacings consistent
with the major reflections for lepidocrocite [242] (3.28 Å (120) 2.46 Å (031), 1.93
Å (200), 1.72 Å (151), 1.51 Å (231), 1.39 Å (251)). The electron diffraction from
the mineralized Hsp however indicated the presence of a poorly crystalline ferric
oxide phase having d-spacings consistent with ferrihydrite [243] (2.50 Å (200),
1.93 Å (113), and 1.55 Å (115)). While the high order reflections for ferrihydrite
48
and lepidocrocite are very similar, the absence of large d-spacing reflections in
the mineralized Hsp diffraction suggests that the nanomineral is ferrihydrite
rather than lepidocrocite. Clearly, the Hsp protein shell acts as a constrained
A.
B.
500 nm
Figure 2.7. Transmission Electron Microscope (TEM) Images
A. TEM of iron oxide cores inside Hsp cages; 50 nm scale bar.
B. TEM of control sample illustrating the bulk precipitation of iron oxide in the
absence of Hsp protein cages; 500 nm scale bar. The insets show the
Electron Diffraction from each sample.
reaction vessel for spatially selective mineralization of Fe(O)OH. From the
crystal structure of the assembled Hsp, it can be seen that the interior surface of
the cage has a slight excess of acidic residues which we have previous shown to
be sufficient to direct oxidative hydrolysis and mineralization of iron oxides [27,
75]. As we have shown with other protein cages, this spatially confined
mineralization suggests that Hsp could serve as a template for the synthesis of
other mineral phases with magnetic properties important for biomedical [31] and
nano-electronic purposes.
49
80
80
Intensity
C. 100
Intensity
A. 100
60
40
40
20
20
0
0
0
10
20
30
40
50
60
Diameter (nm)
70
80
90
0
100
10
20
30
40
50
60
D iam eter (nm )
70
80
90
100
D. 100
B. 100
HspG41C FlMal Mean Dia. 12.6nm
80
Intensity
80
Intensity
60
60
40
60
40
20
20
0
0
0
10
20
30
40
50
60
Diameter (nm)
70
80
90
100
0
10
20
30
40
50
60
D iam eter (nm )
70
80
90
100
Figure 2.8. Dynamic Light Scattering (DLS) Data
A. Hsp G41C no experiment, Mean Diameter = 12.2 nm
B. Hsp G41C FlMal labeled, Mean Diameter = 12.6 nm
C. Hsp Wt 5(6-)FAM labeled, Mean Diameter = 12.1 nm
D. Hsp G41C with Iron Oxide Core, Mean Diameter = 12.9 nm
Summary
These results demonstrate the versatility of the MjsHsp16.5 (Hsp) cage as a
template for the synthesis of hard (inorganic) or soft (organic) materials.
Significantly, the symmetry of the self assembly architecture affords precise
molecular level control over the spatial distribution of attached fluorescent ligands
resulting in an ordered multivalent material. The differences between the interior
and exterior surfaces of the Hsp cage allowed mineralization of iron oxide in a
spatially controlled manner. Chapters 3 and 4 describe additional genetic and
chemical modifications of the Hsp architecture including the introduction of
biomedical functionalities.
50
CHAPTER 3
SELECTIVE ATTACHMENT AND RELEASE OF A
CHEMOTHERAPEUTIC AGENT FROM THE INTERIOR
OF PROTEIN CAGE ARCHITECTURE
Reproduced by permission of The Royal Society of Chemistry,
Chem. Commun., 4:447-449, 2005.
Abstract
Protein cage architectures, including viral capsids, heat shock protein cages,
and ferritins, have potential to serve as cell-specific therapeutic delivery agents.
Toward realization of this goal, the antitumor agent doxorubicin was covalently
bound to the interior surface of a genetically modified small heat shock protein
(Hsp) cage. Each of the 24 subunits which comprise the Hsp architecture bound
a single molecule of doxorubicin. The release of doxorubicin from the interior of
Hsp cages could be triggered within the physiologically relevant pH range of the
endosome (pH ~5) [244]. Importantly, doxorubicin remained associated with Hsp
cages when incubated in conditions similar to extracellular in vivo conditions (pH
7.4 and 37˚C). This chapter focuses on encapsulation and release of an
anticancer agent utilizing the Hsp cage architecture. These are two important
components required for the development of protein cage architectures for
biomedical applications. This work is further described in Chapter 6, in which the
efficacy of doxorubicin containing Hsp cages is investigated in melanoma cell
culture.
51
Introduction
As described in Chapter 1, there are many types of nanometer scale drug
delivery systems. Protein cage architectures have attributes that make them
suitable for development as therapeutic delivery systems. As demonstrated in
Chapter 2, the Hsp cage can be chemically derivatized with small molecules at
precise locations, including within its interior cavity. Limiting the bioavailability of
therapeutics via sequestration within a protein cage architecture is especially
beneficial when compounds have toxic side effects (Discussed in Chapter 6).
Doxorubicin is a clinically used antitumor agent. It is an anthracycline
antibiotic that was originally isolated from the fungus Streptomyces peucetius.
The mechanism of action of doxorubicin is not completely understood, but it does
intercalate within DNA and in turn interferes with DNA replication. In addition, it
damages DNA via free radical generation and has been shown to inhibit DNA
topoisomerase II [19, 245, 246]. Although commonly used, it is sometimes
discontinued in patients due to its cardiotoxic side-effects [19, 246]. Therefore,
tumor specific delivery and release of doxorubicin is desirable.
Previously, tumor targeted antibody delivery of doxorubicin was shown to be
efficacious in animal models [24, 213, 247]. The release of doxorubicin from
antibodies was triggered under acidic conditions [5]. This release strategy is
biologically relevant, since antibodies that enter cells via a receptor mediated
endocytic pathway are exposed to the acidic endosomal and lysosomal
compartments. Lysosomes are membrane bound organelles within eukaryotic
52
cells; they contain digestive enzymes responsible for degradation of
macromolecules and particles taken inside the cells by endocytosis [244]. The
pH of lysosomes is approximately 5, therefore doxorubicin bound to a drug
delivery agent via an acid labile hydrazone linkage can be released [244].
Success of this strategy utilizing antibody platforms made the (6maleimidocaproyl) hydrazone derivative of doxorubicin an ideal candidate for
encapsulation and release studies utilizing the Hsp cage architecture. In
addition, initial studies on the uptake of protein cages by mammalian cells grown
in vitro indicated that an endosomal uptake strategy was feasible (Chapter 6).
The objective of this chapter is to describe the use of a protein cage
architecture for controlled encapsulation and release of a chemotherapeutic
agent. Specifically, doxorubicin, an antitumor agent, was covalently bound and
selectively released from the interior surface of the small heat shock protein
(Hsp) cage from the hyperthermophilic archaeon, Methanococcus jannaschii
(Figure 3.1). As previously described, the Hsp cage is assembled from 24
-SH
+
-SH
SH-
-SH
H3C
(6-Maleimidocaproyl)
hydrazone of Doxorubicin
HspG41C
= HspG41C-Dox
Figure 3.1. Reaction Schematic.
The (6-Maleimidocaproyl)hydrazone of doxorubicin [5] was attached to the 12 nm
diameter heat shock protein (Hsp) cage genetically modified to have 24 interiorly exposed
cysteines (blue)(space filling model – interior view). Arrow indicates acid labile bond
cleavage site.
53
subunits into a spherical architecture 12 nm in diameter [4, 84] (Chapter 1). In
Chapter 2, a genetic variant with internally exposed cysteines, HspG41C, was
chemically modified with fluorescein molecules. In this chapter, HspG41C was
derivatized with the antitumor agent doxorubicin, using similar coupling chemistry
to that developed for fluorescein attachment.
Results and Discussion
Derivatization of HspG41C with Doxorubicin
The (6-maleimidocaproyl) hydrazone derivative of doxorubicin [5] (Mal-Dox)
was linked to the interior surface of the HspG41C protein cage via coupling of the
maleimide and thiol functionalities (Figures 3.1-3.6). HspG41C cages (2.5 mg/ml,
151.5 μM subunit) were reacted with a 3-fold molar excess of Mal-Dox in HEPES
(100 mM, pH 6.5) for 1 hour at room temperature. Immediately following the
reaction, derivatized cages were separated from free Mal-Dox by size exclusion
chromatography (Superose 6, Amersham-Pharmacia) (Figure 3.2A). Co-elution
of HspG41C protein cages (280 nm) and doxorubicin (495 nm) after 21 minutes
indicates that doxorubicin is associated with the protein cages. Both
transmission electron microscopy (TEM) (Leo 912 AB) and dynamic light
scattering (Brookhaven 90 Plus) analysis confirm that HspG41C-Dox maintains
the 12 nm diameter of underivatized Hsp cage (Figure 3.2B).
54
Absorbance (a.u.)
250
A.
B.
200
Abs280 Protein Cage
Abs495 Doxorubicin
150
100
50
50 nm
0
5
10
15
20
Time (minutes)
25
30
35
Figure 3.2. Characterization of Doxrubicin Labeled HspG41C.
A. Size exclusion chromatography elution profile of HspG41C reacted with a 3 fold (3x)
molar excess of (6-Maleimidocaproyl) hydrazone of doxorubicin per subunit. The profile
illustrates the co-elution of doxorubicin (A495) and Hsp protein cage (A280). B. TEM of
HspG41C cages containing doxorubicin (stained with uranyl acetate).
The covalent linkage of doxorubicin to HspG41C is demonstrated by SDSpolyacrylamide gel electrophoresis (SDS-PAGE). Fluorescence imaging of an
SDS-PAGE gel reveals a mobility shift of doxorubicin (excitation 488 nm;
emission 520 nm) bound to HspG41C compared to that of free doxorubicin
(Figure 3.3). Subsequent staining of the protein bands with Coomassie blue
illustrates the co-migration of the HspG41C subunit band with the fluorescent
(doxorubicin) band indicating their covalent linkage (Figure 3.3 A, B Lane 1). The
HspG41C-Dox conjugate migrates slightly slower than underivatized HspG41C
due to its increased molecular weight (Figure 3.3 A, B Lanes 1, 3).
Covalent attachment of doxorubicin to all 24 HspG41C subunits within the
cage was confirmed by liquid chromatography / electrospray mass spectrometry
(LC/MS) analysis (Esquire 3000, Bruker). HspG41C-Dox (10 μl, 0.5mg/ml) was
55
injected onto a C8 column and eluted with a H2O / acetonitrile gradient; both
solvents contained 2% acetic acid at pH 2.3. Analysis of the single LC peak by
electrospray mass spectrometry detected two protein components (Figure 3.4).
Deconvolution of the electrospray mass spectrum detected complete HspG41C
subunit + linker + doxorubicin (experimental mass 17,251; calculated mass
17,249) and HspG41C subunit + linker without doxorubicin (experimental mass
16,725; calculated mass 16,724). This result is consistent with partial acid
hydrolysis of the hydrazone linkage at pH below 5.0. The detection of only two
protein components, both at similar concentrations, is indicative of complete
labeling of HspG41C with Mal-Dox. No unlabeled HspG41C was detected by
mass spectrometry (Figure 3.4).
Quantitative analysis by absorbance spectroscopy also indicates that 24
doxorubicin molecules are linked to the 24 subunit HspG41C cage. The
concentration of HspG41C protein in purified doxorubicin derivatized cage
preparations was determined by subtracting a normalized doxorubicin spectrum
from the labeled Hsp spectrum (Hsp ε = 9322 M-1 cm-1) [4]. The concentration of
doxorubicin covalently attached to HspG41C was determined from the
absorbance maxima at 495 nm (Mal-Dox ε = 8030 M-1 cm-1) [185]. Control
reactions utilizing the wild type (wt) Hsp cage, lacking thiols, did not show
appreciable labeling with Mal-Dox (Figure 3.3A Ln 2).
56
________________________________________________________________
A.
Hsp -Dox
Free Dox
1 2 3 4 5 6
1 2
3 4 5 6
24
2020
Figure 3.3. SDS-PAGE of Doxorubicin (Dox) Linked to
HspG41C imaged by (A) Fluorescence and the same gel
(B) Coomassie stained
Lane: 1. HspG41C-Dox; 2. WtHsp-Dox; 3. HspG41C;
4. WtHsp; 5. Molecular Weight Standards (kDa); 6. Dox.
Analysis of HspG41C reacted with the (6-maleimidocapryoyl)
hydrazone of dox illustrates covalent linkage of dox to the protein
subunit (A, B Ln 1) whereas after identical reaction conditions
wtHsp does not show labeling (A, B Ln 2). The gel illustrates the
shift in mobility of dox when linked to HspG41C subunit (compare
A. Lanes 6 & 1) as well as a shift in mobility of HspG41C-Dox (B
Ln 1) compared to unlabeled HspG41C (B Ln 3).
14.2
6.5
B. Coomassie
HspG41C and Linker and Dox
Hsp subunit
H5C
Experimental Mass = 17,251
Calculated Mass = 17,249
HspG41C and Linker
Hsp subunit
Experimental Mass = 16,725
Calculated Mass = 16,724
600
800
1000
m/z
1200
1400
1600
Figure 3.4 Mass Spectrometry of HspG41C-Doxorubicin (Dox)
Deconvolution of the above electrospray mass spectrum illustrates the presence of two protein
components: the complete HspG41C subunit + linker + doxorubicin (red) (experimental mass 17,251;
calculated mass 17,249) and HspG41C subunit + linker (blue) without the doxorubicin moiety
(experimental mass 16,725; calculated mass 16,724).
57
Release of Doxorubicin from HspG41C in Acidic Conditions
We quantified the selective release of doxorubicin from HspG41C protein
cages through hydrolysis of the hydrazone linkage under acidic conditions
(Figure 3.5) [5, 248]. Doxorubicin release studies were performed at pH 4.0, 4.5,
and 5.0 (370C). After incubation for times ranging from 0.25 to 24 hours,doxorubicin labeled HspG41C was separated from free doxorubicin by chromatography (Micro Bio-Spin P-30 Columns, Bio-Rad). The amount of doxorubicin that
remained associated with the HspG41C protein was quantified by absorbance
spectroscopy. As shown in Figure 3.5, 50% of the doxorubicin bound to the
interior of HspG41C was released after 1.5, 3.9, and 5.1 hours at pH 4.0, pH 4.5
and pH 5.0 respectively. Acid triggered release of doxorubicin from other
systems has been shown effective both in vivo and in vitro [219, 248]. Our data
% Dox in HspG41C
100
80
pH
pH
pH
pH
60
40
7.4
5.0
4.5
4.0
20
0
0
5
10
Time (hours)
15
20
Figure 3.5 Acid Triggered Release of Doxorubicin from HspG41C Cages
50% of the doxorubicin bound to the interior of HspG41C cages was released after 1.5, 3.9,
and 5.1 h at pH 4.0, 4.5, and 5.0, respectively. At pH 7.4, dox remained within the cage.
58
show the potential for doxorubicin release from Hsp cages under biologically
relevant (lysosomal) conditions.
Summary
We have demonstrated the ability to load, quantify, and selectively release
doxorubicin from a heat shock protein cage architecture. Housing a therapeutic
agent within the interior of the Hsp cage restricts its bioavailability until release.
These results illustrate the potential of protein cage architectures to serve as
versatile drug delivery systems.
59
CHAPTER 4
MELANOMA AND LYMPHOCTYE CELL SPECIFIC TARGETING
INCORPORATED INTO A HEAT SHOCK PROTEIN
CAGE ARCHITECTURE
Reproduced by permission of Elsevier, Chem.Biol.,
13(2):161-170, 2006.
Abstract
Protein cages, including viral capsids, ferritins, and heat shock proteins, are
genetically and chemically malleable nanoplatforms with potential to serve in
biomedical applications. In this chapter, both genetic and chemical strategies
were used to impart mammalian cell specific targeting to the 12 nm diameter
small heat shock protein (Hsp) architecture from Methanococcus jannaschii. The
tumor vasculature targeting peptide RGD-4C (CDCRGDCFC) was genetically
incorporated onto the exterior surface of the Hsp cage. This tumor targeting
protein cage bound to αvβ3 integrin expressing C32 melanoma cells in vitro. In a
second approach, cellular tropism was chemically imparted by conjugating antiCD4 antibodies (Ab) to the exterior surface of Hsp cage architectures. These
Ab-Hsp cage conjugates bound to CD4+ cells present in a population of murine
splenocytes. Protein cages have the potential to simultaneously incorporate
multiple functionalities including cell specific targeting, imaging and therapeutic
agent delivery. In this chapter the simultaneous incorporation of two
60
functionalities, imaging and cell specific targeting, onto the Hsp protein cage
architecture is demonstrated.
Introduction
In the previous chapter, the ability of the Hsp cage architecture to house and
release a chemotherapeutic agent was demonstrated. The introduction of cell
specific targeting into protein cages is another component toward their
development as biomedical platforms. Incorporation of cell and tissue targeting
allows for the potential of enhanced imaging capacity and precise therapeutic
delivery. The objective of this chapter is to demonstrate that a single protein
cage platform can simultaneously incorporate cell specific targeting and imaging
agent delivery.
Cell Specific Targeting Peptides
Identified by the Phage Display Technique
The field of cell specific targeting has been significantly advanced by the in
vivo use of phage display techniques to identify targeting peptides [24, 183, 187,
194, 195, 249-252]. Phage display libraries are “living” libraries of infectious
Escherichia coli bacteriophage that have been genetically engineered to express
a library of small peptides on their exterior surface. A simplified overview of the
phage display technique is illustrated in Figure 4.1. In brief, a ‘library’ of
bacteriophage displaying random peptides (100 million or more different
sequences) are exposed to a target, some phage bind the target whereas non-
61
binding phage are washed away. The bound phage are subsequently isolated,
amplified in their bacterial hosts (there is usually several rounds of this selection
process), and the DNA sequences encoding the substrate binding peptides are
identified. A consensus sequence is generated based on the amino acids
presented by the binding phage and the ability of this sequence to target a
specific cell and/or tissue can be further characterized. Phage display library
analysis was first described in 1990 as a method to identify peptide-protein
interactions and map the epitope binding sites of antibodies [253]. Its expanded
utility for discovery of tumor and organ homing peptides in animal models was
described later by the Ruoslahti lab [7, 8, 23].
In vivo phage display identified short peptides that target to the vasculature of
a variety of tissues, organs and tumors (Figure 4.1 B.) [7, 8, 254, 255]. Targeting
peptides linked to specific cargo molecules such as therapeutic agents, proapoptotic peptides, and quantum dots were able to localize the cargo to the
desired in vivo target [24, 187, 254, 256-259]. One characterized example is the
targeting peptide RGD-4C (CDCRGDCFC) which binds αvβ3 and αvβ5 integrins
that are prevalently expressed within tumor vasculature [23, 24, 260-262]. Work
by Arap et al. demonstrated that RGD-4C targeted doxorubicin enhanced tumor
regression at therapeutic concentrations less than that required to demonstrate
therapeutic efficacy with non-targeted doxorubicin [24]. Subsequently, many
researchers have utilized the RGD-4C peptide motif for tumor targeting of
62
liposomes, radiolabels, therapeutics, and adenoviral gene therapy vectors [51,
221, 258, 259, 263-275]. The effects of RGD-4C targeted therapeutics are
A.
B.
Phage Display
with Electronic Materials
Figure 4.1. Phage Display Technique for Discovering Peptides
That Interact with Particular Targets
The 109 -1013 random peptide sequences are exposed to the different targets (A – inorganic
(ZnS) surface [2]; B – in vivo targeting including organs and xenograft tumors); nonspecific
peptide interactions are removed with extensive washes or perfusion [7, 8]. The phage that bind
are eluted by lowering the pH and disrupting the surface interaction or isolated from various
organs and tumors. The eluted or isolated phage are amplified by infecting the E. coli ER2537
host, producing enriched populations of phage, displaying peptides that interact with the specific
substrate. The amplified phage are isolated, titered, and re-exposed to a freshly prepared
substrate surface, thereby enriching the phage population with substrate-specific binding phage.
This procedure is repeated three to five times to select the phage with the tightest and most
specific binding. The DNA of phage that show specificity is sequenced to determine the peptide
binding sequence.
augmented due to the anti-angiogenic property of RGD-4C itself [24, 221, 259,
263, 274, 275]. Due to the prior success of RGD-4C, it was chosen as a “proof
of concept” targeting peptide for genetic incorporation into a small heat shock
protein (Hsp) cage architecture.
63
In this chapter, the ability to introduce cell targeting capacity to protein cage
architectures is described. Both peptides and antibodies were incorporated on
the exterior surface of Hsp cages and tested for their ability to bind cell surface
ligands. Chemically and genetically modified Hsp cages were simultaneously
capable of cell targeting and imaging. Genetic incorporation of other cell-specific
targeting peptides on Hsp and results obtained investigating their ability to bind
target cells are presented in Chapter 6. In addition, a genetic variant analogous
to HspG41C-RGD4C was generated in human H-chain ferritin (RGD4C-Fn)
[276]. Data characterizing the ability RGD4C-Fn to bind cells in vitro is also
presented in Chapter 6.
Results and Discussion
We have demonstrated that genetic addition of the RGD-4C peptide or
chemical conjugation of an anti-CD4 monoclonal antibody (mAb) onto the
exterior surface of the small heat shock protein (Hsp) cage architecture confers
specific cell targeting capacity. In addition, we were able to load a cargo
molecule, a fluorescent imaging agent (fluorescein), within the interior cavity of
the Hsp cage. These results demonstrate the multifunctional capacity of protein
cage architectures and their potential utility in medicine.
Tumor Targeted HspG41C-RGD4C
The RGD-4C peptide was genetically incorporated into the Hsp cage to endow
it with tumor vasculature targeting capability. The RGD-4C peptide binds αvβ3
64
and αvβ5 integrins which are prevalently expressed on angiogenic tumor
vasculature [24, 261]; therefore it was genetically incorporated into HspG41C
cages in order to endow them with tumor targeting capability.
As described in
Chapter 3, HspG41C is a genetic variant of Hsp with interior cysteine residues
employed for attachment of cargo molecules [37, 38]. Protein modeling, based
on crystallographic data, indicate that C-terminal amino acid residues 140-146
are found on the exterior surface of the Hsp cage [4, 84]. Therefore, the RGD4C peptide was genetically incorporated at the C-terminus of Hsp in order to
present RGD-4C loops on the exterior of the protein cage. In addition, glycine
residues (SGGCDCRGDCFCG) were added both before and after the RGD-4C
insert to extend the peptide away from the C-terminus and allow for some
structural flexibility. The insert was confirmed by DNA sequencing and the new
tumor targeting Hsp cages were expressed and purified from an E. coli
expression system. Mass spectrometry verified the average subunit mass of
HspG41C-RGD4C to be 17814.3 compared to the predicted mass of 17814.6.
The HspG41C-RGD4C mutant assembled as efficiently as the wild-type protein
cage (Figure 4.2). HspG41C-RGD4C protein cage purification did not require
reducing agents to prevent inter-cage aggregation, suggesting that the four
cysteines present in each RGD-4C loop are disulfide bonded. Additional tumor
targeting genetic variants of Hsp, including HspG41C-RGD2C are described in
Chapter 7.
65
Characterization of recombinant HspG41C-RGD4C protein cages by size
exclusion chromatography, dynamic light scattering, and transmission electron
microscopy (TEM) demonstrated that the overall spherical structure of the Hsp
cage was not compromised due to the incorporation of the RGD-4C peptide
(Figure 4.2). Size exclusion chromatography elution profiles of HspG41CA
Absorbance (a.u.)
500
HspG41CRGD-4C
400
Abs 280 nm
300
200
100
0
0
5
10
15
20
Elution Volume (mL)
100
Relative Intensity
B
25
30
80
60
40
20
0
0
20
40
60
Diameter ( nm)
80
100
C
50 nm
Figure 4.2 Characterization of the αvβ3 IntegrinTargeted
HspG41CRGD-4C Protein Cage
A. Size exclusion chromatography elution profile of HspG41CRGD-4C cages (Abs 280 nm).
B. Dynamic light scattering analysis of HspG41CRGD-4C (average diameter 15.4 + 0.3 nm).
C. Transmission electron micrograph of the HspG41CRGD-4C cages negatively stained
with 2% uranyl acetate.
66
RGD4C cages indicated that the cages are slightly larger than the HspG41C
parent cages lacking the targeting peptide (Figure 4.2) [38]. This observation was
supported by dynamic light scattering (DLS) which indicated a larger average
diameter for the HspG41C-RGD4C cages (15.4 + 0.3 nm) as compared to
HspG41C (12.7 + 0.5 nm) (Figure 4.2). Transmission electron microscopy
images of the HspG41C-RGD4C cages and HspG41C cages were
indistinguishable (Figure 4.2) [38].
Chemical Derivatization of HspG41C-RGD4C
HspG41C-RGD4C cages were labeled with fluorescein in order to study their
cell-specific targeting capability. This demonstrated that both cell targeting and
imaging functionalities could be simultaneously incorporated into the Hsp protein
cage architecture. The HspG41C-RGD4C has a total of 120 cysteines per cage
(5 cysteines per subunit). Sub-stoichiometric labeling with fluorescein-5maleimide ensured that every cage displayed a significant fraction of unmodified
cysteines within the RGD-4C sequence. The original RGD-4C peptides
discovered by phage display were in a cyclic conformation due to intrapeptide
disulfide bond formation [24, 187]. Likewise, we predicted that RGD-4C peptides
presented on HspG41C-RGD4C architectures would also cyclize due to
intrapeptide disulfide bond formation. Hsp-fluorescein conjugated cages were
purified from free fluorescein via size exclusion chromatography and the covalent
67
nature of the fluorescein-Hsp subunit linkage was demonstrated by SDSpolyacrylamide gel electrophoresis (SDS-PAGE) (Figure 4.3).
Absorbance spectroscopy determined that on average there were 26 out of
the 120 possible cysteines per cage labeled with fluoresceins; therefore only a
fraction of the cysteines within the RGD-4C were bound to fluorescein.
Absorbance (a.u.)
A.
B.
HspG41CRGD4C-Fluorescein
Fluorescent Image
Hsp Protein (280 nm)
Fluorescein (495 nm)
40
30
1
20
2 3 4
Protein Image
50 nm
10
20
0
14.2
0
5
10
15
20
Elution Volume (mL)
25
30
1 2
3 4
Figure 4.3. Characterization of Fluorescein Labeled HspG41CRGD-4C Cages
A. Size exclusion chromatography elution profile of HspG41CRGD-4C cages. Co-elution
of the absorbance at 495 nm (fluorescein) and 280 nm (protein) illustrate that fluorescein
bound to intact Hsp cages. TEM micrograph (inset) also demonstrates the presence of
intact Hsp cages.
B. SDS-PAGE of HspG41CRGD-4C subunits demonstrates covalent linkage of
fluorescein. Lane (1) molecular weight standard; (2) HspG41C; (3) HspG41C-RGD4C;
(4) HspG41C-RGD4C-fluorescein. The same gel was fluorescently imaged (top) and
subsequently stained with Coomassie blue (bottom) to visualize the proteins. The gel
illustrates covalent linkage of fluorescein to HspG41CRGD-4C monomers (Lane 4)
whereas unlabeled Hsp subunits exhibit no fluorescence (Lanes 2 & 3).
Qualitative mass spectrometry analysis of fluorescein labeled HspG41C-RGD4C
subunits indicated that there are between one and five fluoresceins per subunit
(Figure 4.4).
Some of the RGD-4C peptides were presumably in ‘loop’
conformation whereas others were potentially linearized depending on the
number of conjugated fluorescein-5-maleimides per subunit [38]. Data from mass
68
spectrometry indicated the presence of both the disulfide and the free thiol form
of the RGD-4C peptide. Previous structural studies of synthetic RGD-4C peptides
in solution revealed that there are two predominant cyclic conformations both of
which bind αvβ3 integrin although one (RGD-A) exhibited higher binding affinity
Figure 4.4. Deconvoluted Mass spectrum of HspG41CRGD-4C–Fluorescein
Deconvoluted mass spectrum of HspG41CRGD-4C–Fluorescein showing from 1 to 5
fluoresceins covalently attached to one subunit indicated by the number. These
deconvoluted data were produced from the raw data by the software MaxEnt1 (Waters).
[277]. However, in a second report, synthetically produced acyclic RGD-4C was
shown to have higher binding affinity than the cyclic form [269]. Since the
fluorescein labeled HspG41C-RGD4C constructs most likely had a mixed
presentation of ‘loop’ and linearized RGD-4C targeting peptides we hypothesized
that they would bind αvβ3 integrin expressing cells.
69
HspG41C-RGD4C Cages Bind C32 Melanoma Cells
Epifluorescence microscopy was used to visualize fluorescently labeled
HspG41C-RGD4C cages bound to αvβ3 integrin expressing C32 melanoma cells
in vitro [278, 279]. For all microscopy studies, C32 melanoma cells were grown
on glass coverslips. Cell surface expression of αvβ3 integrin on C32 cells was
verified by immunofluorescence microscopy utilizing a fluorescein-conjugated
anti-αvβ3 monoclonal antibody (mAb) (LM609 - Chemicon). HspG41C-RGD4Cfluorescein cages were observed to efficiently bind to C32 cells as compared to
control samples of Hsp cages without the RGD-4C peptide (Figure 4.5). For
direct comparison of epifluorescence data the concentration of fluorescein was
normalized (2.5 μM), and the illumination intensity and the camera exposure
were held constant. HspG41C-RGD4C-fluorescein protein cages were observed
to efficiently bind to C32 melanoma cells as compared to controls of Hsp protein
cages lacking the RGD-4C peptide (Figure 4.5). In order to ensure that
HspG41C-RGD4C-fluorescein interaction with C32 cells was not mediated solely
by exterior RGD-4C bound fluorescein, an additional mutant with surface
exposed cysteine residues (HspS121C) was also tested. The HspS121Cfluorescein cages bind fluorescein-5-maleimide via externally facing cysteines
but lack the RGD-4C targeting sequences. These control cages did not bind C32
cells at a level detectable by epifluorescence microscopy, indicating the cell
binding observed for HspG41C-RGD4C was due to the presence of the RGD-4C
peptide (Figure 4.5).
70
HspG41C-Fl
A.
Light Image
Fluorescent Image
HspG41CRGD4C-Fl
B.
Light Image
Fluorescent Image
HspS121C-Fl
C.
Light Image
Fluorescent Image
Figure 4.5. Epifluorescence Microscopy of C32 Melanoma Cells with
Hsp Cage-Fluorescein Conjugates
Cells were incubated with (A) non-targeted HspG41C-Fl cages with interiorly bound fluorescein (Fl),
(B) αvβ3 integrin targeted HspG41CRGD4C-Fl cages and (C) non-targeted HspS121C-Fl cages with
exteriorly bound fluorescein. C32 melanoma cells grown on cover-slips were incubated with Hsp-Fl
conjugates and imaged by both light (top) and fluorescence microscopy (bottom). The Fl concentration for cage-cell incubations was 2.5 μM and all fluorescent images were taken at a standardized
camera exposure time of 50 ms. Scale bar = 50 μm.
Fluorescence activated cell sorting (FACS) was used to quantify the ability of
fluorescein labeled HspG41C-RGD4C cages to bind to C32 melanoma cells.
Adherent C32 melanoma cells were non-enzymatically disassociated from cell
culture dishes and suspended in DPBS + Ca2+ / Mg2+. Fluorescently labeled Hsp
cage preparations were incubated with cells on ice at a normalized fluorescein
concentration of 2 μM and cells were washed prior to FACS. C32 melanoma cell
associated fluorescence was dramatically increased after incubation with
HspG41C-RGD4C-fluorescein cages. The geometric mean (geo. mean)
fluorescence intensity value of 1410 clearly indicated that targeted HspG41CRGD4C cages cages exhibited cell binding in vitro (Figure 4.6A). C32 melanoma
71
cells do exhibit a background level of auto-fluorescence with geometric mean
fluorescence intensity value of 66 (Figure 4.6A). A positive control experiment, in
which C32 melanoma cells were incubated with a fluorescein conjugated antiαvβ3 mAb, also showed the expected increase in fluorescence (geo. mean 1037)
associated with cell specific binding of the antibody (Figure 4.7). FACS analysis
of non-targeted cages (HspG41C-fluorescein inside; HspS121C-fluorescein
outside) indicated low levels of non-specific binding of protein cages and
fluorescein to C32 melanoma cells (geo. mean intensities of 129 and 216
respectively, Figure 4.6A-B). In all cases, these results were independent of
incubation times, which ranged from 20 minutes to 2 hours (data not shown).
72
A
129
1410
C
353
66
216
D
714
353
1235
Counts
B
Fluorescence Intensity (a.u.)
Figure 4.6 Specific Binding of HspG41CRGD4C-Fluorescein Labeled Cages
to C32 Melanoma Cells
The FACS data of C32 melanoma cells incubated with Hsp-fluorescein cages are
plotted as histograms labeled with their corresponding geometric mean fluorescence
intensity values (geo. mean). The background level of C32 cell associated fluorescence
(blue solid line; geo. mean 66) and the increased level of C32 cell associated fluorescence
due to binding of HspG41CRGD4C-Fl cages (green filled plot; geo. mean 1410) are shown
in all panels. (A) non-targeted cages HspG41C-fluorescein labeled interiorly (red dashed
line; geo. mean 129); (B) non-targeted HspS121C-Fl cages with exteriorly bound
fluorescein (orange dotted line; geo. mean 216); (C) Anti-αvβ3 integrin mAb blocked C32
melanoma cells subsequently incubated with HspG41CRGD4C-Fl cages (dashed purple
line; geo. mean 353) (D) mAb concentration dependent blocking of αvβ3 integrin on C32
cells demonstrated by subsequent incubation of blocked cells with HspG41CRGD4C-Fl
(2.4 μM subunit in the assembled cage); 1.3 μM mAb (purple solid line; geo. mean 353),
cells blocked with 0.65 μM (green solid line; geo. mean 714), 0.065 μM mAb (solid black
line; geo. mean 1235).
73
The binding specificity of RGD-4C targeted Hsp cages was tested with
competition experiments. For these studies, C32 melanoma cells were preincubated with either unlabeled anti-αvβ3 mAb or unlabeled HspG41C-RGD4C
prior to incubation with fluorescently labeled HspG41C-RGD4C cages. FACS
analysis demonstrated a reduction of RGD-4C targeted cage binding (Figures 4.7
and Figure 4.8). As stated above, for FACS analysis the fluorescein
concentration was normalized to 2 μM. This amount of fluorescein corresponds
to 50 μg/ml (2.4 μM) HspG41C-RGD4C subunit in the assembled cage
(corresponding to 0.1 μM cage) and the ratio of fluorescein per subunit is near
one to one. The degree of both antibody and HspG41C-RGD4C blocking was
concentration dependent. In order to compare the effectiveness of different
blocking proteins, we compared them on a molar basis. At anti-αvβ3 integrin mAb
concentrations of 0.065 μM (10 μg/ml) minimal inhibition was observed. At
65 μM (100 μg/ml) some inhibition was evident, and finally at 1.3 μM (200 μg/ml)
binding was inhibited to levels corresponding to that of non-targeted cages
(Figure 4.6D). Likewise, unlabeled HspG41C-RGD4C reduces the binding of
HspG41C-RGD4C-Fl in a concentration dependent manner (Figure 4.8).
74
1037
1410
66
Figure 4.7. Binding of HspG41CRGD4C-Fluorescein Labeled Cages to C32 Melanoma
Cells Relative to an Anti-αvβ3 Antibody-Fluorescein Positive Control
FACS analysis of HspG41CRGD-4C-fluorescein and anti-αvβ3 antibody-fluorescein
interaction with C32 melanoma cells. The FACS data of C32 cells incubated with Hspfluorescein cages are plotted as histograms labeled with their corresponding geometric
mean fluorescence intensity values (geo. mean). The background level of C32 cell
associated fluorescence (blue solid line; geo. mean 66) and the increased level of C32 cell
associated fluorescence due to binding of HspG41CRGD4C-Fl cages (green filled plot; geo.
mean 1410) and anti-αvβ3 antibody-fluorescein positive control (black line; geo. mean 1037).
A
748
905
501
87
B
95
1813
1830
C
1469
95
1546
1140
1685
1685
Figure 4.8 HspG41CRGD4C-Fluorescein Cage Interaction with C32 Melanoma Cells is
Blocked by Unlabeled HspG41CRGD-4C but not by HspG41C or Albumin
FACS data from C32 melanoma cells incubated with an unlabeled blocking protein (HspG41CRGD-4C,
HspG41C, or bovine albumin) prior to incubation with HspG41C-RGD4C-fluorescein cages are plotted
as histograms labeled with their corresponding geometric mean fluorescence intensity values (geo.
mean). The background level of C32 cell associated fluorescence (blue solid line; geo. mean 87 & 95)
and the increased level of C32 cell associated fluorescence due to binding of HspG41CRGD4C-Fl cages
(50 μg/ml) (green filled plots; geo. mean 1140 & 1685), in the absence of any blocking protein, are
shown. (A) Unlabeled HspG41C-RGD4C concentration dependent blocking; 5000 μg/ml (light purple
solid line; geo. mean 501), 500 μg/ml (dark purple dashed line; geo. mean 748), cells blocked with 50
μg/ml (solid black line; geo. mean 905). (B) Unlabeled HspG41C does not block HspG41CRGD4C-Fl
cage interaction with C32 cells; 5000 μg/ml (light purple solid line; geo. mean 1546), 500 μg/ml (dark
purple dotted line; geo. mean 1813), cells blocked with 50 μg/ml (solid black line; geo. mean 1830). (C)
Unlabeled bovine albumin (5000 μg/ml) does not block HspG41CRGD4C-Fl cage interaction with C32
cells (light purple solid line; geo. mean 1469). Note: All FACS machine settings were identical, but data
set A was collected on a different day than sets B. and C.
75
Blocking of the RGD-4C-αvβ3 integrin interaction was demonstrated to be specific
since equivalent amounts of either HspG41C or bovine albumin (fraction V) did
not decrease the binding of HspG41C-RGD4C-Fl to C32 melanoma cells (Figure
4.8). Together these data illustrate the effective genetic introduction of αvβ3
integrin targeting capacity to the Hsp cage architecture, coupled with attachment
of fluorescein imaging agents.
Chemical Introduction of Cell-Specific
Targeting to Hsp via Antibody Conjugation
The versatility of protein architectures for cell specific targeting was
demonstrated by the chemical introduction of lymphocyte targeting to Hsp
architectures. An anti-CD4 mAb (ATCC GK1.5) was conjugated to fluorescein
labeled HspG41C cages via the heterobifunctional cross-linker SMCC
(sulfosuccinimidyl 4-(N-maleimidomethyl) cyclohexane-1-carboxylate). Size
exclusion chromatography was utilized to purify anti-CD4-HspG41C-fluorescein
cage conjugates (Ab-Hsp-Fl) from HspG41C-fluorescein cages. The elution
profile indicated that the Ab-Hsp-Fl cage conjugates were larger than HspG41CFl cages and DLS confirmed the size increase from 12.2 nm to 22 + 0.1 nm
diameter (Figure 4.9).
76
Relative Intensity
100
80
Figure 4.9. Dynamic Light Scattering Analysis
of Anti-CD4 mAb HspG41C Cage Conjugates
Dynamic Light Scattering (DLS) indicates that antiCD4 mAb conjugated HspG41C-fluorescein cages
have an average diameter of 22.4 + 0.1 nm.
60
40
20
0
0
20
40
60
Diameter (nm)
80
100
Cell Specific Binding of Ab-Hsp Conjugates
FACS and epifluorescence microscopy were used to investigate binding of
anti-CD4-HspG41C-fluorescein (Ab-Hsp-Fl) cages to CD4+ cells within a murine
splenocyte population. A comparison of epifluorescence microscope images in
which Ab-Hsp-Fl, Hsp-Fl, or Ab-Fl were each combined individually with murine
splenocytes illustrated that Ab-Hsp-Fl bound to a similar number of cells as the
Ab-Fl positive control (Figure 4.10). FACS was used for quantitative analysis of
this interaction. Ab-Hsp-Fl cages specifically bound to 21% of the total cells
within this population (Figure 4.11A). This level of binding is consistent with the
percentage of CD4+ cells within this murine splenocyte population as determined
by binding of a fluorescein conjugated anti-CD4 mAb to 19% of this murine
splenocyte population (Figure 4.11B). Further confirmation of binding specificity
was obtained from antibody blocking experiments. Splenocytes were first
incubated with unlabeled anti-CD4 mAb, washed to remove unbound blocking
antibody, and subsequently incubated with Ab-Hsp-Fl cage conjugates. FACS
77
analysis of this blocking experiment demonstrated that only 2% of the population
exhibited cell associated fluorescence (Figure 4.11C). This percentage of cell
fluorescence corresponds to that observed with non-targeted HspG41Cfluorescein cages (2%) (Figure 4.11D). The observed binding of Ab-Hsp-Fl
cages to 21% of the murine splenocyte population encompasses both the
specific binding to CD4+ lymphocytes (19%) on top of a small non-specific level
of background association (2%). This indicates that immuno-targeted protein
cages effectively target to specific cells within a mixed population.
Splenocytes
A
+ Ab-Hsp-Fl
B
+ Ab-Fl
C
+ Hsp-Fl
D
Figure 4.10. Eplifluorescence Microscope Images Illustrating the Specific Binding of antiCD4 mAb Conjugated HspG41C-Fluorescein Cages (Ab-HspG41C-Fl) to CD4+ Lymphocytes
Bright field and corresponding epifluorescence images of (A) murine splenocytes incubated with
(B) anti-CD4 mAb HspG41C-fluorescein cage conjugates (Ab-Hsp-Fl), (C) anti-CD4 mAbfluorescein (Ab-Fl), or (D) non-targeted HspG41C-fluorescein cages (Hsp-Fl). The fluorescein
concentration for cage-cell incubations was 2 μM and all fluorescent images were taken at a
standardized camera exposure time of 500 ms. Scale bar = 50 μm.
78
C
Red Fluorescence Intensity (a.u.)
A
2%
21%
D
B
19%
2%
FITC Fluorescence Intensity (a.u.)
Figure 4.11. Specific Binding of anti-CD4 mAb Conjugated HspG41C-Fluorescein
Cages (Ab-HspG41C-Fl) to CD4+ Lymphocytes
FACS analysis of murine splenocytes incubated with: (A) anti-CD4 mAb HspG41Cfluorescein cage conjugates (Ab-Hsp-Fl) bound 21% of cells within this population
(B) anti-CD4 mAb-fluorescein demonstrated that 19% of this splenocyte population is CD4+,
(C) CD4+ cells were blocked with unlabeled anti-CD4 mAb, then subsequently exposed to
Ab-Hsp-Fl cage conjugates demonstrated a low level of non-specific binding (2%),
corresponding to (D) non-targeted HspG41C-fluorescein cages (2%).
Summary
In this chapter, the 12 nm diameter Hsp cage was both genetically and
chemically modified to incorporate cell specific targeting properties. Genetic
incorporation of αvβ3 integrin binding RGD-4C peptide onto the exterior surface of
Hsp cages conferred cell specific targeting capability. It is expected that many
other cell targeting peptides, especially those discovered by in vivo phage display
79
library techniques, could also be incorporated into this and other protein cage
architectures, some of which are described in Chapter 6. Chemical linkage of an
anti-CD4-antibody to Hsp cages and subsequent targeting of CD4+ cells within a
mixed population of murine splenocytes demonstrated the success and
possibilities of immuno-targeted Hsp cages. However, an attempt to link another
less readily available antibody (anti-collagen) to the Hsp cage was not successful
(see Materials & Methods). Fluorescein, an imaging agent, was covalently linked
to protein cages demonstrating that these architectures can simultaneously
incorporate multiple functionalities including cell specific targeting and imaging
agent delivery. This opens exciting avenues for the incorporation and cell
specific delivery of other cargo molecules such as magnetic resonance (MR)
imaging and therapeutic agents.
80
CHAPTER 5
INVESTIGATION OF THE INTERACTIONS BETWEEN MAMMALIAN
CELLS AND PROTEIN CAGE ARCHITECTURES
INCLUDING THE EFFICACY OF THERAPEUTIC
CONTAINING Hsp CAGES
Abstract
Protein cage architectures have potential to serve as cell and/or tissue
targeted therapeutic and imaging agent delivery systems. Housing and
delivering therapeutic molecules within protein cage architectures provides the
following potential advantages as compared to administration of ‘naked’
therapeutics: (1) increased therapeutic payload per delivery event, (2) cell
specific targeting capability, (3) a decrease in the amount of systemic drug and in
turn (4) a decreased level of drug associated side effects, and (5) the ability to
restrict the bioavailability of therapeutics until triggered release at a specific
location or time.
The interactions between protein cage architectures (CCMV and Hsp) and
mammalian cells were investigated via microscopy (epifluorescence and
confocal) and a cell proliferation assay. Together, these results demonstrated
that it was necessary to incorporate cell specific targeting ligands within protein
cage architectures. The ability of protein cage architectures to restrict the
bioavailability of therapeutics was confirmed with cell proliferation assays,
following the incubation of cells with doxorubicin (an antitumor agent) derivatized
heat shock protein (Hsp) cages (described in Chapter 3 and below). In vitro
81
cancer cell proliferation assays demonstrated that doxorubicin (dox) containing
HspG41C cages (non-targeted) were unable to induce cancer cell death; in
contrast to parallel assays in which free doxorubicin caused cell death. A tumor
targeting genetic variant, HspG41C-RGD4C, was derivatized with the (6maleimidocaproyl) hydrazone of doxorubicin (described in Chapter 3) [5, 37].
Interestingly, although the same chemistry (thiol-maleimide) that proved
successful for the conjugation of fluorescein to HspG41C-RGD4C cages was
employed for dox conjugation, the cages were less tolerant of the dox reaction.
Therefore, the yields of dox derivatized HspG41C-RGD4C cages were low, but
sufficient for in vitro evaluation. The efficacy of HspG41CRGD4C-Dox cages
was tested utilizing C32 melanoma cells grown in culture. The results from initial
cell proliferation assays are variable; sometimes the HspG41CRGD4C-Dox
cages seem to kill cells whereas in some assays they do not seem as promising.
Introduction
Each year more than 11 million people, including 1.4 million Americans, are
diagnosed with cancer [20, 245]. It is estimated that there will be 16 million new
cases every year by 2020 [245]. Annually, 7 million deaths (12.5%) are attributed
to cancer [245]. Recently, the National Institutes of Health and the National
Cancer Institute have developed strong interest in the development of new nanotechnological platforms for diagnosis, imaging and treatment of cancer [18, 20].
Their goals include the development of imaging agents and diagnostics to detect
82
cancer in the earliest stage and multi-functional targeted devices to deliver
multiple therapeutic agents directly to cancer cells [18]. Protein cages potentially
provide unique nano-templates for use in combined targeted therapeutic delivery
and imaging applications.
Cancer treatment often involves the use of compounds that kill actively
dividing cells including the cancerous cells which are dividing out of control.
Unfortunately, these drugs also kill ‘normal’ dividing cells, such as those of the
bone marrow, skin, hair follicle, and gastrointestinal tract. Doxorubicin, also
known as adriamycin, is one such commonly used compound. Its mechanism of
action is not completely understood, but it involves the inhibition of important
cellular enzymes including RNA polymerase, reverse transcriptase, and
topoisomerase-II. In addition, it intercalates into DNA and is thought to disrupt
cell membranes [19, 184, 231, 245, 246, 280]. The clinical use of doxorubicin is
associated with many side effects, including nausea, neurotoxicity, and
cardiotoxicity. These effects can be dose dependent, and therefore could
potentially be relieved via tumor specific targeting of dox. The most common
reason for discontinuing the use of doxorubicin in cancer patients is the onset of
cardiotoxicity [19, 246].
Doxorubicin was chosen as a model therapeutic to test the efficacy of protein
cage architectures as drug delivery systems for the reasons stated above and
because antibody and peptide mediated doxorubicin delivery had proven
efficacious both in vitro and in vivo [24, 213]. In these studies, doxorubicin was
83
conjugated to either the anticarcinoma antibody BR64 [213], specific for an
epitope of the Lewis Y antigen, which is abundantly expressed on cancer cells, or
the RGD-4C peptide, which targets αvβ3 and αvβ5 integrins expressed on tumor
vasculature [24]. The antibody conjugates were loaded with a maximum of 8
molecules of dox whereas the peptides were each associated with two dox
molecules [24, 213]. Both constructs were efficacious in cell culture. In addition,
they proved more effective than free doxorubicin in human tumor xenograft
animal models [24, 213].
As described in Chapter 1, there are many types of nanometer scale drug
delivery systems including antibody and peptide conjugates (described above),
micelles, silica nanoparticles, liposomes, and hydrogel dextran nanoparticles [19,
24, 196, 219]. In this chapter, protein cage architectures are presented as
additional platforms for the development of new targeted therapeutic delivery
systems. The potential advantages of protein cages have been described
previously (Abstract and Chapter 1). In brief, protein cage architectures can be
genetically and chemically modified for incorporation of novel amino acids, small
molecules (including therapeutics) and cell targeting peptides and antibodies. All
of which are incorporated at precise locations on the protein cages in a spatially
selective manner. The bioavailability of therapeutics housed within cage
architectures may be restricted until release is selectively triggered by external
stimuli. Lastly, the multivalent nature of protein cage architectures allows the
84
incorporation of multiple functionalities (cell specific targeting, therapeutic agent
delivery, and imaging) on a single platform.
The objective of this chapter is to demonstrate the ability of protein cage
architectures to limit the bioavailability of therapeutics housed within their
interiors and to describe the therapeutic efficacy of doxorubicin derivatized tumor
targeted protein cage architectures in vitro. Two genetic variants of the heat
shock protein (Hsp) cage were derivatized with doxorubicin; non-targeted
HspG41C and ‘tumor-targeted’ HspG41C-RGD4C. As described in the previous
chapter, HspG41C-RGD4C was genetically modified to present the αvβ3 and αvβ5
integrin targeting peptide RGD-4C (CDCRGDCFC) on its exterior. Angiogenic
blood vessels including those that feed tumors are known to express αvβ3 and
αvβ5 integrins and RGD-4C mediated tumor targeting has proven successful [24,
250, 263, 264, 273]. The ability of fluorescein labeled HspG41C-RGD4C protein
cages to bind αvβ3 expressing C32 melanoma cells in vitro was also described in
Chapter 3 [281].
Results and Discussion
As described in Chapter 3, the (6-maleimidocaproyl) hydrazone derivative of
doxorubicin [5] (Mal-Dox) was linked to the interior surface of the HspG41C
protein cage via coupling of the maleimide and thiol functionalities (schematic
Figure 3.1 page 49). The ability of this protein cage to limit the bioavailability of
doxorubicin was evaluated in cell culture assays described later in this chapter.
85
Derivatization of HspG41C-RGD4C with Doxorubicin
Utilizing a similar protocol as described for the conjugation of Mal-Dox to the
non-targeted HspG41C cage, Mal-Dox was conjugated to cysteines of the ‘tumor
targeting’ HspG41C-RGD4C protein cage (Materials and Methods). The best
reaction conditions determined to date are as follows: HspG41CRGD4-C cages
(2 mg/ml; 112 μM subunit) in pH 6.5 buffer (HEPES 100 mM, NaCl 50 mM)
reacted with a substoichiometric amount of the (6-maleimidocaproyl) hydrazone
of doxorubicin (0.5 molar equivalents per subunit) (56 μM) for 10 minutes at room
temperature. As described in Chapter 4, substoichiometric labeling ensured that
every cage displayed a significant fraction of unmodified cysteines within the
RGD-4C αvβ3 integrin targeting sequence. Centrifugation (13k rpm; 5 minutes) is
required to eliminate protein that has disassembled and / or aggregated during
the short course of the reaction. Derivatized HspG41C-RGD4C cages were
separated from free Mal-Dox by size exclusion chromatography (Superose 6,
Amersham-Pharmacia) (Figure 5.1). Co-elution of HspG41C-RGD4C protein
cages (280 nm) and doxorubicin (495 nm) after an elution volume of 12.5 ml –
17.5 ml indicates that doxorubicin is associated with the assembled HspG41CRGD4C cages (Figure 5.1). Both transmission electron microscopy (Figure 5.1)
and dynamic light scattering analysis confirm that HspG41CRGD-Dox maintains
the ~15 nm diameter of underivatized cage.
86
The covalent linkage of doxorubicin to HspG41C-RGD4C is demonstrated by
SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Fluorescence imaging of
an SDS-PAGE gel reveals a mobility shift of doxorubicin (excitation 488 nm;
emission 520 nm) bound to HspG41C-RGD4C compared to that of free
doxorubicin (Figure 5.2). Subsequent staining of the protein with Coomassie
blue illustrates the co-migration of the HspG41C-RGD4C subunit with
doxorubicin (fluorescent band) indicating their covalent linkage (Figure 5.2 Lane
4).
Absorbance (a.u.)
HspG41CRGD4C-Doxorubicin
Hsp Protein (280 nm)
Doxorubicin (495 nm)
300
200
100
50 nm
0
0
5
10
15
20
Elution Volume (mL)
25
30
Figure 5.1. Characterization of HspG41CRGD-4C Cages Derivatized with Doxorubicin
Size exclusion chromatography elution profile of HspG41CRGD-4C cages. Co-elution of the
absorbance at 495 nm (doxorubicin) and 280 nm (protein) illustrate that doxorubicin bound to
intact Hsp cages. TEM micrograph (inset) also demonstrates the presence of intact Hsp
protein cages.
87
Fluorescent
Image
1 2 3 4
20
Protein
Image 14.2
1 2 3 4
Figure 5.2. SDS-PAGE of HspG41CRGD-4C Subunits Demonstrates Covalent
Linkage of Doxorubicin
Lane 1- Molecular weight standard, Lane 2- HspG41C no experiment,
Lane 3- HspG41C-RGD4C no experiment, Lane 4- HspG41CRGD4C-Doxorubicin
The fluorescently imaged gel illustrates covalent linkage of doxorubicin (Lane 4) to
HspG41CRGD-4C monomers whereas unlabeled Hsp subunits exhibit no fluorescence
(Lanes 2 & 3). After fluorescent imaging, the proteins in the gel were stained with
Coomassie Brilliant blue. HspG41CRGD-4C,17816.6 Daltons (Lane 3) runs higher than
HspG41C,16498.2 Daltons (Lane 2) due to the incorporated RGD-4C peptide. The
molecular weight standard (Lane 1) indicates the relative positions of 20 and 14.2 kDa
proteins.
Quantitative analysis by absorbance spectroscopy indicates that there are 10
doxorubicins per 24 subunit cage (an average of 0.42 dox per subunit). Liquid
chromatography / electrospray mass spectrometry (LC/MS) analysis (Esquire
3000, Bruker) of doxorubicin labeled Hsp cages revealed that subunits contained
a range from 0 to 5 doxorubicins per subunit (Figure 5.3).
Derivatization of HspG41C-RGD4C with Mal-Dox proved more difficult than
labeling HspG41C (Chapter 3). In general, the HspG41C-RGD4C genetic variant
was less stable during the Mal-Dox derivatization. This result was unexpected,
since the HspG41C-RGD4C cages are stable to fluorescein-5-maleimide
(Chapter 4).
88
Intensity
0
1
2
Mass (Daltons)
Figure 5.3. Deconvoluted Mass spectrum HspG41CRGD-4C–Doxorubicin
Deconvoluted Mass spectrum HspG41CRGD-4C–Doxorubicin showing from 0 to 2
hydrazone linkers covalently attached to one subunit indicated by the number. The hydrazone
linker attached to the subunit, with the loss of doxorubicin, is expected since the liquid
chromatography conditions are acidic and the doxorubicin is connected through an acid labile
bond. The extra peaks are 2-mercaptoethanol adducts of the sample. These deconvoluted
data was produced from the raw data by the software MaxEnt1, (Waters).
Both reactions employ identical thiol-maleimide chemistry to covalently link
fluorescein and doxorubicin. In addition, the results from Mal-Dox, labeling
reactions of HspG41C-RGD4C were variable even under identical reaction
conditions. In general, the yield of HspG41C-RGD4C-Dox from various reaction
conditions increased as the molar equivalents of Mal-Dox per Hsp subunit were
decreased. But even with low concentrations of Mal-Dox some protein would
precipitate. Consequently, the yields of dox derivatized HspG41CRGD4C cage
were low (~50%). In spite of the difficulties, sufficient quantities of doxorubicin
derivatized ‘tumor targeted’ Hsp cages were produced for their evaluation in
mammalian cell culture. These results are described later within this chapter.
89
Evaluation of the Stability of HspG41C-RGD4C
The cage stability problems associated with HspG41CRGD4C-Dox reaction
pointed to the need to investigate the stability of both the derivatized and
underivatized HspG41C-RGD4C cages. These cages were routinely
characterized by size exclusion chromatography (SEC) and transmission
electron microscopy (TEM) following purification. But, it was also necessary to
evaluate protein cage stability in conditions similar to those required for
subsequent cell culture studies (pH 7.4; 37˚C). The HspG41C-RGD4C cage
stability at 37˚C in Dulbeco’s phosphate buffered saline (DPBS), pH 7.4, was
semi-quantitatively assessed at various time points by SEC. As stated
previously, size exclusion chromatography (SEC) enables one to discern the
nature of protein cage architectures based on their elution profile (large, intact
cages elute much earlier than subunits; aggregated proteins often have a highly
variable and dispersed elution profile). The SEC eluant is monitored at 280 nm,
therefore enabling protein quantification. This type of quantification is only as
reliable as the ability to load and inject an accurate amount of sample onto the
SEC column, thus there is sometimes variability in the results. In order to get an
accurate picture of cage stability from this study, all time points were analyzed 3
times in order to minimize variability. The stability of HspG41C-RGD4C cages
(37˚C, DPBS pH 7.4) was monitored by SEC hourly from 1 to 24 hours. The
results indicate that cages remain intact for the duration of the experiment
(Figure 5.4; representative data) [282]. The stability of the HspG41C-RGD4C
90
cages was also monitored at 4˚C. SEC data demonstrated that the cages could
be stored at 4˚C (DPBS pH 7.4) for as long as 7 days. HspG41C-RGD4C cages
are routinely stored at -80˚C as a precaution against aggregation observed by
transmission electron microscopy (TEM). SEC analysis of thawed cages
indicated that the cages withstood freezing and thawing.
A.
1 hr 37 deg
190.00
24 ht at 37 deg
135.00
115.00
95.00
75.00
55.00
35.00
15.00
-5.00
A b s o rb a n c e
140.00
A b s o rb a n c e
B.
90.00
40.00
0 .3 9
3 .5 4
6 .6 9
9 .8 3
1 2 .9 8
1 6 .1 3
1 9 .2 7
2 2 .4 2
2 5 .5 7
2 8 .7 1
3 1 .8 6
3 5 .0 1
3 8 .1 5
4 1 .3 0
4 4 .4 5
4 7 .5 9
5 0 .7 4
0 .0 0
3 .5 4
7 .0 8
1 0 .6 2
1 4 .1 6
1 7 .7 0
2 1 .2 4
2 4 .7 8
2 8 .3 2
3 1 .8 6
3 5 .4 0
3 8 .9 4
4 2 .4 8
4 6 .0 2
4 9 .5 6
5 3 .1 0
5 6 .6 4
-10.00
Time
Time
Figure 5.4. The stability of HspG41C-RGD4C (37˚C, DPBS pH 7.4)
The stability of HspG41C-RGD4C (37˚C, DPBS pH 7.4) at (A) 1 hour and (B) 24 hours,
demonstrated by size exclusion chromatography (SEC) elution profiles in which the protein
peak elutes at the time characteristic of intact Hsp cages. X-axis is time in minutes.
Studies of Protein Cage-Mammalian Cell Interactions in vitro
The ability to grow a variety of mammalian cell types in culture enabled the
investigation of interactions between protein cage architectures and mammalian
cells. These interactions include cell binding (as described in Chapter 4), subcellular localization, and cell proliferation in the context of therapeutic containing
Hsp cages.
91
The initial protein cage – cell interaction studies involved the use of
fluorescently labeled cages (HspG41C and CCMV S102C) and two readily
available mammalian cell lines (HeLa and CV-KL). HeLa cells are an epithelial
cell line derived from a human cervical carcinoma from a patient named Henrietta
Lacks. HeLa cells are commonly used for nanoparticle – cell interaction studies,
but epifluorescence analysis was difficult because of their small, compact cellular
morphology (Figure 5.5). Therefore, another cell line (CV-KL) with a more
elongated morphology was also utilized. The CV-KL cell line is a post-crisis
subline of CV-1 African green monkey kidney cells [283]. Cells were grown in 4well chamber slides or on glass coverslips in order to facilitate microscopic
observation of protein cage – cell interactions. Initial studies involved the
incubation of fluorescently labeled cages with cells in the presence and absence
of transfection agents. Transfection agents are synthetic molecules or polymers
that are designed to help shuttle proteins, DNA, and viruses into mammalian
cells. The use of several transfection agents including; poly-L-lysine, Transfectin® (Bio-Rad), Lipofectamine® (Invitrogen), and Pro-ject® (Pierce) were
explored [53, 284]. Most studies were analyzed by epifluorescence microscopy
and some using confocal microscopy. The results were variable, but in general it
seemed that the most cell associated fluorescence resulted from studies in which
cages were pre-incubated with poly-L-lysine prior to overnight incubation with
cells (37˚C; 5% CO2) (Materials and Methods). Punctate cell-associated
fluorescence was observed above the level seen with cells incubated with
92
fluorescein-5-maliemide-cysteine at equivalent fluorescein concentrations (Figure
5.5).
A. FITC filter
B. DAPI filter
C. Light
D. Control FITC filter
E.
E.
Figure 5.5. Epifluorescence Microscope Images of HeLa cells Transfected with
HspG41C-FlMal
A.-C. HeLa cells were transfected with HspG41C- FlMal (50 μg/ml), poly-L-lysine (1.5
μg/ml); the same frame viewed with the indicated filters. D. Control Image of Only HeLa
cells. E. Composite confocal microscope image of a 0.3μm ‘slice’ of Hela Cells transfected
with HspG41C-FlMal (green), fixed, and treated with Lyso-Tracker Red (Molecular Probes).
Yellow suggests co-localization of protein cages and lysosomes.
Subcellular Localization of Protein Cage Architectures
In order to try to verify that some fluorescently labeled cages were internalized
by cells, a commercially available lysosomal marker (LysoTracker Red,
Molecular Probes) was used to visualize lysosomes. Lysosomes are membrane
bound organelles within eukaryotic cells, containing digestive enzymes,
responsible for degradation of macromolecules and particles taken inside the
cells by endocytosis [244]. Co-localization (yellow) of cage associated green
fluorescence and LysoTracker red was observed by both epifluorescence and
confocal microscopy (Figures 5.5 & 5.6). Co-localization is difficult to prove with
epifluorescence microscopy because it provides a two dimensional image.
Overlapping signals could occur if one signal was from the cell exterior and the
93
other from its interior. Confocal microscopy address this problem via it’s ability to
image through a series of focal planes, enabling one to look at a cell in very thin
slices (0.3 μM) (Figure 5.5).
A.
B.
C.
Figure 5.6. Epifluorescence Microscope Images of HeLa cells Transfected with
CCMV- FlMal
HeLa cells were transfected with CCMV- FlMal (50 μg/ml), Poly-L-Lysine (1.5 μg/ml); the
same frame viewed with different filters. The fluorescent images were overlayed onto the
light Image. A. Lysosomes illuminated utilizing LysoTracker Red (Rhodamine filter).
B. Fluorescein labeled CCMV cages (FITC filter). C. Composite image of A & B; yellow is
indicative of areas that are both red and green
Both confocal and epifluorescence microscopy are useful for evaluating
protein cage internalization and localization. However, the data collected to date
are insufficient to make definitive statements regarding the cellular localization of
protein cages. Instead, the question of the ability of protein cages to enter
mammalian cells was addressed by studying a more downstream consequence
of protein cage entry, cell viability. For these studies, doxorubicin derivatized
cages were incubated with cells and the relative number of living cells was
assessed via the MTS cell proliferation assay (CellTiter96 Aqueous NonRadioactive Cell Proliferation Assay, Promega, Madison, WI).
94
Cell Proliferation Assay
The MTS cell proliferation assay is used to quantify viable cells (Materials and
Methods) [285, 286]. It is a colorimetric assay that can be performed in a 96-well
cell culture format. This facilitates simultaneous evaluation of multiple
experimental conditions. Briefly, the relative number of viable cells is determined
after addition of MTS (3-(4,5-dimethythiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2(4-sulfopheny)-2H-tetrazolium) to the growth medium followed by incubation for
1.5 hours (37˚C, 5% CO2). During the incubation, MTS is bioreduced to
formazan (Abs 490 nm) by dehydrogenase enzymes of metabolically active cells.
The absorbance at 490 nm is quantified and comparisons between the values
obtained from experimental conditions and controls (i.e., cells grown under
identical conditions and medium only) are made. There are many different types
of cell proliferation assays such as the sulforhodamine cell proliferation assay,
which quantifies the relative number of cells based on the amount of
sulforhodamine stain that binds the proteins of fixed adherent cells (Materials and
Methods). Each technique offers its own advantages and disadvantages but the
MTS assay was selected due to its ability to relatively quantify metabolically
active cells. In future studies, it might be advantageous to compare its results to
those obtained with other techniques.
The applicability of the MTS assay for the relative quantification of HeLa and
C32 melanoma cells was tested before its use in experiments. This was done by
seeding wells with a varying number of cells and evaluating their proliferation
95
after 2 days of growth with the MTS assay. Figures 5.7 and 5.8, demonstrate the
effectiveness of the MTS assay in determining the relative number of
metabolically active HeLa and C32 melanoma cells. In addition, the efficacy of
free doxorubicin to kill both HeLa and C32 melanoma cells grown in culture at
various concentrations, and two time points (4 or 24 hour incubations) was
assessed by the MTS assay (Figures 5.7 & 5.8). These studies found that a
greater amount of doxorubicin is needed to kill C32 melanoma cells than required
to kill HeLa cells. Together these data indicated that the MTS cell proliferation
assay was suitable for use in the investigation of the efficacy protein cage
mediated doxorubicin delivery.
4hrs Dox Only
O/N Dox Only
0.8
120
0.7
% Control Abs 490 nm
Absorbance 490nm
HeLa Cells + Doxorubicin
MTS Cell Proliferation Assay
B.
HeLa Cell Quantification
MTS Cell Proliferation Assay
A.
0.6
0.5
0.4
0.3
0.2
0.1
0
100
80
60
40
20
0
0
500
1000
1500
2000
# Cells Seeded
2500
3000
3500
0
200
400
Doxorubicin (nM)
Figure 5.7 MTS Assay of HeLa Cell Proliferation
A. Varying numbers of HeLa cells (x-axis) were seeded and grown in a 96-well plate.
Cell proliferation was assayed after 3 days by the MTS cell proliferation assay.
Absorbance at 490 nm (y-axis) is indicative of the number of viable cells since the MTS
precursor is bioreduced to formazan, with a characteristic absorbance at 490 nm, by
dehydrogenase enzymes of metabolically active cells.
B.
HeLa cells were exposed to doxorubicin for 4 hours (red) or overnight (~24 hrs, orange).
Cell proliferation was assayed after 2 days by the MTS cell proliferation assay.
The data were compared to the number of cells in the control wells (=100%).
600
96
C32 Cell Quantification
MTS Proliferation Assay
A.
Abs 490 nM
(Correlates with # of Cells)
0.6
0.5
0.4
0.3
0.2
0.1
0
0
5000
10000
15000
# of Cells Seeded
B.
C32 Cells + Doxorubicin
MTS Cell Proliferation Assay
C32 Cells + Doxorubicin
MTS Cell Proliferation Assay
Dox 24hrs
MalDox 24hrs
Dox 4hrs
MalDox 4hrs
0.8
0.8
0.7
0.7
Abs 490 nm (corrected)
Abs 490 nm (corrected)
Dox 4hrs
MalDox 4hrs
C.
0.6
0.5
0.4
0.3
0.2
0.1
Dox 24hrs
MalDox 24hrs
0.6
0.5
0.4
0.3
0.2
0.1
0
0
0
1000
2000
3000
4000
5000
1
Concentration (nM)
10
100
1000
10000
Concentration (nM)
Figure 5.8 C32 Melanoma MTS Cell Proliferation Assays
A. Varying numbers of C32 cells (x-axis) were seeded and grown in a 96-well plate.
Cell proliferation was assayed after 3 days by the MTS cell proliferation assay.
Absorbance at 490 nm (y-axis) is indicative of the number of viable cells since the MTS
precursor is bioreduced to formazan, with a characteristic absorbance at 490 nm, by
dehydrogenase enzymes of metabolically active cells.
B. C32 cells were exposed to doxorubicin for 4 hours (red) or 24 hrs (orange);
or the maleimide derivative of doxorubicin (Mal-Dox) for 4 hrs (dark purple) or 24 hrs
(purple). C32 cell proliferation was assayed after 3 days by the MTS cell proliferation
assay. The data was compared to the absorbance at 490 nm in the control wells (0
nm Dox).
C. Same data as in B., plotted on a log scale.
97
Efficacy of Doxorubicin Containing Hsp Cages
in vitro Non-targeted HspG41C-Dox_________
The first in vitro efficacy experiments were performed using doxorubicin
loaded HspG41C protein cages incubated with HeLa cells. Like the fluorescein
labeled cage experiments described above, Poly-L-Lysine was used as a
transfection agent. Assessment of cell viability 2 days after exposure to
doxorubicin housed within non-targeted HspG41C cages indicated that an
insufficient amount of doxorubicin was released to cause cell death (Figure 5.9).
These results demonstrated that the bioavailability of doxorubicin housed within
HeLa Cells + HspG41C-Dox, Dox, & Mal-Dox
MTS Cell Proliferation Assay
% Control (Abs 490 nm)
120
100
HspDox 4hrs
HspDox 16hrs
Doxorubicin 4hrs
Doxorubicin 16hrs
MalDox 4hrs
MalDox 16hrs
80
60
40
20
0
0
1000
2000
3000
4000
Doxorubicin in Overlay (nM )
Figure 5.9. Efficacy of HspG41C-Dox, Free Dox, and Mal-Dox Evaluated in HeLa Cells
Equally seeded HeLa cell culture plates were exposed to conditions in the legend including HspDoxorubicin in the presence of transfection agent, free doxorubicin (dox), and free Mal-Dox
compound. Subsequently, the MTS cell proliferation assay (Promega) was used to indicate the
number of viable cells compared to the number of cells in control wells (100%).
98
HspG41C protein cage architectures is restricted. The result is beneficial toward
their development as targeted therapeutic delivery systems because it is an
indication that therapeutic molecules housed within protein cages would remain
encapsulated until their release is selectively triggered. In addition, these results
underscored the need to incorporate specific targeting into protein cage
architectures, since transfection agents were unable to get a sufficient amount of
dox-cages into cells to achieve cancer cell killing and are not practical for in vivo
applications.
Tumor Targeted HspG41CRGD4C-Dox
As described in Chapter 4, the tumor targeting HspG41C-RGD4C protein cage
was shown to bind to αvβ3 integrin expressing C32 melanoma cells in vitro. This
was demonstrated by two methods. The first approach utilized epifluorescence
microscopy to probe the ability of fluorescein labeled HspG41C-RGD4C cages to
bind adherent C32 melanoma cells. The second evaluated their ability to bind
C32 cells in suspension by fluorescence activated cell sorting (FACS). The
interaction of HspG41C-RGD4C was also investigated in a reverse fashion,
testing the ability of C32 cells in suspension to bind protein cages adsorbed onto
a plastic Petri dish. With this approach, preferential binding of C32 cells to
HspG41C-RGD4C protein spots, as compared to controls, was observed. The
C32 cell adhesion assay was performed by spotting multiple proteins (HspG41CRGD4C, HspG41CRGD4C-Dox, HspG41C, BSA, and 2% FBS) onto bacterial
99
Petri dishes (see Materials & Methods) prior to flooding the dish with a C32
melanoma cell suspension [287]. Adherent C32 cells were visualized by
Coomassie Blue staining (Figure 5.10). Figure 5.10 shows increased Coomassie
staining of cells near the HspG41C-RGD4C protein spot as compared to
controls. This data indicates that C32 cells preferentially bound to HspG41CRGD4C cages and further supports the binding data presented in Chapter 4. It
also indicates that derivatization of HspG41C-RGD4C with doxorubicin doesn’t
eliminate its cell binding capability (Figure 5.10C).
B.
A.
HspG41C
HspG41C-RGD4C
HspG41C-RGD4C
C.
HspG41C
HspG41C-RGD4C
HspG41CRGD4C-Dox
DPBS +2% FBS
DPBS +
2% BSA
BSA
MEME +
10% FBS
Figure 5.10. Preferential Adherence of C32 Melanoma Cells to
HspG41C-RGD4C Protein Spots
Proteins including: HspG41C*, HspG41C-RGD4C*, HspG41CRGD4C-Dox*, bovine serum
albumin (BSA)*, and fetal bovine serum (FBS), were spotted onto bacterial Petri dishes
(*= 50 μL spots of 2mg/ml). The ability of C32 melanoma cells to adhere after a 16 hour
incubation was assessed by washing the dishes, fixing the cells, and staining them with
Coomassie blue.
After demonstrating that HspG41C-RGD4C cages could bind to target C32
melanoma cells (Chapter 4) and vice versa (above), the next step was to test the
ability of doxorubicin loaded HspG41C-RGD4C cages to kill C32 cells. The
efficacy of HspG41CRGD4C-Dox was evaluated using the MTS cell proliferation
100
assay described above (Materials & Methods). In brief, C32 cells were equally
seeded in a 96-well cell culture plate, allowed to grow for 48 hours, and then
exposed to HspG41CRGD4C-Dox, HspG41C-Dox, free Dox, or medium only
(control). The results from initial cell proliferation assays were variable; sometimes the HspG41CRGD4C-Dox cages seemed to kill cells whereas in some
assays they did not seem as promising (Figures 5.11 & 5.12). Data presented
here exemplify what has been seen to date; this line of investigation is ongoing.
C32 Melanoma MTS Cell Proliferation Assay
4hrs Hsp-Dox
4hrs Free Dox
4hrs Mal-Dox
16hrs Hsp-Dox
16hrs Free Dox
16hrs Mal-Dox
120
Cell Proliferation
C e l l P r o l i fe r a t i o n
(p(%
e r cof
e n Control)
t o f C o n tr o l )
Cell Proliferation
100
80
60
40
20
0
C e l l P (%
r o l i fe
ti o n (% o f C o n tr o l )
ofr aControl)
120
100
80
60
40
20
0
0
1000
2000
3000
4000
5000
0
)
Doxorubicin (nM
Doxorubicin
(nM)
500
1000
1500
2000
# of
# ofDelivery
Delivery EventsEvents
(a.u.)
Figure 5.11 C32 Melanoma Cell Proliferation Assay
A. C32 cells were exposed to ‘tumor targeted’ HspG41CRGD4C-Dox (Hsp-Dox in
legend), free dox, and Mal-dox, for either 4 or 16 hours. Cell proliferation was assayed
after 3 days by the MTS assay. The results are plotted as the percent of viable cells
compared to control (y-axis) vs. the concentration of doxorubicin (x-axis).
B. The same data from A are plotted as the percent of viable cells compared to control
(y-axis) versus the number of delivery events (x-axis). Since each HspG41CRGD4Cdox cage corresponds to 10 molecules of doxorubicin, for every cage delivery event 10
doxorubicins are delivered, whereas for free doxorubicin only a single doxorubicin
molecule is delivered per delivery event. Therefore to account for this discrepancy the
concentrations of free doxorubicin were divided by 10, in order to make a “per delivery
event” comparison.
2500
101
C32 Cells + HspG41CRGD-Dox,
Free Dox, & Mal-Dox
MTS Cell Proliferation Assay
4hrs Hsp-Dox
4hrs Free Dox
4hrs Free MalDox
24hrs Hsp-Dox
24hrs Free Dox
24hrs MalDox
0.5000
Abs 490 nM
(Indicative of Cell #)
0.4500
0.4000
0.3500
0.3000
0.2500
0.2000
0.1500
0.1000
0.0500
0.0000
0
200
400
600
800
1000
1200
Doxorubicin (nM)
Figure 5.12 C32 Melanoma Cell Proliferation Assay
C32 cells were exposed to ‘tumor targeted’ HspG41CRGD4C-Dox (Hsp-Dox in
legend), free dox, and mal-dox, for either 4 or 24 hours. Cell proliferation was
assayed after 3 days by the MTS assay. The results are plotted as the absorbance
at 490 nm (indicative of cell viability) versus the concentration of doxorubicin (x-axis).
As described previously, peptide mediated delivery of doxorubicin was more
efficacious than free doxorubicin in animal models [24], therefore doxorubicin
was selected as the first therapeutic molecule to test in protein cage mediated
delivery. The cell culture data from RGD-4C peptide mediated delivery were not
provided [24], but other studies utilizing nanoparticle drug delivery systems often
find that the delivery agents are not as effective as free drug in cell culture
assays [19]. Future HspG41CRGD4C-Dox efficacy studies in animal models
may provide more positive results. In addition, for future cell culture studies it
102
might be necessary to conjugate more doxorubicin or other more potent
therapeutic molecules to the protein cages. To date approximately twenty, 96well proliferation assay have been completed. These experiments utilized
concentrations of doxorubicin associated with HspG41C-RGD4C up to 4 μM; this
might need to be increased since appreciable C32 cell killing with free
doxorubicin requires 1 μM.
Subcellular Localization of Doxorubicin in vitro
Doxorubicin is known to traffic to the nucleus. Nuclear trafficking of
doxorubicin in C32 melanoma cells was verified by epifluorescence microscopy
(Figure 5.13). Since drug delivery systems have potential to alter therapeutic
trafficking, it is important to verify that doxorubicin delivered to C32 cells in
HspG41C-RGD4C cages also localizes to the nucleus. Provided that HspG41CRGD4C cages are taken into cells by receptor mediated endocytosis, nuclear
localization of doxorubicin would require its ‘endosomal escape’. Endosomal
escape is an important consideration in the development of most gene and
Figure 5.13 Nuclear Localization of Doxorubicin within C32 Melanoma Cells
C32 melanoma cells grown on glass coverslips were incubated for 30 minutes with free
doxorubicin (20 μM). Cells were subsequently washed, fixed and mounted. The cells
were imaged with epifluorescence microscopy with FITC filter [exposure time 200 ms]
(left) and with light (right); the middle panel is a composite of those 2 images.
103
therapeutic delivery systems [132]. In order to test potential endosomal escape
and nuclear localization of doxorubicin, C32 melanoma cells were incubated with
either free doxorubicin or doxorubicin within HspG41CRGD-4C cages and
imaged with epifluorescence microscopy (Materials and Methods). The
concentration of doxorubicin was normalized at concentrations sufficient to cause
cell death (500 nM or 1 μM). Unfortunately, the data were inconclusive. Free
doxorubicin is detectable in the nucleus, but although there may be a slight hint
of fluorescence in the nuclei of C32 cells exposed to HspG41CRGD4C-Dox, it is
not definitive (Figures 5.14 & 5.15).
A. HspG41CRGD4C-Dox
B. Free Doxorubicin
C. Medium Only - Control
Figure 5.14 Subcellular Localization of Doxorubicin within C32 Melanoma Cells
C32 Melanoma Cells were grown on glass coverslips, incubated with (A) HspG41CRGD4CDox, (B) free doxorubicin, (C) medium only; for A & B the concentration of doxorubicin was
normalized to 500 nM. After incubation the cells were washed, fixed, and mounted medium
with DAPI nuclear stain (blue). The cells were imaged with epifluorescence microscopy both
with FITC filter [exposure time 200 ms] (top) and DAPI filter as well as with light; the bottom
panel is a composite of those three images.
104
A. Free Doxorubicin
B. HspG41CRGD4C-Dox
50 μm
50 μm
50 μm
50 μm
50 μm
50 μm
Figure 5.15 Subcellular Localization of Doxorubicin within C32 Melanoma Cells
C32 melanoma cells grown on glass coverslips were incubated for 30 minutes with either
(A) free or (B) doxorubicin containing HspG41C-RGD4C cages. Doxorubicin concentration
was normalized to 1 μM. Cells were subsequently washed, fixed and mounted in medium
containing DAPI nuclear stain (blue). The cells were imaged with epifluorescence
microscopy both with FITC filter [exposure time 200 ms] (top) and DAPI filter as well as with
light (bottom); the middle panel is a composite of those 3 images.
105
Importantly, C32 cells alone did not exhibit autofluorescence with the
experimental exposure time (200 ms). This study is worth repeating, after the
C32 melanoma cell killing parameters are determined.
Summary
In this chapter, mammalian cell culture was used as a tool to investigate the
cell-protein cage (CCMV and Hsp) interaction via fluorescence and confocal
microscopy. These results underscored the importance cell targeting ligand
incorporation into protein cage architectures, since introduction of cages into
cells with transfection agents was only moderately successful and not practical
for most in vivo applications.
The development of the Hsp cage architecture for tumor specific delivery of
anticancer agents was also furthered by investigations with mammalian cells
grown in culture. The tumor targeted Hsp genetic variant, HspG41C-RGD4C,
which binds αvβ3 integrins present on angiogenic vascular endothelium of
tumors, was derivatized with doxorubicin (dox). The cell killing efficacy of
HspG41C-RGD4C-dox and non-targeted HspG41C-dox was compared to that of
free doxorubicin. The data obtained to date demonstrate that the bioavailability
of dox housed within protein cage architectures is limited and that the
HspG41CRGD4C-dox construct exhibits potential. The ability of
HspG41CRGD4C-dox to mediate cell killing both in vitro and in vivo requires
further investigation. Studies utilizing a higher concentration of doxorubicin
106
containing HspG41C-RGD4C should be performed and evaluated in the near
future. It is important to note that drug delivery systems may not out-compete
free drug in cell culture assays, but may prove superior in animal models. In
future experiments, the use of alternative therapeutics, such as singlet oxygen
generators or more potent cytotoxic agents such as monomethyl auristatin E
(MMAE) [216], might be required for protein cage mediated cancer cell killing.
107
CHAPTER 6
CHARACTERIZATION OF THE BIOCOMPATIBILITY OF PROTEIN CAGE
ARCHITECTURES IN MURINE MODELS: INITIAL TUMOR LOCALIZATION,
BIODISTRIBUTION, AND IMMUNE RESPONSE DATA
Abstract
Protein cage architectures, including the small heat shock protein (Hsp) from
Methanococcus jannaschii and the Cowpea chlorotic mottle viral (CCMV) capsid,
have the potential to serve as targeted therapeutic and imaging agent delivery
systems. These proteinaceous nano-architectures are genetically and
chemically malleable. They have been chemically derivatized with fluorescent
dye molecules and antibodies, and genetically altered to present site-specific
reactive groups and cell targeting peptides [39, 76, 281, 288]. In this chapter, we
provide initial in vivo characterization of these particles. The ‘tumor targeting’
HspG41C-RGD4C and HspG41C are shown to localize in the vasculature of a
human tumor xenograft mouse model. The overall biodistribution of radiolabeled
Hsp and CCMV in naïve and immunized mice provides us with useful data on the
trafficking of these proteinaceous nanoparticles in mice. In addition, initial
characterization of how the murine immune system responds to both Hsp and
CCMV provides insight on their biocompatibility. Together these initial in vivo
characterization data will prove useful in the design of protein cage architectures
for biomedical applications. Importantly, the work described herein also
108
addresses a growing concern regarding the potential toxicity of new biomedical
nanoplatforms.
Introduction
Current efforts to develop nanotechnologies to diagnose, image, treat, and
prevent diseases and disorders have provided the scientific community with
numerous biomedical nanoplatforms [18-20]. These nanoplatforms exhibit
potential in vitro but, in general their in vivo characterization has lagged behind.
At this time a high priority should be placed on the thorough understanding of
circulation, clearance half-life, stability, immunogenicity, and organ biodistribution
of proposed nano-container delivery vehicles [289].
The potential use of protein cage platforms as targeted therapeutic and
imaging agent delivery systems requires their in vivo characterization. Typically,
biomedically important molecules such as therapeutics are first tested in vitro
utilizing cell culture techniques, followed by investigation in animal models, and
finally in human trials. Experiments utilizing cell culture techniques to investigate
the cell-protein cage interaction and therapeutic efficacy were described in
Chapters 4 & 5. In parallel with on-going studies in mammalian cells,
experiments within animal models were initiated. The goals of these studies
were: (1) to test the capability of ‘tumor targeted’ HspG41C-RGD4C in vivo, (2) to
determine the bio-distribution of both Hsp and CCMV in both naïve and
immunized mice, and (3) to characterize the immune response generated against
109
protein cage architectures (Hsp & CCMV). In order to carry out these
experiments, two types of mouse models were required: a human tumor
xenograft mouse model in which human tumors are grown in mice lacking
immune systems, and in BALB/c mice, a widely used inbred strain with a normal
immune system.
Background Information on the Biocompatibility and/or
Biodistribution of other Nanometer Scale Delivery Systems
In general, research on nanoparticles has demonstrated that they may
preferentially localize to tumor vasculature due to the enhanced permeability and
retention (EPR) effect (described in Chapter 1) [19]. One of the most relevant
systems from which to gain an understanding of the biocompatibility of protein
cage architectures is the Cowpea mosaic viral (CPMV) platform (described in
Chapter 1). CPMV is also a 28 nm diameter icosahedral plant virus, but
importantly it does not have sequence homology with CCMV.
Initial investigation of the biocompatibility of CPMV and CCMV protein cage
architectures was in the context of their use as potential vaccine constructs.
Antigenic peptides were displayed on the surfaces of CPMV and CCMV, and the
degree of antibody response and/or vaccine efficacy (in the case of CPMV) was
determined. In the majority of these studies, viruses were administered in
conjunction with immune system boosting adjuvants (see Chapter 1).
110
Additional information on the biocompatibility of protein cage architectures was
more recently provided by Dr. M. Manchester’s Lab at the Scripps Research
Institute. Their work involves the investigation of CPMV as a platform for
biomedical applications including vascular imaging, vaccine generation, and
potentially as a therapeutic delivery system [157, 170]. One potential benefit of
the use of plant viral particles for vaccines is the ability to administer them
through food sources – “edible vaccines” [170]. Although, oral inoculation of
CPMV in the absence of adjuvant did not result in an appreciable immune
response, and pre-existing humoral immunity to CPMV did not alter trafficking of
orally delivered cages [170]. These data demonstrate that the route of antigen
delivery and context in which it is delivered have great effect on the immune
response to the same antigen (CPMV). The dose also has an affect on the
degree of immune response mounted against a particular antigen [169].
Investigation of the biodistribution of CPMV after oral or intravenous
inoculation demonstrated systemic distribution [170]. CPMV was found
throughout the mouse body including the spleen, kidney, liver, lung, stomach,
small intestine, lymph nodes, brain and bone marrow for several days following
inoculation [170]. The detection of CPMV genomic RNA by RT-PCR was
indicative of the presence of intact CPMV particles for up to 3 days after
inoculation; no CPMV was detected on days 5 or 7 post-inoculation [170]. Rae et
al. compared CPMV RNA levels in blood-containing organs and those that had
been perfused [170]. They found that there was less detectable CPMV RNA in
111
the gastrointestinal (GI) tract on day 2 and 3 post-inoculation in the perfused
animals, suggesting that the CPMV RNA that was detected in the GI tract was in
the blood [170]. Therefore CPMV remained in the circulation for up to 3 days
post-inoculation [170]. In a parallel study where mice were fed CPMV (1 mg), 2 of
3 mice had detectable CPMV RNA in their spleen, liver, lungs, stomach, ileum
and bone marrow and 1 out of 3 was CPMV RNA positive in the kidney,
duodenum, jejunum and brain [170]. In this study CPMV RNA was only detected
up to 24 hours after inoculation [170]. Since it is possible that the above studies
detect only free CPMV RNA and not actually intact viral particles, Rae et al. also
investigated the biodistribution of fluorescently (Oregon Green-488) labeled
CPMV particles.
Fluorescently labeled CPMV proved useful as intravital imaging agent in live
mice, chick embryos, and human fibrosarcoma (HT1080)-mediated tumor
angiogenesis within onplants on the chick embryo chorioallantoic membrane
(CAM) [157]. Studies of the biodistribution of Alexa Fluor 555 (A555) labeled
CPMV particles in adult mice, demonstrated that CPMV capsids remained in the
vasculature up to 72 hours after injection; they subsequently accumulated in the
liver and spleen [157]. No deleterious effects (thrombosis, malaise, death) were
observed in mice that received fluorescently labeled CPMV [157]. This result is
noteworthy, since pre-immunized mice responded poorly (labored breathing /
hypopnea, fatigue, ruffled fur) after an intravascular inoculation of 125I labeled
CCMV (described further in this Chapter) [290].
112
CPMV-A555 enabled the visualization and distinction of arteries and veins
within chick embryos up to 72 hours after injection [157]. In addition, this study
revealed that CPMV particles were endocytosed by vascular endothelial cells
and localized in perinuclear vesicles associated with lysosomes [157]. In
contrast, PEG-decorated CPMV particles were not internalized by human
umbilical vein endothelial cells (HUVEC) in culture [157]. PEGylation of CPMVFITC resulted in a decreased liver and spleen uptake as compared to CPMVFITC [157]. Prevention of internalization is a desirable property for imaging
agents, but obviously not for therapeutic delivery agents. Therefore in future
experiments utilizing PEGylated Hsp and CCMV (described in this Chapter) cell
specific targeting peptides will be incorporated into the PEG moiety [291].
Background Information Relevant to the
Characterization of the Immune Response to Protein Cage Architectures
One potential hurdle to the use of protein cage architectures for therapeutic
and imaging agent delivery is the mammalian immune system [169]. Obviously,
an adverse immune response to the protein cage architectures is undesirable for
their development as targeted therapeutic and imaging agent delivery vehicles. A
pilot study investigating the immunogenicity of CCMV was performed by Ligocyte
Pharmaceuticals (Bozeman, MT). This study revealed that there was increased
antibody production in mice inoculated with CCMV. The details of these studies
113
were unclear, therefore it was necessary to repeat them and perform initial
characterization of the immune response against Hsp cages.
Since the CCMV capsid and Hsp architecture are composed of proteins, an
antibody response was expected. Characterization of the degree of this
response was required as it may or may not affect either the animal models or
protein cages in a detrimental manner. For example, if the antibodies produced
against the protein cages do not alter their trafficking, then therapeutic and
imaging agents could still be delivered. On the other hand, some research has
focused on the use of these same protein cage architectures (CCMV, CPMV,
Hsp) to boost the production of antibodies (see Chapter 1).
Although protein cages elicit antibody responses, there are methods to
mitigate those responses. One such method, PEGylation, involves chemically
decorating protein cages with polyethylene glycol (PEG). PEGylation of CPMV
resulted in a more rapid decrease in the anti-CPMV antibody titer in mice
measured over a time course as compared to the persistence of the antibody
titers from mice inoculated with wt CCMV [47]. Interestingly, CPMV derivatized
with a fluorescein-PEG moiety was more successful than PEG alone in reducing
the anti-CPMV antibody titer as a whole [47]. The results described below using
PEGylated CCMV are more similar to the CPMV-fluorescein-PEG results.
114
Biodistribution of Other Types of Nanoparticles
In addition to previous work describing the biodistribution of CPMV, other
types of nanoscale therapeutic and imaging agent platforms have been
characterized. Direct comparisons between these studies and those described
herein are difficult, due to the different natures of the particles (i.e., micelles,
liposomes, magnetic particles). Some results of these studies are described in
this section in order provide background information and typical localization data
which can be loosely compared to the biodistribution data of protein cage
architectures presented in this chapter.
Radiolabeled nanoparticle data are typically presented in terms of percent
injected dose (ID) per gram of organ or tissue (%ID/g). For example, the
biodistribution of pancreatic cancer cell targeted versus non-targeted 111Inlabeled magnetic nanoparticles was investigated in mice. The results
demonstrate the ability of targeted nanoparticles to localize in cancer cells (3.7 +
0.14 %ID/g) as compared to non-targeted particles (2.2 + 0.39 %ID/g, P <
0.0001) [237]. These particles were similarly distributed in other organs,
including the liver (targeted 40.4 + 5.3 % ID/g, non-targeted 40.2 + 3.6 %ID/g)
[237].
In another study, the biodistribution of four different shell cross-linked
micelle derived nanoparticles (SCKs) was investigated [292]. In vivo tracking of
the cages was facilitated by copper-64 (64Cu) labeling. In general, it was found
that particle flexibility, composition, and size were important factors contributing
to their biodistribution [292]. SCKs with a polystyrene core showed the most
115
favorable biodistribution. They exhibited prolonged circulation and low liver
accumulation [292]. PEGylation (described below) of these particles did not
seem to appreciably alter their overall biodistribution [292]. In addition, smaller
particles ranging in size from 12 – 27 nm exhibited longer blood retention then
their larger, 33 - 41 nm counterparts. The amount of SCKs in the spleen and
kidney ranged from 3 - 15 %ID/g and 3-10 % ID/g respectively, depending on
particle composition [292]. Interestingly, shortly after administration, PEGdecorated SCKs were taken up into the lung, but these results were dependent
on the strain of rats utilized in addition to the type of SCK [292].
The data presented in this chapter focus on the in vivo characterization the
Cowpea chlorotic mottle virus (CCMV) capsid and a small heat shock protein
(Hsp) from the hyperthermophilic archaeon Methanococcus jannaschii. Data
presented on the biodistribution of CCMV and Hsp in both naïve and immunized
mice provide essential information for the further development of these
nanoscale architectures for biomedical applications. In addition, this work
characterizes the immune response generated against CCMV and Hsp in
BALB/c mice. Modulation of the immune response was demonstrated by
chemically modifying the protein cage architectures.
116
Results and Discussion
Localization of Hsp Cages within Tumor Vasculature [293]
There are many mouse models (syngeneic, genetically engineered, and
xenograft) in which to study cancer [263]. In general, spontaneous tumor models
are primarily used for the study of cancer biology, whereas transplanted animal
tumor models and human tumor xenograft models are used for the evaluation of
experimental therapeutics [28]. Human tumor xenografts have been valuable
tools for the study of anticancer drug development for over 25 years [294]. A
retrospective review conducted by the National Cancer Institute (NCI) for 39
agents for which Phase II activity data were available showed that, where
compounds were active in at least one-third of xenograft models tested, there
was a statistically significant correlation with activity in at least two human tumor
types [294, 295].
The commonly used MDA-MB-435 human breast cancer xenograft model was
selected as an ideal model system for many reasons including that it was used to
demonstrate the efficacy of RGD-4C peptide mediated delivery of doxorubicin
[183]. In addition, Dr. Ruoslahti’s lab provided the mouse model and the required
in vivo expertise for these studies [183]. Protein cage architectures exhibited
tumor localization in vivo (Figure 6.1).
The tumor homing capability of fluorescently labeled HspG41C-RGD4C cages
(tumor targeted) and HspG41C (non-targeted) was tested in mice bearing MDA-
117
MB-435 human breast cancer xenografts (Materials and Methods). Fluorescently
labeled Hsp cages were administered by intravenous injection into the tail vein
and allowed to circulate for 1 – 4 hours [8]. At designated time points,
intracardiac perfusion was used to clear the vasculature of remaining circulating
cages and the tumor and various control organs (brain, skin, kidney, lung, heart,
pancreas, liver, and spleen) were harvested and processed for histology.
Examination of these tissues by fluorescence microscopy determined protein
cage localization and homing capability. Interestingly, both genetic variants of
Hsp cages, HspG41C and HspG41C-RGD4C, home to tumor vasculature. It was
found that tumor specificity was best with a dose in the range of 0.25-0.5 mg
(630 μM – 1260 μM) of cage per animal (Figure 6.1). The pattern of nanoparticle
A.
B.
Figure 6.1. Tumor Vasculature Localization of Hsp
Epifluorescence images of tumor tissue slices from MDA-MB-435 xenograft breast
tumors from mice injected with either fluorescently labeled (A) tumor targeted
HspG41C-RGD4C or (B) HspG41C. Green = fluorescein; Blue = DAPI nuclear
distribution within the tumor is consistent with vascular localization. The particles
were also taken up by the reticuloendothelial system (RES) as expected (liver
118
and spleen), but did not accumulate significantly in any of the other control
organs tested (Figure 6.2). Particulate drug carriers often have increased
accumulation in liver, spleen, and bone marrow due to uptake in mononuclear
phagocytic system (MPS) cells. In many pre-clinical trials, this has not been a
big problem since MPS cells (such as Kupffer cells) are rapidly renewed.
In these preliminary experiments, the Hsp protein cages were differentially
labeled with fluorescein (HspG41C - 11 Fl per cage vs. HspG41C-RGD4C – 19.3
Fl per cage). For in vivo work, equivalent amounts of fluorescein were injected;
therefore more HspG41C cages were injected as compared to HspG41CRGD4C. This could have masked any small differences in RGD-4C peptide
mediated homing of HspG41C-RGD4C cages. Further experiments will be
performed in order to address this question. It is possible that the homing
Figure 6.2. HspG41C-RGD4C is Taken Up by RES System; Not Other Organs
Epifluorescent images of organ and tumor sections from mice bearing MDA-MB-435 xenograft
breast tumors which were injected with fluorescently labeled HspG41C-RGD4C (0.25 mg).
Green = fluorescein; Blue = DAPI nuclear stain.
119
peptide will confer advantages to HspG41C-RGD4C mediated therapeutic
delivery.
Tumor homing of Hsp cages might also be a consequence of their size.
There are many reports that particulate drug carriers (including nanoparticles)
are ‘passively targeted’ to tumor tissue by a mechanism known as the enhanced
permeability and retention (EPR) effect [19]. Passively targeted drug containing
nanoparticles extravasate from the blood and enter tumor tissue through gaps in
vascular endothelial cells [19, 50]. Therefore the EPR effect may account for the
fact that both targeted and non-targeted Hsp cages were found in tumors.
However, the Ruoslahti lab did not find passive targeting of < 10 nm nanometer
sized peptide derivatized quantum-dots (inorganic nanocrystals of ZnS and CdSe
that possess unique luminescent properties) [257]. The current data illustrating
the localization of Hsp cages within tumor vasculature is a significant step toward
their development as therapeutic and imaging agent delivery systems, but clearly
more investigation is warranted.
Biodistribution of Protein Cage Architectures (HspG41C, HspG41C-GFE, CCMV)
in Naïve and Immunized Mice [296]
One of the key requirements for biodistribution studies is the ability to quantify
the degree of localization in a wide variety of tissues. Toward that goal, the
ability to quantify fluorescently labeled cages in tissue homogenates was
evaluated, but proved unreliable. It was difficult to get consistent fluorescent
values from complex tissue samples, because the fluorescent signal could be
120
blocked by tissue. In a second approach, protein cage architectures were
radiolabeled with 125I (iodine), which is a gamma emitter and has a signal that
can penetrate through tissue. Protein radiolabeling is a common method for
studying biodistribution because it facilitates precise quantification of the signal
from each tissue without homogenization.
125
I is a gamma emitter with a half-life
(t1/2) of 60.14 days. It was selected for biodistribution studies of HspG41C,
CCMV K42R, and HspG41C-GFE a ‘lung targeting’ genetic variant described in
Chapter 7.
Iodination of Protein Cages (HspG41C, HspG41C-GFE, CCMV K42R)
125
I was incorporated into the tyrosines of HspG41C, HspG41C-GFE, and
CCMV K42R protein cage architectures (Figure 6.3). Iodination was performed
utilizing IODO-BEADS Iodination Reagent (Pierce) according to manufacturer’s
instructions (Materials and Methods). IODO-BEADS are non-porous polystyrene
beads (1/8’’ diameter) derivatized with N-chloro-benzenesulfonamide (sodium
salt).
The protocol was initially optimized utilizing “cold” NaI to facilitate
characterization of the iodinated protein cages by absorbance spectroscopy,
size exclusion chromatography (Superose 6, Amersham Pharmacia), dynamic
light scattering (Brookhaven 90Plus, Brookhaven, NY), and transmission electron
microscopy (TEM) (Leo 912 AB) (Figure 6.4) (Materials and Methods).
Characterization of the protein cages before and after iodination indicated that
121
A.
B.
Oxidation agent =
N-chloro-benzenesulfonamide (sodium salt)
Figure 6.3 Iodination Reaction
A. Reaction schematic of iodine incorporation into tyrosine
B. Cartoon of IODO-BEAD coated with the N-chloro-benzenesulfonamide oxidation
agent (Pierce).
they remained intact and tolerated the reaction conditions (Figure 6.4). A brief
description of the iodination protocol follows: more details can be found in
Appendix A (Materials and Methods). IODO-BEADs were prepared immediately
before the reaction and added to DPBS and carrier free Na125I (MP Biomedicals,
Inc.) Incubation of Na125I with IODO-BEADS results in oxidation of Na125I and in
turn the creation of electrophilic iodine species for incorporation into tyrosines of
protein cages added to the reaction vessel (Figure 6.3). After 15 minutes, the
protein cage solution was transferred into a new microfuge tube and
unincorporated Na125I was removed utilizing Zeba Desalt Spin Columns for two
rounds of purification (Pierce). Experiments utilizing cold NaI resulted in >90%
protein yield. The yields utilizing 125I were assumed to be equivalent.
122
Figure 6.4 Characterizations of HspG41C, HspG41C-GFE, and CCMV K42R
Before (A, C, E) and After (B, D, F) Iodination
A. HspG41C
Size Exclusion Chromatography
HspG41C
100
Relative Intensity
20
Absorbance 280 nm (a.u.)
Dynamic Light Scattering
15
10
5
TEM
HspG41C
12.7 nm
80
60
40
20
0
0
0
5
10
15
20
25
Elution Volume (mL)
0
30
20
40
60
Diameter (nm)
80
100
B. HspG41C-I
14
100
HspG41C-I
HspG41C-I
15.4 nm
Relative Intensity
Absorbance 280 nm (a.u.)
12
10
8
6
4
2
80
60
40
20
0
0
0
5
10
15
20
25
50 nm
0
30
20
C. HspG41C-GFE
Size Exclusion Chromatography
80
100
Dynamic Light Scattering
TEM
100
HspG41C-GFE
Relative Intensity
25
Absorbance 280 nm (a.u.)
60
Diameter (nm)
Elution Volume (mL)
30
40
20
15
10
HspG41C-GFE
14.4 nm
80
60
40
20
5
0
0
0
5
10
15
20
Elution Volume (mL)
25
30
50 nm
0
20
40
60
80
Diameter (nm)
100
123
Figure 6.4 – continued. Characterizations of HspG41C, HspG41C-GFE, and CCMV K42R
Before (A, C, E) and After (B, D, F) Iodination
D. HspG41C-GFE-I
14
100
HspG41C-GFE-I
Relative Intensity
Absorbance 280 nm (a.u.)
12
10
8
6
4
HspG41C-GFE-I
18 nm
80
60
40
20
2
50 nm
0
0
0
0
5
10
15
20
25
Elution Volume (mL)
E. CCMVK42R
Size Exclusion Chromatography
CCMV K42R
220
20
40
80
Diameter (nm)
100
Dynamic Light Scattering
Abs 280 nm
Abs 260 nm
100
200
Relative Intensity
160
140
120
100
80
60
40
TEM
CCMV K42R
29.9 nm
180
Absorbance (a.u.)
60
80
60
40
20
20
0
0
0
5
10
15
Elution Volume (mL)
20
25
0
30
20
40
60
80
Diameter (nm)
100
50 nm
F. CCMVK42R-I
CCMV K42R
Iodonated
140
Abs 280 nm
Abs 260 nm
100
Relative Intesity
Absorbance (a.u.)
120
100
80
60
40
CCMV K42R-I
30.7 nm
80
60
40
20
20
0
0
0
5
10
15
Elution Volume (mL)
20
25
30
0
20
40
60
80
Diameter (nm)
100
50 nm
124
Specific Activity of Radiolabeled Protein Cage Architectures
The specific activity of the 125I labeled protein cages was determined from
counts associated with the sample compared to a linear regression fit to the
counts from 125I standards (ranging from 0 – 1000 nCi) (x-axis) plotted with their
corresponding counts per minute (CPM) (y-axis) (Beckman Gamma 4000
January
11, 2005 I-125 Standard Curve
125I Standard
Curve
1600000.0
y = 1336.2x + 11204
1400000.0
1200000.0
CPM
1000000.0
800000.0
600000.0
400000.0
200000.0
0.0
0
200
400
600
800
1000 1200
nCi
Figure 6.5. 125I Standard Curve
A standard curve was made from dilutions of 125I source (MP Biomedicals) and gamma
counts were quantified on Beckman Counter; curves of this type we used to estimate the
specific activity of radiolabeled protein cages.
Counting System; Materials and Methods) (Figure 6.5). Typically, specific
activity for Hsp cage samples was ~20 nCi/μg and ~250 nCi/μg for CCMV
samples (Table 6.1). The incorporation of 125I, in terms of specific activity
(nCi/μg), was about 10-fold greater in CCMV as compared to Hsp cages. This
125
discrepancy is most likely because CCMV has more surface accessible tyrosines
(5 per subunit; 900 per cage; exterior Tyr 50, 157, 159, 190; interior Tyr 138)
than Hsp (2 per subunit; 48 per cage; interior Tyr 96, subunit interface Tyr 106
subunit) (Figure 6.6).
Table 6.1 Summary of 125I Labeling Reaction Results
DATE
REACTION CONDITIONS
A – # beads : mCi 125I
B – mCi 125I : mg protein
11/2/05
A – 1 bead : 0.5 mCi
B – 1 mCi : 1 mg
HspG41C
5.0
11/2/05
A – 1 bead : 0.5 mCi
B – 1 mCi : 1 mg
HspG41C-GFE
4.4
11/8/05
A – 2 beads : 0.5 mCi
B – 1 mCi : 0.5 mg
HspG41C
11/8/05
A – 2 beads : 0.5 mCi
B – 1 mCi : 0.5 mg
HspG41C-GFE
11/8/05
A – 2 beads : 0.5 mCi
B – 1 mCi : 0.1 mg
HspG41C
40.0
11/8/05
A – 2 beads : 0.5 mCi
B – 1 mCi : 0.1 mg
HspG41C-GFE
23.5
12/14/05
A – 2 beads : 0.5 mCi
B – 1 mCi : 0.5 mg
HspG41C-GFE
22.6
12/14/05
Iodogen Tube Reaction
HspG41C-GFE
2.0
1/10/06
A – 2 beads : 0.5 mCi
B – 1 mCi : 0.5 mg
CCMV K42R
298.0
1/10/06
A – 2 beads : 0.5 mCi
B – 1 mCi : 0.5 mg
CCMV K42R
232.0
1/10/06
Iodogen Tube Reaction
CCMV K42R
258.0
3/1/06
A – 4 beads : 0.5 mCi
B – 1 mCi : 2 mg
HspG41C
32.5
3/1/06
A – 2 beads : 0.5 mCi
B - ? mCi (rinsed 125I vial) : 0.5 mg
CCMV K42R
62.1
PROTEIN
CAGE
SPECIFIC
ACTIVITY
(nCi/μg)
18.6
7.5
126
Hsp
exterior
interior
CCMV
exterior
interior
Figure 6.6. Differential Presentation of Tyrosines (red)
on the Hsp and CCMV Protein Cage Architectures
Percent 125I Incorporation Studies
As described above, free 125I was separated from radio-labeled proteins by
two rounds of desalting column purification (in some cases 3 rounds of ‘clean-up’
were performed). According to the manufacture’s specifications, this degree of
‘clean-up’ should have been enough to remove all free 125I. Two different
strategies were used in attempt to verify that all the gamma counts from the
radio-labeled protein samples were due to incorporated 125I and not free 125I. In
the first strategy, radio-labeled protein cages were spotted onto nitrocellulose
paper followed by washing with DPBS. The number of counts associated with
the nitrocellulose before and after a series of washes was monitored. This
approach did not seem to work. In the first trial, four different pieces of
nitrocellulose were spotted with samples taken at different time points (A- prior to
127
addition of protein; 1– at time = 0 just after protein was added; 2– after 12
minutes; and 3– after protein had been ‘cleaned up’ with de-salting columns).
Free 125I should have washed away while radio-labeled protein should have
bound to the nitrocellulose. After washing (5 x 50 ml) the counts from samples A,
1, or 2 were not close to ‘background’ levels. The counts kept decreasing with
more washes (and based on the volume of waste generated and the time
required, further washes were not performed). The ‘cleaned-up’ samples (3) had
variable results; the signal decreased about 10-fold for the HspG41C sample
after washing whereas the HspG41C-GFE sample decreased only 3.6% after
washing, but it started lower. Based on those results, the radio-labeled HspG41C
sample was subjected to another round of desalting column ‘clean-up’ because
of the possibility that free 125I had been inadvertently introduced into the purified
sample tube. After the 3rd round of column clean-up the HspG41C data were
more similar to that of the HspG41C-GFE data (Table 6.1). This initial labeling
experiment was a learning experience; conclusions based solely on these results
are not warranted.
In the following experiment, a modified version of this protocol was utilized.
Only the ‘cleaned up’ radiolabeled protein samples were spotted onto
nitrocellulose and subsequently washed and counted. The counts associated
with the HspG41C preparation spots on nitrocellulose decreased slightly after
washing whereas the nitrocellulose associated counts for the HspG41C-GFE
preparation actually increased. These results raised some concerns regarding
128
these data including the possibility that the protein cages (which were intact and
not denatured) did not bind to nitrocellulose well and/or that an insufficient
amount of time was allotted for the protein spots to dry onto the nitrocellulose
paper. This method was abandoned for another which took advantage of the
ability to precipitate proteins with 10% trichloroacetic acid (TCA).
In order to precipitate protein cages with TCA, equal volumes of radiolabeled
protein solution and 20% TCA (5 μL each) were mixed and incubated on ice for 1
hour. The protein was pelleted by centrifugation (13,000 rpm for 5 minutes) and
the number of gamma counts associated with the pellet, supernatant and one
wash (15 μL 10% TCA) was determined. This analysis revealed that for the Hsp
samples the majority of the counts (76.4%) were associated with the protein
pellet. Unfortunately, the CCMV data are less clear because although several
different TCA precipitation protocols were attempted, consistent precipitation was
not obtained; usually less than 50% of the protein would precipitate as
determined by SDS-PAGE gels of parallel ‘cold’ cage experiments (Figure 6.7).
TCA precipitation of radio-labeled CCMV resulted in 53.5% of the counts in the
pellet. Based on this data, it seems that the majority of 125I in the radio-labeled
protein preparations is incorporated into the protein, but the possibility of free 125I
in these preparations can not be ruled out (Table 6.2).
129
Sample
125I
% cpm
Sample
125I
% cpm
Hsp Pellet
76.4
CCMV Pellet
53.5
Hsp Sup
21.6
CCMV Sup
43.3
Hsp Wash
2.0
CCMV Wash
3.2
Table 6.2. TCA Precipitation Data from 125I Labeled Hsp and CCMV
Radiolabeled Hsp and CCMV protein cages were TCA precipitated as described
in the text. The gamma counts associated with the pellet, supernatant (sup), and
wash are expressed as a percent of the total counts of either the Hsp or CCMV sample.
In addition, trial iodination experiments utilizing IODO-GEN (1,3,4,6tetrachloro-3α,6α-diphenylglycouril) Pre-Coated Iodination Tubes (Pierce)
resulted in ~10-fold less iodine incorporation in Hsp cages, but resulted in
similar levels of 125I incorporation, as IODO-BEAD reactions, in CCMV cages
(Table6.1). Both Hsp and CCMV cages were less stable in IODO-GEN Tube
reactions, tolerating reaction times of < 10 minutes before cage disassembly.
Lane 16: CCMV wash (10 of 50ul)
Lane 15: CCMV sup
Lane 14: CCMV pellet
Lane 13: Hsp wash
Lane 12: Hsp sup
Lane 11: Hsp pellet
Lane 10: CCMV No Exp
Land 9: Ladder
Lane 8: Hsp No Exp
Lane 7: CCMV wash (10 ul of 50)
Lane 6: CCMV sup
Lane 5: CCMV pellet
Lane 4: Hsp wash 1
Lane 3: Hsp sup
Lane 2: Hsp pellet
Lane 1: Ladder
130
A.
B.
Figure 6.7. SDS-PAGE Analysis of TCA Precipitation of CCMV and Hsp
Proteins were precipitated by either Protocol 1 (lanes labeled green): 5 μl protein (1mg/ml)
sample + 5 μl 20% TCA, on ice for 30 min., spun 5 min 13k rpm, washed with 15 μl 10%
TCA, or by Protocol 2 (black): 5 μl protein (1mg/ml) sample + 5 μL 20% TCA, on ice for 30
min., spun 5 min 13k rpm, washed with 50 μl 10% TCA. The pellet, supernatant (sup), and
wash were analyzed by SDS-PAGE; the lanes are labeled accordingly. Two representative
gels in which the proteins are stained with (A) Coomassie Flour Orange (Molecular Probes;
UV light image) and (B) Coomassie Blue. Molecular weight ladder from top to bottom 66,
45, 36, 29, 24, 20, 14.2, and 6.5 kDa.
131
Biodistribution of 125I Labeled Protein Cages (HspG41C,
HspG41C-GFE, and CCMV K42R) in Naïve and Immunized Mice [296]
Pilot Study Results – Biodistribution of ‘Lung Targeted’ HspG41C-GFE vs.
Non-Targeted HspG41C in Naïve Mice. In pilot studies utilizing a small number
of mice (3-4 per experimental condition), the route of administration, dosage, and
lung specific targeting capability of HspG41C-GFE (described in Chapter 7) were
examined. The biodistribution of protein cages was determined from gamma
counts of organs and tissues removed from sacrificed mice. The list of organs
and tissues from which individual count data were obtained includes bladder,
thyroid, liver, lungs, salivary gland, spleen, kidney, gastrointestinal tract (GI),
blood, tail, tegument, heart, head, reproductive, olfactory bulbs, and brain. The
results are reported in multiple formats: (1) as a percent of the counts per minute
(CPM) per gram of tissue (%CPM/gm); (2) as a percent of injected dose (%ID)
and (3) as percent injected dose per gram of tissue (% ID/gram). Data from the
pilot studies indicated that a dose of 50 μg protein (0.5 μCi 125I) was sufficient for
obtaining counts from a variety of organs; therefore this dose was used in
subsequent experiments. The pilot studies also revealed a difference in the
trafficking of cages dependent on their route of administration intranasal: (i.n.)
versus intravascular (i.v.). For example, more cages were found in the lung
following i.n. inoculation compared to i.v.
132
The i.v. route of administration was used for all further investigations. Initial
data suggested that the ‘lung targeted’ HspG41C-GFE cages homed to the lung,
since ~ 2-fold more counts were found in the lung tissues of mice given
HspG41C-GFE as compared to the control cages HspG41C (Figure 6.8).
B.
7
5
6
%CPM/gm of tissue
%CPM/gm of tissue
A.
5
4
3
2
1
0
4
3
2
1
0
0
10
20
Hours
30
0
10
20
30
Hours
Figure 6.8. Lung Localization Data of HspG41C-GFE
Data graphed as percent 125I counts per minute (CPM) per gram of lung tissue (% CPM/gm) at
1 and 24 hour time points. Lung targeted HspG41C-GFE (red); non-targeted HspG41C
(green). (A) In the first study, HspG41C-GFE exhibited increased lung localization, (B) but in
the second study the opposite results were obtained, namely more HspG41C in the lungs.
Unfortunately, a follow up study did not concur with the previous results (Figure
6.8). This could be due to differential presentation of the GFE peptide on the
Hsp cage architecture as compared to its presentation on the M13 phage.
Although, Ackerman et al.[257] reported successful in vivo targeting of quantum
dots conjugated to GFE peptides in BALB/c mice, but this study differed from the
one described above: (1) different nanoparticles (quantum dots vs. protein
cages), (2) time of organ harvest (after 5 minutes in [257] vs. after 1 hour), and in
the (3) method of detection (epifluorescence microscopy in [257] and by
133
radioactive signal above). At this time, the focus of the initial biodistribution
studies was shifted from targeted vs. non-targeted Hsp cages, to a comparison of
the biodistribution of two different protein cage architectures, Hsp and CCMV.
Biodistribution Different Protein Cage Architectures (HspG41C & CCMV
K42R) in Naïve and Immunized Mice. As described in Chapter 1, the Hsp and
CCMV protein cage platforms, while similar, do have distinct differences in size
(Hsp 12 nm diameter; CCMV 28 nm diameter) and amino acid composition.
The biodistribution of 125I labeled HspG41C and CCMV K42R cages was
determined at 1 and 24 hour time points following i.v. injection into BALB/c mice.
The doses of HspG41C and CCMV K42R cages were 50 μg protein (0.5 μCi
125
I). For CCMV experiments, 125I cages were combined with non-radioactive
cages to make a 50 μg protein (0.5 μCi) dose. Normalization of the cage dose to
50 μg corresponds to differing number of Hsp and CCMV cages, 7.6 x 1013 and
8.227 x 1012 cages, respectively.
Ideally, therapeutic and imaging agent delivery systems could be administered
multiple times without altering their biodistribution and effectiveness. As
described later in this chapter, antibodies specific to Hsp and CCMV were
generated in mice that received multiple doses of cage. Therefore, it was
important to determine whether these antibodies altered the biodistribution of
subsequently administered 125I-labeled cages. Mice were injected with two 50 μg
doses of either Hsp or CCMV separated by a two week time period. One week
after their second “cold” cage dose, 125I-labeled cages were administered. The
134
biodistribution of protein cages was determined from gamma counts of organs
and tissues removed from sacrificed mice. The results are reported and
discussed in the following formats: (#1) as a percent of the injected dose per
gram of tissue (%ID/g), (#2) as a percent of the injected dose per organ
(%ID/organ), and (#3) as percent injected dose (% ID) (Figures 6.8 & 6.9; Table
6.3). The values reported in Table 6.3 as %ID/g, were calculated utilizing the
following equation:
%ID/g = [(cpm in tissue – background cpm) / organ weight in grams] x 100
(cpm in dose – background cpm)
Other formats of numerical reporting, including normalization of the cpm obtained
from organs to a given weight, are currently being evaluated. Regardless of
format, the numerical results are more meaningful when direct comparisons are
made between various tissue types, mouse immune status, and cage type as
described below and in Table 6.3. The total recovered dose from all experiments
ranged from 64% to 89%.
In general, the two cages had similar biodistribution profiles one hour after
injection (Figure 6.9). Notably, the majority of 125I-cage dose was excreted in the
urine by 24 hours post-administration (Figure 6.10C). The total percent of
injected dose recovered from the feces and urine combined ranged from 56.6%
to 63.3%. The %ID recovered in the urine alone ranged from 51-60%.
Interestingly, an analysis of the RNA viral community isolated from human feces
found that most (97.1%) of the sequences corresponded to plant viruses [297].
135
Of those viruses, the most abundant was the pepper mottle virus (PMMV), which
remained infectious after passing through the human digestive system [297].
A.
CCMV K42R
Percent Injected Dose/Gram Tissue
40
Naive 1hr
Naive 24hr
Immunized 1hr
Immunized 24hr
30
20
10
y
le
en
Sa
l
iv
ar
Sp
nd
G
la
Lu
ng
ve
r
K
Li
id
ne
y
rt
H
ea
ct
G
It
ra
n
B
ra
i
B
lo
od
0
B.
50
Naive 1hr
Naive 24hr
Immunized 1hr
Immunized 24hr
40
30
20
10
Sp
le
en
G
la
nd
Lu
ng
Sa
liv
ar
y
Li
ve
r
K
id
ne
y
H
ea
rt
G
It
ra
ct
B
ra
in
0
B
lo
od
Percent Injected Dose/Gram Tissue
Hsp G41C
Figure 6.9. Selected Organ Distribution of 125I-Labeled Cage.
136
Table 6.3. Percent of Injected Dose per Gram of Tissue (%ID/g); T-test Results.
Cages
Tissue
(Mean
± SEM)
Naïve
Liver
Spleen
Thyroid
Lung
Salivary
Gland
Feces &
Urine
CCMV K42R
(n = 8)
1 hr
Immunized
45.73
±4.45
13.95
±3.05
37.52
±11.62
30.78
±6.1
4.43
±0.5
16.89
±4.01
Pvalue
0.0008
0.79
0.006
0.0009
0.007
0.42
Naïve
24 hr
Immunized
0.59
±0.07
1.12
±0.33
483.0
±184.7
0.13
±0.02
0.19
±0.01
.1425
±0.05
66.28
±2.3
1.05
±0.17
0.99
±0.08
76.39
±17.13
0.24
±0.03
0.17
±0.01
0.17
±0.02
56.63
±2.2
Pvalue
0.03
0.68
0.04
0.008
0.18
0.58
Naïve
1 hr
Immunized
22.79
±2.63
14.5
±1.65
193.0
±58.52
4.76
±0.62
9.61
±1.15
12.36
±1.89
36.81
±1.8
1.11
±1.28
8.66
±1.09
6.82
±0.42
6.21
±0.25
11.88
±0.77
Pvalue
0.0006
0.05
0.02
0.02
0.01
0.82
0.09
Naïve
24 hr
Immunized
5.07
±0.48
2.45
±0.28
126.4
±25.63
0.37
±0.15
1.44
±0.14
0.28
±0.04
63.31
±3.5
3.94
±0.17
1.49
±0.1
81.61
±16.4
0.15
±0.01
0.43
±0.01
0.12
±0.02
59.26
±2.3
Pvalue
0.05
0.006
0.16
0.16
<0.0001
0.004
0.09
P – values
1 hr
24 hr
Tissue
Hsp Naïve
vs.
CCMV Naïve
Hsp Immunized
vs.
CCMV Immunized
Hsp Naïve
vs.
CCMV Naïve
Hsp Immunized
vs.
CCMV Immunized
Spleen
0.86
0.26
0.008
0.002
Liver
0.64
0.08
< 0.0001
< 0.0001
Thyroid
0.36
0.03
0.06
0.83
Lung
0.67
0.002
0.16
0.02
Kidney
0.13
0.007
< 0.0001
< 0.0001
Salivary Gland
0.02
0.24
0.08
0.09
0.09
0.09
Urine & Fecesa
a
Urine & Feces were analyzed as percentage of injected dose.
135
Kidney
20.56
±3.89
15.02
±2.37
132.6
±26.99
5.09
±0.41
7.34
±0.79
22.04
±2.96
% of
dose
Hsp G41C
(n = 7 for naïve 24 hr mice; n = 8 for immunized mice;
Salivary gland n = 4 for naïve 1 & 24 hr mice)
137
A.
Percent Injected Dose/Organ
40
Naive 1hr
Naive 24hr
Immunized 1hr
30
Immunized 24hr
20
10
0
Liver
Thyroid
Sal Gland
Kidney
Spleen
Lung
GI Tract
Percent Injected Dose/ Organ
B.
40
Naive 1hr
Naive 24hr
Immunized 1hr
30
Immunized 24hr
20
10
0
Liver
Thyroid
Sal Gland
Kidney
Spleen
Lung
GI Tract
Percent of Injected Dose
C.
100
Naive Hsp
75
50
Naive CCMV
Immunized Hsp
Immunized CCMV
25
0
Figure 6.10. Percentage of Dose per Organ in Selected Tissues.
A. 125I-labeled CCMV K42R, B. 125I-labeled Hsp G41C, C. Total %ID re-covered from feces and
urine combined. No significant difference in excretion was found for cage type (P > 0.05).
138
A comparison between the biodistribution of cages in naïve versus immunized
mice showed that, in general, the trafficking of the cages was not dramatically
altered at the one hour time point (Figures 6.9 & 6.10; Table 6.3). There were a
few exceptions, including the increase of CCMV in the kidney, liver, and thyroid
of CCMV immunized mice compared to that found in naïve mice, and an
increased amount of Hsp found in the lungs of immunized mice as compared to
that found in naïve mice (Figures 6.9 & 6.10; Table 6.3). A comparison of the
biodistribution of CCMV and Hsp after one hour illustrates that the average
%ID/g found in the lungs, thyroids, and kidneys of CCMV dosed mice compared
to Hsp immunized mice differed significantly (P < 0.05; Table 6.3). The
deposition of Hsp in the lung of immunized mice at the 1 hour time point was 4.5
times greater than CCMV immunized mice (30.8 %ID/g vs. 6.8 %ID/g). The
%ID/g found within the kidney was 1.4 times greater in CCMV immunized mice
compared to the amount found in Hsp immunized mice at 1 hour (Table 6.3).
The amounts of Hsp and CCMV in the liver and spleen of immunized and naïve
mice did not significantly differ in terms of %ID/g of tissue at 1 hour. However, as
noted above, at 24 hours post injection, mice immunized with CCMV cage had
higher %ID/g in the liver, spleen, lung and kidney whereas Hsp cages were found
more prevalently in the lungs of immunized mice (Table 6.3).
In general, the dispersion of cage in terms of %ID per organ in both treatments
(Hsp and CCMV) was similar in the immunized and naïve mice in the tested
organs except for the salivary gland and lung (Figure 6.10; Table 6.3). In the
139
salivary glands of CCMV mice at 1 hour, 54% more of the injected cage was
distributed in the glands of immunized mice than in naïve (2.15 ±0.34 % vs. 1.4
±0.34%). In contrast, the Hsp immunized mice had 62% less deposition of dose
in the salivary gland than naïve mice at 1 hour. There is a difference in the
amount of cage, Hsp vs. CCMV, found in the salivary gland of naïve mice at 1
hour (P < 0.05; Table 6.3), but not in the salivary glands of immunized mice. In
the lung, Hsp immunized mice had a higher %ID than CCMV immunized mice
(Table 6.3).
Naïve mice injected with CCMV had 2.3 times greater %ID in the GI tract as
compared to immunized mice (15.55 ±7.95% vs. 6.76 ±0.81%), and there was
64% and 60% more CCMV injected cage respectively in the liver and lungs of
immunized mice versus naïve at the 1 hour time point (33.8 ±3.85% vs. 20.61
±4.86% and 1.35 ±0.32% vs. 0.84 ±.27%, respectively). The thyroid had the
second highest %ID for both Hsp and CCMV, naïve and immunized mice. The
%ID in the livers of mice from both types of cage and immune status was the
third highest location of dose at 24 hours, but was as low as 0.67 %ID in Hsp
injected mice. All other tissues were less than 1% of the total injected dose.
Also noted was that the %ID in the brain was 0% at 24 hours and never more
than 0.15% at 1 hour (range 0.08-0.15%).
Biodistribution of Free 125I in Mice. Previous work utilizing 125I in mice indicate
that free iodine traffics to the thyroid and stomach [298-300]. Since the presence
of free 125I in the radiolabeled cage preparations was not completely be ruled out,
140
it was important to investigate the biodistribution of free 125I in BALB/c mice using
identical methods to those used in the protein cage biodistribution. The results of
this study indicate that at 1 hour post i.v. injection, the majority of 125I is found in
the tegument (24.6 %ID) and gastrointestinal tract (16.44 %ID). The thyroid
contained 4.33 %ID and the bladder 2.4 %ID. Since these organs are very small
this number is low, but when the counts per minute (CPM) values are presented
on a per gram of tissue basis (see formula described above), the thyroid and
bladder had the highest values (221.30 %CPM/gm and 208.93 %CPM/gm
respectively). These values exceed 100%, due to the manner in which they were
calculated, but these results clearly demonstrate that thyroid had the highest 125I
load on a gram of tissue basis. In general, the pattern of radioactivity associated
with the other organs differed from that obtained from the 125I labeled cage
experiments. This experimental result lends validity to the previous experiments
analyzing the biodistribution of 125I labeled cages. It indicates that those
experiments were indeed tracking 125I labeled cage and not just free 125I.
The only value that was similar between the free 125I and the 125I-labeled cage
biodistribution studies was the amount of radioactive signal excreted within the
first 24 hours post-injection (63.72 %ID). At 24 hours post-injection the thyroid
contained the majority of the organ associated counts, all of the other organs
contained between 0 – 0.14 %ID (excepting the tegument which contained 0.86
%ID).
141
State of Excreted Cages – Intact or Not? Since the majority of 125I signal was
excreted after 24 hours, experiments were designed in order to determine the
state of the excreted cages (i.e., were they intact or degraded). There are more
possibilities for analyzing the state of the CCMV protein cage architectures since
the CCMV used for biodistribution studies was isolated from plants and in turn
contained its RNA genome. The premise of the CCMV experiments was that
identification of CCMV RNA in mouse urine samples would be indicative of the
presence of intact cages. This was based on the idea that RNA is relatively
fragile in vivo and would presumably be degraded by RNAse enzymes in its free
state, but not if it were housed within CCMV viral capsids. RNA dependent
studies included: RT-PCR analysis of mouse urine after CCMV injection and
attempts to generate a CCMV infection in cowpea plants from samples prepared
from mouse urine. These studies are currently underway. Similar strategies
were also utilized by Rae et al. for the detection of CPMV virus in tissue (liver
and spleen) homogenates. Although they obtained positive results from CPMVspecific RT-PCR, they were unable to initiate cowpea plant infection with these
samples [170]. The inability of CPMV isolated from mice to cause infection was
attributed to virus inactivation by components within the circulation [170]. They
demonstrated that pre-incubation of CPMV with either plasma or serum
significantly decreased the degree of infectivity [170].
Another avenue for exploring the nature of protein cage architectures
excreted by mice was to use size selective filters designed to prevent the
142
passage of proteins of a given size. These types of filters are routinely used for
concentrating intact protein cage architectures. The molecular weight of intact
CCMV cages is 3600 kDa and 396 kDa for Hsp cages. Therefore, a filter with a
molecular weight cut off (MWCO) of 100 kDa should prevent the passage of
intact cages while allowing free 125I, protein cage subunits (CCMV 20.3 kDa; Hsp
16.5 kDa), and further degraded protein fragments to pass through. The results
from this study suggest that the majority of Hsp and CCMV protein cage
architectures were retained by the filters, suggesting that the cages remain intact
(Table 6.4). Urine samples from mice injected with 125I-CCMV or 125I-Hsp
collected after 24 hours were analyzed by filtration (Microcon centrifugal filter
device, MWCO = 100 kDa). Gamma counts of both the retentate and flow
through after filtration indicated that 80% of the CCMV cages were intact (larger
than 100 kDa) and 78.4% of the Hsp cages were intact (larger than 100 kDa)
(Table 6.3 Materials and Methods). Control experiments verified the intact nature
of the 125I-labeled cages in both the Hsp and CCMV doses. Additional analyses
of urine samples from mice injected with free 125-I could be performed (these
samples were not available at the time of analysis).
143
Sample
% Intact *
(cpm)
% Free or
Broken (cpm)
1. CCMV Urine
80
20
2. Hsp Urine
78.4
21.6
3. CCMV Dose
84.3
15.6
4. Hsp Dose
82
18
Table 6.4. Percentage of Intact* Protein Cage (CCMV & Hsp) Excreted by Mice
Mice were injected with 125I labeled CCMV and Hsp. Urine samples were collected 24
hours later and analyzed by Microcon filtration (MWCO 100 kDa). The gamma counts
associated with either the retentate (intact) or filtrate (free or broken) were quantified. The
results are expressed as percentages of the total counts. *Percent intact as indicated by
filtration (size >100 kDa); Hsp Cage = 396 kDa; CCMV 3,600 kDa. Samples 1 & 2 are
urine samples from mice that received 125I-CCMV and 125I-Hsp respectively; the
percentages listed were corrected to exclude the amount of free 125I or broken cages in
the injected dose. Samples 3 & 4 are control samples, analyzing an equivalent amount of
125
I-CCMV and 125I-Hsp as the injected dose in DPBS buffer.
Initial Evaluation of the Immune Response
to Protein Cage Architectures [301]
_
Characterization of the Immune Response Generated Against CCMV and Hsp
in Mouse Models. Initial characterization of the immune response against Hsp
and CCMV generated in mice sought to: (1) determine the extent and type of
antibody response generated against the cages, (2) determine if injection of Hsp
and CCMV elicited a hypersensitivity reaction, (3) determine if antibody
production alters the trafficking of subsequently administered doses of protein
cages (described earlier in this chapter).
Antibody Response to CCMV and Hsp Protein Cage Architectures. In order to
determine the extent and type of antibody response generated against the cages,
144
BALB/c mice were given 50 μg i.v. injections of protein cage on week #1 and
then boosted with 50 μg of the same protein 2 weeks later (week #3) followed by
antigen-specific antibody titer quantification (week #4). These studies confirmed
that a substantial amount of IgG antibody was generated against both CCMV and
Hsp cages (Figure 6.11). PEGylation of CCMV and Hsp diminished the amount
of antibody produced as compared to the titer generated against ‘naked’ CCMV
and Hsp [302] (Figure 6.11; Materials and Methods).
A.
CCMV
Specific
Antibody Titer
Titer
CCMV
Specific
Antibody
3.5
3
2.5
2
1.5
1
0.5
B.
Po
siti
ve
Ne
ga
tive
G
-P
E
Po
st
CC
MV
G
Po
st
CC
MV
Po
st
Hs
p-P
E
EG
Po
st
Hs
p
Pre
CC
MV
-P
G
Pr
eC
CM
V
Hs
p-P
E
Pre
Pre
Hs
p
0
Specific Antibody Titer
HspHsp
Specific
Antibody Titer
3
2.5
2
1.5
1
0.5
Po
sit
ive
Ne
ga
t iv
e
G
-P
E
Po
st
CC
MV
Po
st
CC
MV
G
Po
st
Hs
p-P
E
Po
st
Hs
p
G
Pr
eC
CM
V
Pr
eC
CM
VPE
G
Pr
eH
sp
-P
E
Pr
eH
sp
0
Figure 6.11. CCMV and Hsp Specific Antibody Titer Reduced by PEGylation
Mice were given 50 μg injections of protein cages and PEGylated (PEG) protein cages on
week #1 and then boosted with 50 μg of the same protein 2 weeks later (week #3) followed
by antigen-specific antibody (Ab) titer quantification (week #4). (A) CCMV specific Ab titer
and (B) Hsp specific Ab titer from different samples as indicated on x-axis. PEGylation
decreases the Ab titer (stars). Note: PEGylated cages made by Joynal Abedin, Douglas &
Young Lab; Ab titer by Diana Buckner, Jutila Lab.
145
Hypersensitivity to CCMV and Hsp Evaluated in Mice. The most common side
effect of particulate drug carriers is the induction of a hypersensitivity reaction
after i.v. administration [231]. This effect can sometimes be lessened by slowing
the infusion rate and/or pre-medicating the patient [19, 303]. The hypersensitivity
response to CCMV and Hsp was tested in mice. Mice (4 per group) were given
repeated intranasal doses of CCMV and Hsp. The weight of these mice was
similar to control mice through the one month study period. Mouse weight is a
general indicator of health, since sick mice tend to weigh less than healthy mice.
In addition, the numbers of CD4+ and CD8+ lymphocytes within bronchoalveolar
lavage fluid (BALF) samples were analyzed by fluorescence activated cell sorting
(FACS). This analysis revealed an increased number of lymphocytes in Hsp and
CCMV immunized mice as compared to control mice. In addition, lymphoid-like
aggregates were evident in the histological analysis of lungs from Hsp and
CCMV immunized mice.
In general, minimal to no adverse reactions to the protein cages were
observed from all mice in given test groups, except in mice given their third dose
of 50 μg 125I-CCMV. These mice exhibited signs of distress including labored
breathing (hypopnea), fatigue, and ruffled fur within 5-7 minutes of 125I labeled
CCMV injection [290]. Signs of recovery began at approximately 1.5 hours post
injection and a full recovery was apparent at 2.5 hours post i.v. administration.
The histological data did not evidence any remarkable differences in liver, kidney
or spleen H&E stained sections. Lung sections from a CCMV immunized mouse
146
were suggestive of pulmonary vascular congestion and may account for some of
the distress observed in these mice after challenged with the 125I cage. Spleen
sections from an Hsp and a CCMV mouse appeared different. The areas of
white pulp in the Hsp mouse spleen were greater in size with well defined
germinal centers perhaps indicating more of an immune response, but no
adverse affect was observed in mice immunized with Hsp cages. Additional
analysis of these samples is ongoing and further characterization of the immune
response mounted against Hsp and CCMV cages is warranted.
Summary
The data presented in this chapter are very important to the development of
protein cage architectures as therapeutic and imaging agent delivery systems. In
addition, the biodistribution data in particular may provide insight into the field of
nano-medicine as a whole. There are some nanoplatform biodistribution in the
literature, but less than the number of platforms currently being brought forward
for biomedical applications. It is very interesting that two different protein cage
architectures (Hsp 12 nm, CCMV 28 nm; different amino acid composition) traffic
similarly in both naïve and immunized mice. The possibility that these cages are
excreted intact is also exciting since one current problem with nanoparticles
(including quantum dots) in vivo is that they are sequestered in the liver [257].
Although the data on the initial characterization of the immune response were
somewhat disappointing, especially the adverse reaction seen upon the third
147
dose of CCMV, there were also positive results. There was no adverse reaction
to Hsp cages and the antibody response generated against both cages could be
reduced by PEGylation. Tumor localization of both non-targeted and tumor
targeted Hsp cages warrants further investigation. Future experiments will
investigate the efficacy of HspG41CRGD4C-Dox cages as therapeutic delivery
agents in vivo. The data presented herein begin to answer some questions
regarding the biocompatibility of protein cage architectures while at the same
time generating many more questions. The continued in vivo study of these
cages will be of great importance to the development of these systems for
biomedical applications.
148
CHAPTER 7
ADDITIONAL GENETIC VARIANTS OF THE SMALL HEAT SHOCK PROTEIN
(Hsp) FROM M. jannaschii AND THE HUMAN H-CHAIN FERRITIN (HFn)
WITH CELL-SPECIFIC TARGETING POTENTIAL
Abstract
In addition to the ‘tumor targeted’ HspG41C-RGD4C genetic variant and the
CD4+ cell targeting Ab-Hsp chemical conjugate (described in Chapter 4), other
genetic variants with cell specific targeting potential were created. The majority
were generated utilizing the Hsp cage platform (Table 7.1). Some of these
variants have been characterized, derivatized, and tested in cell and/or tissue
assays. Other variants have only been initially characterized by DNA sequencing
and verification of subunit protein expression (Table 7.1).
The ‘lung targeting’ variant of the Hsp cage architecture, HspG41C-GFE, was
generated by genetic incorporation of the GFE peptide as a C-terminal fusion of
the Hsp monomer. The GFE peptide was discovered by in vivo phage display
and was previously shown to bind membrane dipeptidase expressed on lung
endothelial cells [257]. HspG41C-GFE exhibited ex vivo lung tissue binding
capability, whereas the results on the ability of radiolabeled HspG41C-GFE
cages to target lung tissue in vivo were mixed (Chapter 6).
Two additional tumor targeting Hsp variants, HspG41C-RGD2C and
HspG41C-RGDWLE, self-assembled in the E.coli heterologous expression
systems from which they were easily purified. Both of these variants were
149
designed to target αvβ3 integrins known to be up-regulated on angiogenic tumor
vasculature. HspG41C-RGD2C and HspG41C-RGDWLE were labeled with
fluorescein. Initial epifluorescence microscopy data indicated that they were not
as effective at binding adherent C32 melanoma cells as HspG41C-RGD4C.
The other protein cage platform utilized for cell-specific targeting was the
human H-chain ferritin (HFn) architecture (described in Chapter 1). Analogous to
the ‘tumor targeted’ HspG41C-RGD4C construct, a ‘tumor targeted’ human Hchain ferritin was created, RGD-HFn, via genetic incorporation of the RGD-4C
peptide. The RGD-HFn architecture has exhibited promise towards its
development as a tumor targeted imaging agent. RGD-HFn cages bound C32
melanoma cells and served as size constrained reaction vessels for the
synthesis of magnetite nanoparticles (a specific crystal phase of iron oxide useful
as a magnetic resonance imaging (MRI) contrast agent) [276].
Introduction
The focus of this work was to generate and characterize many different cell
targeting variants of the Hsp protein cage architecture and the ‘tumor targeting’
human H-chain ferritin, RGD-HFn [276]. All the cell targeting variants were
genetically engineered to incorporate sequences encoding cell-specific targeting
peptides [281]. As described in Chapter 4, these cell-specific targeting peptides
were discovered by the in vivo application of the phage display library technique
[7, 8, 24, 251, 256, 257, 304]. Table 7.1, provides an overview of the library of
150
different Hsp genetic variants [305]. Most target tumors, but notably target via
different routes (Table 7.1). Simply put, a tumor is a mass of rapidly dividing
cells fed by a newly formed (angiogenic) vasculature system and drained by a
lymphatic system. Therefore three useful strategies for targeting drug delivery
systems to tumors are: (1) target the tumor cells themselves, (2) target the
vasculature that feeds the tumor (i.e., the RGD-4C constructs), and (3) target the
lymphatic vessels. Analysis of 7.1, illustrates that Hsp genetic variants were
created to target tumors via each of these routes.
In addition to tumor targeting Hsp architectures, one lung targeting variant was
created, HspG41C-GFE. This variant is especially of interest to our collaborator,
Dr. Allen Harmsen (Montana State University), since much of his work is focused
on the lung. The ultimate goal of the lung targeted protein cage work is to target
a protein cage housing a therapeutic agent (anti-asthma) to the lung and trigger
the release of the therapeutic agent at the early stages of an asthma attack. In
addition, specific lung targeted imaging agents are desirable.
151
Mammalian
Cellular Target
Peptide
Sequence
Introduced
Status
*
5’-GGA ATA CAA ATT TCC GGA AAA TGC TTC ATG CCA ATC TC-3’
5’-GA GAT TGG CAT GAA GCA TTT TCC GGA AAT TTG TAT TCC-3’
5’ G GAA GAA AAT GCA TGC GCA AAG TTT GAA AAT GG 3’
5’ CC ATT TTC AAA CTT TGC GCA TGC ATT TTC TTC C 3’
5’GATACATTAGAGATTAAAGCTAAGAAAAGCCCATTAATGATCACTG
AGAGTGAAAAAATTATCTACTCAG3’
5’CTGAGTAGATAATTTTTTCACTCTCAGTGATCATTAATGGGCTTTTC
TTAGCTTTAATCTCTAATGTATC3’
5' GA TCT GGA GGA TGC GAC TGC CGC GGA GAC TGC TTC TGC
GGA TAA GGA 3'
5’ TCC TTA TCC GCA GAA GCA GTC TCC GCG GCA GTC GCA
TCCTCC AGA TC 3’
CRGDWLEC
Mass
Ref.
None
none
1&2
16,498
[288]
S121C
None
none
1&2
16,468
[288]
None
none
1&2
16,368
[306]
1&2
17,814
[24,
281]
5’ GA TCT GGA GGA TGT CGC GGA GAC TGG CTC GAG TGC GGA
TAA GGA 3’
5’ G ATC TCC TTA TCC GAC CTC GAG CCA GTC TCC GCG ACA TCC
TCC A 3’
1&2
17,776.5
RGDWLE
5’ GA TCT GGA GGA CGA GGA GAC TGG CTC GAG GGA TAA GGA 3’
5’ G ATC TCC TTA TCC CTC GAG CCA GTC TCC GCG TCC TCC A 3’
1&2
17,570.2
[307]
CGFECVR
QCPERC
5’GATCTGGAGGATGCGGTTTCGAATGCGTTCGTCAGTGTCCGGAAC
GCTGTGGATAAGGA 3’
5’GATCTCCTTATCCACAGCGTTCCGGACACTGACGAACGCATTCGAA
ACCGCA TCCTCCA 3’
1&2
18,325
[257]
R80\83\93K
G41CRGD4C
G41CRGD2C
G41CRGDWXE
G41C-GFE
αvβ3 & αvβ5
integrins on tumor
vasculature
αvβ3 & αvβ5
integrins on tumor
vasculature
αvβ3 & αvβ5
integrins on tumor
vasculature
Membrane
dipeptidase on
lung endothelial
cells
CDCRGDCFC
Table 7.1 Hsp Genetic Variants Summarized
*(1) = Cloned & Sequenced; (2) = Purified / Assembled; Note: All targeting sequences have GSGG before the peptide sequence
and all but CGKRK and CREKA a G is added before the stop codon.
150
Mutagenesis Primers
(+) 1st (-) 2nd
Name of
Hsp
Genetic
Variant
G41C
152
Name of
Hsp
Genetic
Variant
G41CCREKA
G41C-Lyp1
Mammalian
Cellular Target
Peptide
Sequence
Introduced
Mutagenesis Primers
(+) 1st (-) 2nd
Status
*
Mass
Ref.
CREKA
5’ GATC T GGA GGA TGC CGC GAA AAA GCT TAA GGA 3’
5’ GATC TCC TTA AGC TTT TTC GCG GCA TCC TCC A 3’
1
17,344
[308]
CGNKRTRGC
1
17,789
[195]
G41C-F3
tumor endothelium
& tumor cell nuclei
KDEPQRR
SARLSAKP
APPKPEPK
PKKAPAKK
1
20,228
[257,
309]
G41CCGKRK
G41C-NGR
shorter alternative
to F3
angiogenic
endothelium of
tumors
CGKRK
5’GATCTGGAGGATGTGGCAACAAACGCACCCGGGGCTGCGGATAA
GGA 3’
5’GATCTCCTTATCCGCAGCCCCGGGTGCGTTTGTTGCCACATCCTC
CA 3’
5’GATCTGGAGGAAAAGATGAACCACAAAGAAGGAGTGCTCGCTTAA
GCGCAAAGCCTGCTCCACCTAAACCAGAACCTAAGCCAAAAAAGGC
ACCTGCTAAAAAGGGATAAGGA 3’
5’GATCTCCTTATCCCTTTTTAGCAGGTGCCTTTTTTGGCTTAGGTTCT
GGTTTAGGTGGAGCAGGCTTTGCGCTTAAGCGAGCACTCCTTCTTTG
TGGTTCATCTTT TCCTCCA 3’
5’ GATC CGGAGGA TGTGGAAAAAGAAAG TAAGGA 3’
5’ GATC TCCTTA CTTTCTTTTTCCACA TCCTCCG 3’
1
17,329
5’GATCTGGAGGATGCAACGGCCGCTGTGTGTCCGGATGCGCGGGT
CGTTGTGGA TAAGGA 3’
5’GATCTCCTTATCCACAACGACCCGCGCATCCGGACACACAGCGGC
CGTTGCATCCTCCA 3’
1
18,080
[257,
309]
[310]
CNGRCVS
GCAGRC
Table 7.1. – Continued. Hsp Genetic Variants Summarized
*(1) = Cloned & Sequenced; (2) = Purified / Assembled; Note: All targeting sequences have GSGG before the peptide sequence
and all but CGKRK and CREKA a G is added before the stop codon.
151
tumor extracellular
matrix
tumor lymphatic
vessels
153
Results and Discussion
HspG41C-GFE Engineered to Target Membrane
Dipeptidase Expressed on Lung Endothelium
The “GFE peptide” discovered from in vivo phage display library technique has
been shown to target membrane dipeptidase on lung endothelial cells [257]. In
order to endow protein cage architectures with lung specific targeting capability,
the lung targeted HspG41C-GFE protein cage was created. Genetic introduction
of DNA encoding the GFE peptide (CGFECVRQCPERC) was incorporated into
HspG41C as a C-terminal extension [257, 311] (Materials and Methods).
Characterization by size exclusion chromatography (SEC), dynamic light
scattering (DLS), and transmission electron microscopy (TEM) demonstrate that
HspG41C-GFE subunits assemble into protein cage architectures similar to wild
type Hsp and other genetic variants (Figure 7.1A - TEM). HspG41C-GFE cages
were labeled with fluorescent dyes (fluorescein, Texas-red, Molecular Probes –
Invitrogen, see Materials and Methods). The co-elution of Texas-red (580 nm)
and the protein cage from size exclusion chromatography demonstrate the
association of Texas-red with the protein cage (Figure 7.1B).
In order verify the ability of HspG41C-GFE to bind lung tissue, Texas-red
labeled Hsp cages were incubated with ex vivo lung tissue (Figure 7.2, Materials
and Methods). Epifluorescence microscopy demonstrated that ‘lung targeted’
HspG41C-GFE cages adhered to lung tissue, whereas non-targeted HspG41C-
154
B.
Absorbance (mAu)
A.
50 nm
HspG41C-GFE TxRed
300
Abs 582 nm
Abs 280 nm
250
200
150
100
50
0
0
10
20
30
40
Time (minutes)
50
60
Figure 7.1. Characterization of HspG41C-GFE Cage Architectures
A. TEM of HspG41C-GFE B. SEC elution profile of Texas-red derivatized HspG41C-GFE.
Figure 7.2. Lung Targeted HspG41C-GFE Binds ex vivo Lung Tissue
Texas-red labeled (A) HspG41C-GFE and (B) HspG41C cages were incubated with ex
vivo murine lung tissue. Epifluorescence microscopy illustrates the enhanced ability of
HspG41C-GFE to bind lung tissues as compared to HspG41C. Lung tissue exhibits a
high level of green autofluorescence.
Panel C is an image of lung tissue that was not incubated with protein cages.
Texas red cages did not (Figure 7.2). These studies were promising, but
biodistribution studies examining the lung targeting ability of HspG41C-GFE in
vivo exhibited mixed results (Chapter 6). Studies testing HspG41C-GFE cages
and their targeting capabilities in mouse models will continue.
155
Tumor Targeting HspG41C-RGD2C and HspG41C-RGDWLE
In addition to the tumor targeting HspG41C-RGD4C described in Chapters 4 6, two other variants (HspG41C-RGD2C & HspG41C-RGDWLE) with the same
cellular target, αvβ3 and αvβ5 integrins, were genetically engineered (Table 7.1;
Materials and Methods) [307]. As described previously, αvβ3 and αvβ5 integrins
have increased expression on angiogenic endothelial cells of tumor vasculature
and are therefore suitable targets for therapeutic and imaging agent delivery.
HspG41C-RGD2C and HspG41C-RGDWLE cages were purified from E. coli in
the same manner as HspG41C-RGD4C (Materials and Methods). The intact
nature of the cages was determined by their characteristic elution profile from
size exclusion chromatography and by TEM (Figure 7.3 -TEM).
A.
50 nm
B.
50 nm
Figure 7.3 TEM Images of (A) HspG41C-RGD2C-fluorescein and (B) HspG41C-RGDWXLE
HspG41C-RGD2C cages were labeled with fluorescein-5-maleimide (Fl);
reactions with 2-fold molar excess per Hsp subunit resulted in attachment of 17.4
Fl per cage. Similarly, reactions with 6-fold molar excess per subunit resulted in
labeling 16.6 Fl per cage. There are 3 cysteines per subunit (72 per cage),
156
therefore not all cysteines were labeled in these reactions. Substoichiometric
labeling of cages was desired for in vitro cell targeting experiments to ensure that
not all targeting peptides were labeled. It is not known whether or not fluorescein
labeling of targeting peptide disrupts their integrin binding capability, thus
substoichiometric labeling was done as a precaution. Similarly, HspG41CRGDWLE cages were fluorescently labeled with fluorescein maleimide. The
yields of labeled cages were similar, approximately 75% of the starting protein.
HspG41C-RGD2C and HspG41C-RGDWLE cages were assayed for their ability
to bind adherent C32 melanoma cells (as described in Chapter 4). In two initial
experiments the HspG41C-RGD2C and HspG41C-RGDWLE did not
demonstrate cell binding.
Tumor Targeted Human H-chain Ferritin (RGD-HFn) [276]
Analogous to the ‘tumor targeted’ HspG41C-RGD4C construct (described in
Chapter 4), a ‘tumor targeted’ human H-chain ferritin was created, RGD-HFn.
The gene encoding human H-chain ferritin was cloned into an E.coli plasmid, and
subsequently genetically altered via PCR mediated mutagenesis to present the
RGD-4C tumor targeting peptide (Materials and Methods) [276]. The RGD-4C
peptide was incorporated into HFn as an N-terminal fusion, since in the crystal
structure of ferritin the N-terminal amino acid residues are exposed to the exterior
surface. Four glycine residues were added after the RGD-4C peptide to allow for
some structural flexibility. This flexibility might facilitate intra-peptide disulfide
157
bond formation between the four cysteines in the RGD-4C sequence (as
described previously in Chapter 4). Both HFn and RGD4C-Fn were expressed in
E. coli where they self-assembled into the 24 subunit cages. The protein cages
were characterized using SEC, dynamic light scattering (DLS: Brookhaven,
90Plus particle size analyzer) and transmission electron microscopy (TEM: LEO
912AB) (see [276]).
Cell Targeting Assay of Mineralized RGD4C-Fn. Iron oxide mineralized
(magnetite) RGD4C-Fn (RGD4C-Fn-Mag) cages demonstrated cancer cell
targeting capability. Magnetite containing RGD4C-Fn-Mag and non-targeted
HFn (HFn-Mag) cages were made as described in Materials and Methods
(Figure 7.4A). Their C32 melanoma cell (αvβ3 integrin) binding capability was
tested. Adherent C32 cells were incubated with either HFn-Mag or RGD4C-FnMag (200µg/ml) in DPBS. After incubation, the cells harvested and prepared for
thin sectioning (Materials and Methods). The increased amount of RGD4C-FnMag cage association with C32 cells, as compared to HFn-Mag cages, was
imaged by TEM (Figure 7.4B).
Cell Targeting of Fluorescein Labeled RGD4C-Fn. Cysteine residues on both
the HFn and the RGD4C-Fn were labeled with fluorescein-5-maleimide as
described previously (Materials and Methods). This facilitated the investigation
of their C32 melanoma cell binding capability by FACS (Materials and Methods).
Similar to the HspG41C-RGD4C C32 cell binding assay described in Chapter 4,
158
RGD4CFn-Fl
C.
B.
A.
HFn-Fl 1388
C32 Cells
Only
34
TEM
100
100 nm
nm
TEM
453
200 nm
Fluorescence Intensity
Figure 7.4. C32 Melanoma Cell Binding Capability of Derivatized RGD4C-Fn Protein
Cage Architectures
A. Unstained TEM image of RGD4C-Fn encapsulated magnetite nanoparticles.
B. TEM image of a C32 melanoma thin section illustrates binding of magnetite containing
RGD4C-Fn (arrow).
C. FACS data from C32 melanoma cells incubated with fluorescently labeled non-targeted
HFn (red line; geo. mean 453) or αvβ3 integrin targeted RGD4C-Fn (green filled plot; geo.
mean 1388) are plotted as histograms labeled with their corresponding geometric mean
fluorescence intensity values (geo. mean). The background level of C32 cell associated
fluorescence (blue solid line; geo. mean 34).
C32 cell suspensions were incubated with fluorescently labeled RGD4C-Fn and
HFn protein cages. After incubation, the cells were washed and suspended in
DPBS + 1% FBS for FACS analysis. The results demonstrate the superior
binding ability of RGD4C-Fn-Fl (geometric mean fluorescent intensity (geo.
mean) = 1388) as compared to HFn-Fl (geo. mean = 453) (Figure 7.4C).
Notably, the HFn cages exhibit an increased level of non-specific C32 melanoma
cell association compared to non-targeted HspG41C cages (Chapter 4, Figure
4.6).
159
Summary
In conclusion, the data presented in this chapter demonstrate the potential of
lung targeted HspG41C-GFE cages and tumor targeted RGD4C-Fn cages to
serve as cell specific therapeutic and imaging agent delivery systems. In
addition, the data presented illustrate that some protein cages designed for cellspecific targeting do not perform as well as others (i.e., HspG41C-RGD4C is
better than HspG41C-RGD2C at binding C32 melanoma cells). These types of
results are often not published, but represent important pieces of the puzzle. The
work described herein demonstrates our ability to evaluate different protein cage
architectures for mineralization and cell targeting purposes.
160
CHAPTER 8
CHEMICAL DERIVATIZATION OF GENETIC VARIANTS OF COWPEA
CHLOROTIC MOTTLE VIRAL (CCMV) CAPSIDS
Introduction and Abstract
The Cowpea chlorotic mottle viral capsid has been studied for many purposes
including: (1) understanding the assembly, disassembly and structural transitions
of viral capsids, (2) its utilization as a size constrained reaction vessel for
mineralization, (3) its ability to be genetically manipulated, and (4) its ability to
serve as a platform for chemical derivatization (described in detail in Chapter 1).
It was originally thought that the endogenous cysteines (2 per subunit; 360 per
capsid) could not be labeled with activated dye molecules [39] . Other
experiments suggested they could be labeled [312]. Therefore, there was a need
to create a genetic variant of CCMV which lacked endogenous cysteines, but
expressed novel cysteines either on the exterior or interior surface. Genetic
variants of CCMV with only internally exposed cysteines would allow chemical
conjugation via thiol chemistry solely within the capsid. The generation of CCMV
mutants is largely and empirical process based on prior successes and failures.
This chapter describes efforts to create a genetic variant with only internally
exposed cysteines. During this work, it was discovered that the CCMV structure
could not tolerate the simultaneous substitution of both endogenous cysteines (2
per subunit, C59 and C108). Further investigation determined that the CCMV
capsid structure could tolerate one (either C59 or C108) cysteine substitution at a
161
time. Additional investigation indicated that endogenous cysteines could not be
readily labeled with fluorescent dyes; therefore, internal cysteines were
introduced on the CCMV K42R ‘salt stable’ genetic background. The new
genetic variant with internally exposed thiols, CCMV K42R T48C, was
characterized and derivatized with fluorescent dyes.
Results and Discussion
CCMV Virus-Like Particle (VLP) Production in Pichia pastoris
As described in Chapter 1, CCMV virus-like particles (VLPs) are produced in
the Pichia pastoris heterologous expression system (Invitrogen) [55]. This
eukaryotic yeast based system facilitates genetic manipulation and large scale
protein production. The use of the P. pastoris heterologous expression system
for the production of wild type (wt) and genetically altered CCMV VLPs was
recently described by Brumfield et al. (see [55]). In brief, the DNA encoding the
CCMV coat protein was amplified from an infectious cDNA and cloned into the
pPICZA plasmid vector (Invitrogen EasySelect Pichia Expression Kit). The
pPICZA plasmid is a shuttle vector that can be utilized in both E.coli and P.
pastoris; designed for intracellular protein expression in P. pastoris [72]. This
plasmid was utilized as a template for introduction of mutations in the coat
protein by polymerase chain reaction (PCR) mediated mutagenesis (Stratagene).
As described below and listed in Table 8.1, two overlapping primers were
designed that contain the DNA sequence corresponding to the desired change
162
flanked by wild type (or other genetic variant) template sequence. The altered
plasmid DNA can be amplified in E.coli and then sequenced in order to ensure
that the correct mutations have been introduced. Then this plasmid DNA can be
introduced into P. pastoris via electroporation. Yeast containing the plasmid can
be selected by their ability to grow on Zeocin containing medium (see [72] for
more information). The CCMV coat protein is then expressed within the yeast,
where it self assembles into VLPs that can be purified as described in Materials
and Methods (Appendix A). The P. pastoris system also facilitates the “scale-up”
of VLP production, since yeast can be grown to high cell densities in fermentation
systems [55, 72].
Genetic Engineering of CCMV Variants
with Internally Exposed Cysteines
Genetic variants of CCMV with only internally exposed cysteines would allow
chemical conjugation solely on the interior of the capsid. As described below, the
first step in the generation of these mutants was to genetically replace the two
endogenous cysteines (C59 and C108) with serines. It was thought that the
cysteine (side chain = -CH2-SH) to serine (side chain = -CH2-OH) would be a
conservative change that the viral capsid would tolerate. The goal was to
genetically engineer a CCMV variant with no endogenous cysteines and having
introduced internal cysteines (CCMV T48C; CCMV S108C). These cysteines
could then be derivatized via the same thiol-maleimide chemistry described for
Hsp labeling. The starting construct for this work was CCMV NΔ16 in which
163
amino acids 8-23 were deleted [313]. Two variants of this construct were utilized
as templates for PCR mediated mutagenesis: CCMV NΔ16 R26C and CCMV
NΔ16 K42R. CCMV K42R is termed the ‘salt stable’ mutant since it is resistant
to disassembly in high ionic strength (described in Chapter 1) [62-66].
The first step toward generating CCMV variants with only internal cysteines
was to make CCMV constructs lacking wild type cysteines (CCMV NΔ16 C59S
C108S and CCMV NΔ16 K42R C59S C108S). The primers for PCR mediated
mutagenesis are listed in Table 8.1. Analysis of the shuttle vector plasmid DNA
with restriction digests indicated that mutagenesis was successful. DNA
sequencing confirmed that the following constructs were accurately made: (1)
CCMV NΔ16 R26C C59S, (2) CCMV NΔ16 R26C C108S, (3) CCMV NΔ16 R26C
K42R C59S, and (4) CCMV NΔ16 R26C K42R C108S. The next round of
mutagenesis served to generate double mutants with both endogenous cysteines
mutated to serines (CCMV NΔ16 R26C C59S C108S; CCMV NΔ16 R26C K42R
C59S C108S). DNA sequence analysis confirmed that the following clones had
the correct sequence: CCMV NΔ16 R26C C59S C108S (clones designated
M1#1 and M1#2) and CCMV NΔ16 R26C K42R C59S C108S (clone designated
M4#1). In addition, PCR mediated mutagenesis was employed for reversion of
the R26C mutation back to wild type with the primers listed in Table 8.1. Lastly,
internal cysteines were introduced utilizing the primers listed in Table 8.1.
All of the above constructs were obtained and verified in E. coli, and then
transformed into the Pichia pastoris expression system (as described above).
164
Purification of the new CCMV genetic variants was performed as described in
Materials and Methods. Transmission electron microscopy was used to screen
purified virus-like particles (VLPs) from small scale expression (1 L) of CCMV
NΔ16 R26C K42R C59S C108S T181C; CCMV NΔ16 C59S C108S T48C; and
CCMV NΔ16 R26C K42R. VLPs are routinely purified using cesium chloride
gradients and ultracentrifugation. Normally VLPs form blue-tinted bands; in
some cases the bands appear more diffuse and contain a lot of debris which is
an indication of a poor quality VLP preparation. Unfortunately, the latter poor
quality type band was obtained from the T181C preparation. TEM visualization
determined that this band contained very few viral particles of varying sizes. The
T48C preparation seemed more promising in that more VLPs were seen by TEM,
but again of differing sizes. A VLP preparation of CCMV NΔ16 R26C was done
in parallel and yielded the most VLPs that looked typical by TEM.
In some cases, fermentation scale-up of yeast production in a 4 L BioFlow
3000 fermentation vessel results in an increase in VLP yield [55]. With that
objective, two different 4 L fermenter batches of yeast containing either the
CCMV NΔ16 C59S C108S T48C construct or the ‘parent’ construct CCMV NΔ16
K42R C59S C108S were grown. Although more particles were visualized by
TEM from these preparations, the VLP yield was not good. Dot blot analysis
utilizing an anti-CCMV IgG revealed that the majority of CCMV protein was
insoluble and associated with the yeast pellet. I did not attempt to obtain this
protein from the pellet and try to ‘re-fold’ it into capsids, instead I focused on the
165
Hsp system. The bottom line was that the CCMV mutants described above did
not result in high enough yields to be used for applications, so little virus was
found from preparations that this work was discontinued.
Later, Debbie Willits (Young Lab – MSU) determined that the CCMV
structure can tolerate only a single endogenous cysteine change at a time (i.e.,
CCMV C59S or CCMV C108S, but not CCMV C59S C108S). She changed C59
and/or C108 to alanine, methionine or serine and expressed all the different
forms in yeast. Dot blots were performed to check for expression and she found
that none of the double mutants had high levels of expression, but all of the
single mutants did. It was decided that the native cysteines did not interfere that
much (i.e., do not label efficiently) so additional cysteines were added to the wild
type (wt) CCMV background. One such genetic variant with internally exposed
cysteines is CCMV K42R T48C. The yield of this genetic variant is less than that
obtained with other genetic variants, but a sufficient amount of intact VLPs are
regularly obtained (Materials and Methods).
Characterization of CCMV K42R T48C
CCMC K42R T48C is a genetic variant with internally exposed thiol groups
(Figure 8.1). It is purified from the P. pastoris heterologous expression system
as described in Materials and Methods. Interestingly, the cesium gradient step of
the CCMV K42R T48C purification process usually results in two viral bands; a
166
CCMV Genetic
Variant
NΔ16 R26C C59S
CCMV Template
Used for PCR
CCMV NΔ16 R26C
NΔ16 R26C K42R
C59S
Purpose of
Mutagenesis
Primer Set Designed top (+); bottom (-)
Status*
5’-GG ACC GCC TCT TCC GCG GCT GCC GAG-3’
5’-CTTC GGC AGC CGC GGA AGA GGC GGT-3’ (SacII site
underlined)
1
CCMV NΔ16 R26C
K42R
Replace
Endogenous
C59 with S
5’-GG ACC GCC TCT TCC GCG GCT GCC GAG-3’
5’-CTTC GGC AGC CGC GGA AGA GGC GGT-3’ (SacII site
underlined)
1
NΔ16 R26C
C108S
CCMV NΔ16 R26C
Replace
Endogenous
C108 with S
1
NΔ16 R26C K42R
C108S
CCMV NΔ16 R26C
K42R
Replace
Endogenous
C108 with S
NΔ16 R26C C59S
C108S
NΔ16 R26C K42R
C59S C108S
NΔ16 R26C C59S
Double Mutant C59S
C108S
5’-GT GTT AGT GGC ACA GTG AAG AGC TCT GTT ACA GAG
ACG CAG-3’ (also changed codon encoding serine 107 from TCC to
AGC)
C108S(-) 5’-CTG CGT CTC TGT AAC AGA GCT CTT CAC TGT
GCC ACT AAC AC-3’ (Introduced SacI site underlined)
5’-GT GTT AGT GGC ACA GTG AAG AGC TCT GTT ACA GAG
ACG CAG-3’ (also changed codon encoding serine 107 from TCC to
AGC)
C108S(-) 5’-CTG CGT CTC TGT AAC AGA GCT CTT CAC TGT
GCC ACT AAC AC-3’ (Introduced SacI site underlined)
Used primers listed above.
NΔ16 R26C K42R
C108S
Double Mutant C59S
C108S
Used primers listed above
1
1
1
Table 8. 1 CCMV Mutagenesis Primers
Note: CCMV NΔ16 = construct in which amino acids 8-23 were deleted
*1 = Positive E.coli;obtained construct in E.coli, confirmed by restriction digest and DNA sequencing.
*2 = Small scale 1 liter P. pastoris yeast prep with junky bands with some virus like particles (VLPs) seen by TEM in very low yields
*3 = Larger scale 4 liter fermenter P. pastoris yeast preparation, still resulted in low yield
165
Replace
Endogenous
C59 with S
167
CCMV Genetic
Variant
CCMV Template
Used for PCR
NΔ16 R26C K42R
Purpose of
Mutagenesis
Primer Set Designed top (+); bottom (-)
5’- GTC GGA ACC GGT AAC ACT CGT GTG GTC CAA CCT GTT
ATT G-3’
5’C AAT AAC AGG TTG GAC CAC ACG AGT GTT ACC GGT TCC
GAC-3’
5’-GGC AAG GCT ATT AAA GCT TGG TGC GGT TAC AGC GTA
TCG-3’
5’-CGA TAC GCT GTA ACC GCA CCA AGC TTT AAT AGC CTT
GCC-3’
(Introduced HindIII site underlined)
5’-GGC AGG GCT ATT AAA GCA TGG TGC GGT TAC AGC GTA
TCG-3’
5’CGA TAC GCT GTA ACC GCA CCA TGC TTT AAT AGC CCT
GCC-3’
(screen for loss of HindIII site)
5’-GTT GAG CAT GTC AGG CCT TGC TTT GAC GAC TCT TTC-3’
T181C(-) 5’-GAA AGA GTC GTC AAA GCA AGG CCT GAC ATG
CTC AAC-3’ (Introduced StuI site underlined)
5’-GTT GAG CAT GTC AGG CCT TGC TTT GAC GAC TCT TTC-3’
T181C(-) 5’-GAA AGA GTC GTC AAA GCA AGG CCT GAC ATG
CTC AAC-3’ (Introduced StuI site underlined)
NΔ16 C59S
C108S T48C
NΔ16 C59S C108S
Add Internal Cys
T48C
NΔ16 K42R C59S
C108S T48C
NΔ16 K42R C59S
C108S
Add Internal Cys
T48C into K42R “salt
stable” background
NΔ16 C59S
C108S T181C
NΔ16 C59S C108S
Add Internal Cys
T181C
NΔ16 R26C K42R
C59S C108S
T181C
CCMV NΔ16
K42R C59S
C108S
NΔ16 K42R C59S
C108S
Add Internal Cys
T181C
No Cys Construct
3
2
166
Reverted the R26C
to wt
Status*
2
3
Table 8. 1 – Continued. CCMV Mutagenesis Primers
Note: CCMV NΔ16 = construct in which amino acids 8-23 were deleted
*1 = Positive E.coli;obtained construct in E.coli, confirmed by restriction digest and DNA sequencing.
*2 = Small scale 1 liter P. pastoris yeast prep with junky bands with some virus like particles (VLPs) seen by TEM in very low yields
*3 = Larger scale 4 liter fermenter P. pastoris yeast preparation, still resulted in low yield
168
Interior View
Exterior View
Figure 8.1 Space Filling Model of CCMV K42R T48C
Internally exposed cysteines at position 48 colored blue.
Absorbance (a.u.)
A.
B.
800
CCMV T48C
600
AAbs
bs280 280 nm
Abs495
400
200
0
0
10
20
30
40
50
Time (minutes)
Time (minutes)
Figure 8.2. Characterization of CCMV K42R T48C
Intact nature of the cage demonstrated by (A) size exclusion chromatography
elution profile (protein monitored at 280 nm) and (B) transmission electron
microscopy (TEM).
169
top band which seems ‘empty’ (free of yeast RNA) by TEM analysis and a lower
band which seems to contain yeast RNA. The lower band contains particles of
varying size and contains more debris. Characterization by size exclusion
chromatography (SEC), dynamic light scattering (DLS), and transmission
electron microscopy (TEM) demonstrate the intact nature of these cages (Figure
8.2). TEM also revealed that the CCMV K42R T48C particles have diameters
that range from approximately 23 nm to 32 nm (Figure 8.3). In addition, some
“double shell” particles and large oval particles were observed (Figure 8.3A).
Figure 8.3. TEM Characterization of CCMV K42R T48C
A. The diameters of CCMV K42R T48C particles from cesium gradient viral bands measured
on TEM images ranged from approximately 23 nm to 32 nm. In addition, there are some
“double shell” particles noticeable (arrows).
B. Size exclusion purification seems to narrow the size range of the diameter to
approximately the range of 24-28 nm, but some ovals (89 nm dia.) were also visualized.
C. Glutaraldehyde crosslinked CCMV K42R T48C. Scale bars 50 nm.
Derivatization of CCMV K42R T48C
CCMV K42R T48C cages were labeled with fluorescent dyes (fluorescein,
Texas-red (Materials and Methods). The co-elution of fluorescein-5-maleimide
(495 nm) and the protein cage (280 nm) from size exclusion chromatography
170
illustrate the association of fluorescein to the CCMV protein cages (Figure 8.4).
Glutaraldehyde intra-particle crosslinked CCMV architectures are more stable
under certain reaction conditions (Materials and Methods). The reactivity of
glutaraldehyde crosslinked CCMV K42R T48C cages was examined utilizing
fluorescein-5-maleimide (Fl-Mal) labeling experiments.
A series of reactions comparing the reactivity of wild type (wt) CCMV,
crosslinked CMV K42R T48C and CCMV K42R T48C with fluorescein-5maleimide were performed. The Fl-Mal concentrations ranged from 5 to 500-fold
molar excess per subunit. The highlights from this experiment, summarized in
Table 8.2, include: (1) that wt CCMV from the lower virus band does not
appreciably label with Fl-Mal, (2) 154 of the 180 introduced cysteines of CCMV
K42R T48C are labeled at 500x molar equivalents of Fl-Mal, (2) non-crosslinked
CCMV
T48C
CCMV T48C
X-Linked
Wt CCMV
FlMal per subunit
Molar Excess
5
24
unknown
0
15
30
unknown
2
15 (no TCEP)
23
unknown
2
30
33
unknown
4
45
45
unknown
unknown
60
71
28
1
180
61
34
2
500
154
38
7
Table 8.2. Degree of CCMV Labeling with Fluorescein-5-Maleimide
CCMV K42R T48C, glutaraldehyde cross-linked (X-linked) CCMV K42R T48C,
and wt CCMV were reacted with the listed molar excesses of fluorescein-5maleimide (FlMal) as compared to protein subunit concentration.
Unless specified all cages were pre-treated with 5x TCEP prior to FlMal reaction.
171
CCMV K42R T48C is labeled to a higher degree than crosslinked at equivalent
dye concentrations (i.e., at 500x Fl-Mal non-crosslinked 154 fluoresceins per
cage, compared to crosslinked 38 fluoresceins per cage). In addition, these
types of experiments revealed that it is very important to work with the upper
(empty) band from CCMV K42R T48C VLP preparations for cysteine labeling
reactions. Fl-Mal reactions utilizing the lower band exhibited a maximum labeling
of 10 cysteines with 45-fold molar equivalents. This result indicates that the RNA
and/or protein, housed within cages from the lower viral band, blocks access to
the cage interior and interferes with cysteine derivatization.
In addition, CCMV K42R T48C could be labeled with Texas Red C2 maleimide
(Figure 8.5). Unfortunately, a few trials of CCMV K42R T48C labeling with the 6maleimidocaproyl) hydrazone derivative of doxorubicin (Mal-Dox), 5 molar
equivalents Mal-Dox per subunit, did not seem to work. Mal-Dox derivatization of
CCMV was attempted, but is an area for further investigation.
172
Absorbance (a.u.)
A.
50
CCMV K42R T48C : 5x FlMal
12-18-03 Rxn1 - T48C 5xtcep 5xFl-Mal
40
30
Abs
280280nm
495495nm
Abs
20
10
0
-10
0
5
10
15
20
25
Time (minutes)
Absorbance (a.u.)
B.
2-18-03
WtHsp
5xtcep
5xFl-Mal
wtRxnA
CCMV
: 500x
FlMal
80
60
Abs
280280
nm
495495
nm
Abs
40
20
0
0
5
10
15
20
25
Time (minutes)
Figure 8.4. Size Exclusion Chromatography Elution Profiles of CCMV Cages
A. CCMV K42R T48C reacted with fluorescein-5-maleimide (FlMal); 5 molar
equivalents FlMal per CCMV subunit. The co-elution of CCMV K42R T48C
protein cage (280 nm) and fluorescein (495) nm is demonstrated.
B. Wt CCMV was reacted with 500 molar equilants of FlMal per subunit. No
fluorescein (495 nm) co-elution is demonstrated.
173
Absorbance (a.u.)
400
A.
B.
Abs
Abs 280
280 nm
nm CCMV-T48C
-Protein CageProtein Cage
Abs 580 nm Texas
- TexasRed
Red
300
200
100
50 nm
0
0
10
20
30
40
Figure 8.5. Texas-red Derivatized CCMV T48C
A. Size exclusion chromatographySummary
(SEC) elution profile of CCMV illustrating the
co-elution of CCMV (T48C) protein cage (Abs 280 nm) and Texas Red (Abs 580 nm).
B. TEM of CCMV(T48C)-Texas red (stained with uranyl acetate). Scale bar 50 nm.
Summary
Summary
In conclusion, these data demonstrate that CCMV structure could not tolerate
the simultaneous substitution of both endogenous cysteines, but could tolerate
single cysteine (either C59 or C108) substitutions. Additional investigation on the
reactivity of endogenous cysteines indicated that they could not be readily
labeled with fluorescent dyes. These results may be because C59 and C108 are
located at the subunit interface of the assembled CCMV cage. Internal
cysteines were introduced on the CCMV K42R ‘salt stable’ genetic background.
The new genetic variant with internally exposed thiols, CCMV K42R T48C, was
characterized and derivatized with fluorescent dyes.
174
CHAPTER 9
SUMMARY AND CONCLUSIONS
The long-term objective of this research is to develop nanometer scale protein
cage architectures for in vivo targeted delivery of therapeutic and imaging
agents. The present work describes efforts toward the achievement of that goal
utilizing three different protein cage architectures; the Cowpea chlorotic mottle
viral capsid (CCMV), a small heat shock protein (Hsp) architecture originally
isolated from the hyperthermophilic archaeon Methanococcus jannaschii, and
human H-chain ferritin (HFn). In the process of tailoring these protein cage
architectures for particular applications, fundamental information about the
architectures themselves was gained.
The data presented herein demonstrate that the Hsp cage is a versatile
nanoscale platform whose exterior and interior surfaces are amenable to both
genetic and chemical modification (described in Chapters 2-7). Wild type and
genetic variants of the Hsp cage were shown to react with fluorescein and the
anti-tumor agent doxorubicin in a site specific manner. Hsp cages were also
employed as size constrained reaction vessels for the formation of monodisperse
9 nm iron oxide nanoparticles. Doxorubicin was selectively released from the
Hsp cage in vitro via a pH dependent trigger. In addition, tumor and lymphocyte
cell-specific targeting was imparted to the Hsp cage by both genetic and
chemical strategies (Chapter 4). The versatility of the Hsp architecture was
175
demonstrated by the generation of many different cell-specific targeting variants,
including the ‘lung targeted’ HspG41C-GFE (Chapter 7).
Likewise, the human H-chain ferritin (HFn) proved useful for both organic and
inorganic chemical modification (Chapter 7). It was fluorescently derivatized and
utilized as a size constrained reaction vessel for the synthesis of magnetite. A
tumor targeted genetic variant, RGD4C-Fn demonstrated cancer cell binding
capability. In addition, the Cowpea chlorotic mottle viral capsid was genetically
and chemically modified (Chapter 8).
A great deal of work remains to be done toward the development of these
architectures for biomedical applications. The initial results from the in vivo
biodistribution studies of Hsp and CCMV in a human tumor xenograft mouse
model and in naïve and pre-immunized mice might help guide this development.
Cell culture and animal studies will become increasingly important.
Although similar in that Hsp, HFn, and CCMV are all biological, self-assembled
proteinaceous architectures, these cages differ in diameter and in amino acid
and subunit composition. The analogous use of three protein cage platforms
illustrates that the methodologies described herein can be employed on a wide
variety of protein cage architectures. While this is generally true, there are
distinct advantages and disadvantages associated with the use of each these
cage architectures for particular applications.
176
Comparison of Hsp, HFn, and CCMV Protein Cage Platforms for Their
Development as Targeted Therapeutic and Imaging Agent Delivery Systems
A comparison of the amenability of the Hsp, HFn, and CCMV protein cage
architectures for potential biomedical applications is difficult. There is still a lot to
learn about each of these protein cage systems and many of the current
chemical and genetic strategies utilized for their modification can be improved
upon, in order to make various platforms more suitable for specific applications.
That said, it is important to present an overview of what is known, what has
succeeded, as well as what has not. Table, 9.1 provides an overview of the
types of chemical modifications performed on various protein cage platforms.
One immediate conclusion that can be made from the data presented in this
table is that not all genetic variants of a particular cage platform perform in a
similar fashion. In addition, although many of the chemical modifications utilize
thiol-maleimide chemistry, the success of these reactions seem to be dependant
on which molecule is being conjugated. For example, although labeling
HspG41C-RGD4C cages with fluorescein-5-maliemide is facile, labeling with the
maleimide derivative of doxorubicin results in much lower yields of derivatized
cages.
Genetic modification of all three cage architectures via PCR mediated
mutagenesis is routine at this time, since many genetic variants have been made
and characterized. Some general points about genetic modification include: (1)
single amino acid changes have been very successful in all three platforms, but
177
each change must be evaluated individually since the structures do not tolerate
all changes; (2) at this time, peptide incorporation in CCMV is more difficult than
in Hsp and HFn, especially within the ‘loop’ regions; (3) N-terminal peptide
extensions of CCMV have proven successful; (4) peptide insertions as C-terminal
fusions have been effective in the Hsp platform whereas insertions in the Nterminal resulted in cleavage of the insert; (5) N-terminal peptide insertions have
proven successful in HFn, but the ease of purification depends on the particular
insert, (6) it is much quicker to create genetic variants of Hsp and HFn than of
CCMV. This is because Hsp and HFn cages are produced from an E. coli
expression system whereas CCMV genetic variants are first made in plasmids
housed within E.coli and successful constructs must then be introduced and
produced in the P. pastoris system which requires longer growing times as
compared to E. coli. Lastly, purification of protein cages (Hsp and HFn) from E.
coli is easier than from yeast (Materials and Methods). It is for these reasons
that the majority of my work utilized the Hsp architectures. In addition, Hsp
cages are stable over a broader pH range (pH 5 – 10) than CCMV (pH 4 – 8).
Ferritin cages are the most pH and thermal stable of these three architectures.
In turn, HFn is the best of these cages for biomineralization reactions such as
magnetite synthesis, which requires elevated temperature (85˚C) and pH 8.5
(Materials and Methods). HFn and RGD-Fn cages have served as size
constrained reaction vessels for magnetite nanoparticle formation. In contrast,
178
the HspG41C-RGD4C construct was not stable under the conditions necessary
for magnetite reactions.
Chemical derivatization of all three cages is quite facile and the addition of
“click chemistry” techniques (described in Chapter 1) to our lab will open up more
possibilities. The extent of derivatization is dependant on both the cage and the
conjugate (molecule, peptide, or antibody) (Table 9.1). The numbers regarding
the extent of derivatization of different cage architectures with a subset of
molecules are reported as a general guideline in order to facilitate comparison
(Table 9.1). Specific results are dependent on the precise reaction conditions. In
addition, most of these labeling reactions could be further optimized.
179
CAGE
Fluorescein
Texas Red
Doxorubicin
Mineralization
Iron Oxide
lepidocrocite
Gd3+
DOTA/DTPA
both; max.
5 per cage
Other Notes
Refs.
Ab linkage for cell
targeting low yield
[37, 41,
95, 97,
281,
288]
[281]
HspG41C
Routine –
24 per cage
Routine 10/cage
Routine –
24 per cage
HspG41CRGD4C
Routine –
12-29 per
cage
Routine –
20 per cage
unknown
More Difficult
12-20
per cage
Unknown
More Difficult
unknown
Stability of cage for Dox
and Min. rxns not great
Unknown
unknown
Ex vivo targeting
In vivo mixed results
[311]
Unknown
Magnetite
Unknown
Magnetite
More Difficult
40 per cage
very low yield
CCMV SubE
Stable under magnetite
min. conditions
Stable under magnetite
min. conditions
Cell targeting more
difficult
[41,
276]
unknown
each subunit
1 DOTA
unknown
HspG41CGFE
HFn
DOTA up to 3
per subunit;
more are bad
Table 9.1. Summary of the Utility of Different Protein Cage Architectures (Hsp, HFn, CCMV)
The table summarizes the ability to deriviatize various cages with fluorescein, Texas red, and doxorubicin;
their abililty to serve as size constrained reaction vessels for mineralization; and the linkage of gadolinium chelators.
This table is for general comparison purposes, since actual results are dependent on the specific reaction conditions.
[276]
[27, 29,
39, 41,
49, 55,
61, 71,
74, 76,
314]
178
RGD4CFn
CCMV
Routine –
24 per cage
Routine –
24 per cage
Routine
Depends on
variant
More Difficult
12-15
per cage
unknown
180
Future Studies
There is great potential for the development of protein cage architectures for
biomedical applications. The results presented herein substantiate this claim. In
addition, recent results from mammalian cell culture experiments indicate that
mineralized H-chain ferritin architectures may prove useful as imaging agents.
RGD-4C mediated delivery of therapeutics to tumors is another avenue that can
be further developed utilizing both HFn and Hsp cages. Initial biocompatibility
data for Hsp cages is promising, and similar studies will be performed utilizing
HFn in the near future. CCMV has also exhibited great promise as a contrast
agent for medical imaging. Future work may involve different specific
components (i.e., more potent therapeutic molecules and/or other targeting
ligands) than those described herein, but the general concepts will remain the
same. “Ultimately, the utility of viral capsids (and other protein cage
architectures) is limited only by our own creativity,” from Douglas & Young,
Science 2006 [315].
181
REFERENCES CITED
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Boonyaratanakornkit, B., and Clark, D.S.C., Chemical Engineering, G.
Vrdoljak. Electron Microscope Lab, University of California Berkeley, vol.
2006: Berkeley, CA.
Seeman, N.C., and Belcher, A.M. (2002). Emulating biology: Building
nanostructures from the bottom up. PNAS 99, 6451-6455.
Rice, G., Stedman, K., Snyder, J., Wiedenheft, B., Willits, D., Brumfield,
S., McDermott, T., and Young, M.J. (2001). Viruses from extreme thermal
environments. Proc Natl Acad Sci USA 98, 13341-13345.
Kim, K.K., Kim, R., and Kim, S.H. (1998). Crystal structure of a small heatshock protein. Nature 394, 595-599.
Willner, D., Trail, P.A., Hofstead, S.J., King, H.D., Lasch, S.J., Braslawsky,
G.R., Greenfield, R.S., Kaneko, T., and Firestone, R.A. (1993). (6Maleimidocaproyl)hydrazone of doxorubicin--a new derivative for the
preparation of immunoconjugates of doxorubicin. Bioconjug Chem 4, 521527.
Khayat, R., Tang, L., Larson, E.T., Lawrence, C.M., Young, M., and
Johnson, J.E. (2005). Structure of an archaeal virus capsid protein reveals
a common ancestry to eukaryotic and bacterial viruses. Proc Natl Acad
Sci U S A 102, 18944-18949.
Pasqualini, R., and Ruoslahti, E. (1996). Tissue targeting with phage
peptide libraries. Mol Psychiatry 1, 423.
Pasqualini, R., and Ruoslahti, E. (1996). Organ targeting in vivo using
phage display peptide libraries. Nature 380, 364-366.
Fox, J.M., Zhao, X., Speir, J.A., and Young, M.J. (1996). Analysis of a salt
stable mutant of cowpea chlorotic mottle virus. Virology 222, 115-122.
Speir, J.A., Munshi, S., Wang, G., Baker, T.S., and Johnson, J.E. (1995).
Structures of the native and swollen forms of cowpea chlorotic mottle virus
determined by X-ray crystallography and cryo-electron microscopy.
Structure 3, 63-78.
Prasad, B.V., Hardy, M.E., Dokland, T., Bella, J., Rossmann, M.G., and
Estes, M.K. (1999). X-ray crystallographic structure of the Norwalk virus
capsid. Science 286, 287-290.
Shepherd, C.M., Borelli, I.A., Lander, G., Natarajan, P., Siddavanahalli, V.,
Bajaj, C., Johnson, J.E., Brooks, C.L., 3rd, and Reddy, V.S. (2006).
VIPERdb: a relational database for structural virology. Nucleic Acids Res
34, D386-389.
Lawson, D.M., Artymiuk, P.J., Yewdall, S.J., Smith, J.M.A., Livingstone,
J.C., Treffry, A., Luzzago, A., Levi, S., Arosio, P., Cesareni, G., Thomas,
C.D., Shaw, W.V., and Harrison, P.M. (1991). Solving the Structure of
Human H-Ferritin by Genetically Engineering Intermolecular Crystal
Contacts. Nature 349, 541-544.
182
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Gauss, G., Wiedenheft, B., Douglas, T., and Lawrence, M. (2006). in
press.
Wiedenheft, B., Mosolf, J., Willits, D., Yeager, M., Dryden, K.A., Young,
M., and Douglas, T. (2005). From The Cover: An archaeal antioxidant:
Characterization of a Dps-like protein from Sulfolobus solfataricus. Proc
Natl Acad Sci U S A 102, 10551-10556.
Ilari, A., Stefanini, S., Chiancone, E., and Tsernoglou, D. (2000). The
dodecameric ferritin from Listeria innocua contains a novel intersubunit
iron-binding site. Nat Struct Biol 7, 38-43.
Namba, K., and Stubbs, G. (1986). Structure of tobacco mosaic virus at
3.6 A resolution: implications for assembly. Science 231, 1401-1406.
(2006). National Cancer Institute's Nanotechnology Initiative, vol. 2006,
National Cancer Institute (NCI).
Allen, T.M., and Cullis, P.R. (2004). Drug delivery systems: entering the
mainstream. Science 303, 1818-1822.
Service, R.F. (2005). Materials and biology. Nanotechnology takes aim at
cancer. Science 310, 1132-1134.
Crowley, C., Willits, D., Douglas, T., and Young, M. (2005). unpublished
data.
Douglas, T., and Young, M. (1998). Host-guest encapsulation of materials
by assembled virus protein cages. Nature (London) 393, 152-155.
Pasqualini, R., Koivunen, E., and Ruoslahti, E. (1995). A peptide isolated
from phage display libraries is a structural and functional mimic of an
RGD-binding site on integrins. J Cell Biol 130, 1189-1196.
Arap, W., Pasqualini, R., and Ruoslahti, E. (1998). Cancer treatment by
targeted drug delivery to tumor vasculature in a mouse model. Science
279, 377-380.
Harrison, P.M., and Arosio, P. (1996). The ferritins: molecular properties,
iron storage function and cellular regulation. Biochim Biophys Acta 1275,
161-203.
Narberhaus, F. (2002). Alpha-crystallin-type heat shock proteins:
socializing minichaperones in the context of a multichaperone network.
Microbiol Mol Biol Rev 66, 64-93; table of contents.
Allen, M., Willits, D., Mosolf, J., Young, M., and Douglas, T. (2002).
Protein Cage Constrained Synthesis of Ferrimagnetic Iron Oxide
Nanoparticles. Adv. Mater. 14, 1562-1565.
Allen, M., Willits, D., Young, M., and Douglas, T. (2003). Constrained
synthesis of cobalt oxide nanomaterials in the 12-subunit protein cage
from Listeria innocua. Inorg Chem 42, 6300-6305.
Allen, M.A., Leipold, L.O., Bulte, J.W.M., Basu, G., Zywicke, H.A., Frank,
J.A., Young, M., and Douglas, T. (2005). Paramagnetic Viral
Nanoparticles as Potential High-Relaxivity MRI Contrast Agents. Magn
Reson Med 54, 807-812.
183
30.
Bulte, J.W.M., Douglas, T., Mann, S., Frankel, R.B., Moskowitz, B.M.,
Brooks, R.A., Baumgarner, C.D., Vymazal, J., and Frank, J.A. (1993).
Magnetoferritin: Biomineralization as a novel approach in the design of
iron oxide-based MR contrast agents. Inv. Rad., S214-S216.
31.
Bulte, J.W.M., Douglas, T., Mann, S., Frankel, R.B., Moskowitz, B.M.,
Brooks, R.A., Baumgarner, C.D., Vymazal, J., Strub, M.-P., and Frank,
J.A. (1994). Magnetoferritin: Characterization of a novel
superparamagnetic MR contrast agent. J. Magn. Res. Imag. 4, 497-505.
32.
Bulte, J.W.M., Douglas, T., Witwer, B., Zhang, S.-C., Lewis, B.K., van
Gelderen, P., Zywicke, H., Duncan, I.D., and Frank, J.A. (2001).
Monitoring Stem Cell Therapy in Vivo Using Magnetodendrimers
as a New Class of Cellular MR Contrast Agents. Nature Biotechnology
submitted.
33.
Chatterji, A., Burns, L.L., Taylor, S.S., Lomonossoff, G.P., Johnson, J.E.,
Lin, T., and Porta, C. (2002). Cowpea mosaic virus: from the presentation
of antigenic peptides to the display of active biomaterials. Intervirology 45,
362-370.
34.
Chatterji, A., Ochoa, W., Shamieh, L., Salakian, S.P., Wong, S.M., Clinton,
G., Ghosh, P., Lin, T., and Johnson, J.E. (2004). Chemical conjugation of
heterologous proteins on the surface of Cowpea mosaic virus. Bioconjug
Chem 15, 807-813.
35.
Chatterji, A., Ochoa, W.F., Paine, M., Ratna, B.R., Johnson, J.E., and Lin,
T. (2004). New addresses on an addressable virus nanoblock; uniquely
reactive Lys residues on cowpea mosaic virus. Chem Biol 11, 855-863.
36.
Douglas, T., and Stark, V.T. (2000). Nanophase Cobalt Oxyhydroxide
Mineral Synthesized within the Protein Cage of Ferritin. Inorg. Chem. 39,
1828-1830.
37.
Flenniken, M.L., Liepold, L.O., Crowley, B.E., Willits, D.A., Young, M.J.,
and Douglas, T. (2005). Selective attachment and release of a
chemotherapeutic agent from the interior of a protein cage architecture.
Chem Commun (Camb), 447-449.
38.
Flenniken ML, W.D., Brumfield S, Young MJ, Douglas T (2003). The Small
Heat Shock Protein Cage from Methanococcus jannaschii Is a Versatile
Nanoscale Platform for Genetic and Chemical Modification. NanoLetters
3, 1573-1576.
39.
Gillitzer, E., Willits, D., Young, M., and Douglas, T. (2002). Chemical
modification of a viral cage for multivalent presentation. Chem Commun
(Camb), 2390-2391.
40.
Hooker, J.M., Kovacs, E.W., and Francis, M.B. (2004). Interior surface
modification of bacteriophage MS2. J Am Chem Soc 126, 3718-3719.
41.
Liepold, L.O., Willits, D., Oltrogge, L., Allen, M., Young, M., and Douglas,
T. (2005). Protein cages as MRI contrast agents. in preparation.
184
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
Raja, K.S., Wang, Q., and Finn, M.G. (2003). Icosahedral virus particles
as polyvalent carbohydrate display platforms. Chembiochem 4, 13481351.
Schlick, T.L., Ding, Z., Kovacs, E.W., and Francis, M.B. (2005). Dualsurface modification of the tobacco mosaic virus. J Am Chem Soc 127,
3718-3723.
Wang, Q., Kaltgrad, E., Lin, T., Johnson, J.E., and Finn, M.G. (2002).
Natural supramolecular building blocks. Wild-type cowpea mosaic virus.
Chem Biol 9, 805-811.
Wang, Q., Lin, T., Johnson, J.E., and Finn, M.G. (2002). Natural
supramolecular building blocks. Cysteine-added mutants of cowpea
mosaic virus. Chem Biol 9, 813-819.
Wang, Q., Lin, T.W., Tang, L., Johnson, J.E., and Finn, M.G. (2002).
Icosahedral virus particles as addressable nanoscale building blocks.
Angewandte Chemie-International Edition 41, 459-462.
Raja, K.S., Wang, Q., Gonzalez, M.J., Manchester, M., Johnson, J.E., and
Finn, M.G. (2003). Hybrid virus-polymer materials. 1. Synthesis and
properties of PEG-decorated cowpea mosaic virus. Biomacromolecules 4,
472-476.
Allen, M., Willits, D., Young, M., and Douglas, T. (2003). Constrained
Synthesis of Cobalt Oxide Nano-Materials In the 12-subunit Protein Cage
From Listeria innocua. Inorg. Chem. 42, 6300-6305.
Basu, G., Allen, M.A., Willits, D., Young, M., and Douglas, T. (2003). Metal
binding to Cowpea Chlorotic mottle Virus using Tb(III) Fluorescence.
Journal of Bioinorganic Chemistry 8, 721-725.
Winter, P.M., Caruthers, S.D., Kassner, A., Harris, T.D., Chinen, L.K.,
Allen, J.S., Lacy, E.K., Zhang, H., Robertson, J.D., Wickline, S.A., and
Lanza, G.M. (2003). Molecular imaging of angiogenesis in nascent Vx-2
rabbit tumors using a novel alpha(nu)beta3-targeted nanoparticle and 1.5
tesla magnetic resonance imaging. Cancer Res 63, 5838-5843.
Winter, P.M., Morawski, A.M., Caruthers, S.D., Fuhrhop, R.W., Zhang, H.,
Williams, T.A., Allen, J.S., Lacy, E.K., Robertson, J.D., Lanza, G.M., and
Wickline, S.A. (2003). Molecular imaging of angiogenesis in early-stage
atherosclerosis with alpha(v)beta3-integrin-targeted nanoparticles.
Circulation 108, 2270-2274.
Allen, M.A., Bulte, J.W.M., Basu, G., Zywicke, H.A., Frank, J.A., Young,
M., and Douglas, T. (2004). Gd(III)-bound to a Viral Protein Cage: An MR
Contrast Agent with Unusually High Relaxivity. Inorg. Chem. submitted.
Bulte, J., Douglas, T., Strable, E., Moskowitz, B., Zywicke, H., and Frank,
J. (2001). Magnetodendrimers as a New Class of MR Contrast Agents for
Universal, Non-Specific Magnetic Labeling of Mammalian (Stem) Cells.
Nature Biotechnology submitted - see enclosed copy.
185
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
Klem, M.T., Willits, D.A., Young, M.J., and Douglas, T. (2003). 2-D Array
Formation of Genetically Engineered Viral Cages on Au Surfaces and
Imaging by Atomic Force Microscopy. J. Am. Chem. Soc. 125, 1056-1057.
Brumfield, S., Willits, D., Tang, L., Johnson, J.E., Douglas, T., and Young,
M. (2004). Heterologous expression of modified Cowpea chlorotic mottle
bromovirus coat protein results in the assembly of protein cages with
altered architectures and function. Journal of General Virology 85, 10491053.
Flenniken M, A.M., Young M, Douglas T (2004). Viruses as Host
Assemblies. In Encyclopedia of Supramolecular Chemistry, A.J. Steed W,
ed. (New York City: Marcel Dekker, Inc.), pp. 1563-1568.
Francki, R.I.B. (1985). The viruses and their taxonomy. In The Plant
Viruses. (New York, NY: Plenum Press).
Zhao, X. (1997). Studies on the Assembly of Cowpea Chlorotic Mottle
Virus, Purdue University, Lafayette, Indiana.
Lin, T.W., Chen, Z.G., Usha, R., Stauffacher, C.V., Dai, J.B., Schmidt, T.,
and Johnson, J.E. (1999). The refined crystal structure of cowpea mosaic
virus at 2.8 angstrom resolution. Virology 265, 20-34.
Speir, J.A., Munshi, S., Baker, T.S., and Johnson, J.E. (1993). Preliminary
X-ray data analysis of crystalline cowpea chlorotic mottle virus. Virology
193, 234-241.
Liepold, L.O., Revis, J., Allen, M., Oltrogge, L., Young, M., and Douglas,
T. (2005). Structural transitions in Cowpea chlorotic mottle virus (CCMV).
Phys Biol 2, S166-172.
Bancroft, J.B., Hiebert, E., Rees, M.W., and Markham, R. (1968).
Properties of Cowpea Chlorotic Mottle Virus Its Protein and Nucleic Acid.
Virology 34, 224-&.
Bancroft, J.B., Mcdonald, J.G., and Rees, M.W. (1976). Mutant of Cowpea
Chlorotic Mottle Virus with a Perturbed Assembly Mechanism. Virology 75,
293-305.
Bancroft, J.B., Mclean, G.D., Rees, M.W., and Short, M.N. (1971). Effect
of an Arginyl to a Cysteinyl Replacement on Uncoating Behaviour of a
Spherical Plant Virus. Virology 45, 707-&.
Bancroft, J.B., Rees, M.W., Johnson, M.W., and Dawson, J.R.O. (1973).
Salt-Stable Mutant of Cowpea Chlorotic Mottle Virus. Journal of General
Virology 21, 507-513.
Bancroft, J.B., Short, M.N., Rees, M.W., Dawson, J.R.O., and Mclean,
G.D. (1972). Some Properties of a Temperature-Sensitive Mutant of
Cowpea Chlorotic Mottle Virus. Journal of General Virology 16, 69-&.
Zhao, X., Fox, J.M., Olson, N.H., Baker, T.S., and Young, M.J. (1995). In
vitro assembly of cowpea chlorotic mottle virus from coat protein
expressed in Escherichia coli and in vitro-transcribed viral cDNA. Virology
207, 486-494.
186
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
Johnson, J.E., and Speir, J.A. (1997). Quasi-equivalent viruses: a
paradigm for protein assemblies. J Mol Biol 269, 665-675.
Basu, G., Allen, M., Willits, D., Young, M., and Douglas, T. (2003). Metal
binding to cowpea chlorotic mottle virus using terbium(III) fluorescence. J
Biol Inorg Chem 8, 721-725.
Schneemann, A., and Young, M.J. (2003). Viral assembly using
heterologous expression systems and cell extracts. Adv Protein Chem 64,
1-36.
Douglas, T., and Ripoll, D. (1998). Electrostatic gradients in the iron
storage protein ferritin. Protein Science 7, 1083-1091.
Invitrogen EasySelect Pichia Expression Kit Manual. Catalog no. K174001 Version A.
Flenniken M, A.M., Young M, Douglas T (2004). Viruses as Host
Assemblies. In Encyclopedia of Supramolecular Chemistry, Volume 2, A.J.
Steed W, ed. (New York City: Marcel Dekker, Inc.), pp. 1563-1568.
Klem, M.T., Willits, D., Young, M., and Douglas, T. (2003). 2-D array
formation of genetically engineered viral cages on Au surfaces and
imaging by atomic force microscopy. Journal of the American Chemical
Society 125, 10806-10807.
Douglas, T., Strable, E., Willits, D., Aitouchen, A., Libera, M., and Young,
M. (2002). Protein Engineering of a Viral Cage for Constrained NanoMaterials Synthesis. Advanced Materials 14, 415-418.
Gillitzer, E., Suci, P., Young, M., and Douglas, T. (2006). Controlled ligand
display on a symmetrical protein cage architecture through mixed
assembly. Small in press.
Fox, J.M., Wang, G., Speir, J.A., Olson, N.H., Johnson, J.E., Baker, T.S.,
and Young, M.J. (1998). Comparison of the native CCMV virion with in
vitro assembled CCMV virions by cryoelectron microscopy and image
reconstruction. Virology 244, 212-218.
Johnson, J.M., Willits, D.A., Young, M.J., and Zlotnick, A. (2004).
Interaction with capsid protein alters RNA structure and the pathway for in
vitro assembly of cowpea chlorotic mottle virus. J Mol Biol 335, 455-464.
Intamaso, U. (2003). Dissertation - Development of a Mimotope-Based
Synthetic Peptide Vaccine Against HIV Using Plant Viruses, Montana
State University, Bozeman, MT.
Arnold, T. (2000). CCMV and Cancer, Montana State Univeristy,
Bozeman, MT.
Le Clainche, L., Plancque, G., Amekraz, B., Moulin, C., PradinesLecomte, C., Peltier, G., and Vita, C. (2003). Engineering new metal
specificity in EF-hand peptides. J Biol Inorg Chem 8, 334-340.
Jones, W.J., Leigh, J.A., Mayer, F., Woese, C.R., and Wolfe, R.S. (1983).
Methanococcus-Jannaschii Sp-Nov, an Extremely Thermophilic
Methanogen from a Submarine Hydrothermal Vent. Archives of
Microbiology 136, 254-261.
187
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
Bult, C.J., White, O., Olsen, G.J., Zhou, L., Fleischmann, R.D., Sutton,
G.G., Blake, J.A., FitzGerald, L.M., Clayton, R.A., Gocayne, J.D.,
Kerlavage, A.R., Dougherty, B.A., Tomb, J.F., Adams, M.D., Reich, C.I.,
Overbeek, R., Kirkness, E.F., Weinstock, K.G., Merrick, J.M., Glodek, A.,
Scott, J.L., Geoghagen, N.S., and Venter, J.C. (1996). Complete genome
sequence of the methanogenic archaeon, Methanococcus jannaschii.
Science 273, 1058-1073.
Kim, K.K., Yokota, H., Santoso, S., Lerner, D., Kim, R., and Kim, S.H.
(1998). Purification, crystallization, and preliminary X-ray crystallographic
data analysis of small heat shock protein homolog from Methanococcus
jannaschii, a hyperthermophile. J Struct Biol 121, 76-80.
Kim, R., Lai, L.H., Lee, H.H., Cheong, G.W., Kim, K.K., Wu, Z., Yokota, H.,
Marqusee, S., and Kim, S.H. (2003). On the mechanism of chaperone
activity of the small heat-shock protein of Methanococcus jannaschii.
Proceedings of the National Academy of Sciences of the United States of
America 100, 8151-8155.
Haley, D.A., Bova, M.P., Huang, Q.L., Mchaourab, H.S., and Stewart, P.L.
(2000). Small heat-shock protein structures reveal a continuum from
symmetric to variable assemblies. Journal of Molecular Biology 298, 261272.
Bova, M.P., Ding, L.L., Horwitz, J., and Fung, B.K. (1997). Subunit
exchange of alphaA-crystallin. J Biol Chem 272, 29511-29517.
Arrigo, A.P. (2005). Heat shock proteins as molecular chaperones. M SMedecine Sciences 21, 619-625.
Fu, X., Zhang, H., Zhang, X., Cao, Y., Jiao, W., Liu, C., Song, Y.,
Abulimiti, A., and Chang, Z. (2005). A dual role for the N-terminal region of
Mycobacterium tuberculosis Hsp16.3 in self-oligomerization and binding
denaturing substrate proteins. J Biol Chem 280, 6337-6348.
Brady, J.P., Garland, D., Duglas-Tabor, Y., Robison, W.G., Jr., Groome,
A., and Wawrousek, E.F. (1997). Targeted disruption of the mouse alpha
A-crystallin gene induces cataract and cytoplasmic inclusion bodies
containing the small heat shock protein alpha B-crystallin. Proc Natl Acad
Sci U S A 94, 884-889.
Horwitz, J. (1992). Alpha-crystallin can function as a molecular chaperone.
Proc Natl Acad Sci U S A 89, 10449-10453.
Kim, R., Kim, K.K., Yokota, H., and Kim, S.H. (1998). Small heat shock
protein of Methanococcus jannaschii, a hyperthermophile. Proc Natl Acad
Sci U S A 95, 9129-9133.
Kim, R., Lai, L., Lee, H.H., Cheong, G.W., Kim, K.K., Wu, Z., Yokota, H.,
Marqusee, S., and Kim, S.H. (2003). On the mechanism of chaperone
activity of the small heat-shock protein of Methanococcus jannaschii. Proc
Natl Acad Sci U S A 100, 8151-8155.
188
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
Tolgyesi, E., Bode, C.S., Smelleri, L., Kim, D.R., Kim, K.K., Heremans, K.,
and Fidy, J. (2004). Pressure activation of the chaperone function of small
heat shock proteins. Cell Mol Biol (Noisy-le-grand) 50, 361-369.
Klem, M., Willits, D., Solis, D.J., Belcher, A., Young, M., and Douglas, T.
(2005). Bio-Inspired Synthesis of Protein-Encapsulated CoPt
Nanoparticles. Advanced Functional Materials 15.
Liepold, L., Young, M., Douglas, T., and al., e. unpublished data.
Varpness, Z., Peters, J.W., Young, M., and Douglas, T. (2005).
Biomimetic synthesis of a H-2 catalyst using a protein cage architecture.
Nano Letters 5, 2306-2309.
Whaley, S.R., English, D.S., Hu, E.L., Barbara, P.F., and Belcher, A.M.
(2000). Selection of Peptides with Semiconductor Binding Specificity for
Directed Nanocrystal Assembly. Nature 405, 665-668.
Liepold, L., Willits, D., Young, M., and Douglas, T. Lanthanide binding to
an EF-hand motif engineered within protein cage architectures. in
preparation.
Arosio, P., and Levi, S. (2002). Ferritin, iron homeostasis, and oxidative
damage. Free Radical Biology and Medicine 33, 457-463.
Harrison, P.M., and Arosio, P. (1996). Ferritins: Molecular properties, iron
storage function and cellular regulation. Biochimica Et Biophysica ActaBioenergetics 1275, 161-203.
Harrison, P.M., Banyard, S.H., Hoare, R.J., Russell, S.M., and Treffry, A.
(1976). The structure and function of ferritin. Ciba Found Symp, 19-40.
Klem, M., Young, M., and Douglas, T. (2005). Biomimetic magnetic
nanoparticles. Materials Today, 28-37.
Mann, S. (1996). Biomimmetic Materials Chemistry (New York City: VCH).
Mann, S., Archibald, D.D., Didymus, J.M., Douglas, T., Heywood, B.R.,
Meldrum, F.C., and Reeves, N.J. (1993). Crystallization at inorganicorganic interfaces: biominerals and biomimetic synthesis. Science
(Washington, D. C., 1883-) 261, 1286-1292.
Yamashita, I., Hayashi, J., and Hara, M. (2004). Bio-template synthesis of
uniform CdSe nanoparticles using cage-shaped protein, apoferritin.
Chemistry Letters 33, 1158-1159.
Ensign, D., Young, M.J., and Douglas, T. (2004). Photocatalytic Synthesis
of Copper Colloids from Cu(II) by the Ferrihydrite Core of Ferritin. Inorg.
Chem. 43, 3441-3446.
Hosein, H.A., Strongin, D.R., Allen, M., and Douglas, T. (2004). Iron and
cobalt oxide and metallic nanoparticles prepared from ferritin. Langmuir
20, 10283-10287.
Okuda, M., Iwahori, K., Yamashita, I., and Yoshimura, H. (2003).
Fabrication of nickel and chromium nanoparticles using the protein cage of
apoferritin. Biotechnol Bioeng 84, 187-194.
Su, M., Cavallo, S., Stefanini, S., Chiancone, E., and Chasteen, N.D.
(2005). The so-called Listeria innocua ferritin is a Dps protein. Iron
189
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
incorporation, detoxification, and DNA protection properties. Biochemistry
44, 5572-5578.
Reddy, V.S., Natarajan, P., Okerberg, B., Li, K., Damodaran, K.V., Ryan,
T.M., III, C.L.B., and Johnson, J.E. (2001). Virus Partilce Explorer
(VIPER), a Website for Virus Capsid Structures and Their Computational
Analyses. Journal of Virology 75, 11943-11947.
Hermanson, G.T. (1996). Bioconjugate Techniques (San Diego: Academic
Press).
Douglas, T., and Young, M. (1999). Virus particles as templates for
materials synthesis. Advanced Materials 11, 679-681.
Douglas, T., Allen, M., and Young, M. (2002). Self-assembling Protein
Cage Systems and Applications in Nanotechnology. In Polyamides and
Complex Proteinaceous Materials I, Volume 7, S.R. Fahnestock and A.
Steinbuchel, eds. (Weinheim: Wiley-VCH), p. 517.
Maeda, H., Wu, J., Sawa, T., Matsumura, Y., and Hori, K. (2000). Tumor
vascular permeability and the EPR effect in macromolecular therapeutics:
a review. J Control Release 65, 271-284.
Hashizume, H., Baluk, P., Morikawa, S., McLean, J.W., Thurston, G.,
Roberge, S., Jain, R.K., and McDonald, D.M. (2000). Openings between
defective endothelial cells explain tumor vessel leakiness. Am J Pathol
156, 1363-1380.
Koo, M., Bendahmane, M., Lettieri, G.A., Paoletti, A.D., Lane, T.E.,
Fitchen, J.H., Buchmeier, M.J., and Beachy, R.N. (1999). Protective
immunity against murine hepatitis virus (MHV) induced by intranasal or
subcutaneous administration of hybrids of tobacco mosaic virus that
carries an MHV epitope. Proc Natl Acad Sci U S A 96, 7774-7779.
Yusibov, V., Hooper, D.C., Spitsin, S.V., Fleysh, N., Kean, R.B.,
Mikheeva, T., Deka, D., Karasev, A., Cox, S., Randall, J., and Koprowski,
H. (2002). Expression in plants and immunogenicity of plant virus-based
experimental rabies vaccine. Vaccine 20, 3155-3164.
Portney, N.G., Singh, K., Chaudhary, S., Destito, G., Schneemann, A.,
Manchester, M., and Ozkan, M. (2005). Organic and inorganic
nanoparticle hybrids. Langmuir 21, 2098-2103.
Ewers, H., Smith, A.E., Sbalzarini, I.F., Lilie, H., Koumoutsakos, P., and
Helenius, A. (2005). Single-particle tracking of murine polyoma virus-like
particles on live cells and artificial membranes. Proc Natl Acad Sci U S A
102, 15110-15115.
Gleiter, S., and Lilie, H. (2003). Cell-type specific targeting and gene
expression using a variant of polyoma VP1 virus-like particles. Biol Chem
384, 247-255.
Singh, P., Destito, G., Schneemann, A., and Manchester, M. (2006).
Canine parvovirus-like particles, a novel nanomaterial for tumor targeting.
J Nanobiotechnology 4, 2.
190
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
Peabody, D.S. (2003). A Viral Platform for Chemical Modification and
Multivalent Display. J Nanobiotechnology 1, 5.
Henning, P., Andersson, K.M., Frykholm, K., Ali, A., Magnusson, M.K.,
Nygren, P.A., Granio, O., Hong, S.S., Boulanger, P., and Lindholm, L.
(2005). Tumor cell targeted gene delivery by adenovirus 5 vectors carrying
knobless fibers with antibody-binding domains. Gene Ther 12, 211-224.
Hong, S.S., Magnusson, M.K., Henning, P., Lindholm, L., and Boulanger,
P.A. (2003). Adenovirus stripping: a versatile method to generate
adenovirus vectors with new cell target specificity. Mol Ther 7, 692-699.
Douglas, J.T., Miller, C.R., Kim, M., Dmitriev, I., Mikheeva, G., Krasnykh,
V., and Curiel, D.T. (1999). A system for the propagation of adenoviral
vectors with genetically modified receptor specificities. Nat Biotechnol 17,
470-475.
Wickham, T.J., Segal, D.M., Roelvink, P.W., Carrion, M.E., Lizonova, A.,
Lee, G.M., and Kovesdi, I. (1996). Targeted adenovirus gene transfer to
endothelial and smooth muscle cells by using bispecific antibodies. J Virol
70, 6831-6838.
Hedley, S.J., Chen, J., Mountz, J.D., Li, J., Curiel, D.T., Korokhov, N., and
Kovesdi, I. (2006). Targeted and shielded adenovectors for cancer
therapy. Cancer Immunol Immunother.
McMillan, R.A., Howard, J., Zaluzec, N.J., Kagawa, H.K., Mogul, R., Li,
Y.F., Paavola, C.D., and Trent, J.D. (2005). A self-assembling protein
template for constrained synthesis and patterning of nanoparticle arrays. J
Am Chem Soc 127, 2800-2801.
McMillan RA, P.C., Howard J, Chan SL, Zaluzec NJ, Trent JD (2002).
Ordered nanoparticle arrays formed on engineered chaperonin protein
templates. Nature Materials 1, 247-252.
Mastrobattista, E., Schoen, P., Wilschut, J., Crommelin, D.J., and Storm,
G. (2001). Targeting influenza virosomes to ovarian carcinoma cells.
FEBS Lett 509, 71-76.
Mastrobattista, E., van der Aa, M.A., Hennink, W.E., and Crommelin, D.J.
(2006). Artificial viruses: a nanotechnological approach to gene delivery.
Nat Rev Drug Discov 5, 115-121.
Dudek, T., and Knipe, D.M. (2006). Replication-defective viruses as
vaccines and vaccine vectors. Virology 344, 230-239.
Kaliberov, S.A., and Buchsbaum, D.J. (2006). Gene delivery and gene
therapy of prostate cancer. Expert Opin Drug Deliv 3, 37-51.
Ralph, G.S., Binley, K., Wong, L.F., Azzouz, M., and Mazarakis, N.D.
(2006). Gene therapy for neurodegenerative and ocular diseases using
lentiviral vectors. Clin Sci (Lond) 110, 37-46.
Young, L.S., Searle, P.F., Onion, D., and Mautner, V. (2006). Viral gene
therapy strategies: from basic science to clinical application. J Pathol 208,
299-318.
191
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
153.
Roberts, D.M., Nanda, A., Havenga, M.J., Abbink, P., Lynch, D.M., Ewald,
B.A., Liu, J., Thorner, A.R., Swanson, P.E., Gorgone, D.A., Lifton, M.A.,
Lemckert, A.A., Holterman, L., Chen, B., Dilraj, A., Carville, A., Mansfield,
K.G., Goudsmit, J., and Barouch, D.H. (2006). Hexon-chimaeric
adenovirus serotype 5 vectors circumvent pre-existing anti-vector
immunity. Nature 441, 239-243.
Harrop, R., and Carroll, M.W. (2006). Viral vectors for cancer
immunotherapy. Front Biosci 11, 804-817.
Nettelbeck, D.M., and Curiel, D.T. (2003). Tumor-Busting. Scientific
American, 68-75.
Thorne, S.H., Negrin, R.S., and Contag, C.H. (2006). Synergistic
antitumor effects of immune cell-viral biotherapy. Science 311, 1780-1784.
Aghi, M., and Martuza, R.L. (2005). Oncolytic viral therapies - the clinical
experience. Oncogene 24, 7802-7816.
Chernajovsky, Y., Layward, L., and Lemoine, N. (2006). Fighting cancer
with oncolytic viruses. Bmj 332, 170-172.
Geletneky, K., Herrero, Y.C.M., Rommelaere, J., and Schlehofer, J.R.
(2005). Oncolytic potential of rodent parvoviruses for cancer therapy in
humans: a brief review. J Vet Med B Infect Dis Vet Public Health 52, 327330.
Kaplan, J.M. (2005). Adenovirus-based cancer gene therapy. Curr Gene
Ther 5, 595-605.
O'Shea, C.C. (2005). Viruses - seeking and destroying the tumor program.
Oncogene 24, 7640-7655.
Parato, K.A., Senger, D., Forsyth, P.A., and Bell, J.C. (2005). Recent
progress in the battle between oncolytic viruses and tumours. Nat Rev
Cancer 5, 965-976.
Rein, D.T., Breidenbach, M., and Curiel, D.T. (2006). Current
developments in adenovirus-based cancer gene therapy. Future Oncol 2,
137-143.
Relph, K.L., Harrington, K.J., and Pandha, H. (2005). Adenoviral
strategies for the gene therapy of cancer. Semin Oncol 32, 573-582.
Shen, Y., and Nemunaitis, J. (2006). Herpes simplex virus 1 (HSV-1) for
cancer treatment. Cancer Gene Ther.
Thorne, S.H., Hermiston, T., and Kirn, D. (2005). Oncolytic virotherapy:
approaches to tumor targeting and enhancing antitumor effects. Semin
Oncol 32, 537-548.
Gromeier, M. (2001). Viruses as therapeutic agents against malignant
disease of the central nervous system. J Natl Cancer Inst 93, 889-890.
Gromeier, M., and Wimmer, E. (2001). Viruses for the treatment of
malignant glioma. Curr Opin Mol Ther 3, 503-508.
Ochiai, H., Campbell, S.A., Archer, G.E., Chewning, T.A., Dragunsky, E.,
Ivanov, A., Gromeier, M., and Sampson, J.H. (2006). Targeted therapy for
192
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
glioblastoma multiforme neoplastic meningitis with intrathecal delivery of
an oncolytic recombinant poliovirus. Clin Cancer Res 12, 1349-1354.
Porta, C., Spall, V.E., Loveland, J., Johnson, J.E., Barker, P.J., and
Lomonossoff, G.P. (1994). Development of cowpea mosaic virus as a
high-yielding system for the presentation of foreign peptides. Virology 202,
949-955.
Brennan, F.R., Jones, T.D., Gilleland, L.B., Bellaby, T., Xu, F., North, P.C.,
Thompson, A., Staczek, J., Lin, T., Johnson, J.E., Hamilton, W.D., and
Gilleland, H.E., Jr. (1999). Pseudomonas aeruginosa outer-membrane
protein F epitopes are highly immunogenic in mice when expressed on a
plant virus. Microbiology 145 ( Pt 1), 211-220.
Brennan, F.R., Jones, T.D., and Hamilton, W.D.O. (2001). Cowpea
mosaic virus as a vaccine carrier of heterologous antigens. Molecular
Biotechnology 17, 15-26.
Lewis, J.D., Destito, G., Zijlstra, A., Gonzalez, M.J., Quigley, J.P.,
Manchester, M., and Stuhlmann, H. (2006). Viral nanoparticles as tools for
intravital vascular imaging. Nat Med 12, 354-360.
Yasawardene, S.G., Lomonossoff, G.P., and Ramasamy, R. (2003).
Expression & immunogenicity of malaria merozoite peptides displayed on
the small coat protein of chimaeric cowpea mosaic virus. Indian Journal of
Medical Research 118, 115-124.
Porta, C., Spall, V.E., Lin, T., Johnson, J.E., and Lomonossoff, G.P.
(1996). The development of cowpea mosaic virus as a potential source of
novel vaccines. Intervirology 39, 79-84.
Porta, C., and Lomonossoff, G.P. (1998). Scope for using plant viruses to
present epitopes from animal pathogens. Rev Med Virol 8, 25-41.
Brennan, F.R., Jones, T.D., and Hamilton, W.D. (2001). Cowpea mosaic
virus as a vaccine carrier of heterologous antigens. Mol Biotechnol 17, 1526.
Brennan, F.R., Jones, T.D., Longstaff, M., Chapman, S., Bellaby, T.,
Smith, H., Xu, F., Hamilton, W.D., and Flock, J.I. (1999). Immunogenicity
of peptides derived from a fibronectin-binding protein of S. aureus
expressed on two different plant viruses. Vaccine 17, 1846-1857.
Langeveld, J.P., Brennan, F.R., Martinez-Torrecuadrada, J.L., Jones,
T.D., Boshuizen, R.S., Vela, C., Casal, J.I., Kamstrup, S., Dalsgaard, K.,
Meloen, R.H., Bendig, M.M., and Hamilton, W.D. (2001). Inactivated
recombinant plant virus protects dogs from a lethal challenge with canine
parvovirus. Vaccine 19, 3661-3670.
Mechtcheriakova, I.A., Eldarov, M.A., Nicholson, L., Shanks, M., Skryabin,
K.G., and Lomonossoff, G.P. (2006). The use of viral vectors to produce
hepatitis B virus core particles in plants. J Virol Methods 131, 10-15.
Usha, R., Rohll, J.B., Spall, V.E., Shanks, M., Maule, A.J., Johnson, J.E.,
and Lomonossoff, G.P. (1993). Expression of an animal virus antigenic
site on the surface of a plant virus particle. Virology 197, 366-374.
193
166.
167.
168.
169.
170.
171.
172.
173.
174.
175.
176.
177.
178.
179.
Xu, F., Jones, T.D., and Rodgers, P.B. (1996). Potential of chimaeric plant
virus particles as novel, stable vaccines. Dev Biol Stand 87, 201-205.
Porta, C., Spall, V.E., Findlay, K.C., Gergerich, R.C., Farrance, C.E., and
Lomonossoff, G.P. (2003). Cowpea mosaic virus-based chimaeras.
Effects of inserted peptides on the phenotype, host range, and
transmissibility of the modified viruses. Virology 310, 50-63.
Taylor, K.M., Lin, T., Porta, C., Mosser, A.G., Giesing, H.A., Lomonossoff,
G.P., and Johnson, J.E. (2000). Influence of three-dimensional structure
on the immunogenicity of a peptide expressed on the surface of a plant
virus. J Mol Recognit 13, 71-82.
Goldsby, R.A., Kindt, T.J., and Osborne, B.A. (2000). Kuby Immunology
(New York, NY: W.H. Freeman and Company).
Rae, C.S., Khor, I.W., Wang, Q., Destito, G., Gonzalez, M.J., Singh, P.,
Thomas, D.M., Estrada, M.N., Powell, E., Finn, M.G., and Manchester, M.
(2005). Systemic trafficking of plant virus nanoparticles in mice via the oral
route. Virology 343, 224-235.
Meloen, R.H., Hamilton, W.D., Casal, J.I., Dalsgaard, K., and Langeveld,
J.P. (1998). Edible vaccines. Vet Q 20 Suppl 3, S92-95.
Liu, L., Canizares, M.C., Monger, W., Perrin, Y., Tsakiris, E., Porta, C.,
Shariat, N., Nicholson, L., and Lomonossoff, G.P. (2005). Cowpea mosaic
virus-based systems for the production of antigens and antibodies in
plants. Vaccine 23, 1788-1792.
Strable, E., Johnson, J.E., and Finn, M.G. (2004). Natural Nanochemical
Building Blocks: Icosahedral Virus Particles Organized by Attached
Oligonucleotides. Nano Lett 4, 1385-1389.
Blum, A.S., Soto, C.M., Wilson, C.D., Cole, J.D., Kim, M., Gnade, B.,
Chatterji, A., Ochoa, W.F., Lin, T.W., Johnson, J.E., and Ratna, B.R.
(2004). Cowpea mosaic virus as a scaffold for 3-D patterning of gold
nanoparticles. Nano Letters 4, 867-870.
Medintz, I.L., Sapsford, K.E., Konnert, J.H., Chatterji, A., Lin, T.W.,
Johnson, J.E., and Mattoussi, H. (2005). Decoration of discretely
immobilized cowpea mosaic virus with luminescent quantum dots.
Langmuir 21, 5501-5510.
Punna, S., Kuzelka, J., Wang, Q., and Finn, M.G. (2005). Head-to-tail
peptide cyclodimerization by copper-catalyzed azide-alkyne cycloaddition.
Angew Chem Int Ed Engl 44, 2215-2220.
Wang, Q., Chan, T.R., Hilgraf, R., Fokin, V.V., Sharpless, K.B., and Finn,
M.G. (2003). Bioconjugation by copper(I)-catalyzed azide-alkyne [3 + 2]
cycloaddition. J Am Chem Soc 125, 3192-3193.
Tilley, S.D., and Francis, M.B. (2006). Tyrosine-selective protein alkylation
using pi-allylpalladium complexes. J Am Chem Soc 128, 1080-1081.
Warren, J.D., Miller, J.S., Keding, S.J., and Danishefsky, S.J. (2004).
Toward fully synthetic glycoproteins by ultimately convergent routes: a
solution to a long-standing problem. J Am Chem Soc 126, 6576-6578.
194
180.
181.
182.
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
Gurdag, S., Khandare, J., Stapels, S., Matherly, L.H., and Kannan, R.M.
(2006). Activity of Dendrimer-Methotrexate Conjugates on MethotrexateSensitive and Resistant Cell Lines. Bioconjug Chem 17, 275-283.
Duncan, R. (1999). Polymer conjugates for tumour targeting and
intracytoplasmic delivery. The EPR effect as a common gateway? Pharm.
Sci. Technol. Today 2, 441-449.
Roy, I., Ohulchanskyy, T.Y., Pudavar, H.E., Bergey, E.J., Oseroff, A.R.,
Morgan, J., Dougherty, T.J., and Prasad, P.N. (2003). Ceramic-based
nanoparticles entrapping water-insoluble photosensitizing anticancer
drugs: a novel drug-carrier system for photodynamic therapy. J Am Chem
Soc 125, 7860-7865.
Arap, W., Pasqualini, R., and Ruoslahti, E. (1998). Chemotherapy
targeted to tumor vasculature. Curr Opin Oncol 10, 560-565.
Braslawsky, G.R., Edson, M.A., Pearce, W., Kaneko, T., and Greenfield,
R.S. (1990). Antitumor activity of adriamycin (hydrazone-linked)
immunoconjugates compared with free adriamycin and specificity of tumor
cell killing. Cancer Res 50, 6608-6614.
Braslawsky, G.R., Kadow, K., Knipe, J., McGoff, K., Edson, M., Kaneko,
T., and Greenfield, R.S. (1991). Adriamycin(hydrazone)-antibody
conjugates require internalization and intracellular acid hydrolysis for
antitumor activity. Cancer Immunol Immunother 33, 367-374.
Brigger, I., Dubernet, C., and Couvreur, P. (2002). Nanoparticles in cancer
therapy and diagnosis. Adv Drug Deliv Rev 54, 631-651.
Ellerby, H.M., Arap, W., Ellerby, L.M., Kain, R., Andrusiak, R., Rio, G.D.,
Krajewski, S., Lombardo, C.R., Rao, R., Ruoslahti, E., Bredesen, D.E.,
and Pasqualini, R. (1999). Anti-cancer activity of targeted pro-apoptotic
peptides. Nat Med 5, 1032-1038.
Emerich, D.F., and Thanos, C.G. (2003). Nanotechnology and medicine.
Expert Opin Biol Ther 3, 655-663.
Fundaro, A., Cavalli, R., Bargoni, A., Vighetto, D., Zara, G.P., and Gasco,
M.R. (2000). Non-stealth and stealth solid lipid nanoparticles (SLN)
carrying doxorubicin: pharmacokinetics and tissue distribution after i.v.
administration to rats. Pharmacol Res 42, 337-343.
Hallahan, D., Geng, L., Qu, S., Scarfone, C., Giorgio, T., Donnelly, E.,
Gao, X., and Clanton, J. (2003). Integrin-mediated targeting of drug
delivery to irradiated tumor blood vessels. Cancer Cell 3, 63-74.
Jana, S.S., Bharali, D.J., Mani, P., Maitra, A., Gupta, C.M., and Sarkar,
D.P. (2002). Targeted cytosolic delivery of hydrogel nanoparticles into
HepG2 cells through engineered Sendai viral envelopes. FEBS Lett 515,
184-188.
Janes, K.A., Calvo, P., and Alonso, M.J. (2001). Polysaccharide colloidal
particles as delivery systems for macromolecules. Adv Drug Deliv Rev 47,
83-97.
195
193.
194.
195.
196.
197.
198.
199.
200.
201.
202.
203.
204.
205.
206.
Janes, K.A., Fresneau, M.P., Marazuela, A., Fabra, A., and Alonso, M.J.
(2001). Chitosan nanoparticles as delivery systems for doxorubicin. J
Control Release 73, 255-267.
Laakkonen, P., Akerman, M.E., Biliran, H., Yang, M., Ferrer, F., Karpanen,
T., Hoffman, R.M., and Ruoslahti, E. (2004). Antitumor activity of a homing
peptide that targets tumor lymphatics and tumor cells. Proc Natl Acad Sci
U S A 101, 9381-9386.
Laakkonen, P., Porkka, K., Hoffman, J.A., and Ruoslahti, E. (2002). A
tumor-homing peptide with a targeting specificity related to lymphatic
vessels. Nat Med 8, 751-755.
Lanza, G.M., Abendschein, D.R., Yu, X., Winter, P.M., Karukstis, K.K.,
Scott, M.J., Fuhrhop, R.W., Scherrer, D.E., and Wickline, S.A. (2002).
Molecular imaging and targeted drug delivery with a novel, ligand-directed
paramagnetic nanoparticle technology. Acad Radiol 9 Suppl 2, S330-331.
Muller, K., Nahde, T., Fahr, A., Muller, R., and Brusselbach, S. (2001).
Highly efficient transduction of endothelial cells by targeted artificial viruslike particles. Cancer Gene Ther 8, 107-117.
Na, K., Lee, E.S., and Bae, Y.H. (2003). Adriamycin loaded pullulan
acetate/sulfonamide conjugate nanoparticles responding to tumor pH: pHdependent cell interaction, internalization and cytotoxicity in vitro. J
Control Release 87, 3-13.
Na, K., and Bae, Y.H. (2002). Self-assembled hydrogel nanoparticles
responsive to tumor extracellular pH from pullulan derivative/sulfonamide
conjugate: characterization, aggregation, and adriamycin release in vitro.
Pharm Res 19, 681-688.
Nahde, T., Muller, K., Fahr, A., Muller, R., and Brusselbach, S. (2001).
Combined transductional and transcriptional targeting of melanoma cells
by artificial virus-like particles. J Gene Med 3, 353-361.
Wickline, S.A., and Lanza, G.M. (2002). Molecular imaging, targeted
therapeutics, and nanoscience. J Cell Biochem Suppl 39, 90-97.
Yoo, H.S., Lee, K.H., Oh, J.E., and Park, T.G. (2000). In vitro and in vivo
anti-tumor activities of nanoparticles based on doxorubicin-PLGA
conjugates. J Control Release 68, 419-431.
Choi, Y., Thomas, T., Kotlyar, A., Islam, M.T., and Baker, J.R., Jr. (2005).
Synthesis and functional evaluation of DNA-assembled polyamidoamine
dendrimer clusters for cancer cell-specific targeting. Chem Biol 12, 35-43.
Damge, C., Aprahamian, M., Humbert, W., and Pinget, M. (2000). Ileal
uptake of polyalkylcyanoacrylate nanocapsules in the rat. J Pharm
Pharmacol 52, 1049-1056.
Damge, C., Michel, C., Aprahamian, M., and Couvreur, P. (1988). New
approach for oral administration of insulin with polyalkylcyanoacrylate
nanocapsules as drug carrier. Diabetes 37, 246-251.
Brayden, D.J., and Baird, A.W. (2001). Microparticle vaccine approaches
to stimulate mucosal immunisation. Microbes Infect 3, 867-876.
196
207.
208.
209.
210.
211.
212.
213.
214.
215.
216.
Mathiowitz, E., Jacob, J.S., Jong, Y.S., Carino, G.P., Chickering, D.E.,
Chaturvedi, P., Santos, C.A., Vijayaraghavan, K., Montgomery, S.,
Bassett, M., and Morrell, C. (1997). Biologically erodable microspheres as
potential oral drug delivery systems. Nature 386, 410-414.
Dubowchik, G.M., Radia, S., Mastalerz, H., Walker, M.A., Firestone, R.A.,
Dalton King, H., Hofstead, S.J., Willner, D., Lasch, S.J., and Trail, P.A.
(2002). Doxorubicin immunoconjugates containing bivalent, lysosomallycleavable dipeptide linkages. Bioorg Med Chem Lett 12, 1529-1532.
Hurwitz, H., Fehrenbacher, L., Novotny, W., Cartwright, T., Hainsworth, J.,
Heim, W., Berlin, J., Baron, A., Griffing, S., Holmgren, E., Ferrara, N.,
Fyfe, G., Rogers, B., Ross, R., and Kabbinavar, F. (2004). Bevacizumab
plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal
cancer. N Engl J Med 350, 2335-2342.
Jendreyko, N., Popkov, M., Beerli, R.R., Chung, J., McGavern, D.B.,
Rader, C., and Barbas, C.F., 3rd (2003). Intradiabodies, bispecific,
tetravalent antibodies for the simultaneous functional knockout of two cell
surface receptors. J Biol Chem 278, 47812-47819.
King, H.D., Dubowchik, G.M., Mastalerz, H., Willner, D., Hofstead, S.J.,
Firestone, R.A., Lasch, S.J., and Trail, P.A. (2002). Monoclonal antibody
conjugates of doxorubicin prepared with branched peptide linkers:
inhibition of aggregation by methoxytriethyleneglycol chains. J Med Chem
45, 4336-4343.
King, H.D., Yurgaitis, D., Willner, D., Firestone, R.A., Yang, M.B., Lasch,
S.J., Hellstrom, K.E., and Trail, P.A. (1999). Monoclonal antibody
conjugates of doxorubicin prepared with branched linkers: A novel method
for increasing the potency of doxorubicin immunoconjugates. Bioconjug
Chem 10, 279-288.
Trail, P.A., Willner, D., Lasch, S.J., Henderson, A.J., Greenfield, R.S.,
King, D., Zoeckler, M.E., and Braslawsky, G.R. (1992). Antigen-specific
activity of carcinoma-reactive BR64-doxorubicin conjugates evaluated in
vitro and in human tumor xenograft models. Cancer Res 52, 5693-5700.
Cortez-Retamozo, V., Backmann, N., Senter, P.D., Wernery, U., De
Baetselier, P., Muyldermans, S., and Revets, H. (2004). Efficient cancer
therapy with a nanobody-based conjugate. Cancer Res 64, 2853-2857.
Doronina, S.O., Toki, B.E., Torgov, M.Y., Mendelsohn, B.A., Cerveny,
C.G., Chace, D.F., DeBlanc, R.L., Gearing, R.P., Bovee, T.D., Siegall,
C.B., Francisco, J.A., Wahl, A.F., Meyer, D.L., and Senter, P.D. (2003).
Development of potent monoclonal antibody auristatin conjugates for
cancer therapy. Nat Biotechnol 21, 778-784.
Francisco, J.A., Cerveny, C.G., Meyer, D.L., Mixan, B.J., Klussman, K.,
Chace, D.F., Rejniak, S.X., Gordon, K.A., DeBlanc, R., Toki, B.E., Law,
C.L., Doronina, S.O., Siegall, C.B., Senter, P.D., and Wahl, A.F. (2003).
cAC10-vcMMAE, an anti-CD30-monomethyl auristatin E conjugate with
potent and selective antitumor activity. Blood 102, 1458-1465.
197
217.
218.
219.
220.
221.
222.
223.
224.
225.
226.
227.
Hellstrom, I., Hellstrom, K.E., and Senter, P.D. (2001). Development and
activities of the BR96-doxorubicin immunoconjugate. Methods Mol Biol
166, 3-16.
Rader, C., Sinha, S.C., Popkov, M., Lerner, R.A., and Barbas, C.F., 3rd
(2003). Chemically programmed monoclonal antibodies for cancer
therapy: adaptor immunotherapy based on a covalent antibody catalyst.
Proc Natl Acad Sci U S A 100, 5396-5400.
Trail, P.A., Willner, D., Lasch, S.J., Henderson, A.J., Hofstead, S.,
Casazza, A.M., Firestone, R.A., Hellstrom, I., and Hellstrom, K.E. (1993).
Cure of xenografted human carcinomas by BR96-doxorubicin
immunoconjugates. Science 261, 212-215.
Ludwig, D.L., Pereira, D.S., Zhu, Z., Hicklin, D.J., and Bohlen, P. (2003).
Monoclonal antibody therapeutics and apoptosis. Oncogene 22, 90979106.
Brooks, P.C., Montgomery, A.M., Rosenfeld, M., Reisfeld, R.A., Hu, T.,
Klier, G., and Cheresh, D.A. (1994). Integrin alpha v beta 3 antagonists
promote tumor regression by inducing apoptosis of angiogenic blood
vessels. Cell 79, 1157-1164.
Brooks, P.C., Stromblad, S., Klemke, R., Visscher, D., Sarkar, F.H., and
Cheresh, D.A. (1995). Antiintegrin alpha v beta 3 blocks human breast
cancer growth and angiogenesis in human skin. J Clin Invest 96, 18151822.
Rader, C., Popkov, M., Neves, J.A., and Barbas, C.F., 3rd (2002). Integrin
alpha(v)beta3 targeted therapy for Kaposi's sarcoma with an in vitro
evolved antibody. Faseb J 16, 2000-2002.
Kobayashi, H., Kawamoto, S., Saga, T., Sato, N., Hiraga, A., Ishimori, T.,
Akita, Y., Mamede, M.H., Konishi, J., Togashi, K., and Brechbiel, M.W.
(2001). Novel liver macromolecular MR contrast agent with a
polypropylenimine diaminobutyl dendrimer core: comparison to the
vascular MR contrast agent with the polyamidoamine dendrimer core.
Magn Reson Med 46, 795-802.
Kobayashi, H., Kawamoto, S., Saga, T., Sato, N., Hiraga, A., Ishimori, T.,
Konishi, J., Togashi, K., and Brechbiel, M.W. (2001). Positive effects of
polyethylene glycol conjugation to generation-4 polyamidoamine
dendrimers as macromolecular MR contrast agents. Magn Reson Med 46,
781-788.
Kobayashi, H., Sato, N., Kawamoto, S., Saga, T., Hiraga, A., Ishimori, T.,
Konishi, J., Togashi, K., and Brechbiel, M.W. (2001). Novel intravascular
macromolecular MRI contrast agent with generation-4 polyamidoamine
dendrimer core: accelerated renal excretion with coinjection of lysine.
Magn Reson Med 46, 457-464.
Kobayashi, H., Shirakawa, K., Kawamoto, S., Saga, T., Sato, N., Hiraga,
A., Watanabe, I., Heike, Y., Togashi, K., Konishi, J., Brechbiel, M.W., and
Wakasugi, H. (2002). Rapid accumulation and internalization of
198
228.
229.
230.
231.
232.
233.
234.
235.
236.
237.
238.
radiolabeled herceptin in an inflammatory breast cancer xenograft with
vasculogenic mimicry predicted by the contrast-enhanced dynamic MRI
with the macromolecular contrast agent G6-(1B4M-Gd)(256). Cancer Res
62, 860-866.
Sato, N., Kobayashi, H., Saga, T., Nakamoto, Y., Ishimori, T., Togashi, K.,
Fujibayashi, Y., Konishi, J., and Brechbiel, M.W. (2001). Tumor targeting
and imaging of intraperitoneal tumors by use of antisense oligo-DNA
complexed with dendrimers and/or avidin in mice. Clin Cancer Res 7,
3606-3612.
Portney, N.G., and Ozkan, M. (2006). Nano-oncology: drug delivery,
imaging, and sensing. Anal Bioanal Chem 384, 620-630.
Cloninger, M. (2004). Dendrimers and protein cages as nanoparticles in
drug delivery. Drug Discov Today 9, 111-112.
Alberts, D.S., and Garcia, D.J. (1997). Safety aspects of pegylated
liposomal doxorubicin in patients with cancer. Drugs 54 Suppl 4, 30-35.
Vasey, P.A., Kaye, S.B., Morrison, R., Twelves, C., Wilson, P., Duncan,
R., Thomson, A.H., Murray, L.S., Hilditch, T.E., Murray, T., Burtles, S.,
Fraier, D., Frigerio, E., and Cassidy, J. (1999). Phase I clinical and
pharmacokinetic study of PK1 [N-(2-hydroxypropyl)methacrylamide
copolymer doxorubicin]: first member of a new class of chemotherapeutic
agents-drug-polymer conjugates. Cancer Research Campaign Phase I/II
Committee. Clin Cancer Res 5, 83-94.
Lee, R.J., and Low, P.S. (1994). Delivery of liposomes into cultured KB
cells via folate receptor-mediated endocytosis. J Biol Chem 269, 31983204.
Lee, R.J., and Low, P.S. (1995). Folate-mediated tumor cell targeting of
liposome-entrapped doxorubicin in vitro. Biochim Biophys Acta 1233, 134144.
Stephenson, S.M., Low, P.S., and Lee, R.J. (2004). Folate receptormediated targeting of liposomal drugs to cancer cells. Methods Enzymol
387, 33-50.
Hamblett, K.J., Senter, P.D., Chace, D.F., Sun, M.M., Lenox, J., Cerveny,
C.G., Kissler, K.M., Bernhardt, S.X., Kopcha, A.K., Zabinski, R.F., Meyer,
D.L., and Francisco, J.A. (2004). Effects of drug loading on the antitumor
activity of a monoclonal antibody drug conjugate. Clin Cancer Res 10,
7063-7070.
Weissleder, R., Kelly, K., Sun, E.Y., Shtatland, T., and Josephson, L.
(2005). Cell-specific targeting of nanoparticles by multivalent attachment
of small molecules. Nat Biotechnol 23, 1418-1423.
Lee, S.W., Mao, C., Flynn, C.E., and Belcher, A.M. (2002). Ordering of
quantum dots using genetically engineered viruses. Science 296, 892895.
199
239.
240.
241.
242.
243.
244.
245.
246.
247.
248.
249.
250.
251.
252.
253.
254.
255.
Kim, I., Hosein, H.A., Strongin, D.R., and Douglas, T. (2002).
Photochemical Reactivity of Ferritin for Cr(VI) Reduction. Chem. Mater.
14, 4874 -4879.
Haughland, R. (1996). Handbook of Fluorescent Probes and Research
Chemicals, 6th Edition (Eugene OR: Molecular Probes).
Chasteen, N.D., and Harrison, P.M. (1999). Mineralization in ferritin: an
efficient means of iron storage. J Struct Biol 126, 182-194.
Lepidocrocite (g-FeOOH) JCPDS file 08-0098.
Ferrihydrite (Fe5O7(OH)-4H2O JCPDS file 29-0712.
Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K., and Walter, P.
(2002). Molecular Biology of the Cell, 4th Edition (New York, NY: Garland
Science).
(2006). Cancer. In World Health Organization, vol. 2006.
Laurent, S., Colbert, N., Izrael, V., and Motte, G. (1984). [Cardiotoxicity of
anthracyclines]. Ann Med Interne (Paris) 135, 464-472.
Trail, P.A., Willner, D., Knipe, J., Henderson, A.J., Lasch, S.J., Zoeckler,
M.E., TrailSmith, M.D., Doyle, T.W., King, H.D., Casazza, A.M.,
Braslawsky, G.R., Brown, J., Hofstead, S.J., Greenfield, R.S., Firestone,
R.A., Mosure, K., Kadow, K.F., Yang, M.B., Hellstrom, K.E., and
Hellstrom, I. (1997). Effect of linker variation on the stability, potency, and
efficacy of carcinoma-reactive BR64-doxorubicin immunoconjugates.
Cancer Res 57, 100-105.
Kaneko, T., Willner, D., Monkovic, I., Knipe, J.O., Braslawsky, G.R.,
Greenfield, R.S., and Vyas, D.M. (1991). New hydrazone derivatives of
adriamycin and their immunoconjugates--a correlation between acid
stability and cytotoxicity. Bioconjug Chem 2, 133-141.
Brown, K.C. (2004). Hitting the target with bifunctional phage. Chem Biol
11, 1033-1035.
Ruoslahti, E. (1996). RGD and other recognition sequences for integrins.
Annu Rev Cell Dev Biol 12, 697-715.
Ruoslahti, E. (2000). Targeting tumor vasculature with homing peptides
from phage display. Semin Cancer Biol 10, 435-442.
Ruoslahti, E. (2002). Specialization of tumour vasculature. Nat Rev
Cancer 2, 83-90.
Scott, J.K., and Smith, G.P. (1990). Searching for peptide ligands with an
epitope library. Science 249, 386-390.
Arap, W., Haedicke, W., Bernasconi, M., Kain, R., Rajotte, D., Krajewski,
S., Ellerby, H.M., Bredesen, D.E., Pasqualini, R., and Ruoslahti, E. (2002).
Targeting the prostate for destruction through a vascular address. Proc
Natl Acad Sci USA 99, 1527-1531.
Rajotte, D., and Ruoslahti, E. (1999). Membrane dipeptidase is the
receptor for a lung-targeting peptide identified by in vivo phage display. J
Biol Chem 274, 11593-11598.
200
256.
257.
258.
259.
260.
261.
262.
263.
264.
265.
266.
267.
268.
Ruoslahti, E. (2002). Drug targeting to specific vascular sites. Drug Discov
Today 7, 1138-1143.
Akerman, M.E., Chan, W.C., Laakkonen, P., Bhatia, S.N., and Ruoslahti,
E. (2002). Nanocrystal targeting in vivo. Proc Natl Acad Sci U S A 99,
12617-12621.
de Groot, F.M., Broxterman, H.J., Adams, H.P., van Vliet, A., Tesser, G.I.,
Elderkamp, Y.W., Schraa, A.J., Kok, R.J., Molema, G., Pinedo, H.M., and
Scheeren, H.W. (2002). Design, synthesis, and biological evaluation of a
dual tumor-specific motive containing integrin-targeted plasmin-cleavable
doxorubicin prodrug. Mol Cancer Ther 1, 901-911.
Wermuth, J., Goodman, S.L., Jonczyk, A., and Kessler, H. (1997).
Stereoisomerism and biological activity of the selective and superactice
avb3 integrin inhibitor cyclo(-RGDfV-) and its retro-inverso peptide.
Journal of the American Chemical Society 119, 1328-1335.
Koivunen, E., Wang, B., and Ruoslahti, E. (1995). Phage libraries
displaying cyclic peptides with different ring sizes: ligand specificities of
the RGD-directed integrins. Biotechnology (N Y) 13, 265-270.
Brooks, P.C., Clark, R.A., and Cheresh, D.A. (1994). Requirement of
vascular integrin alpha v beta 3 for angiogenesis. Science 264, 569-571.
Friedlander, M., Brooks, P.C., Shaffer, R.W., Kincaid, C.M., Varner, J.A.,
and Cheresh, D.A. (1995). Definition of two angiogenic pathways by
distinct alpha v integrins. Science 270, 1500-1502.
Kim, J.W., and Lee, H.S. (2004). Tumor targeting by doxorubicin-RGD-4C
peptide conjugate in an orthotopic mouse hepatoma model. Int J Mol Med
14, 529-535.
Chen, X., Plasencia, C., Hou, Y., and Neamati, N. (2005). Synthesis and
biological evaluation of dimeric RGD peptide-paclitaxel conjugate as a
model for integrin-targeted drug delivery. J Med Chem 48, 1098-1106.
Wickham, T.J., Carrion, M.E., and Kovesdi, I. (1995). Targeting of
adenovirus penton base to new receptors through replacement of its RGD
motif with other receptor-specific peptide motifs. Gene Ther 2, 750-756.
Dmitriev, I., Krasnykh, V., Miller, C.R., Wang, M., Kashentseva, E.,
Mikheeva, G., Belousova, N., and Curiel, D.T. (1998). An adenovirus
vector with genetically modified fibers demonstrates expanded tropism via
utilization of a coxsackievirus and adenovirus receptor-independent cell
entry mechanism. J Virol 72, 9706-9713.
Chen, L., Zurita, A.J., Ardelt, P.U., Giordano, R.J., Arap, W., and
Pasqualini, R. (2004). Design and validation of a bifunctional ligand
display system for receptor targeting. Chem Biol 11, 1081-1091.
Grifman, M., Trepel, M., Speece, P., Gilbert, L.B., Arap, W., Pasqualini,
R., and Weitzman, M.D. (2001). Incorporation of tumor-targeting peptides
into recombinant adeno-associated virus capsids. Mol Ther 3, 964-975.
201
269.
270.
271.
272.
273.
274.
275.
276.
277.
278.
279.
280.
281.
282.
283.
Burkhart, D.J., Kalet, B.T., Coleman, M.P., Post, G.C., and Koch, T.H.
(2004). Doxorubicin-formaldehyde conjugates targeting alphavbeta3
integrin. Mol Cancer Ther 3, 1593-1604.
Su, Z.F., Liu, G., Gupta, S., Zhu, Z., Rusckowski, M., and Hnatowich, D.J.
(2002). In vitro and in vivo evaluation of a Technetium-99m-labeled cyclic
RGD peptide as a specific marker of alpha(V)beta(3) integrin for tumor
imaging. Bioconjug Chem 13, 561-570.
Zitzmann, S., Ehemann, V., and Schwab, M. (2002). Arginine-glycineaspartic acid (RGD)-peptide binds to both tumor and tumor-endothelial
cells in vivo. Cancer Res 62, 5139-5143.
Fahr, A., Muller, K., Nahde, T., Muller, R., and Brusselbach, S. (2002). A
new colloidal lipidic system for gene therapy. J Liposome Res 12, 37-44.
DeNardo, S.J., Burke, P.A., Leigh, B.R., O'Donnell, R.T., Miers, L.A.,
Kroger, L.A., Goodman, S.L., Matzku, S., Jonczyk, A., Lamborn, K.R., and
DeNardo, G.L. (2000). Neovascular targeting with cyclic RGD peptide
(cRGDf-ACHA) to enhance delivery of radioimmunotherapy. Cancer
Biother Radiopharm 15, 71-79.
Burke, P.A., DeNardo, S.J., Miers, L.A., Lamborn, K.R., Matzku, S., and
DeNardo, G.L. (2002). Cilengitide targeting of alpha(v)beta(3) integrin
receptor synergizes with radioimmunotherapy to increase efficacy and
apoptosis in breast cancer xenografts. Cancer Res 62, 4263-4272.
Smith, J.W. (2003). Cilengitide Merck. Curr Opin Investig Drugs 4, 741745.
Uchida, M., Willits, D., Flenniken, M.L., Allen, M., Crowley, B., Young, M.,
and Douglas, T. (2006). Cancer cell targeted ferritin-iron oxide
nanocompoiste for hyperthemic treatment and as an MR imaging agent. in
preparation.
Assa-Munt, N., Jia, X., Laakkonen, P., and Ruoslahti, E. (2001). Solution
structures and integrin binding activities of an RGD peptide with two
isomers. Biochemistry 40, 2373-2378.
Gao, A.G., Lindberg, F.P., Dimitry, J.M., Brown, E.J., and Frazier, W.A.
(1996). Thrombospondin modulates alpha v beta 3 function through
integrin-associated protein. J Cell Biol 135, 533-544.
Hellwage, J., Kuhn, S., and Zipfel, P.F. (1997). The human complement
regulatory factor-H-like protein 1, which represents a truncated form of
factor H, displays cell-attachment activity. Biochem J 326 ( Pt 2), 321-327.
Cancer Medicine e.5, 5th Edition (Hamilton, Ontario: BC Decker Inc. ).
Flenniken, M.L., Willits, D.A., Harmsen, A.L., Liepold, L.O., Harmsen,
A.G., Young, M.J., and Douglas, T. (2006). Melanoma and lymphocyte
cell-specific targeting incorporated into a heat shock protein cage
architecture. Chem Biol 13, 161-170.
Crowley, B., and Flenniken, M.L. (2005). unpublished data.
Roth, M. The CV-KL cell line was derived from the African Green Monkey
Kidney cell line CV-1. .
202
284.
285.
286.
287.
288.
289.
290.
291.
292.
293.
294.
295.
296.
Bulte, J.W., Arbab, A.S., Douglas, T., and Frank, J.A. (2004). Preparation
of Magnetically Labeled Cells for Cell Tracking by Magnetic Resonanace
Imaging, Volume 386 (SanDiego, CA: Elsevier Academic Press).
Cory, A.H., Owen, T.C., Barltrop, J.A., and Cory, J.G. (1991). Use of an
aqueous soluble tetrazolium/formazan assay for cell growth assays in
culture. Cancer Commun 3, 207-212.
Riss, T.L., and Moravec, R.A. (2004). Use of multiple assay endpoints to
investigate the effects of incubation time, dose of toxin, and plating density
in cell-based cytotoxicity assays. Assay Drug Dev Technol 2, 51-62.
Aumailley, M., Mann, K., von der Mark, H., and Timpl, R. (1989). Cell
attachment properties of collagen type VI and Arg-Gly-Asp dependent
binding to its alpha 2(VI) and alpha 3(VI) chains. Exp Cell Res 181, 463474.
Flenniken, M.L., Willits, D.A., Brumfield, S., Young, M.J., and Douglas, T.
(2003). The Small Heat Shock Protein Cage from Methanococcus
jannaschii Is a Versatile Nanoscale Platform for Genetic and Chemical
Modification. NanoLetters 3, 1573-1576.
Service, R.F. (2005). Nanotechnology. Calls rise for more research on
toxicology of nanomaterials. Science 310, 1609.
Kaiser, C., Harmsen, A.L., Flenniken, M.L., Willits, D., Douglas, T., Young,
M.J., and Harmsen, A.G. (2006). Bio-distribution of Nano-meter Sized
Protein Cage (Hsp and CCMV) Architectures in Mice. unpublished
observations.
Abedin, M.J. (2006). Click Chemistry Protocols. unpublished observations.
Sun, X., Rossin, R., Turner, J.L., Becker, M.L., Joralemon, M.J., Welch,
M.J., and Wooley, K.L. (2005). An assessment of the effects of shell
cross-linked nanoparticle size, core composition, and surface PEGylation
on in vivo biodistribution. Biomacromolecules 6, 2541-2554.
Duza, T., Flenniken, M.L., Teesalu, T., Sugahara, K., Douglas, T., Young,
M., and Ruoslahti, E. (2005-6). Data Obtained in Collaboration with Dr. E.
Ruoshahti's Lab at The Burnham Research Institute for Medical Research.
unpublished data.
Kelland, L.R. (2004). Of mice and men: values and liabilities of the
athymic nude mouse model in anticancer drug development. Eur J Cancer
40, 827-836.
Johnson, J.I., Decker, S., Zaharevitz, D., Rubinstein, L.V., Venditti, J.M.,
Schepartz, S., Kalyandrug, S., Christian, M., Arbuck, S., Hollingshead, M.,
and Sausville, E.A. (2001). Relationships between drug activity in NCI
preclinical in vitro and in vivo models and early clinical trials. Br J Cancer
84, 1424-1431.
Kaiser, C., Harmsen, A., Flenniken, M.L., Douglas, T., Young, M., and
Harmsen, A. (2005-6). Biodistribution of 125I-Labeled HspG41C,
HspG41C-GFE, and CCMV K42R in BALB/c Mice. Data obtained in
203
297.
298.
299.
300.
301.
302.
303.
304.
305.
306.
307.
308.
collaboration with Dr. A. Harmsen's Lab in the Veterinary Molecular
Biology Department at Montana State University. unpublished data.
Zhang, T., Breitbart, M., Lee, W.H., Run, J.Q., Wei, C.L., Soh, S.W.,
Hibberd, M.L., Liu, E.T., Rohwer, F., and Ruan, Y. (2006). RNA viral
community in human feces: prevalence of plant pathogenic viruses. PLoS
Biol 4, e3.
Awasthi, V., Meinken, G., Springer, K., Srivastava, S.C., and Freimuth, P.
(2004). Biodistribution of radioiodinated adenovirus fiber protein knob
domain after intravenous injection in mice. J Virol 78, 6431-6438.
Josefsson, M., Evilevitch, L., Westrom, B., Grunditz, T., and Ekblad, E.
(2006). Sodium-iodide symporter mediates iodide secretion in rat gastric
mucosa in vitro. Exp Biol Med (Maywood) 231, 277-281.
Josefsson, M., Grunditz, T., Ohlsson, T., and Ekblad, E. (2002).
Sodium/iodide-symporter: distribution in different mammals and role in
entero-thyroid circulation of iodide. Acta Physiol Scand 175, 129-137.
Buckner, D., Kaiser, C., Willits, D., Flenniken, M.L., Harmsen, A., Douglas,
T., Young, M., and Jutila, M. (2005- 2006). Initial Evaluation of the
Immune Response to Protein Cage Architectures. Data obtained in
collaboration with Dr. M. Jutila's lab in the Veterinary Molecular Biology
Department at Montana State University. unpublished data.
Abedin, M.J., Buckner, D., Willits, D., Flenniken, M.L., Jutila, M.A.,
Douglas, T., and Young, M.J. (2006). PEGylation of CCMV and Hsp
Results in Decreased Antibody Production unpublished observations.
Sterman, D.H., Molnar-Kimber, K., Iyengar, T., Chang, M., Lanuti, M.,
Amin, K.M., Pierce, B.K., Kang, E., Treat, J., Recio, A., Litzky, L., Wilson,
J.M., Kaiser, L.R., and Albelda, S.M. (2000). A pilot study of systemic
corticosteroid administration in conjunction with intrapleural adenoviral
vector administration in patients with malignant pleural mesothelioma.
Cancer Gene Ther 7, 1511-1518.
Ruoslahti, E. (2002). Antiangiogenics meet nanotechnology. Cancer Cell
2, 97-98.
Willits, D., Flenniken, M.L., Douglas, T., and Young, M. (2005). Tumor
Targeting Hsp Genetic Variants. unpublished observations.
Willits, D., Young, M., and Douglas, T. (2003). Hsp Genetic Variant
(HspR80\83\93K). unpublished observations.
Richards, J., Miller, M., Abend, J., Koide, A., Koide, S., and Dewhurst, S.
(2003). Engineered fibronectin type III domain with a RGDWXE sequence
binds with enhanced affinity and specificity to human alphavbeta3 integrin.
J Mol Biol 326, 1475-1488.
Essler, M., Brown, D.M., and Ruoslahti, E. (2002). CREKA binds
specifically to the extracellular matrix of tumours. In unpublished
observations, vol. Class: 424178100 (USPTO), A61K039/395 (Intl Class),
Mcdermott Will & Emery LLP: USA.
204
309.
310.
311.
312.
313.
314.
315.
Porkka, K., Laakkonen, P., Hoffman, J.A., Bernasconi, M., and Ruoslahti,
E. (2002). A fragment of the HMGN2 protein homes to the nuclei of tumor
cells and tumor endothelial cells in vivo. Proc Natl Acad Sci U S A 99,
7444-7449.
Pasqualini, R., Koivunen, E., Kain, R., Lahdenranta, J., Sakamoto, M.,
Stryhn, A., Ashmun, R.A., Shapiro, L.H., Arap, W., and Ruoslahti, E.
(2000). Aminopeptidase N is a receptor for tumor-homing peptides and a
target for inhibiting angiogenesis. Cancer Res 60, 722-727.
Willits, D., Flenniken, M.L., Kaiser, C., Young, M., and Douglas, T. (2005).
Lung Targeting Hsp Genetic Variant (HspG41C-GFE). unpublished
observations.
Usselman, R.J. (2005). Dissertation - Evaluation of the CCMV Structural
Transition by EPR, Montana State University, Bozeman, MT.
Willits, D., and Young, M.J. (2005). Creation and Characterization of
CCMV Nd16. unpublished observations.
Flenniken, M.L., Young, M.J., and Douglas, T. (2006). A Library of Protein
Cage Architectures as Nanomaterials. In Current Topics in Microbiology
and Immunology, Volume Viruses as nanomaterials for biomedicine and
bioengineering, M. Manchester, ed. (New York Springer-Verlag ).
Douglas, T., and Young, M. (2006 ). Viruses: Making Friends with Old
Foes. Science 312, 873-875.
205
APPENDIX A
MATERIALS AND METHODS
Cloning Hsp:
Methanococcus jannaschii genomic DNA was obtained from the American Type
Culture Collection (ATTC 43067). The gene encoding the small heat shock
protein Hsp16.5 was amplified by polymerase chain reaction (PCR) with the
following primers: Forward primer 5’ GG TGG GGG CAT ATG TTC GGA AGA
GAC CC; Reverse primer 5’ TA GGA TCC TTA TTC AAT GTT GAT TCC TTT C
and subsequently cloned into NdeI / BamHI restriction sites (underlined above) of
the pET-30a(+) vector (Novagen). The vector was transformed into XL2 Blue E.
coli (Stratagene, LaJolla, CA) for selection. Positive colonies were grown
overnight in LB + kanamycin, plasmid DNA was isolated (Eppendorf Mini Prep
Kit), and inserted DNA was sequenced by an ABI 310 automated capillary
sequencer using Big Dye chain termination sequence technology (Applied Bio
Systems, Foster City, CA). After sequence verification, the pET-30a(+)
MjHsp16.5 was subcloned into BL21(DE3) B strain E. coli (Novagen).
Hsp Mutagenesis:
Polymerase chain reaction (PCR) mediated site directed mutagenesis was
employed to engineer a novel cysteine in the MjHsp16.5 clone (Stratagene).
Complementary forward and reverse mutagenesis primers were designed to
contain the two point mutations needed to change a Glycine (GGG) to a Cysteine
(TGC) at position 41 (G41C) of the MjHsp. In addition, the mutagenesis primers
contained a novel BspEI restriction enzyme site (restriction sites underlined
below; base changes in bold) to aid in mutant screening. The MjHsp G41C
mutagenesis primer sequences are as follows: (+) 5’ GGA ATA CAA ATT TCC
GGA AAA TGC TTC ATG CCA ATC TC 3’; (-) 5’ GA GAT TGG CAT GAA GCA
TTT TCC GGA AAT TTG TAT TCC 3’. Likewise in order to make the HspS121C
mutant, Serine121 (TCA) was changed to Cysteine (TGC). The S121C primers
also contained novel restriction sites NsiI and FspI (restriction sites underlined
below; base changes in bold) to aid in mutant screening. The MjHspS121C
mutagenesis primer sequences are as follows: (+) 5’ G GAA GAA AAT GCA
TGC GCA AAG TTT GAA AAT GG 3’; (-) 5’ CC ATT TTC AAA CTT TGC GCA
TGC ATT TTC TTC C 3’. The engineered mutations and the integrity of the
entire MjHsp coding sequence was confirmed by sequencing on an ABI 310
automated capillary sequencer using Big Dye chain termination sequence
technology (Applied Bio Systems, Foster City, CA).
206
Expression, Purification:
One liter cultures of E. coli (BL21(DE3) B strain) containing pET-30a(+)
MjHsp16.5 plasmid were grown overnight in LB + kanamycin medium (370C, 220
rpm). Cells were harvested by centrifugation 3700xg for 15 minutes (Heraeus
#3334 rotor, Sorvall Centrifuge) and re-suspended in 30 milliliters of 100mM
HEPES, pH 8.0. Lysozyme and DNase were added to final concentrations of 1
mg/ml and 1 mg/L respectively. The sample was incubated for 30 minutes at
room temperature and sonicated on ice (Branson Sonifier 250, Power 4, Duty
cycle 50%, 3 x 5minutes with 3 minute intervals). Bacterial cell debris was
removed via centrifugation for 20 minutes at 12,000xg. The supernatant was
heated for 15 minutes at 60°C thereby denaturing many heat labile E. coli
proteins. The supernatant was centrifuged for 20 minutes at 12,000xg and then
purified by gel filtration chromatography (Superose-6, Amersham-Pharmacia;
Biorad Duoflow). The 16.5 kDa subunit molecular weight (MjHsp16.5) was
verified by SDS poly-acrylamide gel electrophoresis (SDS-PAGE). The
assembled protein was imaged by transmission electron microscopy (TEM) (LEO
912 AB) (stained with 2% uranyl acetate on formvar carbon coated grids), and
analyzed by dynamic light scattering (DLS) 90 plus (Brookhaven Instruments).
Protein concentration was determined by absorbance at 280 nm divided by the
published extinction coefficient (9322 M-1cm-1; 0.565 (mg/ml)-1cm-1). It is worth
noting, that my calculations of the Hsp extinction coefficient of Hsp were different
than the literature value, but it was decided that our lab would use the published
value. The value I calculated based on concentrations determined utilizing the
BCA Protein Assay Kit (Pierce) and a bovine serum albumin (BSA) standard was
1.05 (mg/ml)-1cm-1. The percent composition of tyrosine and tryptophan are
similar for Hsp and BSA (Tyr - 6.12% & 5.60%; Trp - 0.68% & 0.494%
respectively). The theoretically predicted extinction coefficient, calculated using
Biology Workbench (www.biowb.sdsc.edu) is 8250 M-1cm-1; 0.5 (mg/ml)-1cm-1;
closer to the published value.
Mass Spectroscopy (Esquire3000, Bruker):
The measured mass of wild type MjHsp was 16,452 Daltons, this result is
comparable to the theoretically expected mass of 16,453 Daltons. The
measured mass of the MjHsp genetic mutants was HspG41C - 16,499 Daltons
and HspS121C – 16,471 Daltons these results are comparable the theoretically
expected masses of 16,498 and 16,468 Daltons respectively.
LABELING Hsp with Activated Fluorescein:
All labeling experiments included many controls such as: all labeling reactions
were performed on both wild type (wt), G41C, and S121C mutant Hsp protein
cages, fluorophore only (no protein) reactions were analyzed in parallel with
reactions including protein to ensure that no unbound fluorophore was interfering
with our analysis, and sulfhydryl reactive groups were blocked with
iodoacetamide (IAA) prior to labeling in some reactions in order to confirm
207
specificity of fluorescein-maleimide reactions for the engineered reactive thiol
groups.
Fluorescein labeling of the interior surface of Methanococcus Hsp G41C:
The engineered M jHspG41C has one internal thiol group per monomer (24 per
protein shell). Step 1: the reduction of disulfide bonds with water soluble tris(2carboxyethyl)phosphine (TCEP). Mj Hsp G41C and wt Hsp (final concentration
0.5 mg/ml; 54 μM) were reduced with a 5.5x molar excess of TCEP (300 μM) in
100 mM HEPES pH 6.5 for one hour at room temperature (Note – the TCEP step
is not necessary). Step 2: Excess ioacetamide (IAA) was added to control
reactions (5 mM final concentration) pH 8.0 and allowed to react for 30 minutes
at room temperature. Step 3: Fluorescein-5-maleimide fluorophore (Molecular
Probes, Inc. F-150) was dissolved in dimethylformamide (DMF) prior to its
addition to each reaction. Fl-Mal working stocks were made in 100x
concentration relative to desired final concentration for each reaction.
Concentration was verified by taking absorbance measurements at 495nm and
calculating concentration Concentrations of Fl-Mal stocks were verified prior to
addition to reactions by taking absorbance spectra and calculating fluorophore
concentration base on ε = 83,000 M-1cm-1.). Labeling reactions were carried out
at pH6.5, incubated at room temperature for 2 hours and then transferred to 4oC
overnight. A range of fluorophore concentrations was used from 30 μM to 1.5
mM. Initial purification of fluorophore labeled Hsp’s (removal of unbound,
excess fluorophore) was performed by loading 50 μL of each reaction onto
prepared Micro Bio-Spin Chromatography Columns P-30 (Bio Rad) and
centrifuging for 4 minutes at 1,000 x g. Since fluorescein emission is sensitive to
pH (see Molecular Probes, Inc. Handbook [240]), 50 μL of Micro Bio-Spin column
eluant was combined with 100 μL of 0.5 M Tris-Hal pH 8.5 and then analyzed on
a fluorimeter (TECAN Safire). The fluorophore alone controls showed no
fluorescein emission, thus confirming that all unbound fluorophore was trapped in
the P-30 polyacrylamide gel suspension of the Micro Bio-Spin column (at
fluorophore concentrations above 150 μM two or three columns were used
sequentially to remove free fluorophore).
To confirm that the fluorophore was indeed covalently linked to HspG41C, the
same material that was analyzed by fluorimetry was analyzed by size exclusion
chromatography (SEC) (Superose 6 - Amersham-Pharmacia). The absorbance
was monitor at 280nm and 500nm for protein and fluorescein respectively. SEC
results confirm that HspG41C is labeled with fluorescein whereas the size wild
type Hsp does not. Free fluorescein elutes from the column at a much later time
then the Fl-Mal conjugated to the Hsp cages.
208
Fluorescein labeling of the exterior surface of Methanococcus Hsp G41C:
The labeling protocol of HspS121C is the same as described above for HspG41C
except for the reduction with TCEP (Step 1.). HspS121C is continuously stored
in the presence of TCEP to prevent disulfide linkage between cages. The Fl-Mal
reactions were performed in the presence of 5mM TCEP, 100 mM HEPES pH
6.5.
Fluorescein labeling of endogenous lysines on the exterior/interior surface
Hsp:
5-(and-6)-carboxyfluorescein, succinimidyl ester (FAM) (Molecular Probes) was
dissolved in dimethylformamide (DMF) in 100x concentrations relative to final
reaction concentrations ranging from 30 μM to 1.5mM. Concentrations of FAM
stocks were verified prior to addition to reactions by taking absorbance spectra
and calculating fluorophore concentration base on ε= 74,000 M-1cm-1. The
reactions were carried out in 100 mM HEPES buffer at pH 8.0 for 2 hours at
room temperature and overnight at 4˚C. Unbound fluorophore was removed from
reactions and fluorescence analysis was carried out as above.
Quantification of Fluorophore Labeling:
Quantification of fluorophore labeling was done spectroscopically. In order to
quantitate the amount of protein in purified fluorophore labeled Hsp preps,
normalized fluorescein absorbance spectrum were subtracted from the labeled
Hsp spectrum, and protein concentration was calculated from the newly
generated spectrum (Hsp ε= 0.565 (mg/ml)-1cm-1), therefore none of the A280
contribution was due to fluorescein. The concentration of fluorescein covalently
attached to Hsp was determined from the absorbance maxima near A495 (we
noticed a shift of ~8 nm in the absorbance maxima of bound fluorescein
compared to free fluorescein) using the literature values of the extinction
coefficients (ε) (fluorescein-5-maleimide ε = 83,000 M-1cm-1, 5-(and-6)carboxyfluorescein, succinimidyl ester ε= 74,000 M-1cm-1).
Hsp Cage Constrained Mineralization of Iron Oxide:
The mineralization of Hsp cages was performed at room temperature in capped
cuvettes, under air oxidation with a theoretical Fe loading of 200 Fe atoms per
Hsp cage. To this end, 20 μl of 25.5 mM deaerated (NH4)2Fe(SO4)2·6H2O
solution was added to a 1 ml Hsp cages (1 mg, 61 μM) in 100 mM HEPES
buffered at pH 6.5. We followed the Fe(II) oxidation activity by monitoring the
increase in the O-Fe charge transfer absorbance in the visible spectrum at 400
nm. Over the course of 2 hours air oxidation resulted in the formation of a
homogenous rust colored solution. When imaged by TEM, small iron oxide
particles were seen with an average diameter of 9 + 1.2 nm (Fig. 5A).
209
Derivatization of HspG41C with Doxorubicin
The (6-maleimidocaproyl) hydrazone derivative of doxorubicin[5] (Mal-Dox) was
linked to the interior surface of the HspG41C protein cage via coupling of the
maleimide and thiol functionalities. HspG41C cages (2.5 mg/ml, 151.5 μM
subunit; Hsp ε = 9322 M-1 cm-1 ) were reacted with an excess (3 molar
equivalents) of Mal-Dox (454.5 μM, Mal-Dox ε = 8030 M-1 cm-1) in HEPES (100
mM, pH 6.5) for 1 hour at room temperature. Immediately following the reaction,
derivatized cages were separated from free Mal-Dox by size exclusion
chromatography (Superose 6, Amersham-Pharmacia). Both transmission
electron microscopy (TEM) (Leo 912 AB) and dynamic light scattering
(Brookhaven 90 Plus) analysis confirm that HspG41C-Dox maintains the 12 nm
diameter of underivatized Hsp cage. The covalent linkage of doxorubicin to
HspG41C is confirmed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE)
and mass spectrometry.
The concentration of HspG41C protein in purified doxorubicin derivatized cage
preparations was determined by subtracting a normalized doxorubicin spectrum
from the labeled Hsp spectrum (Hsp ε = 9322 M-1 cm-1)[4]. The concentration of
doxorubicin covalently attached to HspG41C was determined from the
absorbance maxima at 495 nm (Mal-Dox ε = 8030 M-1 cm-1) [5].
Quantification of Doxorubicin Release from HspG41C
We quantified the selective release of doxorubicin from HspG41C protein cages
through hydrolysis of the hydrazone linkage under acidic conditions[5, 248].
Doxorubicin release studies were performed at pH 4.0, 4.5, and 5.0 (37°C). After
incubation for times ranging from 0.25 to 24 hours, doxorubicin labeled HspG41C
was separated from free doxorubicin by chromatography (Micro Bio-Spin P-30
Columns, Bio-Rad). The amount of doxorubicin that remained associated with
the HspG41C protein was quantified by absorbance spectroscopy.
Genetic Engineering of αvβ3 Integrin Targeting HspG41C-RGD4C
Methanococcus jannaschii genomic DNA was obtained from the American Type
Culture Collection (ATCC 43607). As described previously, the gene encoding
the small heat shock protein (Mj HSP16.5) was polymerase chain reaction (PCR)
amplified and cloned into NdeI / BamHI restriction sites of the PET-30a(+) vector
(Novagen, Madison, WI) for expression of the full-length protein with no
additional amino acids. PCR mediated site-directed mutagenesis was employed
to replace the glycine at position 41 with a unique cysteine residue, therefore
generating the HspG41C mutant [38]. Deletion of the HSP-stop codon directly
upstream of the BamHI site was also accomplished by PCR mediated site
directed mutagenesis. This deletion allowed for the insertion of additional
sequence into the BamHI site to create the RGD-4C (CDCRGDCFC) carboxylterminal fusion protein engineered to present exposed RGD-4C loops on the
210
exterior of the protein cage. The HspG41C-RGD4C fusion protein was
engineered to have extra glycine residues (italicized below) both before and after
the RGD-4C insert to extend the insert away from the C-terminus and allow some
structural flexibility. Complimentary RGD-4C encoding primers with gatc
overhangs for cloning into the BamHI site (+sense primer: 5' ga tct gga gga tgc
gac tgc cgc gga gac tgc ttc tgc gga taa gga 3'; encoding - S G G C D C R G D C
F C G stop) were mixed at a 1:1 molar ratio, annealed and treated with kinase
(Promega, Madison, WI). These inserts were subsequently ligated into an
alkaline phosphatased, BamHI digested vector overnight at 17°C
and transformed into XL-2 ultracompetant E. coli (Stratagene, La Jolla, CA).
Transformants were screened for the presence of the RGD-4C insert and
confirmed by sequencing the PCR amplified product on an ABI 310 automated
capillary sequencer using BigDye Chain termination sequence technology
(Applied Biosystems, Foster City, CA).
HspG41C-RGD4C Cage Purification and Characterization
All small heat shock protein cages (HspG41C, HspS121C, HspG41C-RGD4C)
were purified from an E. coli heterologous expression system as previously
described [38]. One liter cultures of E. coli (BL21 (DE3) B strain) containing pET30a(+) MjHsp16.5 plasmid were grown overnight in LB + kanamycin medium
(370C, 220 rpm). Cells were harvested by centrifugation 3700 x g for 15 minutes
(Heraeus #3334 rotor, Sorvall Centrifuge) and resuspended in 30 ml of 100 mM
HEPES 50 mM NaCl, pH 8.0. Lysozyme, DNase, and RNAse were added to
final concentrations of 50, 60, and 100 μg/ml, respectively. The sample was
incubated for 30 minutes at room temperature, French pressed (American
Laboratory Press Co., Silver Springs, MD) and sonicated on ice (Branson
Sonifier 250, Power 4, Duty cycle 50%, 3 x 5 minutes with 3 minute intervals).
Bacterial cell debris was removed via centrifugation for 20 minutes at 12,000 x g.
The supernatant was heated for 15 minutes at 65°C thereby denaturing many E.
coli proteins. The supernatant was centrifuged for 20 minutes at 12,000 x g and
purified by gel filtration chromatography (Superose-6, Amersham-Pharmacia,
Piscataway, NJ; Bio-Rad Duoflow, Hercules, CA). Recombinant HspG41CRGD4C protein cages were routinely characterized by size exclusion
chromatography (Superose 6, Amersham Pharmacia), dynamic light scattering
(Brookhaven 90Plus, Brookhaven, NY), transmission electron microscopy (TEM)
(Leo 912 AB), SDS polyacrylamide gel electrophoresis (SDS-PAGE) and mass
spectrometry (Acquity/Q-Tof micro, Waters, Milford, MA). Protein concentration
was determined by absorbance at 280 nm divided by the published extinction
coefficient (9322 M-1cm-1)[84].
Labeling Hsp Genetic Variants with Activated Fluorescein Dye
Cysteine containing Hsp cages (100 mM HEPES 50 mM NaCl pH 6.5) were
reacted with fluorescein-5-maleimide (Molecular Probes, Eugene, OR) in
concentrations ranging from 1-6 molar equivalents per Hsp subunit for 30
211
minutes at room temperature, followed by overnight incubation at 4°C.
Fluorescein labeled Hsp cages were purified from free dye by size exclusion
chromatography (DPBS pH 7.4). The number of fluorescein molecules per cage
was calculated from absorbance spectra [37, 38]. For example, HspG41CRGD4C (2 mg/ml; 112 μM subunit) reacted with 2.2 molar equivalents of
fluorescein-5-maleimide (246 μM) per Hsp subunit resulted in HspG41C-RGD4C
cages with an average of 26.2 fluoresceins per cage (or 1.09 fluoresceins per
subunit). The number of fluoresceins per cage was quantified by absorbance
spectroscopy [38].
HeLa Cell Line
HeLa cells, an epithelial cell line derived from human cervical carcinoma (isolated
from Henrietta Lacks), were provided by Dr. Heini Meittinen at Montana State
University. HeLa cells were propagated in Minimum Essential Medium Eagle
(DMEM) (Sigma D-5648) supplemented with 10% fetal bovine serum (Atlanta
Biologicals, Norcross, GA), penicillin (100 units/ml) and streptomycin (100 μg/ml)
(Sigma, St. Louis, MO) at 370C, in a 5% CO2 incubator. Routine passaging of
cells was done every other day. In order to maintain sufficient cell density, 80%
confluent cells were typically split 1:4.
CV-KL African Green Monkey Kidney Cell Line
The CV-KL African Green Monkey kidney cell line was a post crisis sub-line of
CV-1 established by Dr. Mike Roth’s Lab (Utah) and provided by Dr. Bruce
Granger at Montana State University. CV-KL cells were propagated in Minimum
Essential Medium Eagle (DMEM) (Sigma D-5648) supplemented with 10% fetal
bovine serum (Atlanta Biologicals, Norcross, GA), penicillin (100 units/ml) and
streptomycin (100 μg/ml) (Sigma, St. Louis, MO) at 37°C, in a 5% CO2 incubator.
Routine passaging of cells was done every other day. In order to maintain
sufficient cell density, 80% confluent cells were typically split 1:3.
Transfection Mediated ‘Delivery’ of Fluorescently Labeled Hsp & CCMV
cages to HeLa and CV-KL Cells
Poly-L-Lysine Mediated
This protocol was derived from Dr. J. Franks Lab [284]. HspG41C-Fl or CCMVFl cages (final concentrations ranging from 5 – 250 μg/ml; best results were with
50 μg/ml) are mixed with Poly-L-Lysine (Sigma; 1mg/ml stock) at final
concentrations ranging from 1.5- 15 μg/ml for 30 minutes at room temperature in
DMEM with and without 10% FBS. An equivalent volume of medium was added
immediately before overlaying on cells grown in 4-well chamber slides. For some
studies leupeptin (a protease inhibitor) was added during cell overlay. In
addition, commercially available transfection agents including BioPorter were
used according to manufactures instructions, the results with these expensive
reagents was worse than those achieved with Poly-L-Lysine so their use was not
212
pursued. Cells were imaged live and fixed with epifluorescence microscopy as
described below.
C32 Amelanotic Melanoma Cell Culture
Human amelanotic melanoma cell line, C32, was obtained from the American
Type Culture Collection (ATCC CRL-1585) (ATCC Manassas, VA). C32 cells
were propagated in Minimum Essential Medium Eagle (MEME) (ATCC 30-2003)
supplemented with 10% fetal bovine serum (Atlanta Biologicals, Norcross, GA),
penicillin (100 units/ml) and streptomycin (100 μg/ml) (Sigma, St. Louis, MO) at
370C, in a 5% CO2 incubator. Routine passaging of cells was done every other
day. In order to maintain sufficient cell density, 80% confluent cells were typically
split 1:2.
Mass Spectrometry
Hsp samples were analyzed by liquid chromatography / electrospray mass
spectrometry (LC/MS) (Acquity/Q-Tof micro, Waters; Milford, MA). HspG41CRGD4C and dramatized HspG41C-RGD4C (5-15 µL, 0.3 – 0.5 mg/ml) were
injected onto a C8 column (208TP5115, Vida) and eluted with a H2O–acetonitrile
linear gradient; eluant A - 0.1% formic acid in water; eluant B - 0.05%
trifluoroacetic acid in acetonitrile.
Epifluorescence Microscopy of C32 Melanoma Cells
Epifluorescence microscopy was performed on an Axioscope 2-Plus microscope
(Zeiss) utilizing version 4.1 software and an Axiocam High resolution camera
(Hrc). For all microscopy studies, C32 melanoma cells were grown on glass
coverslips to ~60% confluency (MEME +10% FBS), in the presence of penicillin
(100 units/ml) and streptomycin (100 μg/ml) (Sigma, St. Louis, MO). C32
expression of the RGD-4C target receptor, αvβ3 integrin, was verified by
immunofluorescence utilizing a fluorescein-conjugated anti-αvβ3 monoclonal
antibody (mAb) (LM609) (Chemicon MAB1976F, Temecula, CA).
C32 cells grown on coverslips were incubated with Hsp cages in serum free
medium for 30 minutes at 370C, in a 5% CO2 incubator. The fluorescein
concentration of the Hsp-fluorescein preparations was normalized to 2.5 μM to
facilitate comparison. After incubation, the cells were washed 5 times with
Dulbecco’s Phosphate Buffered Saline (DPBS) (Sigma, St. Louis, MO), fixed with
4% paraformaldehyde for 10 minutes, washed with DPBS and then mounted on
slides in Vectashield mounting medium (Burlingame, CA). Illumination intensity
and camera exposure times were held constant. Note: fixing the cells with 4%
paraformaldehyde prior to incubation with cages resulted in a higher level of
background cell associated fluorescence.
213
Fluorescence Activated Cell Sorting (FACS) Analysis of
C32 cells Incubated with Fluorescein Conjugated Hsp Cages
Flow cytometry was performed on a FACSCalibur, (Becton Dickinson, Mountain
View, CA) and analyzed using Cell Quest software (Becton Dickinson, Mountain
View, CA). Adherent C32 melanoma cells were non-enzymatically disassociated
from cell culture dishes with DPBS without Ca2+ or Mg2+ + 1% EDTA (~25 mM)
(for about 2 minutes at room temperature), washed once with serum containing
medium, and finally suspended in DPBS + Ca2+ / Mg2+ at 2.1 x106 cells / ml.
Experiments were carried out both in the presence and absence of 1% FBS in
the buffer, results were similar in both conditions. Fluorescently labeled cage
preparations (normalized to 2 μM fluorescein) were incubated with cells on ice
from 20 minutes to 2 hours. After incubation the cells were washed 5 times with
DPBS (both with and without Ca2+ / Mg2+), and suspended in DPBS + 1% FBS
for FACS analysis; 10,000 events were counted for each condition. Both the antiαvβ3 monoclonal antibody (mAb) (LM609) (Chemicon MAB1976Z, Temecula, CA)
and the corresponding fluorescein-conjugated anti-αvβ3 mAb (Chemicon
MAB1976F) were used for FACS analysis.
HspG41C-Fluorescein Anti-CD4 Antibody Conjugation
Fluorescein labeled HspG41C protein cages were conjugated to anti-CD4
monoclonal antibodies (generated from ATCC GK1.5) via a heterobifunctional
cross-linker SMCC (sulfosuccinimidyl 4-(N-maleimidomethyl) cyclohexane-1carboxylate) (Pierce, Rockford, IL). First, the antibodies (6.5 mg/ml in PBS pH
7.4) were partially reduced with 10 mM TCEP (tris(2-carboxyethyl)phosphine) in
the presence of 10 mM EDTA (ethylenediaminetetraacetic acid) with the final pH
adjusted to 6.5 and incubated for 2 hours at room temperature [112].
Simultaneously, the exposed lysines (amines) of HspG41C-fluorescein cages
(0.25 mg/ml / 15 μM subunit in 500 μL DPBS pH 7.4; 11 fluoresceins per cage)
were reacted with the sulfo-NHS-ester component of the SMCC linker (added in
excess 0.5 mg). The Hsp cage plus linker reaction was incubated at room
temperature for one hour followed by the removal of free SMCC linker by size
exclusion chromatography (Desalting Column, Pierce, Rockford, IL). The
reduced anti-CD4 antibodies were combined with the HspG41C-fluoresceinSMCC cages and incubated for 3 hours before final purification of anti-CD4HspG41C-fluorescein cage (Ab-Hsp) conjugates by size exclusion
chromatography (Superose 6, Amersham-Pharmacia, Piscataway, NJ). This
protocol resulted in a very low yield of antibody-Hsp conjugate. When a less
readily available antibody (anti-collagen; Chemicon AB765P) was reduced in a
scaled down fashion from that described above (50 μL of 1mg/ml Ab, 10mM
TCEP, 10 mM EDTA) and combined with SMCC conjugated Hsp-Fl cages, I did
not obtain Ab-Hsp conjugates. These types of reactions should be easier with
‘click-chemistry’ now being utilized in our lab [46, 291].
214
Murine Splenocyte Preparation
A BALB/c mouse spleen was homogenized in Hanks Balanced Salt Solution
(Mediatech, Herndon, VA) by pushing it through a 60 gauge stainless steel mesh,
the homogenate was filtered through 100 μm nylon mesh, and centrifuged (200 x
g for 10 minutes). The supernatant was discarded and the cell pellet suspended
in 5 ml ACK lysis buffer (150 mM NH4Cl, 1 mM KHCO3, 0.1 mM Na2EDTA) for 5
minutes at room temperature to lyse unwanted red blood cells. Lysis was
stopped via the addition of PBS + 2% donor calf serum (25 ml). The remaining
white blood cells were pelleted by centrifugation (200 x g for 10 minutes) and
suspended in PBS + 2% donor calf serum (2 x 107 cells/ml) containing antimouse Fc receptor antibody (from HB-197, ATCC) in order to prevent nonspecific binding of antibodies to the Fc receptor on lymphocytes. Cells were
incubated on ice until binding assays were performed.
Epifluorescence Microscopy of Splenocytes
Aliquots of murine splenocytes were combined with equal volumes of each of the
following: A) anti-CD4-HspG41C-fluorescein (Ab-Hsp-Fl) cages, B) fluorescein
conjugated anti-CD4 mAb (positive control), and C) HspG41C-fluorescein cages
(non-targeted control), and incubated for 30 minutes on ice. Following incubation
the cells were washed (x3) with PBS containing 2% donor calf serum. The cells
were wet-mounted prior to epifluorescence microscopy using a Nikon Eclipse
E800 microscope, equipped with a Nikon DMX1200 digital camera, utilizing
MetaVue software. Illumination intensity and camera exposure times were held
constant.
Fluorescence Activated Cell Sorting (FACS) Analysis of Splenocytes
Incubated with Anti-CD4 mAb Conjugated Hsp Cages
FACS was performed on a FACSCalibur, (Becton Dickinson, Mountain View, CA)
and analyzed using Cell Quest software (Becton Dickinson, Mountain View, CA).
Aliquots of murine splenocytes were combined with equal volumes of each of the
following: A) anti-CD4-HspG41C-fluorescein (Ab-Hsp-Fl) cages, B) fluorescein
conjugated anti-CD4 mAb (positive control), and C) HspG41C-fluorescein cages
(non-targeted control), and incubated for 30 minutes on ice. Following incubation
the cells were washed in PBS containing 2% donor calf serum in preparation for
FACS. FACS analysis was performed on a gated murine splenocyte cell
population. For anti-CD4 mAb blocking experiments, splenocytes were incubated
for 30 minutes with unlabeled anti-CD4 mAb and washed to remove unbound
blocking antibody prior to incubation with Ab-Hsp-Fl cage conjugates.
HspG41CRGD-4C Doxorubicin Conjugation
HspG41CRGD4-C cages (2 mg/ml; 112 μM subunit) in 100 mM HEPES 50 mM
NaCl pH 6.5 were reacted with a 2.2 fold excess of the (6-maleimidocaproyl)
hydrazone of doxorubicin (246 μM) for 30 minutes followed by purification via
size exclusion chromatography. The average number of doxorubicin molecules
215
per cage was 10 (an average of 0.42 doxorubicin molecules per subunit) as
calculated from absorbance spectra [5, 37]. An increased yield of derivatized
HspG41CRGD4C-Dox is obtained if size exclusion if performed, before Microcon
(MWCO 100 kDa) concentration.
MTS Cell Proliferation Assay
(CellTiter 96 Aqueous Non-Radioactive Cell Proliferation Assay – Promega)
MTS = 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4sulfophenyl)-2H-terazolium
PMS = phenazine methosulfate
Cell Growth in 96-well Plate
Cells were grown in 96-well cell culture plates (Costar 3595). HeLa cells were
seeded 3000 cells per well (200 μL of a 1.5 x 104 cells/ml suspension) and grown
overnight prior to experimentation. C32 cells were seeded 6000 cells per well
(200 μL of a 3 x 104 cells/ml suspension) and require two days of growth prior to
experimentation. In order to make cell suspension, lift adherent cells with either
a trypsin/EDTA solution or just EDTA, evenly suspend cells in ~2 ml of medium
for every 75% confluent plate. Determine the concentration (cells/ml) using a
hemacytometer according to manufacturers instructions. In brief, mix cell
suspension with Trypan Blue (10 μl cells + 90 μl Trypan blue) which stains dead
cells blue (.4%w/v trypan blue in PBS; Sigma T8154). Make sure that cells are
suspended individually (not many clumps) and count 4 ‘big’ quadrants [cells/ml =
total cell count in 4 squares x 2500 x dilution factor (i.e.,10). Use this number to
calculate the appropriate dilution. Remember to put medium only in 6 wells of
the 96-well plate for controls.
Experimental Overlays
After cells were grown for appropriate length of time they were subjected to
various experimental conditions. It was necessary to remove cell culture from
each well individually by pipetting, since the cells dried out very quickly when
medium was aspirated away. Immediately following medium removal from a
single well, cells were overlayed with medium containing the experimental
conditions (i.e., HspG41C-Dox versus Dox only for various time points). Each
condition was tested in triplicate.
MTS Cell Proliferation Assay
The MTS assay was p MTS cell proliferation assay (CellTiter96 Aqueous NonRadioactive Cell Proliferation Assay, Promega, Madison, WI) pperformed
according to manufactures instructions. In brief, an MTS/PMS solution was
prepared immediately before use, 40 μl was added to each well, the plate was
incubated (37˚C, 5% CO2) for 1.5 hours, and then analyzed by absorbance at
490 nm (TECAN). The results were more consistent when cells were grown in
200 μl of medium (as opposed to 100 μl) and if the ‘developed’ assay mixed (by
216
pipetting up and down) and then transferred to a new 96-well plate for the
absorbance reading. Multiple reads of the plate helped ensure accuracy. The
background level of 490 nm absorbance was determined by MTS analysis of
wells with medium only. Note - It is important to control each experiment
internally (i.e., – positive cell only control; negative medium only control). The
MTS assay continuously develops over time; therefore the values obtained aren’t
absolute and should not be directly compared between different experiments.
Sulforhodamine B Colorimetric Assay for Cell Proliferation
Protocol from Jean Starkey’s Lab, Montana State University, Sept. 2004
Cells grown in a 96-well plate are subjected to experimental conditions with
proper controls; multiple wells of each. At the conclusion of the experiment, the
cells are washed with serum free medium (found that when DPBS was used they
lost too many cells) to remove protein in the cell culture medium. (The Starkey
Lab submerges the 96-well plates in DPBS within their plastic packages and then
taps them dry.) The cells are then fixed for 30 minutes with cold 10% TCA
(trichloroacetic acid). Following fixation, the plates are washed 5 times with
water. Next, 70 μL of 0.4% sulforhodamine B (SRB; Biotium Cat. # 80100) in 1%
acetic acid is added to each well and incubated for 20 minutes. The cellular
proteins are stained with SRB. The unincorporated SRB stain is washed away
with 1% acetic acid. Next, 200 μL of 10mM Tris pH 10.4 is added to each well,
and incubated for 10-20 minutes with shaking. The assay is read at 515 nm.
Hsp Cage Localization in Breast Tumor (MDA-MB-435) Xenograft Bearing
Mice
1st Study by Tasmia Duza (Post Doctoral Researcher in Dr. E. Ruoslahti’s Lab) –
June, 2005.
Fluorescently labeled Hsp cages (HspG41C-Fl and HspG41C-RGD4C-Fl) were
prepared as described above and stored at -80˚C in DPBS until use. The cages
were administered by intravenous injection into the tail vein and allowed to
circulate for either 1 or 4 hours in mice bearing breast tumor (MDA MB 435)
xenografts (mice ~25 grams; tumors ~1 gram) [8]. HspG41C-RGD4C-Fl was
administered in the following doses: 1.2, 0.5, 0.25, and 0.12 mg per animal.
HspG41C-Fl was tested at 0.5 mg and 0.25 mg per animal. Intracardiac
perfusion of PBS (10 ml) and 4% PFA (in 10 ml PBS) was used to clear the
vasculature of remaining circulating cages. The tumor and various control
organs (brain, skin, kidney, lung, heart, pancreas, liver, and spleen) were
harvested and processed for histology. Examination of these tissues by
fluorescent microscopy determined protein cage localization and homing
capability.
Both genetic variants of Hsp cages, HspG41C and HspG41C-RGD4C, home to
tumor vasculature. It was found that tumor specificity was best with a dose in the
217
range of 0.25-0.5 mg of protein, corresponds to 15 – 30 mM subunit or 0.6 mM –
1.26 mM of assembled cage, per animal. In this range the particles were also
taken up by the reticulo-endothelial system (RES) as expected (liver and spleen),
but did not accumulate significantly in any of the other control organs tested. The
pattern of nanoparticle distribution within the tumor is consistent with vascular
localization.
In these preliminary experiments the Hsp protein cages were differentially
labeled with fluorescein (HspG41C - 11 Fl per cage vs. HspG41C-RGD4C – 19.3
Fl per cage). For in vivo work, equivalent amounts of fluorescein were injected;
therefore more HspG41C cages were injected as compared to HspG41CRGD4C. This could have masked any small differences in RGD-4C peptide
mediated homing of HspG41C-RGD4C cages. Further experiments will be done
to address this question. It is possible that the homing peptide will confer
advantages to HspG41C-RGD4C mediated therapeutic delivery.
Derivatization of HspG41C-RGD4C with Doxorubicin
The (6-maleimidocaproyl) hydrazone derivative of doxorubicin[5] (Mal-Dox) was
linked to HspG41C-RGD4C protein cage via coupling of the maleimide and thiol
functionalities. The best Mal-Dox labeling conditions to date are as follows:
HspG41C-RGD4C cages (2 mg/ml, 112.3 μM subunit; Hsp ε = 9322 M-1 cm-1 )
were reacted with 0.5 molar equivalents of Mal-Dox per Hsp subunit (454.5 μM,
Mal-Dox ε = 8030 M-1 cm-1) in reaction buffer (100 mM HEPES, 50 mM NaCl, pH
6.5) for 15 minutes at room temperature. Immediately following the reaction,
derivatized cages were separated from free Mal-Dox by size exclusion
chromatography (Superose 6, Amersham-Pharmacia). Both transmission
electron microscopy (TEM) (Leo 912 AB) and dynamic light scattering
(Brookhaven 90 Plus) analysis confirm that HspG41CRGD4C-Dox maintains the
~13 nm diameter of underivatized Hsp cage. The covalent linkage of doxorubicin
to HspG41C is confirmed by SDS-polyacrylamide gel electrophoresis (SDSPAGE) and mass spectrometry.
The concentration of HspG41C-RGD4C protein in purified doxorubicin
derivatized cage preparations was determined by subtracting a normalized
doxorubicin spectrum from the labeled Hsp spectrum (Hsp ε = 9322 M-1 cm-1)[84].
The concentration of doxorubicin covalently attached to HspG41C-RGD4C was
determined from the absorbance maxima at 495 nm (Mal-Dox ε = 8030 M-1 cm-1)
[248]. Quantification of the extent of labeling with doxorubicin via mass
spectrometry and absorbance spectroscopy has not always correlated.
Variability in labeling efficiency has been observed under identical conditions.
218
C32 Melanoma Cell – Protein Interaction Assay
Proteins (HspG41C-RGD4C, HspG41C, BSA – each at 2 mg/ml, and 2% FBS)
were spotted in 50 μL aliquots a bacterial petri dish, incubated at 4˚C for 2 hrs,
and then the protein spots were pipetted off and the spots “washed” by pipetting
up and down on spots with DPBS pH 7.4. The Petri dishes were then flooded
with a C32 melanoma cell suspension and incubated at 37˚C, 5% CO2 incubator
for 16 hours. The cell suspension was removed and the plate was washed with
DPBS 3 times, the remaining adherent C32 cells were fixed with EtOH for 10
minutes, and then visualized by Coomassie Blue staining.
Subcellular Localization of Doxorubicin with C32 Melanoma Cells
Epifluorescence microscopy was performed on an Axioscope 2-Plus microscope
(Zeiss) utilizing version 4.1 software and an Axiocam High resolution camera
(Hrc). For all microscopy studies, C32 melanoma cells were grown on glass
coverslips to ~60% confluency (MEME +10% FBS), in the presence of penicillin
(100 units/ml) and streptomycin (100 μg/ml) (Sigma, St. Louis, MO). Doxorubicin
or HspG41CRGD4C-Dox protein cages, normalized to 1 μM doxorubicin in
serum free MEME, were incubated with C32 melanoma cells for 30 minutes at
370C, in a 5% CO2 incubator. After incubation, the cells were washed 5 times
with Dulbecco’s Phosphate Buffered Saline (DPBS) + 2% bovine serum albumin
(BSA) (Sigma, St. Louis, MO), fixed with 4% paraformaldehyde for 10 minutes,
washed with DPBS+2% BSA and then mounted on slides in Vectashield
mounting medium containing DAPI nuclear stain (Burlingame, CA). Illumination
intensity and camera exposure times were held constant.
Iodination of Protein Cages (HspG41C, HspG41C-GFE, CCMV K42R)
Protein cages were iodinated utilizing IODO-BEADS Iodination Reagent (cat. #
28665, Pierce, Rockford, IL) according to manufacturers instructions. IODOBEADS are non-porous polystyrene beads (1/8’’ diameter) derivatized with Nchloro-benzenesulfonamide (sodium salt). The protocol was initially optimized
utilizing “cold” NaI in order to characterize the iodinated protein cages by
absorbance spectroscopy, size exclusion chromatography (Superose 6,
Amersham Pharmacia), dynamic light scattering (Brookhaven 90Plus,
Brookhaven, NY), and transmission electron microscopy (TEM) (Leo 912 AB). In
brief, immediately before the reaction IODO-BEADS (2 beads per 0.5 ml (0.5 mg
protein) reaction) were washed with DPBS pH 7.2, blotted on filter paper, and
transferred to a 5 ml, cylindrical, flat bottom reaction vial. Next 250 μL of DPBS
pH 7.2 was added to the reaction vial, followed by 1 mCi of carrier free Na125I
(100 mCi/ml, ~17 Ci/mg I, cat. # 016303410, MP Biomedicals, Inc). The buffered
Na125I was incubated with the IODO-BEADS for 5 minutes at room temperature
(slightly tilting the reaction vessel at an angle to allow better mixing of bead,
buffer, and Na125I) for oxidation of Na125I and the creation of electrophilic iodine
species for incorporation into tyrosines. Protein cages, 250 μL of a 2 mg/ml
stock, were added to the reaction vessel and the reaction proceeded at room
219
temperature for 15 minutes. After 15 minutes, the protein cage solution was
transferred into a new microfuge tube; according to Pierce this stops the reaction.
Unincorporated Na125I was removed utilizing Zeba Desalt Spin Columns, 0.5 ml
(Pierce). The columns were prepared according to manufactures instructions (2
washes with 300 μL DPBS pH 7.2, and one final 1,500 x g spin for 1 minute) then
125 μl sample was added to the center of the resin bed, followed by spinning for
2 minutes at 1,500 x g to collect the sample which was then loaded on another
Zeba Desalt Spin Column for a second round of iodine removal (8 columns for 2
rounds of clean up of a 0.5 ml reaction in 125 μL aliquots). Cold NaI experiments
determined >90% recovery after 2 rounds of desalting; therefore a final protein
concentration of 1 mg/ml was assumed after radio-labeling.
The protocol for labeling protein cages with 125I was usually performed as
described above, except for the last radio-labeling on March 1, 2006. On March
1, 2006 I did not receive the amount of radio-label requested; therefore I modified
the protocol as follows: the ratio of IODO-BEADS (N-chloro-benzenesulfonamide
oxidation agent) to NaI was changed from 2 beads per 1 mCi 125I to 8 beads per
1 mCi 125I while maintaining the ratio of IOD0-BEADS (N-chlorobenzenesulfonamide oxidation agent) to protein. Surprisingly, this modified
protocol resulted in a similar level of 125I incorporation (Table 6.1).
Quantification of the Specific Activity
In order to determine the specific activity of the 125I labeled protein cages two of
each 1 μl and 0.5 μl aliquots (including pipette tip) were transferred into vials and
counted on a Beckman Gamma 4000 Counting System. Specific activity of the
samples was determined utilizing a linear regression equation derived from 125I
standards (ranging from 0 – 1000 nCi) (x-axis) plotted with their corresponding
counts per minute (y-axis). Typically, specific activity for Hsp cage samples was
~20 nCi/μg and ~250 nCi/μg for CCMV samples.
Trial iodination experiments utilizing IODO-GEN (1,3,4,6-tetrachloro-3α,6αdiphenylglycouril) Pre-Coated Iodination Tubes, (cat. # 28601, Pierce) resulted in
~10-fold less iodine incorporation in Hsp cages, but resulted in similar levels of
125
I incorporation, as IODO-BEAD reactions, in CCMV cages. Both Hsp and
CCMV cages were less stable in IODO-GEN Tube reactions, tolerating reaction
times of < 10 minutes before cage disassembly.
Trichloroacetic Acid (TCA) Precipitation of Radiolabeled Protein Cages
Equal volumes of radio-labeled protein solution and 20% TCA (5 μL each) were
mixed and incubated on ice for 1 hour. The protein was pelleted by
centrifugation (13,000 rpm for 5 minutes) and the number of gamma counts
associated with the pellet, supernatant and one wash (15 μL 10% TCA) was
determined. This analysis revealed that for the Hsp samples the majority of the
220
counts (76.4%) were associated with the protein pellet. Unfortunately, the CCMV
data are less clear because although several different TCA precipitation
protocols were attempted, I did not get consistent precipitation – usually less that
50% of the protein would precipitate (as determined by SDS-PAGE gels of
parallel ‘cold’ cage experiments) (Figure 6.7). TCA precipitation of radio-labeled
CCMV resulted in 53.5% of the counts in the pellet. In conclusion, I would say
that based on the above data, I think that the majority of 125I in my radio-labeled
protein preparations is incorporated into the protein, but I would not rule out the
possibility of free 125I in these preparations.
Biodistribution of 125I Labeled Protein Cages (HspG41C, HspG41C-GFE, and
CCMVK42R) in Naïve Mice
These experiments were performed by Coleen Kaiser and Ann Harmsen in Allen
Harmsen’s Lab in the Veterinary Molecular Biology Department at Montana State
University.
Protein cages were administered either intra nasally (i.n.) or intra venously (i.v.)
or i.n. in Balb/c mice ((8-10 week females from Charles River). For HspG41C
and HspG41C-GFE experiments 50 μg (0.5 μCi) doses were given and allowed
to circulate for either 1 hour or 24 hours. For CCMV experiments 125I cages were
combined with non-radioactive cages to make a 50 μg protein (0.5 μCi) dose
delivered i.v. and allowed to circulate for either 1 or 24 hours. Normalization of
the dose of cage to 50 μg, results in a differential number of cages injected per
experiment since 50 μg of Hsp corresponds to 7.6 x 1013 cages whereas 50 μg of
CCMV corresponds to 8.227 x 1012 cages. Mice were sacrificed, and the organs
and tissues were removed for gamma counting. Organ count per minutes were
reported on a per gram of tissue basis and the percent of dose per organ was
calculated (see Chapter 6).
Biodistribution of 125I Labeled Protein Cages (HspG41C, HspG41C-GFE, and
CCMVK42R) in Immunized Mice
Prior to injection of 125I labeled protein cage (either HspG41C or CCMVK42R)
Balb/c mice (8-10 week females from Charles River) were exposed to the
following immunization protocol: Week 1: 50 μg protein cage injection, Week 3:
50 μg boost , Week 4: bleed mice for antibody titer determination by ELISA and
125
I labeled protein cage injection. The biodistribution of 125I labeled protein
cages in immunized mice was identical to the protocol described above for naïve
mice.
Filtration Analysis of Mouse Urine
Radiolabeled (125I) CCMV and Hsp (50 μg protein; 229.5 nCi/ug CCMV; 448
nCi/ug Hsp) were i.v. injected into the tail veins of Balb/c mice. Urine was
collected after 24 hours and analyzed by filtration (Microcon centrifugal filter
221
device, MWCO = 100 kDa, Amicon, Millipore Corp., Bedford, MA). 125I associated
with the column retentate and filtrate was counted with a Beckman Gamma
Counter. Analysis of 125I labeled CCMV and Hsp in DPBS at concentrations
equivalent to the dose given to mice was performed in parallel and served as a
control.
PEGylation of CPMV and Hsp (protocol from Joynal Abedin)
Poly(ethylene) glycol (PEG) was conjugated to the exposed lysines of CCMV and
Hsp protein cages. This reaction was carried out as follows: CCMV and Hsp
protein cages were reacted with an excess of mPEG-succinimidyl propionate
(mPEG-SPA; MW 2000 Da; 33 molar excess per subunit) in pH 6.5 buffer
(100mM HEPES, 50mM NaCl) for 24 hours at 4 °C with stirring. Following the
reaction, the PEGylated capsids were purified by size exclusion chromatography.
The recovery of protein cage ranged from 60 - 90%. SDS-PAGE analysis
demonstrated that 50% of the protein cages were labeled with PEG; >1 PEG2000 chains were covalently attached per protein cage subunit (Hsp = 24
subunits; CCMV = 180 subunits).
Hypersensitivity to CCMV and Hsp Evaluated in Mice
Hypersensitivity experiment with a CK017 label; for this experiment Hsp and
CCMV were introduced intranasally into mice (2 doses per week for 4.5 weeks).
The serum samples from the CCMV sensitized mice were analyzed for the
presence of antibodies and in some cases their subtype. The results indicated
that IgG was present in the CCMV sensitized mice serum. Another indicator of
adverse immune responses, such as hypersensitivity, in mice is their bodyweight;
the mice in the above study were similar in weight to control mice.
Genetic Incorporation Targeting Peptides into HspG41C; Table 7.1
Methanococcus jannaschii genomic DNA was obtained from the American Type
Culture Collection (ATCC 43607). As described previously, the gene encoding
the small heat shock protein (Mj HSP16.5) was polymerase chain reaction (PCR)
amplified and cloned into NdeI / BamHI restriction sites of the PET-30a(+) vector
(Novagen, Madison, WI) for expression of the full-length protein with no
additional amino acids. PCR mediated site-directed mutagenesis was employed
to replace the glycine at position 41 with a unique cysteine residue, therefore
generating the HspG41C mutant [38]. Deletion of the HSP-stop codon directly
upstream of the BamHI site was also accomplished by PCR mediated site
directed mutagenesis. This deletion allowed for the insertion of additional
sequence into the BamHI site. Sequences encoding the cell targeting peptides
listed in Table 7.1, were added as carboxyl-terminal fusions in order to present
them as loops on the exterior of the protein cage. The HspG41C-cell targeting
peptide fusion proteins were engineered to have extra sequences (encoding –
GSGG) added before the targeting peptide sequence. In addition, all cell
targeting HspG41C constructs, excepting CGKRK and CREKA, had a G was
222
added after directly before the stop codon (Table 7.1). The extra sequences
served to extend the insert away from the C-terminus and allow some structural
flexibility for disulfide bond formation within the peptides. Complimentary cell
targeting encoding primers with gatc overhangs (see Table 7.1) for cloning into
the BamHI site were mixed at a 1:1 molar ratio, annealed and treated with kinase
(Promega, Madison, WI). These inserts were subsequently ligated into an
alkaline phosphatased, BamHI digested vector overnight at 17°C
and transformed into XL-2 ultracompetant E. coli (Stratagene, La Jolla, CA).
Transformants were screened for the presence of the cell targeting peptide insert
and confirmed by sequencing the PCR amplified product on an ABI 310
automated capillary sequencer using BigDye Chain termination sequence
technology (Applied Biosystems, Foster City, CA).
Fluorescent Labeling of HspG41C-GFE Cages
HspG41C-GFE cages were labeled with fluorescent dyes (fluorescein-5maleimide, Texas Red C2 maleimide (Molecular Probes T-6008). HspG41C-GFE
cages were easily labeled with fluorescein-5-maleimide, under similar reaction
conditions as those described above (100 mM HEPES, 50 mM NaCl, pH 6.5, 3fold molar excess of dye per subunit, 20 minutes at room temperature, overnight
at 4˚C, purification by size exclusion in DPBS pH 7.4, ~50% yield). On the other
hand, HspG41C-GFE was more sensitive to Texas Red labeling yields increased
as reaction time was decreased to 30 – 40 minutes (20 minutes at room
temperature, 10-20 min at 4˚C), but conditions could be further optimized.
HspG41C cages can be easily labeled with Texas Red C2 maleimide.
HspG41C-GFE Cages Bind ex vivo Lung Tissue
Texas-red labeled HspG41C-GFE cages were incubated with mouse lung
sections in order to test their ability to bind lung endothelium. Vibratome sections
of mouse lung (6 slices per condition) were incubated with HspG41C-GFE-Texas
red (TR) (from 9/25/05 prep – Hsp 13.83 μM, TR 13 μM, 22.6 TR/cage) (100 μL
Hsp-GFE-TR in a total volume of 3 ml HBSS, corresponds to 3.47 x 1013 total
cages in 3ml; 7.8 X 1014 TR molecules) or HspG41C-TR (9/26/05 prep – Hsp
62.12 μM, TR 101 μM, 39 TR/cage) (12 μL Hsp-GFE in a total volume of 3 ml
HBSS, corresponds to 1.87 x 1013 total cages in 3ml and 7.3 x 1014 TR
molecules) on ice and covered for 1 hour 15 minutes while shaking at 50 rpm.
After incubation the sections were rinsed with HBSS. Fresh HBSS was added to
the containers (3 ml) and they were incubates as described above for 1 hour.
Epifluorescence microscopy was used to image the lung sections, the exposures
ranged from 100 to150 ms for the Texas Red channel and 1000 ms for the FITC
channel. A similar experiment with less Texas Red was performed, but the 1st
experiment yielded better images. In addition, the use of confocal microscopy for
imaging was investigated, but these images were not great.
Fluorescein labeling of HspG41C-RGD2C and HspG41C-RGDWLE
223
HspG41C-RGD2C and HspG41C-RGDWLE cages were labeled with fluorescein.
The protocol was similar to HspG41C and HspG41C-RGD4C fluorescein labeling
reactions described above. In brief, HspG41C-RGD2C and HspG41C-RGDWLE
cages (100 mM HEPES, 50 mM NaCl, pH 6.5) were labeled with fluorescein-5maleimide (2-fold or 6-fold molar excess per Hsp subunit). The reactions
proceeded for 20 minutes at room temperature and 4˚C overnight (for
convenience). Fluorescently labeled cages were purified from free fluorescein by
size exclusion chromatography (DPBS pH 7.4).
Note – The human H chain ferritin related Materials and Methods (below)
were written by M. Uchida [276].
Cloning of human H chain ferritin (HFn)
Human muscle chromosome DNA was obtained from American Type Culture
Collection (ATCC). The gene encoding the human H chain ferritin was amplified
by polymerase chain reaction (PCR). The forward primer was 5’ A GTC GCC
CAT ATG ACG ACC GCG TCC and reverse primer was 5’ GCC GGA TCC TTA
GCT TTC ATT ATC AC for amplification. The PCR product was subsequently
cloned into NdeI and BamHI restriction sites of the pET-30a(+) vector (Novagen).
The plasmid vector was transformed into XL2 Blue ultracompetent E. coli
(Stratagene, LaJolla, CA) and cultured on LB with kanamycin plate for selection.
Positive colonies were grown overnight in 3 ml of LB with 30 mg/l kanamycin
medium. The HFn plasmid DNA was isolated using Perfectprep Plasmid Mini
(Eppendorf, Hamburg, Germany) and inserted DNA was sequenced using an ABI
310 automated capillary sequencer using Big Dye chain termination sequence
technology (Applied Bio System, Foster City, CA).
Genetic Engineering of αvβ3 Integrin Targeting Human H-Chain Ferritin
It is known that N-terminal amino acid residues are exposed on the exterior
surface of the human H-chain ferritin (HFn) cage. Therefore, the RGD4C peptide
was incorporated at the N-terminus of the HFn subunit. An AatII restriction site
was introduced into the native HFn for subsequent incorporation of the RGD4C
peptide sequence. Primers were designed as follows to change a tyrosine (ACC)
to a serine (TCC) at position 3 of the HFn: (+) 5’ GAA GAA GAT ATA CAT ATG
ACG TCC GCG TCC ACC TCG CAG GTG; (-) 5’ CAC CTG CGA GGT GGA
CGC GGA CGT CAT ATG TAT ATC TCC TTC. To incorporate the RGD4C
peptide sequence onto the N-terminus of the HFn, the complimentary primer GC
GAC TGC CGC GGA GAC TGC GGA GGC GGA ACG T and TCC GCC TCC
GCA GAA GCA GTC TCC GCG GCA GTC GCA CGT was designed and
inserted into the introduced AatII site of the plasmid. Four glycine residues were
added after the RGD4C peptide to allow for some structural flexibility. These
mutageneses resulted in the change of the original HFn amino acid sequence
MTTAS to MTCDCRGDCFCGGGTSAS. The RGD4C-Fn plasmid was isolated
and sequenced as previously described for HFn.
224
Purification and Characterization of Ferritin
The HFn and RGD4C-Fn were expressed in E. coli where they self-assembled
into the 24 subunit cages. One liter cultures of E. coli (BL21 (DE3): Novagen)
containing pET-30a(+) HFn or RGD4C-Fn plasmid were grown overnight in LB
with 30 mg/l kanamycin medium. Cells were collected by centrifugation and then
the pellets were resuspended in 45 ml of a lysis buffer (100mM HEPES, 50mM
NaCl, pH 8.0). Lysozyme, DNAse and RNAse were added to final
concentrations of 50, 60 and 100 μg/ml, respectively. After 30 min incubation at
room temperature, the solution was subjected to French press followed by
sonication on ice. The solution was centrifuged to remove E. coli debris. The
supernatant was heated at 60 °C for 10 min precipitating many of the E. coli
proteins, which were removed by centrifugation. The supernatant was subjected
to size exclusion chromatography (SEC: Amersham-Pharmacia, Piscataway, NJ,
USA) with Superose 6 column to purify HFn or RGD4C-Fn. The protein cages
were characterized using SEC, dynamic light scattering (DLS: Brookhaven,
90Plus particle size analyzer) and transmission electron microscopy (TEM: LEO
912AB). Protein concentration was determined by absorbance at 280 nm.
Fn Cage Mediated Iron oxide Mineralization & Characterization
A degassed solution (8ml of 100 mM NaCl) was added to a jacketed reaction
vessel under N2 atmosphere followed by addition of HFn (1 mg, 1.98 x 10-6
mmoles) or RGD4C-Fn (1.06 mg, 1.98 x 10-6 mmoles). The temperature of the
vessel was kept at 65 °C using circulating water through the jacketed flask. The
pH was titrated to 8.5 using 50 mM NaOH (718 Auto Titrator, Brinkmann). Fe(II)
was added (12.5 mM (NH4)2 Fe(SO4)•6H2O) to attain a theoretical loading factor
of 1000 Fe (158 μL), 3000 Fe (474 μL), or 5000 Fe (790 μL), per protein cage.
Freshly prepared H2O2 (12.5/3 mM; 4.52 μL per 10 ml reaction) was also added
as an oxidant. The Fe(II) And H2O2 solutions were added simultaneously and at
a constant rate of 100 Fe/protein•min using a syringe pump (Kd Scientific). H+
generated during the reaction was titrated dynamically using 50 mM NaOH to
maintain a constant pH of 8.5. The reaction was considered complete 5 min after
addition of all the iron and oxidant solutions. After the completion of the reaction,
200 µl of 300 mM sodium citrate was added to chelate any free iron. The
mineralized sample was analyzed by SEC (Superose 6). Absorbance at 280 nm
and 410 nm were simultaneously monitored for protein and the mineral,
respectively. The sample was imaged by TEM, and electron diffraction and
electron energy loss spectroscopy (EELS) data were collected on all samples.
Magnetic characterizations of the mineralized samples were performed on a
physical properties measurement system (PPMS: Quantum Design). Dynamic
and static magnetic measurements were carried out using an alternating current
magnetic susceptibility (ACMS) and a vibrating sample magnetometer (VSM)
option, respectively. ACMS measurements were performed under 10 Oe field at
225
frequency of 100, 500, 1000, 5000 and 10000 Hz, with no dc background, while
the temperature was varied from X to Y K. The superparamagnetic blocking
temperature (Tb) was determined from susceptibility curves. VSM measurements
were performed under a magnetic field up to 80KOe at 5K.
Cell targeting assay of Mineralized RGD4C-Fn
The mineralized RGD4C-Fn with loading factor of 3000Fe/cage was used to
evaluate cancer cell targeting ability of the protein cage. The mineralized HFn
(3000Fe/cage) were used as control. C32 cells were cultured in six-well
polystyrene plate. After 2 days of incubation, MEME was removed from the well
followed by addition of 1 ml of mineralized protein (200µg/ml) in Dulbecco’s
Phosphate Buffered Saline (DPBS) and incubated at 37°C for 30 min. After the
incubation, the solution was aspirated and washed 2 times with DPBS.
Cells were fixed in glutaraldehyde (3%) for 10 min at room temperature.
Following removal of glutaraldehyde 1ml DPBS was added and cells were
scraped using a rubber policeman and centrifuged (5000 xg, 4 minutes). Cell
pellets were mixed with 50 µl of 4% agar solution, 60-70 nm and thin sectioned
prior to imaging by TEM.
Labeling of Ferritin with Activated Fluorescein Dye
Cysteine residues on both the HFn and the RGD4C-Fn were labeled with
fluorescein-5-maleimide for subsequent FACS analysis. The Fn and RGD4C-Fn
(100mM HEPES, 50 mM NaCl, pH6.5) were reacted with fluorescein-5maleimide (Molecular Probes, Eugene, OR) at room temperature for 30 min
followed by overnight incubation at 4°C. Fluorescein labeled protein cages were
purified from free dye using SEC.
Fluorescence activated cell sorting (FACS) analysis of C32 cells incubated
with fluorescein conjugated RGD4C-Fn
Flow cytometry was performed on a FACSCalibur, (Becton Dickinson, Mountain
View, CA) and analyzed using Cell Quest software (Becton Dickinson, Mountain
View, CA). C32 cells were non-enzymatically removed from cell culture dishes
by DPBS (-) with 1% EDTA, washed once with serum containing MEME and then
suspended in DPBS (+) with 1% BSA. The cell suspension was incubated with
fluorescently labeled protein cages on ice for 20 min. After the incubation, the
cells were washed 3 times with DPBS containing Ca2+ and Mg2+ + 1% FBS and
then resuspended in DPBS + 1% FBS for FACS analysis.
CCMV Mutagenesis:
Described by Brumfield et al. (see [55]). In brief, the DNA encoding the CCMV
coat protein was amplified from an infectious cDNA and cloned into the pPICZA
plasmid vector (Invitrogen EasySelect Pichia Expression Kit). The pPICZA
plasmid is a shuttle vector that can be utilized in both E.coli and P. pastoris, it
was designed for intracellular protein expression [72]. This plasmid was utilized
226
as a template for introduction mutations in the coat protein by polymerase chain
reaction (PCR) mediated mutagenesis (Stratagene). Two overlapping primers
are designed (see Table 8.1) that contain the DNA sequence corresponding to
the desired change flanked by wild type (or other genetic variant) template
sequence. The altered plasmid DNA can be amplified in E.coli and then
sequenced in order to ensure that the correct mutations have been introduced.
Then this plasmid DNA can be introduced into P. pastoris via electroporation.
Yeast containing the plasmid can be selected for by its ability to grow on Zeocin
containing medium (see [72] for more information). The CCMV coat protein is
then expressed within the yeast, where it self assembles into VLPs that can be
purified as described in Materials and Methods (Appendix A). The P. pastoris
system, also facilitates the “scale-up” of VLP production, since yeast can be
grown to high densities in fermentation systems (see below) [55, 72].
CCMV Purification from Yeast
(Protocol from Debbie Willits and Sue Brumfield [55])
The following buffers are needed:
Yeast Homogenization Buffer: 0.2M sodium acetate/acetic acid pH 4.8, 0.01 M
ascorbic acid, 0.01 M EDTA. Store at 4˚C – turns yellow when bad.
Virus Buffer: 0.1M sodium acetate/acetic acid pH4.8, 0.001M EDTA.
Perform the following steps, keep sample on ice as much as possible:
1. Place 100g of yeast cell paste into bead beater container (BioSpec
Products, Inc.) with 235g of glass beads (BioSpec Products, Inc.). Fill
container with homogenization buffer (~1.25 ml of buffer per gram of cell
mass).
2. Assemble apparatus, making sure to keep the sealing area free of all
liquid and glass beads, fill around top of bead beater with ice. Bead beat
for 3 minutes, off for 1 minute. Repeat minimum of three times.
3. Pour homogenate into centrifuge tubes without pouring off beads. Repeat
steps 1-3 with another 100g of yeast – same beads, more buffer.
4. Spin tubes in low speed centrifuge for 20-30 minutes at 10,000xg
(8,000rpm in a GSA rotor)
5. Decant supernatant into 25ml ultra tubes, balance and spin in a T30 rotor
for 2hours at 25000rpm at 4˚C.
6. Dump off supernatant – virus should now be in the pellet. Resuspend
pellets in virus buffer 500ul to 1ml per tube depending on size of pellet.
Pellets range from ~1/2 cm in diameter to 2cm. I usually let the pellets
resuspend at 4˚C overnight unless you’re in a hurry.
7. Combine all pellets in a 15 ml flacon tube and spin in low speed centrifuge
for 5 minutes at 5,000 rpm. If not clear transfer supernatant to
microcentrifuge tubes and spin 5 minutes at 14,000 rpm in a microfuge
tube. The clearer the solution the cleaner the virus off CsCl – the solution
will be yellow.
227
8. Fill each ultracentrifuge tube (Ultra clear Beckman 14 x 89 mm order
#344059) for the SW41 swinging bucket rotor with 10 ml of 39% (w/v)
CsCl in virus buffer.
9. Layer 2 ml of the virus solution onto the top of the CsCl solution and place
in rotor buckets and balance.
10. Spin SW41 rotor for 18-20 hours at 37000 rpm at 4˚C.
11. After run remove buckets from head and tubes from buckets and pull virus
band - will have a blue tinge when illuminated with a flashlight. – full virus
migrates to about 1cm from the bottom of the tube. Empty virus migrates
~ halfway down the tube.
12. Place sample in dialysis tubing (12,000-14,000MWC) and dialyze against
virus buffer, one change, overnight.
Glutaraldehyde Intra-particle Cross Linking of CCMV
(protocol from Mark Allen, Douglas Lab, MSU)
Glutaraldehyde crosslinking of CCMV involves several dialysis steps
(Spectra/Por molecularporous membrane tubing; MWCO 12-14,000). First,
dialyze CCMV into 50 mM HEPES 100 mM NaCl pH 7 (~2 hours). Second
dialyze CCMV into glutaraldehyde containing buffer (50 mM HEPES, 100 mM
NaCl, 10mM Glutaraldehyde (Amersham; 490 mM); therefore add 2ml of
glutaraldehyde to the 100 ml of HEPES) for 24 hours at 4oC. Next, dialyze
CCMV into 100 ml ammonium acetate containing buffer (50 mM HEPES 100 mM
NaCl, 10 mM ammonium acetate) for at least 2 hours at 4oC in order to stop the
reaction. Any primary amine containing buffer such as ethylene diamine or Tris
should serve to quench the reaction. Next, repeat the ammonium acetate
containing buffer dialysis step, this time with the addition of 1ml of 1M Tris (pH 7)
into the dialysis buffer. Lastly, dialyze exhaustively into any buffer of choice in
order to remove any excess glutaraldehyde and ammonium acetate, and Tris.
The success of the crosslinking reaction can by verified by SDS-PAGE; the
crosslinked virus should not even enter the stacking gel. In addition the elution
volume of cross-linked CCMV from size exclusion chromatography (SEC) should
be similar to non-cross-linked CCMV.
228
Fluorescent Labeling of Genetic Variants of CCMV
CCMV cages were labeled with fluorescent dyes (fluorescein-5-maleimide, Texas
Red C2 maleimide (Molecular Probes T-6008). CCMV S102C cages were easily
labeled with fluorescein-5-maleimide, under similar reaction conditions as those
described above (100 mM HEPES, 50 mM NaCl, pH 6.5, 3-fold molar excess of
dye per subunit, 20 minutes at room temperature, overnight at 4˚C, purification
by size exclusion in DPBS pH 7.4, ~50 -75% yield, 62 Fl/cage). In addition,
CCMV K42R T48C was labeled with Texas Red C2 maleimide and fluorescein-5maleimide using the same reaction conditions described above, excepting the
additional pre-treatment of the cages with 10-fold molar excess per subunit of
TCEP and reaction with Texas Red C2 maleimide at a concentration of 10-fold
molar excess compared to CCMV subunit concentration.