© 2010 brittney l. gurda · 1 mapping dominant antigenic regions on the capsid surfaces of...
TRANSCRIPT
1
MAPPING DOMINANT ANTIGENIC REGIONS ON THE CAPSID SURFACES OF ADENO-ASSOCIATED VIRIONS
By
BRITTNEY LYNN GURDA
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2010
2
© 2010 Brittney L. Gurda
3
To my mother, for reminding me that I could when I said “I couldn’t”, and my father, who said that he didn’t care what I grew up to be as long as it made me happy
4
ACKNOWLEDGMENTS
I would like to begin with my mom, Christine Gurda. Not only has she been there
financially, but also spiritually and mentally. I can think of no other who has shown greater
support in all of my endeavors. My father, Thomas Gurda, my sister, CaryBeth, and my brother,
Dean were always there for support. I feel exceptionally lucky and honored to have had an
excellent mentor. Dr. Mavis Agbandje-McKenna’s patience and encouragement make her the
ideal mentor and I could have never found a better place to start. I would also like to thank Dr.
Robert McKenna for insightful conversations and general advice in not only science, but life as
well. I also need to thank Michael DiMattia and Edward Miller for help with computational
programs. The monoclonal work was done in collaboration at Cornell University by Wendy
Weichert under Dr Colin Parrish, and the neutralization data was mostly done by Dr. Michael
Schmidt and Beverly Handelman under Dr. Jay Chiorini at the NIH. As for data collection, all of
our cryoEM data was collected at UCSD under the direction of Dr. Tim Baker by Norm Olson.
My thanks also go to Dr. Baker’s staff especially Drs. Robert Sinkovits and Kristin Parent who
helped with the HBoV structure in Appendix A. The data in Appendix B is almost entirely the
work of Dr. Christina Raupp in Dr. Jurgen Kleinchmidt’s lab in Hieldelberg, Germany and I am
excited to have been able to offer data that helped shaped the efforts of the epitope mapping of
the mAb ADK8. I am honored to have had the expertise of my committee members at my
disposal; Dr. James B. Flanegan, Dr. Nicholas Muzyczka, Dr. Roland Herzog, and Dr. Arun
Srivastava. All of whom are not only experts in their field, but also in the shaping of a project
and the student involved. The IDP graduate student office was invaluable in making my first year
transition smooth, and I would also like to thank the staff in the biochemistry office. Last but not
least, I would like to thank the hordes of undergraduate students that I have taught over the years,
who have unknowingly taught me many valuable lessons in return.
5
TABLE OF CONTENTS
page
ACKNOWLEDGMENTS.................................................................................................................... 4
LIST OF TABLES................................................................................................................................ 9
LIST OF FIGURES ............................................................................................................................ 10
LIST OF ABBREVIATIONS ............................................................................................................ 12
ABSTRACT ........................................................................................................................................ 14
CHAPTER
1 BACKGROUND AND INTRODUCTION .............................................................................. 15
Introduction to Genetic Disease and Gene Therapy ................................................................. 15 Gene Delivery Systems ............................................................................................................... 16
Non-Viral Delivery Methods .............................................................................................. 17 Physical approaches ..................................................................................................... 17 Chemical methods ........................................................................................................ 19
Viral Delivery Methods ....................................................................................................... 22 Retroviruses .................................................................................................................. 23 Adenoviruses ................................................................................................................ 25 Herpesvirus ................................................................................................................... 27 Alphaviruses ................................................................................................................. 29 Adeno-associated viruses............................................................................................. 31
Current Status and Issues in Gene Therapy ............................................................................... 33 Antibody Recognition of Viruses and Viral Neutralization ..................................................... 40 Parvovirus Capsid Structures, Receptor Binding, and Infection .............................................. 44 Parvovirus Antigenic Properties................................................................................................. 47 Parvovirus Antigenic Structures for Autonomous Viruses ...................................................... 48 The Dependoviruses and their Antigenic Properties................................................................. 51 AAV Capsid Structure and Serotypes........................................................................................ 54 Significance ................................................................................................................................. 56
2 MATERIALS AND METHODS ............................................................................................... 59
Production and Purification of Recombinant AAV VLPs for Structural Studies ................... 59 Fv Antibody Modeling Using the Web Antibody Modelling (WAM) Server ........................ 60 Preliminary Neutralization Assays ............................................................................................. 61 Preliminary In-House Green Cell Assays .................................................................................. 61 Dot Blot Analysis to check Cross-Reactivity among VLPs ..................................................... 62 Generation of Fragment Antibodies from Monoclonal Antibodies ......................................... 62 Preparation of Fab:VLP Complexes .......................................................................................... 63
6
CryoEM Data Collection ............................................................................................................ 63 Three-Dimensional Reconstruction of AAV:Fab Complexes .................................................. 63 Fitting of Atomic and Pseudo-Atomic Models into CryoEM Density Maps .......................... 64 Mutagenesis of Proposed Antigenic Regions ............................................................................ 65 Production and Purification of Antigenic Mutants ................................................................... 67
Western Blot Analysis ......................................................................................................... 70 AAV1 Enzyme-Linked Immunosorbent Assay (ELISA) ................................................. 70 Dot Blot Analysis for Antigenic Reactivity ....................................................................... 71 Pixel Intensity Calculations from Native Dot Blots .......................................................... 71 Infectivity Assays ................................................................................................................ 71 DNA Extraction and Real-Time Polymerase Chain Reaction .......................................... 72
3 CHARACTERIZATION OF ANTIBODIES AGAINST AAV1 AND AAV5 CAPSIDS, CROSS-REACTIVITY AND NEUTRALIZATION CAPABILITIES .................................. 74
Introduction ................................................................................................................................. 74 Results .......................................................................................................................................... 76
Production of Anti-AAV Monoclonal Antibodies and Isotyping..................................... 76 Cross-Reactivity among Different AAV Serotypes .......................................................... 77 Neutralization Ability of the Monoclonal Antibody Panel ............................................... 77
Discussion .................................................................................................................................... 78
4 DETERMINATION OF DOMINANT ANTIGENIC SITES ON THE CAPSID SURFACE OF ADENO-ASOCIATED VIRUS SEROTYPES 1, 2, 5, 6, and 8 THROUGH CRYO-ELECTRON MICROSCOPY STUDIES ................................................ 90
Introduction ................................................................................................................................. 90 Results .......................................................................................................................................... 93
Analysis of the Fab-VLP CryoEM Complexes ................................................................. 93 Fab G7 and AAV1 ............................................................................................................... 94 Fab D11 with AAV1 and AAV6 Capsids .......................................................................... 94 Fab C37B and AAV2 .......................................................................................................... 95 Fab F4 and AAV5 Capsids ................................................................................................. 96 Fab ADK8 and AAV8 Capsids ........................................................................................... 97 Fab ADK1a and AAV1 Capsids ......................................................................................... 98 Common Features of Antibody Binding Sites on the AAV Capsid ................................. 99
Discussion .................................................................................................................................. 100 Viruses as Antigens - Features Revealed ......................................................................... 100 Structures of Bound Antibodies and Possible Mechanisms of Neutralization or
Effects on Capsid Functions .......................................................................................... 101
5 MUTATIONAL ANALYSIS OF PROPOSED DOMINANT ANTIGENIC SITES ON THE CAPSID SURFACE OF ADENO-ASSOCIATED VIRUS SEROTYPE 1 ................ 111
Introduction ............................................................................................................................... 111 Results ........................................................................................................................................ 112
Generation of AAV1 Capsid Mutants .............................................................................. 112
7
Analysis of Mutant Protein Expression ............................................................................ 113 Characterization of Mutant Capsid Production................................................................ 113 Quantification of Mutant Particle Titers, Infectivity, and Packaging Ability................ 114 Ability of Antigenic Mutants to Evade Monoclonal Antibody Recognition ................. 115 Direct Effects of Mutations on Neutralization ................................................................. 116
Discussion .................................................................................................................................. 117 Preliminary Analysis on Mutant AAV1 Viruses Reveals a Retained Ability to
Assemble Infectious Capsids ........................................................................................ 117 Ability of Antigenic Escape is Limited to Specific Regions on the Capsid Surface .... 120
6 SUMMARY AND FUTURE DIRECTIONS ......................................................................... 128
Summary .................................................................................................................................... 128 Future Directions ....................................................................................................................... 131
APPENDIX
A HUMAN BOCAVIRUS CAPSID STRUTURE: INSIGHTS INTO THE STRUCTURAL REPERTOIRE OF THE PARVOVIRIDAE ................................................. 133
Introduction ............................................................................................................................... 133 Materials and Methods .............................................................................................................. 136
Expression and Purification .............................................................................................. 136 Negative Stain Transmission Electron Microscopy ........................................................ 137 CryoEM and Image Reconstruction ................................................................................. 138 Generation of Density Maps for Representative Members of the Parvoviridae
Genera ............................................................................................................................. 139 Sequence Analysis and Generation of Unrooted Phylogenic Tree ................................. 140 Generation and Docking of a Homology Model of HBoV VP2 into Cryo-
Reconstructed Density ................................................................................................... 140 Comparison of Parvovirus VP2 Structures ...................................................................... 141
Results and Discussion ............................................................................................................. 141 HBoV Capsid Structure ..................................................................................................... 141 Pseudo-Atomic Model of HBoV VP2 Built Based on the B19 VP2 Crystal
Structure.......................................................................................................................... 143 HBoV VP2 Shares Low Sequence Identity but High Structural Similarity to Other
Parvovirinae VP2s ......................................................................................................... 145 HBoV Capsid Surface Topology is Unique, but Shares Features Common to Other
Vertebrate Parvoviruses ................................................................................................. 146 Summary .................................................................................................................................... 149
B MONOCLONAL ANTIBODY AGAINST THE ADENO-ASSOCIATED VIRUS TYPE 8 CAPSID AND THE IDENTIFICATION OF CAPSID DOMAINS INVOLVED IN THE NEUTRALIZATION OF THE AAV8 INFECTION ........................ 159
Introduction ............................................................................................................................... 159 Materials and Methods .............................................................................................................. 162
8
Animals, Cell Lines and Cell Culture............................................................................... 162 Plasmids and Site-Directed Mutagenesis ......................................................................... 163 Plasmids and Peptide Insertions ....................................................................................... 164 Transfection of 293T Cells and Vector Particle Purification .......................................... 164 Purification and Titration of Mutant AAV Stocks .......................................................... 165 Analysis of Viral Protein Expression ............................................................................... 165 Production and Purification of Monoclonal Antibody ADK8 ........................................ 165 Determination of Antibody Isotypes ................................................................................ 166 AAV Capture ELISA ........................................................................................................ 166 In Vitro and In Vivo Neutralization .................................................................................. 166 Binding Analysis by DNA Dot Blot ................................................................................. 167 Immune Fluorescence ........................................................................................................ 168 Native Dot Blot Assay ....................................................................................................... 168 Epitope Mapping................................................................................................................ 169
Results and Discussion ............................................................................................................. 169 Three-Dimensional Reconstruction and Model Docking of the AAV8:ADK8
Complex.......................................................................................................................... 169 Analysis of the 3D Models Identifies Proposed Monoclonal Antibody ADK8
Binding Sites on the AAV8 Capsid .............................................................................. 170 Characterization of rAAV8- and rAAV2- Capsid Mutants to Elucidate the ADK8
Binding Epitope ............................................................................................................. 170 In Vitro and In Vivo Neutralization of AAV8 by ADK8 ................................................ 172 Neutralization Analysis of ADK8 .................................................................................... 174 Impact on rAAV8 vector generation ................................................................................ 176
LIST OF REFERENCES ................................................................................................................. 184
BIOGRAPHICAL SKETCH ........................................................................................................... 214
9
LIST OF TABLES
Table page 3-1 Mouse monoclonal antibodies selected for further studies.................................................. 85
3-2 Amino acid differences between AAV1-like viruses and AAV1* ..................................... 85
4-1 Summary of AAV-Fab complexes generated for cryoEM data studies ........................... 104
4-2 CryoEM data reconstruction statistics ................................................................................ 104
4-3 Scaling and fitting statistics of cryoEM 3D reconstructions with atomic models ........... 104
4-4 Proposed epitopes for each respective complex including residue numbers and sequence data ........................................................................................................................ 104
5-1 AAV1 mutational nomenclature and sequence differences .............................................. 122
5-2 AAV1 single-site mutant characterization.......................................................................... 122
A-1 Selected properties of representative members of Parvoviridae ...................................... 151
A-2 Sequence alignment statistics for representative members with structural data .............. 151
10
LIST OF FIGURES
Figure page 1-1 The antigenic structure of AAV2. ......................................................................................... 57
1-2 Panel of available AAV atomic structures. .......................................................................... 58
2-1 Coomassiee stained SDS-PAGE and EM images of purified AAV VLPs for structural studies. .................................................................................................................... 73
2-2 Images of the Fab WAM models. ......................................................................................... 73
2-3 Production, purification, and characterization of mAb and Fabs. ....................................... 73
3-1 Basic immunoglobulin structure ........................................................................................... 85
3-2 Monoclonal antibody characterization and cross-reactivity.. .............................................. 86
3-3 Preliminary neutralization assay data for the anti-AAV5 monoclonals. ............................ 87
3-4 Preliminary neutralization assay data for the anti-AAV1 monoclonals. ............................ 87
3-5 The ability of the anti-AAV1 monoclonals to neutralize the AAV1-like viruses.............. 88
3-6 Differnces between the AAV1 and X1 sequences are mapped to the atomic coordinates of AAV1 and cluster at the three-fold axis of symmetry. ............................... 89
4-1 CryoEM 3D reconstructions shown with respective roadmap of epitope positions on the 3D capsid surface ........................................................................................................... 105
4-2 Variable regions of the AAV-VP3 monomer ..................................................................... 106
4-3 Docking of atomic models into cryoEM density ............................................................... 106
4-4 Difference map between the AAV5:F4 reconstruction and a similar scale AAV5-uncomplexed map ................................................................................................................ 107
4-5 Three-dimensional reconstruction and model fitting for the AAV8:ADK8 complex ..... 108
4-6 AAV1:ADK1a reconstruction and model fitting ............................................................... 109
4-7 Illustration of overlapping antigenic regions at the capsid surface level .......................... 109
4-8 Structure based sequence alignment of AAV1, 2, 5, and 6.. ............................................. 110
5-1 Western blot analysis on single-site mutant cell lysates .................................................... 123
5-2 Purification of single-site antigenic escape mutants .......................................................... 123
11
5-3 Electron micrograph images of AAV1 mutant capsids.. ................................................... 124
5–4 ADK1a native dot blot of mutant cell lysates. ................................................................... 124
5-5 Analysis of a native dot blot of single and double-site mutants........................................ 125
5-6 Infectious green cell assay images of NAb treated single-site mutants. ........................... 126
5-7 Histograms for in vitro Nab treated mutant virus. ............................................................. 127
5-8 Heparin-binding analysis. .................................................................................................... 127
A-1 Characterization of recombinant HBoV VLPs................................................................... 152
A-2 HBoV cryo-reconstruction. ................................................................................................. 153
A-3 Comparison of six representative parvovirus capsid structures. ....................................... 154
A-4 Radial density projections of six representative parvoviruses. ......................................... 155
A-5 Pseudo-atomic model of HBoV VP2. ................................................................................. 156
A-6 Superposition of VP2 structures from six representative members of the Parvoviridae. ........................................................................................................................ 158
B-1 Three-dimensional reconstruction and model fitting for the AAV8:ADK8 complex.. ... 178
B-2 Homology alignment of a portion of the G-H loop between AAV2 and AAV8. ............ 179
B-3 Analysis of rAAV2 and rAAV8 mutant vector production. .............................................. 179
B-4 Binding ability of rAAV2 and 8 mutant capsids by the ADK8 monoclonal.................... 180
B-5 in vitro and in vivo neutralization data ................................................................................ 181
B-6 Neutralization analysis of the ADK8 mAb......................................................................... 182
B-7 ADK8 and its impact on the N terminus externalization in the post-entry process. ........ 183
12
LIST OF ABBREVIATIONS
AAV adeno-associated virus
AAV1 adeno-associated virus serotype 1
AAVS1 adeno-associated virus integration site S1
Ab antibody
AC artificial chromosome
Ad adenovirus
Ag antigen
BCR B-cell receptor
CC correlation coefficient
cDNA copy deoxyribonucleic acid
CF cystic fibrosis
CFTR cystic fibrosis transmembrane conductance regulator
Ch19 human chromosome chromosome 19
CNS central nervous system
cryoEM cryo electon microscopy
DMEM Dubelco’s modified eagle media
DNA deoxyribonucleic acid
ELISA enzyme-linked immunosorbent assay
Fab fragment antibody-binding region
FBS fetal bovine serum
Fc fragment crystallizable region
FV foamy vector
HFV human foamy virus
HIV-1 human immunodeficiency virus type 1
13
HSPG heparan-sulfate proteoglycan
HSV herpes simplex virus
LAT latency associated transcript
mAb monoclonal antibody
mRNA messanger ribonucleic acid
mtRNA mitochondrial ribonucleic acid
MHC major histocompatibility complex
PBS phosphate buffered saline
PDGFR platelet derived growth factor receptor
pDNA plasmid dexoyribonucleic acid
pEI polyethleneinimine
PLA2 Phospholipase A2
PLL poly-L-lysine
RNA ribonucleic acid
RT reverse transcriptase
RU resonance units
SFV simian foamy virus
SIN sindbis virus
SPR surface plasmon resonance
TCR T-cell recptor
TMB tetramethylbenzodine
VEE Venezuelan equine encephalitis
VH heavy chain
VL light chain
VLPs viral like-proteins
14
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
MAPPING DOMINANT ANTIGENIC REGIONS ON THE CAPSID SURFACES OF
ADENO-ASSOCIATED VIRIONS
By
Brittney L. Gurda
May 2010 Chair: Mavis Agbandje-McKenna Major: Medical Sciences – Biochemistry and Molecular Biology
The Adeno-associated virus (AAV) is a small ssDNA virus that can package and deliver
non-genomic DNA. Its low toxicity and relative lack of pathology, along with broad tropisms,
has made it popular as a vector for gene delivery trials. However, despite the critical role of
antibodies in the success of gene therapy with AAV vectors, the nature of the antigenic structure
of the AAV capsids is not well understood. Here we defined the binding of monoclonal
antibodies against AAV serotypes 1, 2, 5, 6, and 8 using cryo-electron microscopy and three-
dimensional image reconstructions. Docking of capsid structures with fragment antibody models
showed that the footprints on these viruses all fell on or around the three-fold protrusions of the
capsid regardless of the serotype examined. Significant overlap between the binding sites led to
the conclusion that this is a common region of antigenicity among the AAV capsid surfaces. In
preliminary studies, mutagenesis of the proposed sites on the AAV1 capsid showed that certain
surface loop changes resulted in viruses that were able to evade previous recognition from
parental antibodies. These data support the hypothesis that the three-fold axes are antigenically
important and that mutations in this region may avert a pre-existing immune response. Further
characterization is necessary to better design escape mutants and fully understand if these
mutants are still viable in their original tissue tropisms and transduction efficiencies.
15
CHAPTER 1 BACKGROUND AND INTRODUCTION
Introduction to Genetic Disease and Gene Therapy
In the central dogma of molecular biology the normal flow of biological information is
identified: DNA can be copied to DNA (DNA replication), DNA information can be copied into
mRNA (transcription), and proteins can be synthesized using the information in mRNA as a
template (translation) (Crick, 1970). There are numerous factors involved in each stage of this
flow that assist in the final outcome, which often leaves many opportunities for mistakes to
occur. Evolutionarily we have a distinct advantage in that there are also many checkpoints that
aid in the correct outcome of these processes. Even so, a mistake can occur, but it is often not
evident at what point it occurred, particularily for uncharacterized and multigenic diseases. When
the fault occurs at the level of DNA, it is ultimately our genes that become altered. These
abnormalities are generally referred to as genetic disorders and can range from a small mutation
in a single gene to the addition or subtraction of an entire chromosome or set of chromosomes. In
single gene disorders a mutation causes the protein product of a single gene to be altered or
missing, i.e., in cystic fibrosis. Multiple gene disorders result from mutations in multiple genes,
often coupled with environmental causes, i.e., Alzheimer’s disease. In chromosome
abnormalities, entire chromosomes, or large segments of them, are missing, duplicated, or
otherwise altered, i.e., Down syndrome.
With the completion of the human genome project in 2003 and recent technological and
medical advances, the number of genetic disorders discovered has escalated. There are currently
no cures for these diseases and most are treated symptomatically. In rare cases, replacement
protein therapies can alleviate the issues associated with the knock down or loss of a specific
protein. Another option is gene therapy. In this DNA-based therapeutic, a functional gene is
16
inserted or delivered to the cell in order to replace a dysfunctional one. Although this concept is
still in its infancy, many advances have been made over the last few decades that have identified
this option as a possible mainstream therapy. Gene therapy can be seen in two schools of
thought. The first, germ line therapy, takes place in the germ line cells, i.e., sperm or eggs.
Modifications are created by the introduction of functional genes, which are ordinarily integrated
into their genomes. Therefore, the change due to therapy would be heritable and would be passed
on to later generations. The second is aptly termed somatic gene therapy and is a more accepted
option that involves the passage of genetic material into the somatic cells of a patient. This
allows for alterations to the patient’s genetic material only and does not transfer genetically to
immediate or distant offspring (http://www.dnapolicy.org/science.gm.php). There are a variety of
different methods to replace or repair the genes targeted in gene therapy: (I) An abnormal gene
could be swapped for a normal gene through homologous recombination, (II) The abnormal gene
could be repaired through selective reverse mutation, which returns the gene to its normal
function or, (III) The regulation (the degree to which a gene is turned on or off) of a particular
gene could be altered (http://www.ornl.gov/sci/techresources/Human_Genome/medicine/
genetherapy.shtml). In disorders where mitochondrial DNA (mtDNA) is defective, a new
technique referred to as ‘spindle transfer’ can be used to replace entire mitochondria (Tachibana
et al., 2009). The most common approach is the delivery of a functional gene to a cell in context
of an episome. This approach requires the use of a vector for delivery of the gene into the desired
tissue.
Gene Delivery Systems
The delivery of naked DNA, or DNA-based drugs, has had limited success due to its poor
cellular uptake profile and biological stability, as well as a short half-life resulting in
unpredictable pharmacokinetics (Patil et al., 2005). Furthermore, a broad application of naked
17
DNA–mediated gene transfer to gene therapy may not be conceivable because DNA, being large
in size and highly hydrophilic, is efficiently kept out of the cells in a whole animal by several
physical barriers. These include the blood endothelium, the interstitial matrices, the mucus lining
and specialized ciliate/tight junction of epithelial cells, and the plasma membrane of all cells
(Gao et al., 2007). The use of DNA delivery systems has not only improved the
pharmacokinetics of DNA-based therapeutics but has also achieved efficient targeted
introduction of these molecules into desired tissues (Patil et al., 2005). An ideal vector should be
safe, stable, easy to produce in large quantities, and capable of achieving efficient and tissue
specific gene expression when directly administered in vivo (Li and Ma, 2001). These guidelines
have aided in the development of numerous delivery methods. There are currently two types of
vectors for gene delivery; non-viral and viral vectors. Each system has its own set of advantages
and disadvantages in both the production aspect and clinical counterpart.
Non-Viral Delivery Methods
Non-viral delivery vectors are particularly suitable with respect to simplicity, ease of large-
scale production, and lack of specific immune response. The delivery of therapeutic DNA in the
absence of viral counterparts employs both physical and chemical approaches. The physical
approaches include needle injection, hydrodynamic, electroporation, gene gun, and ultrasound
delivery, which employ a physical force that permeates the cell membrane and facilitates
intracellular gene transfer.
Physical approaches
Direct injection of naked plasmid DNA (pDNA) intramuscularly has been shown to
transfer genes to myofibers, although the levels of gene expression were shown to be low (Wolff
et al., 1990). Intratracheal administration of pDNA in the mouse airway has been shown to be
safe and effective (Meyer et al., 1995), which warrants further study for use in pulmonary
18
diseases where gene transfer could be beneficial. Studies involving intravascular injection
generally show limited transfer in all major organs (Li and Ma, 2001), however Budker et al.
(1998) reported that intravascular injection of naked DNA under high hydrostatic pressure in rats
led to high levels of foreign gene expression throughout all of the muscles of the targeted limb.
These studies led to the observation that levels of gene expression could be increased by
improving the method of injection, such as using hydrostatic pressure and clamping veins near
the major injection vein to increase uptake of pDNA.
Electroporation works through the notion that exposing cells to an intense electrical field
induces a transmembrane potential that facilitates cell permeabilization (Neumann et al., 1982).
A number of tissues have been examined including liver, skin, subcutaneous tumors, and muscle
(Li and Ma, 2001). Most reports showed 10- to 1000-fold differences in gene expression between
naked DNA injection and DNA injection followed by electroporation, and in many cases, a
broad distribution of transfected cells was also achieved (Li and Ma, 2001). Many tissues,
including liver cells, testes, skin, arteries, and tumors responded with high gene expression after
electropartion (Parker et al., 2003), but it is noted that tissues may respond differently to
electroporation due to the variation in their local biological environment (Li and Ma, 2001).
Typical voltages used in studies can be 1300 V/cm with 100 μs pulses (Heller et al., 2000)
however; ultra-short pulses of nanosecond (ns) durations and 10–300 kV/cm have been used for
electroporation (Schoenbach et al., 2001). Since markers of apoptosis, including
phosphatidylserine translocation and caspase activation have been observed with high-intensity
nanopulse electroporation it is thought that these methods could be useful for cancer/gene
therapy (Sundararajan et al., 2009). Besides local tissue damage, other drawbacks include the
19
amount of DNA needed to reach therapeutic levels of transgene expression as well as
translational methodologies to include internal organs (Parker et al., 2003).
The gene gun was originally designed to deliver genes to plant tissues. It has shown
promise in animal tissues as well, albeit limited to superficial tissues. In this method, particles of
varying sizes (1-3µm) are coated in pDNA and expelled out by a powerful burst of air. The
bombardment of the particles through the tissues allows for delivery of the genes. This approach
has been reported to be superior to other methods of plasmid delivery to the skin, but tends to
generate a Th2 (humoral) rather than the Th1 (cytotoxic) immune response that commonly
follows intramuscular administration (Wells, 2004). For this reason, biolistic approaches are
heavily studied in the delivery of DNA vaccines.
Ultrasound is considered a non-invasive technique and has been used in the clinical setting
for diagnostic and therapeutic applications. The application of ultrasound results in acoustic
cavitation that can disrupt tissues and produce transient membrane permeabilization, thus
enhancing the delivery of plasmid to the cytoplasm (Wells, 2004). Nucleation agents such as
ultrasound contrast agents can enhance cavitation, and several studies have examined a range of
such contrast agents, concluding that albumin-coated octa-fluoropropane gas microbubbles
(Optison) is preferable to several other commonly available agents (Wells, 2004). Researchers
have reported successful delivery of naked pDNA at low frequencies in lung (Xenariou et al.,
2010), heart (Bekenedjian et al., 2003), tendon (Delalende et al., 2010), and skeletal muscle
(Taniyama et al., 2002). However, current levels of cellular transfection have not yet reached
therapeutic levels (Parker et al., 2003) using this approach.
Chemical methods
The chemical approaches use synthetic or naturally occurring compounds as carriers to
deliver the transgene into cells (Gao et al., 2007). Efforts to develop deleivery systems that
20
mimic the desired properties of a viral vector has incorporated the use of cationic lipid and
protein carriers, termed lipoplexes and polyplexes, respectively.
Lipoplexes consist of a positively charged or cationic lipid outer layer that encloses
negatively charged DNA molecules. The positively charged lipid layer can self assemble around
the negative DNA giving these lipoplexes more effective transfection efficiency of the negatively
charged host cell membrane. These particles can be ~100-300nm in diameter (Midoux et al.,
2009; Parker et al., 2003) and usually vary in shape depending on the DNA: lipid ratios.
Internalization of the lipoplex is thought to occur through clathrin-mediated endocytosis
(Midoux et al., 2009) and its escape still remains a mystery. Several thoughts about their escape
from the endosome have been proposed, but it seems that a majority of the DNA complexes
become trapped in the endosome and are eventually destroyed when fusion with a lysosome
occurs (Midoux et al., 2009). This has greatly limited the efficiency of lipoplexes in gene
delivery and has led to the use of fusogenic lipids and protein, as well as the incorporation of
histidines and imidazole, in the complexes for improved endosomal escape (Midoux et al., 2009)
to the nucleus. Another drawback is the tendency of the negatively charged particle to become
associated with positively charged proteins and glycans in the extracellular milieu (Parker et al.,
2003). Researchers are also investigating the interaction of lipoplexes with membrane-associated
proteoglycans into order to better characterize the effect of cell surface receptors and trafficking
pathways on this approach (Mislick and Baldeschweieler, 1996).
Polyplexes are cationic polymers that have been mixed with DNA to form stable
complexes. Several polymers have been studied for use in gene delivery including linear, i.e.
poly-lysine (PLL), and branched, i.e. polyethyleneimine (pEI), as well as block and grafted
copolymers (De Smedt et al., 2000). These particles are often nanoparticulate (<100nm) and
21
homogenous (Parker et al., 2003). Many of these studies have also involved the addition of
ligands to aid in targeting specific tissues, i.e. galactose for hepatocyte delivery (Zanta et al.,
2007). The uptake of polyplexes appears to happen through both clathrin-mediated and
cholesterol-dependent pathways, which can also occur simultaneously (Midoux et al., 2008).
Peptides that mimic viral proteins have also been used in conjunction with nanoscopic material
to help target delivery to and aid in passage across the cell membrane (Roy et al., 2009). The
major barrier is, as for lipofection, endosomal escape (De Smedt et al., 2000).
Protocols that use PLL have incorporated the use of endosome-lytic functions, such as
inactivated adenovirus or fusogenic peptides (Li and Ma, 2001). The most promising studies
have used pEI as it appears to have an inherent mechanism for escape into the cytoplasm (Parker
et al., 2003). Although the mechanism of endosomal escape is still not well understood there are
several ideas that have been proposed. These include an endosomal buffering capacity,
quenching of lysosomal acidification through protonation of amine groups, and an increase in
osmolarity leading to eventual rupture of the endosomal membrane (Parker et al., 2003).
Other novel non-viral delivery methods include transposons, where the advantage of virus
and naked DNA is combined, and artificial chromosomes (ACs). Both are highly appealing for
gene therapy. Transposons are naturally occurring with a desirable safety/toxicity profile and
integrate into chromosomes, providing long-term expression of the gene of interest in cells
(Izsvak et al., 2009). Current studies include recombinant forms of a fish transposon, Sleeping
Beauty, and have been fairly successful, but require further optimization of cellular transduction
(Ivics and Izsvak, 2006). On the contrary, ACs are autonomous and do not integrate into the host
chromosome. Although they offer escape from other issues surrounding gene therapy including,
22
insertional mutagenesis and immune responses, they are highly complex, difficult to deliver, and
technically challenging to construct (Macnab and Whitehouse, 2009).
Optimization of non-viral vector delivery methods is a growing field due to the safety of
the vectors, no or low immune response, and their relative ease of production. In summary,
major issues that surround these methods include specific tissue targeting, cellular uptake,
trafficking to the nucleus, and stable translation. Many of these issues are being addressed by
studying and applying viral methods, such as the addition of viral peptides, which will provide
safer forms of delivery and give an advantage over viral vectors.
Viral Delivery Methods
Viruses have spent thousands of years evolving to deliver their genomes into our tissues in
order to express the proteins needed to continue their lifecycle. The application of viruses as
vectors in gene transfer involves members who naturally infect mammalian tissues and can be
used in diseases which are genetic or environmentally acquired (Lundstrom, 2003). Although
progress has been slower than expected, many advances have been made over the past two
decades which have lead to various successes and unfortunately, some failures (Flotte, 2007).
The engineering of integrating (retroviruses) and non-integrating (herpes, adenovirus, and adeno-
associated viruses) has made it increasingly obvious that there are no universally applicable ideal
viral vector systems available (Lundstrom, 2003). Several factors should be reviewed before the
application of a vector including the various features of each vector as well as the type of disease
being treated (Lundstrom, 2003). Brief discussions of several popular viruses currently being
studied as vehicles in gene delivery are given below, as well as the major advantages and
disadvantages of each vector.
23
Retroviruses
Retroviruses have a large 7-10kb single-stranded RNA (ssRNA) genome which must be
reverse-transcribed in order for viral replication to occur. Retroviruses follow a special case in
the central dogma of molecular biology by first transcribing its mRNA to DNA. In order to do
this, the virus encodes for its own reverse transcriptase as well as an integrase, to assist with
integration of the viral genome into the host genome for transcription. These proteins are
translated from the polymerase or pol gene which also encodes for a viral protease. Retroviruses
are enveloped viruses which obtain their membrane from the host during cell budding. This
envelope is decorated with surface antigens that are encoded in the envelope or env gene and
expressed on host cell membranes. The capsid proteins and membrane associated matrix proteins
are found inside the membrane and are encoded by the group-specific antigen or gag gene
(Coffin, 1992). Several of these viruses have been studied for use as gene therapy vectors,
including members of the Gamma-retroviruses (Buchschacher, 2001), Lentiviruses (Valori et al.,
2008) and Spumaviruses (Trobridge, 2009). Pioneering of in vivo gene transfer involved the
retroviruses, namely murine leukemia virus (Lundstrom, 2004). The lentiviruses were used to
show a vector’s potential to transduce cells which have no or very low mitotic activity, and have
further since been developed for clinical application in neurological disorders (Lundberg et al.,
2008). Vectors based on human immunodeficiency virus type 1 (HIV-1) have been developed for
enzyme replacement and/or delivery of neurotrophic factors for neurodegenerative diseases and
CNS manifestations of lysosomal storage diseases, as well as delivery of glial cell line-derived
neurotrophic factor and ciliary neurotrophic factor for Parkinson’s and Huntington’s diseases
(Lundberg et al., 2008). These vectors have also shown great promise in the delivery of small
hairpin RNA for gene therapy in neurological disorders and transgenic knockdown animals
(Singer and Verma, 2008).
24
The retroviruses offer the advantage of large packaging capabilities, the ability to
efficiently transduce non-dividing cells, and to the ability to maintain stable gene expression
(Cockrell and Kafri, 2007). Their ability to integrate into the host genome is appealing for stable
gene expression, but is limited to cell division and has also been shown to randomly integrate (Li
et al., 2002). In comparison to conventional retroviruses, the lentiviruses can infect both dividing
and non-dividing cells (Lundstrom, 2004). One of the greatest successes in gene therapy trials to
date was the ex vivo retroviral mediated transfer of the γc gene into CD34+ cells for X-linked
severe combined immunodeficiency disease (X-SCID) (Hacein-Bey-Abina et al., 2002).
Unfortunately, aberrant viral integration into the LMO-2 locus was found to cause leukemia in a
few patients and the trial was halted (Hacein-Bey-Abina et al., 2003). With the increase in
knowledge of retrovirus biology, especially with the lentivirus, the engineering of non-
integrating vectors continues (Sarkis et al., 2008; Valori et al., 2008). As toxicity of replication
competent vectors have increased, efforts to improve the safety and titers of recombinant
retroviral vectors have included the creation of replication incompetent vectors by the
introduction of cis-acting elements in trans during vector production (Sinn et al., 2005). Tissue
targeting is also an issue with many of the retroviruses due to their strict receptor specificity. To
address this issue, researchers have developed ‘pseudotyping’approaches, which can greatly aid
in re-trageting (Sanders, 2002). In this method, the retrovirus’ native envelope is replaced with a
helper plasmid expressing heterologous envelope glycoproteins, including lyssaviruses,
arenaviruses, hepadnaviridae, flaviviridae, paramyxoviridae, baculovirus, filoviruses, and
alphaviruses (Sinn et al., 2005).
The Foamy virus (FV) is a spumavirus, another subfamily of the retroviruses, and is
unique in that they have evolved to exhibit no pathology in their host during transmission and
25
infection (Trobridge, 2009). The human foamy virus (HFV) was found to have evolved from the
simian foamy virus (SFV) and it appears that humans are an ‘accidental’ host (Meiering and
Linial, 2001). There is no evidence of human-to-human transmission and it also appears to be
contained among individuals working amid non-human primates with active infections (Meiering
and Linial, 2001). Their other promising attributes for gene therapy consist of a desirable safety
profile, a broad tropism, a large transgene capacity, and the ability to persist in quiescent cells
(Trobridge, 2009). FV vectors have been shown to efficiently transduce embryonic stem cells
and neuronal progenitor cells (Trobridge, 2009), and their resistance to human sera provides
further advantage for in vivo usage (Russell and Miller, 1996).
Adenoviruses
The Adenoviruses (Ad) are large, non-enveloped, double-stranded DNA viruses that are
commonly associated with upper respiratory tract infections and have been linked to other
illnesses such as gastroenteritis, conjunctivitis, cystitis, and rash (Horwitz, 1995). Positive factors
making Ad an attractive vehicle for gene transfer include their ubiquitous nature; over 100
serotypes have been isolated from many different species, their ability to transfect a broad range
of human cells, including non-proliferating cells, their ability to yield a high rate of gene transfer,
their relative low pathogenicity; causing mild symptoms associated with the common cold, their
large 7.5 kb genome, and finally, their genome stability that allows for the maintenance of
recombinant genes (Vorburger and Hunt, 2002).
Ad vectors were originally studied for use as vehicles for gene transfer in patients suffering
from cystic fibrosis (CF), a monogenic disease where a mutation disables the cystic fibrosis
transmembrane conductance regulator (CFTR) gene, due to their specificity for pulmonary
epithelial cells (Wilson, 1996). Adenoviral DNA was also found in liver, skeletal muscle, heart,
brain, lung, pancreas, and tumor tissue (Bramson et al., 1995), expanding the range of cell lines
26
for gene delivery. Interestingly, when Ad vectors are administered intravenously, most of the
virus accumulates in the liver (Vorburger and Hunt, 2002).
Preliminary trials were met with several problems, including transient expression and a
host immune response (Wilson, 1996). It was eventually realized that viral products needed to be
limited in order to minimize the immune response and from this realization came the second
generation vectors. The Ad genome contains early genes which encode for regulatory proteins,
and late genes which encode for structural proteins. In order to make the Ad virus less
immunogenic, additional early genes were removed to stop the up-regulation of structural
proteins that would continue the viral lifecycle (Wilson, 1996). Unfortunately, these vectors were
still toxic due to low transcription levels of the remaining viral genes (Yang e al., 1995). One
recombinant strategy to improve the vector included a Cre-loxP helper-dependent (HD-Ad)
vector in which all viral coding sequences were deleted and a helper virus supplied the necessary
functions in trans (Parks and Graham, 1997). Another novel design in Ad vectors incorporated a
transposon system in which the gene of interest could be inserted into the host genome (Yates et
al., 2002). Other strategies include the creation of hybrid vectors, which combine the infectivity
of Ad with the integration processes of retroviruses or adeno-associated viruses, retargeting of
Ad to non-CAR-expressing cells, and promoter regulation at the packaged genome level
(Vorburger and Hunt, 2002).
Although many of these systems have improved toxicity and prolonged gene expression
(Schiedner et al., 1998), re-administration of the vector is still problematic. In most cases, an
initial immune response is seen upon first administration of the vector, and furthermore, most
adults have been exposed to Ad2 and 5, the two serotypes most commonly used in gene therapy
(Vorburger and Hurt, 2002). Because the Ad gene does not integrate into the host genome, vector
27
re-administration is necessary, especially in chronic problems, like CF. However, this transient
display of gene expression has been found to be useful in issues where a patient may benefit
from an induced immune response, especially against cancer cells (Vorburger and Hunt, 2002).
Ad vectors are commonly used in clinical trials for cancer, including head and neck,
prostate, breast cancer, melanomas (Vorburger and Hunt, 2002), and ovarian cancers (Matthews
et al., 2009). Several unique vectors have been designed to target cancer cells and are termed
conditionally replicating adenoviruses or CRAds (Matthews et al., 2009). One of the best
examples is ONYX-015 (Khuri et al., 2000). This system is modified in the early gene region,
specifically, E1b, which normally allows the virus to bind and inactivate the p53 gene (Barker
and Berk, 1987). Ironically, this mutated virus cannot replicate in normal cells with a functioning
p53 gene, but can replicate and kill cancer cells that have a defective p53 gene (Vorburger and
Hunt, 2002). One of the first commercial gene therapy products, Gendicine, which was approved
by the State Food and Drug of China (SFDA) in 2005, is based on an Ad5 vector that expresses
p53 (Wilson, 2009). More than 20 kinds of cancer indications have been treated with Gendicine,
such as head and neck squamous cell carcinoma, lung cancer, breast cancer, and liver cancers
(Peng, 2005). Another Ad-based therapy, Ocorine (H-101) was also released in China alongside
Gendicine in 2005 (Shi and Zheng, 2009).
Herpesvirus
The herpes simplex virus (HSV) is one of the largest viruses being studied for use in viral
mediated gene delivery. The two most commonly studied human species are HSV-1 and 2,
which are usually associated with watery blisters on the lips, mouth, and genitals. The HSV
genome is single stranded DNA and encodes for ~100-200 genes. This relatively large, complex
genome is ~152kb and is encapsulated in an icosahedral shell. The capsid is then surrounded by
the tegument, which contains about 20 different proteins that have structural and regulatory roles
28
(Marconi et al., 2009). The tegument is encompassed by a membrane that is obtained when the
virus leaves the host cell. This membrane is decorated with glycoproteins that are translated from
viral genes at the late stage of infection. The HSV infection is characterized by immediate-early,
early, and late gene translation. The appealing features of the HSVare not only the large
packaging ability, but also its neurotropism and ability to generate latent infections (Lundstrom,
2004). A latent infection is established in the neural ganglia through the expression of a Latency
Associated Transcipt. This allows the virus to maintain the host cell, by-passing cell death
mechanisms, in order to preserve a copy of the virus (Wang et al., 2005).
Full-length HSV are neurotoxic and are generally manipulated to be either replication-
competent attenuated vectors, replication-incompetent recombinant vectors, or defective helper-
dependent vectors known as amplicons. In attenuated vectors, the removal of several non-
essential genes that are involved in the replication, immune evasion, and virulence of HSV
allows for an alteration of the HSV virulence (Marconi et al., 2009). These have been tested as
live viral vaccines, as oncolytic viruses, and as gene therapy vectors to deliver transgenes to the
nervous system (Marconi et al., 2009). The replication-deficient vectors have mutated or deleted
essential genes that render the virus incapable of replication (Marconi et al., 2009). These vectors
are generated in transformed cell lines that offer the missing replication genes in trans which
allows for the generation of ‘second-generation’ vectors that have reduced toxicity associated
with them (Marconi et al., 2009). The removal of these genes also allows for larger foreign gene
products, or multiple genes under separate promoters (Marconi et al., 2009). The HSV-amplicon
system has gained significance as a versatile gene transfer platform due to its extensive transgene
capacity, widespread cellular tropism, minimal immunogenicity, and its amenability to genetic
manipulation (de Silva and Bowers, 2009). Among other components, the amplicon has only two
29
non-coding sequences from the wild-type HSV-1 genome, which are required for replication and
packaging of the amplicon into infectious particles (de Silva and Bowers, 2009). Upon delivery
to cells, the genome remains episomal and does not integrate into the host system (de Silva and
Bowers, 2009). Since this would be lost during cell replication, several factors have been added
to the amplicon to allow for replication-competency or to enhance its episomal retention within
the transduced cell nucleus (Lufino et al., 2008). Packaging systems for the amplicons have also
been developed and included either helper-free or helper-dependent systems based on HSV
cosmids or HSV replication-deficient vectors, respectively (reviewed in de Silva and Bowers,
2009). Re-targeting of the HSV vector’s natural tropism has also been attempted through the
genetic manipulation of HSV envelope glycoproteins, namely gC, gD, and gB (de Silva and
Bowers, 2009). Conversely, restriction of the vector to its natural cell targets has also been
attempted by replacing envelope glycoproteins with ligands (de Silva and Bowers, 2009). HSV-
amplicons have been incorporated into therapeutic studies for disease of the CNS, including
Parkinson’s, Ischemia, and hereditary ataxia, as well as anti-cancer therapies (de Silva and
Bowers, 2009). The limitations for implementation of HSV-amplicons in clinical trials continue
to be vector production, extension of transgene expression (mainly conventional amplicon
vectors), and regulation of transgene expression (de Silva and Bowers, 2009). Overall, the major
issue with HSV vectors is its innate and humoral immune response in the host system (Flotte,
2007).
Alphaviruses
The alphaviruses have been found to infect a wide variety of hosts, including humans,
rodents, birds, and horses. Their overall structure consists of a 9.7-11.8kb ssRNA genome that is
housed in a nucleocapsid, which is surrounded by a membrane (http://www.expasy.org/
viralzone/all_by_species/625.html). They have a relatively simple genome that encodes for both
30
structural and non-structural genes. The structural genes give rise to the capsid protein, as well as
the envelope glycoproteins (E1-3), while the non-structural genes are important for transcription
and replication. Alphaviruses, such as Semliki Forest (SFV) (Liljestrom and Garoff, 1991),
Sindbis (SIN) (Xiong et al., 1989), and one of the encephalitis-producing members, Venezuelan
Equine Encephalitis (VEE) (Davis et al., 1989), have all been used to develop vehicles for gene
transfer. There are several general features that make the alphaviruses appealing for gene transfer
including; positive-sense RNA, which allows for direct translation of the message, mass
production of viral mRNA which shuts down host protein synthesis and eventually leads to
apoptosis (may be beneficial in cancer therapy), their efficient packaging ability, and finally,
their affinity for a broad range of host cells (Ehrengruber and Lundstrom, 2007).
The alphaviruses have been used extensively for recombinant protein expression in cell
lines, in expression studies in neurons, and in vaccine production (Lundstrom, 2003).
Furthermore, it appears that while SFV and SIN have been developed further on the gene therapy
front, VEE has had more promise as a vaccine vector (Pushko et al., 1997; Lundstrom, 2004),
specifically as a DNA vaccine for bioterrorism (Lee et al., 2005; Dupuy and Schmaljohn, 2009).
Animal studies using SFV vectors have shown efficient gene transfer of reporter and therapeutic
genes to quite a few tumor models (Lundstrom, 2003). In the case of SIN vectors, a
morphological mutant vector has been shown to allow for an increase in packaging size from
~12 kb to an 18 kb cDNA insert (Nanda et al., 2009). It has been speculated that this
heterologous vector could possibly package and transcribe an RNA insert as long as 32 kb
(Nanda et al., 2009). Unfortunately, alphavirus is cytotoxic to most cell lines causing cell death
between 12 and 24 hours post-infection (Schlesinger, 2001). On the other hand, studies in insect
cell lines have shown that a persistent, non-cytopathic infection occurs and that virus production
31
can reach up to a 100-fold higher compared to mammalian cells (Schlesinger, 2001). Vector
design has included the use of replicons that eliminates most of the structural proteins, as well as
the use of two defective helpers, which offers the advantage of overcoming the ‘copy-choice’
mechanisms of the replicase in order to avoid infectious particle production (Schlesinger, 2001).
Replicons with reduced cytopathogenicity have been identified that allow replication of the virus
in certain cell lines without causing cell death (Schlesinger, 2001). Further limitations of the
vector include transient transgene expression, a broad host tropism, and a neuron-specific
transgene expression (Lundstrom, 2003). In order to overcome these issues, advances in
targeting have included genetic alteration of the E2 glycoproteins (Schlesinger, 2001), which
determines binding to cell receptors (Dubuisson and Rice, 1993), in addition to encapsulation of
the vector into liposomes (Lundstrom, 2003). A phase I trial using encapsulated SFV-particles
produced no toxicity to the liposome or SFV-particle and allowed for repeat treatment with the
vector (Lundstrom, 2003).
Adeno-associated viruses
The adeno-associated viruses (AAVs) are the smallest of the gene transfer vectors, but
have had a large impact in the field of gene therapy. Although over 100 different genotypes of
AAV have been isolated from human and non-human primate tissues alone, only a handful of
them have been serotyped and studied through structural analysis (Gao et al., 2005; Xie et al.,
2002; Govindasamy et al., 2006; Nam et al., 2007, Agbandje-McKenna unpublished data). The
serotyped members include AAV1-12, with AAV2 being the best characterized and most studied
among them (Mori et al., 2004). The AAVs are members of the genus dependovirus and are
dependent on a larger DNA virus, like adenovirus, for replication. Ironically, they were
originally found as contaminants of adenoviral stocks (Atchison et al., 1965) and have been
continually isolated from a very broad range of hosts, including goats, birds, sheep, snakes, and
32
cows (Olson et al., 2004; Luchsinger et al., 1970; Farkas et al., 2004; Clarke et al., 1979; Yates et
al., 1973). The AAVs are non-enveloped and have a very simple genome, encoding just two
open readings frames, rep and cap. Because of its simplicity, as well as relative lack of
pathogenicity, low toxicity, and ease of manipulation, the AAVs have been extensively studied
as vectors in gene transfer. These vectors are also able to infect dividing and non-dividing cells,
and can offer integration into a unique site, AAVS1, on human chromosome 19 (Ch19) under
certain constraints, i.e. inclusion of rep proteins and specific integration site sequences (Surosky
et al., 1997). The different serotypes have also been identified in animal studies as having
different specificities for different tissues. For example, AAV1 has shown to have the best gene
expression in the muscle, while AAV8 has repeatedly been superior in the liver (Thomas et al,
2004; Davidoff et al., 2005; Jiang et al., 2006; Wang et al., 2005). Although AAV9 has been
found to be best in cardiac tissue (Pacak et al., 2006), a recent study with AAV9 has identified an
ability to cross the blood-brain barrier resulting in widespread transduction in neurons
throughout the brain including the neocortex, hippocampus and cerebellum (Foust et al., 2009).
AAV vectors have already been tested in clinical trials for CF, hemophilia B, Parkinson’s
and Canavan disease, late infantile neuronal ceroid lipofuscinosis, arthritis, and an ex vivo study
to treat prostate cancer (Carter, 2005). The greatest success story involves a recombinant AAV2
(rAAV2) vector packaging the RPE65 gene for treatment in patients with Leber’s Congenital
Amaurosis (LCA). These studies have not only improved eyesight in both animal models and
human patients, but have also produced stable gene expression for over a year for continued
improvement of the disease (Bainbridge et al., 2008; Maguire et al., 2008; Hauswirth et al.,
2008; Cideciyan et al., 2009). Although these vectors have shown relative success in the field of
gene therapy they also have a few disadvantages. First, their small size allows for minimal
33
packaging, and attempts to package more than 5 kb have been quite unsuccessful (Lai et al.,
2010). This limits the size of the therapeutic gene that can be packaged as well as the disease
targets, as many monogenic diseases can also involve large gene defects. Second, many of the
recombinant vectors are difficult to produce in the laboratory setting and efficient yields are
difficult to produce for large scale studies (Flotte, 2007). Third, wild-type AAV is known to have
an integration site on the human chromosome during latent infection (Kotin et al., 1990;
Samulski et al., 1991) and efforts to characterize this phenomenon will aid in the exploitation of
AAV integration (Daya et al., 2009). The development of integrating AAV vectors may also help
deal with the transient expression of AAV delivered therapies in dividing cells. Lastly and one of
the largest concerns, is the immune response against AAV vectors. Fortunately the immune
response is not lethal, but does raise concerns for re-administration of vectors and cell-mediated
death of transduced cells. Several trials in animal models have been able to express therapeutic
genes stably in target tissues often resulting in impressive levels of protein expression, but
eventually are overcome by an immune response in human trials (Wang et al., 2007; Zaiss and
Muruve, 2008). The types and extent of these responses will be addressed further in the
following sections, since the focus of this study is on the antigenic nature of the AAV capsid,
more specifically, the humoral response.
Current Status and Issues in Gene Therapy
To date, the Food and Drug Administration (FDA) has not approved any human gene
therapy products. Although many animal trials have been relatively successful (hemophilia B,
RPE, color-blindness), past human clinical trials have had moderate to little success. However,
success that has occurred throughout the last decade has gained gene therapy a runners-up title in
Science’s 2009 “breakthrough of the year” issue (The News Staff, 2009). Gene therapy is
considered experimental and still has many hurdles to overcome. The field has definitely grown
34
in knowledge since its conception and there are many different approaches to the current issues.
These issues, which are briefly discussed in the section following, include short-lived gene
transfer, insertional mutagenesis, immune responses to the vector and the transgene, vector
toxicity, ability of viral vectors to revert to wild-type virus, horizontal gene transfer, and the
correction of multi-gene disorders.
Retroviral and lentiviral vectors appear to stand at the fore-front of integrating vectors due
to their inherent ability to integrate into the host genome. Although this allows for long-term
gene expression, there are concerns involving the efficiency and effects of integration.
Integration of the vector into host genes or regulatory elements can knock out the gene or cause
truncation products, or even lead to changes in expression patterns (Voigt et al., 2008). These
results can have devastating effects on the host, including causing cancer (Voigt et al., 2008).
The AAVs also have the ability to integrate into the host genome at Ch19 where a persistent
latency is established. Studies have shown that vectors containing rep gene were more efficient
at integration compared to vectors without AAV Rep (McCarty et al., 2004). The mechanisms
and efficiency for AAV integration have been studied by several groups (McCarty et al., 2004;
Daya et al., 2009; Linden et al., 1996) and have had difficulty in defining either the mechanism
or its actual efficiency, which is apparently very low (McCarty et al., 2004). However, these
studies have also identified important regions on Ch19 for integration as well as cellular
components and DNA sequences (McCarty et al., 2004; Daya et al., 2009; Linden et al., 1996).
One such study examines the rate of integration in mouse neonate livers. This study found that at
least 0.05% of hepatocytes contained rAAV integration, while 68% of integrations occurred in
genes, but none of them were near the mir-341 locus, the common rAAV integration site found
in mouse hepatocellular carcinoma (Inagaki et al., 2008). Of course, integration is not limited to
35
viral vectors. Non-viral methods rely on the synthetic fish and frog transposons, Sleeping Beauty
and Frog Prince, respectively, for integration (Voigt et al., 2008). These allow for a DNA ‘cut-
and-paste’ mechanism that has been shown to be relatively safe in animal models as well as
human tissues (Voigt et al., 2008). Non-viral vectors still lack targeting efficiency, but may be
overcome by the combination of viral delivery with non-viral gene transfer. Most vector systems
to date lack well characterized integration and generally exist as episomes in host cells, which
are eventually lost in dividing cells or silenced (Voight et al., 2008). Efforts to stabilize episomes
include the removal of bacterial sequences and the inclusion of elements such as nuclear antigen
I, from the Epstein - Barr virus, and human scaffold or matrix attachment regions (S/MARs) can
prolong persistence of the episome (Wu, Xiao, et al., 2000; Piechaczek et al., 1999).
The use of viral vectors in gene delivery has many advantages, especially in targeting and
gene transfer efficiency, but their disadvantages are rather serious. The most obvious is that of a
host immune response and the toxicity that can ensue. Of course, the types and severity of these
factors will depend on the target tissue, mode and route of delivery, vector being used and doses,
as well as the immunocompetence of the host (Waters and Lillicrap, 2009). For one, delivery of
vector to the eye has shown only an attenuated response to new immunogens while systemic
delivery through the circulatory system results in activation of the innate and adaptive system at
various sites (Waters and Lillicrap, 2009). Although non-viral transfer has shown better tolerance
by the immune system, methods using hydrodynamic protocols have been shown to cause tissue
damage which activates inflammatory responses and subsequent reaction can include the
delivered transgene (Miao et al., 2006). Viral vectors are naturally immunogenic and most have
developed methods to evade detection by the host immune system. At large doses, like those
given in gene therapy settings, these reponses can be detrimental to the host or the therapeutic
36
outcome. The adenovirus is a good example. This vector has great potential as it transduces
many tissues, including antigenic presentic cells (APCs), and although large doses up to 1 x 1013
are generally well tolerated in mice, studies in baboons highlight lethal toxicity (Di Paolo and
Shayakhmetov, 2009). Compared to adenoviral-vectors, the AAVs do not transduce APCs well
and therefore, do not tend to create a robust innate response (Waters and Lillicrap, 2009).
Although studies have shown that a cellular response can be elicited depending upon delivery
methods, transgene, and species (Vandenberghe and Wilson, 2007; Mingozzi et al., 2007; Lu and
Song, 2009). The AAVs have also been shown to generate neutralizing antibodies against the
vector capsid, which can cause issues upon re-administration using the same AAV vector (Zaiss
and Muruve, 2008; Huang and Yang, 2009). Interstingly, responses to lentiviral vectors injected
into animal tissues, such as non-human primate brain, cat eye, and mouse thymus have shown
no, or limited, immune response (Chang, 2009). However, the ability of lenti-vectors to
transduce APCs, such as dendritic cells, can result in immunological responses to the transgene
as it will most likely be displayed in either major histocompatability complexes I (MHCI) or
MHCII (Waters and Lillicrap, 2009).
The next issue would be the possibility of a recombinant virus reverting to its wild-type
disease causing form. An extensive amount of time and research has gone into protecting
ourselves from the very viruses that we have placed such high hopes in for gene delivery. Almost
all vectors studied today use minimal wild-type genome from the infectious viral parent and,
when needed can usually incorporate a helper replication system to provide replication proteins
in trans (Walther and Stein, 2000; de Silva and Bowers, 2009; Valori et al., 2008). Specifically
in the lenti-vectors, removal of as much viral sequence as possible is important because there are
several genes that can cause detrimental effects on cells including, oncogenesis (Tat), apoptosis
37
(Vpr), MHC down regulation (Nef), and differentiation (Vpu) (Azzouz and Mazarakis, 2004).
This ensures that replication of the vector cannot occur in the patient and minimizes the cellular
toxicity seen in response to parental virus (de Silva and Bowers, 2009; Azzouz and Mazarakis,
2004).
Another concern is that of vertical gene transfer, or the transfer of these therapeutic
elements to the next generation. This line of thought breaches the Weismann barrier, proposed by
August Weismann in 1885. Weismann’s theory basically states that the germ cells are
responsible for heredity and that the somatic cells, or other cells of the body, do not influence
this heredity. Apparently there is enough evidence to suggest that there is a selective
permeability to this theory. One such study using retroviral vectors delivered angiotensin II type
I receptor antisense (AT1R-AS) cDNA to spontaneously hypertensive rats (SHR) and found that
subsequent generations had cardiovascular protection against hypertension as a result of a
genomic integration and germ line transmission (Reaves et al., 1999). Another study for
hemophilia using a viral delivery system was halted after viral DNA was identified in a patient’s
semen (Boyce, 2001). However, since semen contains other material, further testing was ordered
and isolated sperm cells were eventually found to be negative for the viral DNA (Boyce, 2001).
Katherine A. High, an M.D. and Howard Hughs Medical Institute investigator suggests that:
Even if foreign DNA gains access to a germ cell, in order for harm to result, it must lodge in a site where it can have an effect. Based on reasonable assumptions about the human genome and on data from a recent AAV trial, the risk of detecting a birth defect caused by gene transfer appears to be less than one in one million—a negligible increase over the baseline risk of birth defects of about one in one hundred (high-risk of gene transfer).
Among concern for vertical transmission is also horizontal gene transfer. In this case, an
organism receives DNA from another organism, without being the offspring of that organism.
Almost all vectors are typically gutted or have minimal elements retained in the vector construct,
38
but this does not rule out the chance of viral integration, or homologous recombination events
leading to residual viral DNA elements. Interestingly enough, researchers have recently
identified the presence of bornavirus, a negative-sense RNA virus incorporated into the genomes
of multiple mammalian lineages and at different times, ranging from more than 40 million years
ago in anthropoid primates to less than 10 million years ago in squirrels (Horie et al., 2010;
Feschotte, 2010). This novel find, along with the data identified in gene transfer trials indicate
that viruses have indeed optimized themselves for future revival and could have profound
implications in viral vector mediated gene therapy.
An additional consideration is one of eco-pharmcovigilance. The excretion of active
pharmaceutical material in the environment has been going on since the introduction of these
compounds, but we have just recently begun to quantify their levels (Boxall, 2004). The best
known example is the effects of birth-control hormones in freshwater fish populations (Kidd et
al, 2007; Boxall, 2004). A workshop, held in October 2007 by the International Conference on
Harmonisation Gene Therapy Discussion Group, aimed to provide a better understanding of the
contribution of viral vector-shedding studies to the benefit-risk assessment of gene therapy
products (Kühler et al., 2009). Topics discussed included advantages and disadvantages of assays
used for detection of viral vector shed into excreta (Kühler et al., 2009). It is expected that more
sensitive assays should be developed to test for viral nucleic acids and among that, the detection
of transmission rates of shed vector to third-party persons or animals should be included in all
trials (Kühler et al., 2009).
Current strategies in gene therapy have focused on monogenic disorders or those resulting
from a single gene disorder, such as CF. These disorders have been well characterized and the
therapeutic target is generally one form of a gene. Unfortunately, most genetic diseases are the
39
result of multigene disorders, such as heart disease, high blood pressure, and diabetes. These
diseases pose multiple issues for gene therapy due to targeting and coordination of multiple gene
delivery. More complicated diseases will most likely need to be well characterized so that the
molecular factors involved can be fully corrected. Regrettably, it seems that many of these
disorders, monogenic and multifactorial alike have environmental factors involved that may not
be reversible with gene therapy. The stage of disease will also make this more difficult, as many
effects of disease, such as tissue damage, are limited in reversibility depending on the extent of
damage. However, for the more characterized diseases, prevention may be the best cure here and
may at some point involve gene therapy of a single disorder at a time.
It is obvious, from even the small amount of information presented here, that the issues in
gene therapy are quite numerous and still require more detailed knowledge before a solution may
be found. Depending on the nature of the disease in question, therapies through gene transfer will
most likely not be able to rely on one method alone. An interesting combination is that of RNAi
and Ad vectors, in which poor in vivo gene transfer of RNAi alone was overcome with the use of
an Ad delivery system (Lundstrom, 2004). Immune responses may be overcome with the
encapsulation of viral vectors in liposomes (Lundstrom, 2003), while pre-existing responses may
be overcome by the use of vectors that have not been seen by the human patient, i.e. Goat-AAV
(Arbetman et al., 2005). It may be that stable inducible, gene expression is achieved through the
combination of well characterized integrating vectors combined with drug-regulating
transcription (Auricchio et al., 2002). There seem to be endless possibilities and multiple
combinations that have yet to be tried or optimized. Unfortunately, beyond the specific concerns
associated with each method still exists the translation of observed therapies in animal models to
the human patient. This major obstacle is currently being tackled as researchers advance to larger
40
animal models and aim to further characterize the toxicity of introducing foreign genes and
proteins into our system. As one last point, beyond the biological concerns, are also the ethical
concerns of human modification and enhancement. These issues are beyond the strictly
biological aspect of this study, but are definitely another barrier in the advancement of gene
therapy and many of the techniques surrounding this modern marvel.
As the focus of this project is on the antigenic nature of viruses as a vehicle for gene
therapy, specifically the AAVs, it is necessary to look further at the events surrounding an
immune response. As these vectors are foreign to our systems, we have adapted to respond to
them for our own good. On the first insult our bodies will naturally work to clear the infection,
generally beginning with a cellular response. The cellular response will then alert the humoral
responses which will aid in continued protection through antibody production and memory
responses. This memory response will be a factor if treatment is needed again. As mentioned
earlier, many of the current therapies are transient and will require a secondary, or even, more
consistent delivery. This type of treatment will require superior engineering to evade the immune
responses and it is our hope that the data generated here will eventually add to the understanding
of the antigenic nature of viral vectors and their further exploitation and manipulation for
successful gene delivery therapies.
Antibody Recognition of Viruses and Viral Neutralization
Antibodies are critical for the control of many viruses and generally target proteins on the
viral surface or on the surface of infected cells. Neutralization of a virus by these antibodies is
defined as “the loss of infectivity which ensues when antibody molecule(s) bind to a virus
particle, and usually occurs without the involvement of any other agency” (Burton, 2002). The
protective effects of neutralizing antibodies can be achieved not only by neutralization of free
particles, but also by several activities directed against infected cells. Although there is much
41
controversy surrounding the production and mechanisms of neutralizing antibodies (Burton,
2002; Dimmock, 1984; Burton et al., 2000; Klasse and Sattentau, 2002), they remain imperative
to most vaccine-mediated protective strategies.
Efficient B-cell induction and generation of B-cell memory requires two signals. The first
is mediated by the B-cell receptor (BCR) while the second signal is usually provided by T-cells.
It has been postulated that extensive BCR cross-linking can activate B-cells to produce IgMs
without the aid of T-cells. Antigens that conatin highly repetitive epitope have been correlated to
these specific B-cell responses (Bachmann and Zinkernagel, 1996). In fact, it has been suggested
that for an antigen to be T-cell independent it must contain a threshold number of appropriately
spaced epitopes in order to stimulate the B-cell receptor (Dintzis et al., 1983). This includes all
non-enveloped viruses, especially those that have a highly organized icosahedral structure, like
AAV. Interestingly, in hepatitis B (which is an enveloped structure) studies identified the
nucleocapsid antigen to generate both T-cell dependent and independent response, while the
surface antigen is T-cell dependent (Milich and McLachlan, 1986). Any antigen may have
several epitopes, or specific sites that are recognized by a lymphocyte or an antibody. Epitopes
for antibodies on proteins can be linear or continuous, like those recognized by T-cell receptors
(TCRs) or they can be made up of different regions of a polypeptide that come together in a
three-dimensional structure. The latter is often referred to as a discontinuous or a conformational
epitope. However, not all conformational epitopes are discontinuous (Harris et al., 2006). Indeed,
linear stretches of residues have been identified in conformational epitopes where the unique
structure of the paratope, like a loop-helix motif, plays an important role in recognition (Harris et
al., 2006). These antibodies are generally raised against native folded proteins, but may
recognize fragments of the protein if the proper binding elements of the structure are retained.
42
When the heavy-chain (Hc) and light-chain (Lc) regions of the antibody’s variable region
are paired their hypervariable loops form the complementarity-detemining regions or CDRs.
These regions determine antigen specificity by forming a surface complementary to the antigen.
The formation of these shape complementary regions is an important determinant of the
antibody-antigen interaction along with solvent accessibility and contact energies (Rapberger et
al., 2007). This interaction can be disrupted by extreme salt concentrations, large pH shifts,
detergents, and even competition with high concentrations of pure epitope. The binding is
therefore a reversible, non-covalent interaction that involves electrostatic forces, hydrogen
bonds, Van der Waals forces, and hydrophobic interactions. A striking difference seen in
antibody-protein interactions is the involvement of aromatic residues (Janeway et al., 2005).
Several groups have addressed if an unusual amino acid composition exists in antigen-antibody
interactions. Studies involving the physical properties of these interfaces have revealed that high
percentages (34%) of the antibody-contacting residues are aromatic, with tyrosines and
tryptophans in the CDRs being more solvent exposed (Davies and Cohen, 1996). In contrast, the
contacting residues on the antigens contain relatively few aromatic amino acids (Davies and
Cohen, 1996). The pattern of binding that emerges is similar to that found in other protein-
protein interactions, with good shape complementarity between the interacting surfaces (Davies
and Cohen, 1996).
Viruses are eliminated by the host through many different mechanisms, including
cytokines, cytotoxic T-cells (CTLs), complement, and antibodies. It follows that viruses that are
efficiently controlled by antibodies are under a direct evolutionary pressure to form serotypes
(Bachmann and Zinkernagel, 1996). These serotypes exist so that neutralizing antibodies against
serotype A do not work against serotype B, thus an individual can be infected by two (or more)
43
serotypes despite the pre-existence of neutralizing antibodies to the first. The existence of many
different antigenic forms of viruses (e.g. Rhinoviruses) makes them harder to control as they
often have the ability to mutate rapidly to generate many different variants in a single host (e.g.
HIV-1 in humans). The ability of viruses to produce structural variation while maintaining
function and the effectiveness of the antibody to neutralize are other factors involved in their
response. The success of a vaccine will be determined by an antibody’s ability to neutralize and
the subsequent protection after a natural infection. However, in the case of viral-vectored gene
therapy the ability of engineered viruses to escape neutralization may be beneficial to patients
with pre-existing immunity or those in need of repeated treatment with the same vector.
Antibodies directed at viruses may be able to completely block infectivity after direct
binding to the virion, while others are less efficient. Mechanisms of neutralization can be
classified according to the event in virus replication that they block. Studies have noted antibody
interference at cellular attachment sites, post-attachment interactions of the virus with receptor
and co-receptor, internalization, and fusion at the cell surface or in endosomal compartments of
enveloped virions (Klasse and Sattentau, 2002). Additional mechanisms can occur in vivo
including Fc-mediated phagocytosis, complement binding and activation, opsonization, and
antibody-dependent cellular cytotoxicity of infected cells (Burton, 2002). It has also been noted
that some antibodies neutralize through mechanisms that have yet to be explored. Apparently
some antibodies to Sindbis virus and poliovirus can eliminate infection or reduce progression in
vitro even when added to cells several hours post-infection (Smith, 2003). There is also evidence
that some antibodies show non-neutralizing activity in vitro, but offer protective immunity in
vivo (Schmaljohn et al., 1982; McCullough et al., 1992). Although neutralizing antibody is one
of the main arms of protective immunity and the type that is often desired in vaccination, there is
44
surprisingly little known about these interactions. Some of the best studied examples of
neutralization include many common human pathogens, such as HIV-1, rhinovirus, influenza
virus, and poliovirus (Reading and Dimmock, 2007; Huber and Trkola, 2007; Smith, 2001). The
list has recently grown to include emerging pathogens like West Nile, Hendra, and Nipah virus
(Reading and Dimmock, 2007; Oliphant and Diamond, 2007).
Parvovirus Capsid Structures, Receptor Binding, and Infection
It appears that the immune responses seen among patients and animals alike are ultimately
dependent upon the structural nature and infectivity profile of the insulting virus, or vector. Since
this work is dedicated to characterizing the immune responses seen in our vector of choice,
AAV, it is best to have a detailed look at the infectious pathway and structural nature of this
virus in comparison with other family members and in context of the host.
Members of the family Parvoviridae infect a variety of hosts, ranging from insects to
primates, predominantly causing disease in young vertebrates. These viruses are non-enveloped
with a capsid shell of ~260 Å in diameter that exhibits a T=1 icosahedral symmetry. They are
made up of 60 copies of between two and four overlapping capsid proteins designated viral
protein (VP) 1-VP4 (Table A-1). The autonomous parvoviruses (including minute virus of mice
(MVM) and canine parvovirus (CPV)) are composed of ~50 copies of VP2, while the adeno-
associated viruses (AAVs), a member of the genus Dependovirus, contain mostly VP3. The VP1
protein makes up about 5% of the capsid and has been reported to contain phospholipase A2
(PLA2) activity in the VP1 unique region, as well as basic sequences which appear to act as
nuclear localization sequences (Girod et al., 2002; Zadori et al., 2001; Grieger et al., 2007). The
AAVs are also composed of ~5% of an alternatively spliced protein, also called VP2, which is
reported to not be required for infectivity. Parvovirus capsids are similar in that there is one
capsid protein that forms the majority of the structure, as well as minor proteins that form the
45
same core structure, but have additional sequences on their amino-termini. Despite differences in
sequence there are certain structural elements that have been identified as common among the
parvoviruses. These include raised regions at the five-fold axes of symmetry (Figure A-3),
depressed regions or canyons surrounding the five-fold axes, one or three protrusions at or
surrounding the three-fold axes of symmetry (three-fold spikes or peaks; Figure A-3), and
depressed or dimpled regions at the two-fold axes of symmetry (Chapman and Agbandje-
McKenna, 2006). All the viruses have a superimposable core eight-stranded β-barrel domain that
forms the interior of the capsid (Figure A-6A).
The parvoviruses use a variety of proteins and carbohydrates as receptors or co-receptors
for infectivity and entry. Sialic acid has been shown to be important for infectivity of MVM and
several of the AAV serotypes. However, studies have identified that an AAV serotypes’ ability
to bind sialic acid depends on the linkages involved in the carbohydrate chemistry; AAV4 binds
2,3-O-linked sialic acids, and AAV1, 5, and 6 bind both 2,3- and 2,6-N-linked sialic acids
(Kaludov et al., 2001; Seiler et al., 2006; Wu et al, 2006). Receptors and co-receptors for AAV2
include heparin sulfate proteoglycan (HSPG), human fibroblast growth factor receptor-1, αvβ5
and α5β1 integrins, and hepatocyte growth factor receptor (c-met) (Qing et al., 1999; Qiu and
Brown, 1999; Summerford et al., 1999; Kashiwakura et al., 2005). Platelet-derived growth factor
receptor (PDGFR) is the protein that mediates infection for AAV5 (Di Pasquale et al., 2003)
while the transferrin receptor has been reported as a receptor for CPV and feline panleukopenia
(FPV) (Parker et al., 2001). The sites for interaction of some of these receptors with the
parvovirus capsid have been mapped either by mutagenesis or through direct interactions at the
structural level. In AAV2 HSPG binds to the three-fold axis of symmetry, whereas for CPV the
transferrin receptor attachment site is near the top of a raised region around this three-fold axis of
46
symmetry (Opie et al., 2003; Kern et al., 2003; Hafenstein et al., 2007; O’Donnell et al., 2009;
Levy et al., 2009). Mutations in AAV2 have been used to alter binding specificity and infectivity
(Rabinowitz et al., 1999; Wu, Xiao, et al., 2000; Xu et al., 2005; Lochrie et al., 2006; Shi et al.,
2006). The receptor attachment sites are still unknown for the other AAV serotypes.
Although the attachment sites vary among different parvoviruses, it is thought that they all
enter by receptor-mediated endocytosis (Weitzman, 2006). Entry of AAV2 and CPV into the cell
occurs by endocytosis through clathrin-coated pits and is inhibited by overexpression of a
dominant interfering mutant of dynamin (Duan et al., 1999; Parker and Parrish, 2000; Bartlett et
al., 2000). After uptake into the cell the pathway is less well characterized. Acidification has
been suggested as being necessary for infectivity by using drugs that disrupt endosomal pH
(Bartlett et al., 2000; Ros et al., 2002; Basak and Turner, 1992). Escape from the endosome
seems less efficient, but studies identify a release of the VP1 unique region and its associated
PLA2 activity, allowing the virion to enter the cytoplasm (Girod et al., 2002; Suikkanen et al.,
2003; Zadori et al., 2001). The capsid is then proposed to move through the cytoplasm by
microtubule-mediated processes to the perinuclear region, the capsid then enters the nucleus
possibly via a NLS that is found in the already exposed VP1 unique region (Ros and Kempf,
2004; Vihinen-Ranta et al., 2002; Girod et al., 2002; Greiger et al., 2007). Modifications to the
capsid during transport may also result in the exposure of sequences that aid in transport through
the nuclear membrane (Cotmore, 1999). Different entry and transport patterns by the different
parvoviruses may also exist. For example, transduction of cells by AAV capsids can be enhanced
by inhibitors of proteosomal degradation proteases, while the MVM and CPV capsids are not
affected by those inhibitors (Denby et al., 2005; Yan et al., 2002, Ros et al., 2002, Yan et al.,
2004).
47
Parvovirus Antigenic Properties
There have been many studies involving antibody binding and neutralization of the
parvoviruses, most notably on MVM, CPV, FPV, AAV2, and human parvovirus B19 (B19).
These viruses differ in their host range, pathogenic nature, and capsid protein composition, but
still have structural similarities that determine the antigenic response due to infection (Agbandje-
McKenna and Chapman, 2006). The mechanisms of antibody induction, binding, and
neutralization are not well understood, but the data presented in this work adds to the knowledge
in this field and may eventually aid in their definition. It appears that the elaborate loop regions
between the strands of the viral β-barrel, which make up the protrusions at or surrounding the
icosahedral threefold axes and the walls between the depressions at the two- and five-fold axes,
are the most immunogenic sites in the parvovirus capsid (Agbandje et al., 1995; Govindasamy et
al., 2003; Bloom et al., 2001; Xie et al., 2002; Simpson et al., 2002; Strassheim et al., 1994).
Studies of CPV and MVM have used escape mutations from neutralizing mAbs to map epitopes
on the CPV, FPV and MVM capsids (Lopez-Bueno et al., 2003; Strassheim et al., 1994), while
others have used peptide scans and ELISA based assays to map immunodominant regions on the
parvovirus capsids, including AAV2 (Wobus et al., 2000; Brown et al., 1992). Structural studies
based on CPV and FPV have also used cryoelectron microscopy (cryoEM) to confirm previously
identified antigenic regions on these capsid surfaces (Hafenstein, 2009), much like this study has
done with AAV. The capsid surface of the insect infecting densoviruses appear quite smooth
compared to those of the viruses infecting vertebrates (Figure A-3) (Simpson et al., 1998;
Bruemmer et al., 2005), suggesting that the raised structures in the vertebrate viruses provide a
benefit when the virus is under antibody selection. These data also confirm the idea that only
viruses that are efficiently controlled by antibodies are under evolutionary pressure to form
serotypes (Bachmann and Zinkernagel, 1996) as described below for the parvoviruses.
48
Parvovirus Antigenic Structures for Autonomous Viruses
FPV isolates infect cats, mink, and other carnivores, while CPV infects dogs and close
relatives, such as wolves, coyotes, and Asiatic raccoon dogs. CPV emerged in 1978 and the
original strain (CPV type-2) was replaced between 1979 and 1981 by a genetically and
antigenically distinct virus (type-2a), which has recently been replaced by further variants
(Parrish, 2006). FPV and CPV differ antigenically in at least 1 epitope, and after emerging in
dogs CPV later gained 3 or 4 additional mutations in the VP2 gene which changed 2 different
epitopes (Parrish et al., 1991; Parrish et al., 1985). There are close connections between the host
ranges of the viruses and their antigenic structures and variation. The effects of antibody binding
on the capsid surface of CPV and FPV have been previously examined (Parrish and Carmichael,
1983; Strassheim et al., 1994; Wikoff et al., 1994). Efforts to map epitope regions on the capsid
have included the use of natural variants or mAb-selected escape mutant analysis (Strassheim et
al., 1994; Wikoff et al., 1994), peptide mapping, and cryoEM analysis with image reconstruction
of Fab:capsid complexes. Most of these studies have identified two major regions of antigenicity
on the virus, while a recent structural study using cryoEM has confirmed these regions
(Hafenstein, 2009). Site A is close to the top of the three-fold spike of the capsid, while site B is
located on the shoulder of the three-fold spike (Chang et al., 1992; Langeveld et al., 1993; Lopez
de Turiso et al., 1991; Strassheim et al., 1994). The capsid regions around epitopes A and B also
show the greatest structural variation between CPV and FPV, and their host range and antigenic
mutants (Agbandje et al., 1993; Govindasamy et al., 2003; Llamas-Saiz et al., 1996; Simpson et
al., 2002). These methods have determined that VP2 residue 93 in epitope A and 299 and 300 in
site B determine the canine host range of the viruses, by controlling the interaction of the capsid
with canine and feline TfRs (Govindasamy et al., 2003; Hueffer et al., 2003). Antibodies to these
regions preclude receptor attachment ultimately causing neutralization.
49
Studies of MVM during long term immune therapy with a single neutralizing mAb in
immunodeficient mice resulted in the isolation of mutant viruses escaping neutralization through
single amino acid changes within the VP2 protein, residues 433-439, which are located on a
surface-exposed loop at the three-fold axes of the capsid (Lopez-Bueno et al., 2003). Recent
studies using cryoEM have led to the image reconstruction, at 7Å resolution, of the structure of
the immunosuppressive strain of MVM (MVMi) bound to the neutralizing antibody, B7
(Kaufmann et al., 2007). The VP2 residues involved in the contact are 225, 228, 229, 425, 426,
and the loop 432-440, which correlate with the escape mutant findings. The structural results,
together with the analysis of B7 binding and neutralization activity, suggest that this antibody
may hinder conformational transitions in the viral capsid required for the MVMi infection
process (Kaufmann et al., 2007).
Antigenic studies on Aleutian mink disease virus (AMDV) have revealed some rather
interesting results. Adult mink infected with AMDV develop a persistent infection associated
with high levels of antiviral antibodies and hypergammaglobulinemia (Aasted et al., 1984;
Bloom et al., 1994). In spite of this robust immune response, virus is not eliminated in vivo and
complications usually arise. Furthermore, antiviral antibody enables AMDV to infect cells such
as macrophages via an Fc-receptor-dependent mechanism termed antibody-dependent
enhancement (ADE) (Bloom et al., 2001). It is also interesting to note that antiviral antibody
does not protect adult mink from infection, but rather, leads to an accelerated form of disease
upon challenge (Aasted et al., 1998). Sites on the AMDV capsid have been mapped through
peptide studies (Bloom et al., 1997; Bloom et al., 2001; McKenna et al., 1999) to regions similar
for AAV (see below) and CPV/FPV, located on the wall between the two-, three-, and five-fold
axes, and on the inner wall of its three-fold mounds. Immunoreactive spans of peptides were
50
located in VP2 residues 429-524, which correlates with a residue span of 428-446 that is
recognized by infected minks (Bloom et al., 1997). This region is located in the wall of the two-
fold depression, and antibodies directed at this peptide, although elicit ADE, are able to
neutralize infectivity in CrFK (feline kidney origin) cells. However, another well characterized
epitope spanning from residues 487-501 at the inner wall of the mounds of the icosahedral three-
fold axes, is capable of inducing ADE, but is not neutralizing. The observation that the same
amino acid sequence can both neutralize virus and induce ADE may explain the inability of
capsid-based vaccines to protect against AMDV infection (Bloom et al., 2001).
B19, a member of the erythrovirus genus, is a human pathogen that predominantly infects
young children with mild flu-like symptoms and a characteristic red rash that gives the
appearance of a “slapped cheek”. The disease usually runs its course with no complications, but
has been known to have adverse effects in pregnant women and those with immune disorders.
The B19 capsid epitopes have been mapped to the VP1 unique region and to the VP1/VP2
overlapping sequence, and most of the VP1/VP2 neutralizing epitopes are located on the loops
that make up the protrusions on the B19 capsid surface (Kaufmann et al., 2004; Rosenfeld et al.,
1992; Sato et al., 1991; Yoshimoto et al., 1991). Antibodies that recognize VP2 residues 57-77,
345-365, and 446-466 inhibit hemagglutination by B19 (Brown et al., 1992), while those
generated for peptides with the residues 253-515 are neutralizing. Mapping these residues onto
the crystal structure of B19 identifies residues on the wall between the two-and five-fold
depressions and at the base of the three-fold protrusions. The loop containing residues 309-319 is
disordered in the B19 structure (Kaufmann et al., 2004), but residues 314-330 structurally
superimpose onto a portion of the C37B epitope for the dependovirus AAV2 (discussed below)
(Wobus et al., 2000).
51
The Dependoviruses and their Antigenic Properties
The genus dependovirus houses the AAVs, which share a common, but not absolute
requirement for an unrelated DNA helper virus (e.g. adenovirus, herpesvirus, papillomavirus,
vaccinia virus) to aid in the completion of their otherwise replication deficient life cycle
(Muzyczka and Berns, 2001; Bowles et al., 2006). Phylogenic analysis now places the
autonomously replicating goose (GPV) and duck parvoviruses into this genus where they cluster
with the avian adeno-associated virus (AAAV) (Tattersal, 2006). Probing studies have lead to the
identification and cloning of 53 new AAV capsid DNA sequences from non-human primates
(Gao et al., 2002; Gao et al., 2003) and 64 from human tissues (Gao et al., 2004). The
identification of so many isolates has brought about the need for a classification system in order
to move the field forward. The isolates have therefore been divided by three different
classification systems based on differences in the capsid: serology, transcapsidation, and clades
based on genetic relatedness. Much of the virus variation is concentrated into the regions of the
sequence that make up the surface exposed loops of the capsids, and they therefore are mostly
antigenically distinct when tested with polyclonal antibodies, being classified as AAV1 through
AAV11. There are currently six clades; A,B,D,E, and F, which house AAV1, AAV2, AAV7,
AAV8, and AAV9 respectively, while the AAV2/3 hybrids belong to Clade C. AAV4, AAV5,
do not fit any of the clades and are regarded as clonal isolates (Gao et al., 2004). Serology is no
longer used to group the AAVs, but the isolates still fall into serologically distinct ‘groups’.
In addition to their genetic and antigenic differences, the AAV serotypes exhibit a wide
range of distinct biological properties including receptor binding and tissue tropism (Chiorini et
al., 1997, 1999; Xiao et al., 1999; Davidson et al., 2000; Rabinowitz et al., 2002, Choi et al.,
2005). Their dependency upon helper viruses has generated a promiscuity among the AAVs that
enables them to be in many different cell lines awaiting helper functions. In the absence of this
52
helper, the virus has the ability to integrate into the host chromosome to set up a latent infection.
This relatively simple strategy is not harmful to the host and has not been associated with any
known disease or pathogenicity. These qualities have enabled the development of recombinant
AAVs for use as vectors in gene therapy. Recombinant gene transfer vectors based on AAV2
have proven effective in animal models for the correction of genetic diseases of the eye, brain,
muscle, liver, and lung (Warrington and Herzog, 2006). However, numerous studies have shown
that other serotypes possess unique tissue tropisms and have improved gene transfer activities
compared to that of AAV2 for some cell types and are also being studied to avoid the pre-
existing antibodies that are widespread in the human population, or which develop after primary
AAV treatments and which reduce the efficacy of subsequent treatments (Jooss and Chirmule
2003; Zaiss and Muruve 2005, Calcedo 2009). AAV1, for example, can transduce rodent skeletal
muscle more efficiently than AAV2 (Xiao et al., 1999). AAV4 has strong tropism for ependymal
cells in the central nervous systems of mice (Davidson et al., 2000; Liu et al., 2005) and for
retinal pigmented epithelium in rats, dogs, and nonhuman primates (Weber et al., 2003). Among
the more recently discovered serotypes, AAV7 has also been shown to have superior muscle
transduction compared to AAV2, while AAV8 and AAV9 are the most efficient serotypes
discovered so far for transducing the liver (Gao et al., 2004; Gao et al., 2002). Studies on AAV2
capsids and their mutants suggests that the polyclonal antibody response is focused on a
relatively small number of sites on the capsid, and that reactivity can be reduced by mutations of
residues within those sites (Lochrie et al., 2006). However, until the developments of new mAbs
to AAV1 and AAV5, discussed in this study, there have been only a few mAbs reported that
react with intact capsids (Kuck et al., 2007; Wobus et al., 2000). There is currently little
53
information available about how the various anti-AAV IgGs recognize capsids, their mechanisms
of neutralization, or the distinction between cross-reactive and specific epitopes.
The antigenic structure of AAV2 is the best characterized among the dependoviruses, since
peptide mapping has been used to identify both linear and conformation-dependent epitopes on
the capsid that are recognized by mAb (Wobus et al., 2000). Linear epitopes (recognized by
antibodies A1, A69, and B1) and conformational epitopes (recognized by antibodies A20, C24-
B, C37B, and D3) have been reported, although the C24-B site has not been mapped. The linear
epitopes are not relevant to the infectious virus as A1 recognizes VP1 only, while A69
recognizes a peptide in the VP1/VP2 common region, and B1 recognizes all three capsid proteins
via an epitope at their C-termini (Wobus et al., 2000). Antibodies to the conformation-dependent
epitopes show multiple mechanisms of neutralization, as the C24-B and C37B antibodies inhibit
receptor attachment, while the A20 antibody apparently neutralizes at a post-attachment step, and
D3 antibodies are non-neutralizing (Wobus et al., 2000; Figure 1-1). The C37B epitope (residues
493-502, 602-610, VP1 numbering) is close to the HSPG binding site on the capsid, which is
consistent with its ability to block receptor binding. Antibodies to the C37B epitope specifically
recognize AAV2, but not AAV1, 3, 4, or 5, likely related to the fact that surface loops near the
heparin binding region are the most variable in the AAV capsid structures (Govindasamy et al.,
2006; Padron et al., 2005). The D3 epitope is fairly conserved in the sequences of most AAV
serotypes (Padron et al., 2005). D3 antibodies recognize serotypes AAV1, 3 and 5, but not 4, and
residues in that epitope are close to structural differences between AAV2 and AAV4 within the
mounds surrounding the icosahedral three-fold axes (Govindasamy et al., 2006; Padron et al.,
2005). The A20 antibody reacts only with AAV2 and AAV3, and this epitope contains three
different regions in the primary sequence (residues 272-281, 369-378, and 566-575) (Wobus et
54
al., 2000) that come together in the capsid structure (Xie et al., 2002). Comparisons of the
AAV4, AAV5, AAV8, and AAV2 coat protein structures identified the A20 binding regions as
some of the most variable between the three viruses (Govindasamy et al., 2006; Padron et al.,
2005; Walters et al., 2004; Nam et al., 2007). The AAV4 capsid protein is not recognized by the
B1 antibody, which recognizes a linear epitope (residues 726-733) of denatured capsids near the
C-terminus of the VP3 protein (Wobus et al., 2000). A comparison of the AAV2 and AAV4 VP3
crystal structures showed that there is no major conformational difference in the B1 epitope
region, suggesting that amino acid sequence differences may play a role in the capsid-antibody
recognition (Govindasamy et al., 2006).
A panel of mutations of surface residues of AAV2 to either conservative substitutions (to
Ala) or to radical changes were prepared and examined for their ability to bind HSPG and to
bind or be neutralized by the A20 mAb, by polyclonal antibodies in 3 human sera, or by a pool
of human IgGs (Huttner et al., 2003; Lochrie et al., 2006). Mutations that reduce binding and
neutralization by polyclonal human and monoclonal murine antibodies were identified as well as
mutations that reduce, have no effect on, or increase transduction. There was not always a direct
relationship between the change in IgG binding and neutralization, indicating that not all IgGs
neutralize equally efficiently as has been reported for other virus families such as rhinovirus
(Smith et al., 2001).
AAV Capsid Structure and Serotypes
The AAV2 capsid structure has been determined by cryoEM and by X-ray crystallography
to ~3 Å (Kronenberg et al., 2001; Xie et al., 2002). Even more recently the structure of AAV2 at
65˚C was solved by cryoEM (Kronenberg et al., 2005). This structure gave some insight into the
elusive location of the VP1 unique region in that it is known to be required for infectivity, but
has not been identified in previous reconstructions. This structural region was described as
55
“globules” located under the icosahedral two-fold axes of symmetry. Other serotype structures
that have been solved by cryoEM and X-ray crystallography include AAV1, AAV4, and AAV5 -
9 by the Agbandje-McKenna lab (Figure 1-2A) (Govindasamy et al., 2006; Nam et al., 2007;
Walters et al., 2004; and Agbandje-McKenna, unpublished data. Sequence analysis has identified
AAV4 and AAV5 as being the most diverse serotypes at the amino acid level and also
antigenically (Gao et al., 2004).
The AAV serotype structures can be placed into two categories based on their surface
topology. AAV2 (Kronenberg et al., 2001; Xie et al., 2002) and AAV1, and AAV6-8 (Nam et
al., 2007 and unpublished data) exhibit pointed “finger-like” protrusions around the threefold
axes, while AAV4, AAV5, and AAV9 (Govindasamy et al., 2006; Padron et al., 2005; Walters et
al., 2004; and unpublished data) have “rounder” mounds surrounding the icosahedral three-fold
axis of symmetry (Figure 1-2B). Sequence alignments and 3D-modeling of AAV3, 6 and 10
shows that they will be similar to the AAV1, 2, and 8 class, while AAV11 will be more like
AAV4, 5, and 9. Comparison of all the available AAV structures or models revealed a conserved
β-barrel motif, while the more variable regions were mapped to the exposed surface loops
(Govindasamy et al., 2006; Figure 1-2C). This has also held true for other parvovirus capsids
(reviewed in Kontou et al., 2005). The surface loop regions become clustered as VP monomers
come together to form the capsids icosahedral symmetry. This assembly creates variations on the
capsid surface at the five-fold axes, the “shoulder” regions and the tips of the three-fold
protrusions, which surround the two-fold axes. These variable regions localize with previously
identified surface areas that affect receptor recognition, capsid recognition by conformational
antibodies, specifically A20 (Wobus et al., 2000), and transduction efficiency for AAV2
(Lochrie et al., 2006). This study has mapped antigenic sites on five serotypes, including the
56
already well characterized AAV2, the structurally similar AAV1 and AAV6, and the more
distinctly related AAV5, as well as hepatrotrophic AAV8, which provides a strong insight on the
recognition of antibodies by the AAV capsids and their resultant binding abilities.
Significance
The generation of neutralizing antibodies during viral infections is often crucial to recovery
and is usually the target response in vaccination strategies. Despite the important role that
immunoglobulins have in animal virology there are still many aspects that are not completely
understood. There are many other viruses, including such important pathogens as HIV,
rhinovirus, hepatitis C, and influenza that induce only partially protective antibodies that do not
aid in clearance of the virus. The rules that govern neutralization mechanisms are even less well
known, as it has not yet been determined how the step in the viral lifecycle to be blocked is
chosen. The evolution of viruses is partially driven by the host’s mechanisms of defense in that
they must choose mutations that still allow for receptor attachment, replication, and immune
evasion in order to survive. Through the data presented here it will be possible to define
dominant antigenic regions that are directly involved in binding and neutralization by mAbs
against the capsid of a simple and well characterized virus, the AAVs.
The parvoviruses are important pathogens in the veterinary sciences, as well as among
humans including the erythrovirus B19, which had been recognized for many years as one of the
classical rash illnesses of childhood. There are many unknown autonomous parvoviruses
identified by PCR (Candotti et al., 2004; Terossi et al., 2007), as well as different AAVs (Chen et
al., 2005; Gao et al., 2004). Because the AAVs are associated with no known pathology, have a
wide range of infectivity, and can establish long-term transgene expression, they have great
potential as a viral vector in gene therapy (Muzyczka, 1994; Choi et al., 2005; Muzyczka and
Warrington, 2005). The AAVs under examination here have distinct tissue tropisms likely due to
57
differences in cellular receptor utilization which is thought to arise from structural differences
between AAV serotypes at the capsid level (Wu, Asokan, et al., 2006). Similar to wild-type
viruses, recombinant vectors are detected by the immune system and generate a response that
becomes effective usually before the virus infects its target cell. This immunity not only acts
against the vector and the infected cells, but also results in a memory response that affects further
efforts to use the same vector or even transgene again (Bessis et al., 2004). Efforts to enhance the
safety and delivery of vectors are ongoing as well as the determination of strategies to sidestep
the antiviral response. This study supports the idea that it would be useful to have well
characterized AAV vectors with predictable tropisms and altered antigenic sites that can be used
in the treatment of individuals with a pre-existing immune response as well as for subsequent
therapies.
Figure 1-1. The antigenic structure of AAV2. Antigenic epitopes of AAV2 (Wobus et al., 2000) displayed on the 3D capsid surface of PDB 1LP3 (Xie et al., 2002). Amino acids that form the A20 (green spheres), C37B (magenta), D3 (orange), and B1 (blue) epitopes are displayed in context of the entire capsid surface depicted in grey cartoon. The icsahedral symmetrical axes are indicated by red arrows and the 5-fold is also highlighted in dark grey for orientation purposes.
58
Figure 1-2. Panel of available AAV atomic structures A) VP3 is shown for each serotype that has been solved to date by X-ray crystallography. B) The two major surface topologies of the existing serotype structures can be grouped into two classes. AAV5 represents those serotypes with “rounder” protrusion. The other class depicts “finger-like” protrusions and is represented by AAV2. C) Superimposed VP3 monomers reveal a common motif with variable regions at the exposed surface loops (I-IX). The symmetrical axes are represented by black shapes (same in A and C); the pentamer represents the five-fold axis, while the triangle and the oval represent the three-fold and the two-fold, repectively. An asymmetric unit is depicted as a white-outlined triangle in B.
59
CHAPTER 2 MATERIALS AND METHODS
Production and Purification of Recombinant AAV VLPs for Structural Studies
The AAV capsid proteins VP1, VP2, and VP3, for serotypes 1, 4, and 5-9, were expressed
in sf9 insect cells under the baculovirus polyhedron promoter as previously reported (Miller et
al., 2006). These proteins self-assemble into virus-like particles (VLPs). After 72 hours
incubation at 27̊ C the cells were spun down for 15 minutes at 1 500 rev min-1 (4ºC), and the
pellet was resuspended in lysis buffer (50mM.Tris pH8.0, 100mM NaCl, 0.2% Triton X-100).
The pellets were then subjected to three freeze/thaws with the addition of benzonase (Novagen)
after the second cycle and then centrifuged for 15-30 minutes at 10 000 rev min-1 (4ºC). The
supernatant was then diluted with TNET buffer (50mM Tris pH 8.0, 100mM NaCl, 1mM EDTA,
0.2% Triton X-100) and pelleted through a 20% sucrose cushion for 3 hours at 45 000 rev min-1
(4ºC). This pellet was resuspended overnight at 4ºC and further purified by ultracentrifugation on
a step gradient (5-40% sucrose) at 35 000 rev min-1 (4 ºC). A visible blue fraction containing
empty (no DNA) viral capsids, sedimenting at ~20-25% sucrose, is expected and will be
extracted and dialyzed.
AAV2 capsids were purified using iodixanol gradients and a heparin agarose affinity
column. Discontinuous iodixanol step gradients were formed in quick-seal tubes (25×89 mm,
Beckman) by underlaying and displacing the less dense cell lysate (15 ml) with iodixanol (5,5′-
[(2-hydroxy-1,3 propanediyl)bis(acetylamino)] bis[N,N′-bis(2,3dihydroxypropyl-2,4,6-triiodo-
1,3-benzenecarboxamide] prepared using a 60% (w/v) sterile solution of OptiPrep (Nycomed)
and PBS-MK buffer (1× PBS containing 1 mM MgCl2 and 2.5 mM KCl). Therefore, each
gradient consists of (from the bottom) 5 ml 60%, 5 ml 40%, 6 ml 25%, and 9 ml of 15%
iodixanol; the 15% density step also contains 1 M NaCl. Tubes were sealed and centrifuged in a
60
Type 70 Ti rotor at 69 000 rev min-1 for 1 h at 18°C. Approximately 5 ml of the 40–25% step
interface was aspirated after side-puncturing each tube with a syringe equipped with an 18-gauge
needle (modified from Zolotukhin et al., 2002). A heparin-agarose (H6508, Sigma) column with
a 2.5ml bed volume was then prepared and equilibrated with 10 ml of PBS-MK buffer (1× PBS
containing 1 mM MgCl2 and 2.5 mM KCl) then 10 ml of PBS-MK/1 M NaCl, followed by 20 ml
of PBS-MK buffer. The AAV2 capsid containing iodixanol fraction(s) were then diluted with
twice the original volume and applied to the column. The column was washed with 3-10 column
volumes of PBS-MK buffer and sample was eluted with PBS-MK/1M NaCl. The high sodium
chloride content was reduced by buffer exchange or dialysis and the eluent was checked for
protein by optical density. Samples were then concentrated on centrifugal filter units (Millipore)
and checked for the VP1, VP2, and VP3 proteins by SDS-PAGE analysis (Figure 2-1). Intact
capsids were verified by negative stain electron microscopy (EM) (Figure 2-1).
Fv Antibody Modeling Using the Web Antibody Modelling (WAM) Server
Sequences (obtained for the Fab variable domain of AA4E4.G7, AA5H7.D11, and
BB3C5.F4 by the Parrish group at Cornell) were were submitted to the WAM server
(http://antibody.bath.ac.uk/index.html). The modeling is done piecewise for the heavy and light
chains and the canonical loops (Figure 2-2), which is based on the WAM algorithm
(http://antibody.bath.ac.uk/algorithm.html). Briefly, homology modeling is used to build the
framework of the antibody structure, i.e., heavy and light chains. The canonical loops are also
built through homology modeling with 5 rounds of minimization to smooth out joint regions.
The more diverse, non-canonical regions are built based on a different algorithm, the Combined
Antibody Modelling Algorithm (CAMAL) (Martin et al., 1989).
61
Preliminary Neutralization Assays
Neutralization assays were conducted at NIH (Bethesda, MD) in collaboration with Dr.
John Chiorini. Briefly, AAV1 and AAV5 capsids that were packaging the green fluorescing
protein (GFP) gene were pre-incubated with serial dilutions of hybridoma supernatant for 30
minutes at room temperature prior to addition to cos7 cells and incubated for 1 hour at 37°C.
Fresh media was then added to the cells and the cultures incubated for 48 hours at 37°C. The
percentage of infected cells was evaluated by measuring the GFP-expressing cells using flow
cytometry.
Preliminary In-House Green Cell Assays
Purified monoclonal antibodies of AA4E4.G7, AA9A8.B12, and BB3C5.F4 were used in
green cell assays to check their neutralization properties (for future studies aimed at deciphering
theirmechanisms of neutralization). Briefly, cos7 cells grown to ~100% confluency in complete
Dulbecco’s Modified Eagle Media (with 10% FBS and 1% P/S; DMEM) were plated at a
dilution of 1:4 in 96-well dishes. Assuming 60 binding sites per capsid, purified virus and
purified mAb were complexed at different ratios. As an example, one capsid to 240mAbs is a
ratio of 1:4 (VP:mAb). Ratios of 1:1 and 1:0.5 VP:mAb were also examined. (To note, a mAb
was considered to have one binding site although the presence of bivalent binding was not ruled
out.) These capsid:mAb mixtures were incubated at 22°C for ~30 minutes. Media was removed
from the cells and the mixture was applied to the wells. The mixture was left on cells for 1 hour
and incubated at 37°C with 5% CO2. The mixtures were then removed and complete DMEM was
added. At ~48 hours post-infection green cells were visualized by UV excitation on an Olympus
scope. These experiments were done in triplicate for each VP:mAb ratio.
62
Dot Blot Analysis to check Cross-Reactivity among VLPs
To examine the reactivity of the mAbs against different AAV serotypes, dot blots of
previously purified VLPs were performed. PVDF membrane (Millipore) was activated in 100%
methanol for 30 seconds and then soaked in PBS for ~15 minutes. Approximately 100-300ngs of
each AAV serotype was applied to the prepared membrane through a vacuum-manifold dot blot
system. PVDF membrane containing samples were then blocked in 10% non-fat milk (Bio-Rad)
in PBS for ~1 hour at 22°C. Membranes were then incubated with primary antibody for 1 hour
and then secondary antibody conjugated to horse radish peroxidase (HRP) for another hour. The
bound antibody-HRP was detected with Supersignal luminescent substrate and exposed to X-ray
film.
Generation of Fragment Antibodies from Monoclonal Antibodies
Intact IgGs were purified using 1 ml Hi TrapTM Protein G HP columns (GE Healthcare).
Briefly, hybridoma sera were run through a 0.45μm syringe filter while protein G columns and
were equilibrated with 20mM sodium phosphate pH7.0. The clarified supernatant was then
diluted and applied to the Protein G column at 1ml/min as specified in the manufactures protocol
(GE Healthcare). Whole IgGs were eluted from the column with Glycine pH 2.5 into prepared
collection tubes containing 200μls 1M Tris pH9.0, to ensure a final pH of 7.5-8.0. The IgGs were
then concentrated and buffer exchanged into 20mM sodium phoshate pH7.0, 10mM EDTA for
papin cleavage. Samples were then incubated with activated immobilized papain (Pierce) at a
suggested enzyme: substrate ratio of 1:160 w/w at 37°C for ~12 hours. The Fc fragment was then
captured on a Hi Trap™ Protein A column (GE Healthcare) and the fragment antibody-binding
portions (Fabs) were collected in the flow-through, which was then subjected to gel filtration on
a Superdex 75 column (GE Healthcare) in PBS. Samples were checked through-out the
63
purification and production process to ensure purity and further characterize products (Figure 2-
3).
Preparation of Fab:VLP Complexes
Purified capsids were mixed together with Fabs at a ratio of approximately 2 Fabs per
potential binding site on the capsid, giving a final complexed ratio of ~1:120 capsid:Fab.
Complexes were left at 4°C for 1 hour and checked by negative stain electron microscopy for the
presence of decorated particles before cryoEM data collection.
CryoEM Data Collection
Sample aliquots of 3.5 µl were vitrified with a manual plunge freezing device either on
grids with Quantifoil films or on grids with continuous carbon films. The samples were
examined at -193°C on a FEI Tecnai G2 Polara electron microscope at an accelerating voltage of
200 keV and at a nominal magnification of 59,000 X (in collaboration with Timothy Baker,
UCSD). Images were recorded with a Gatan Ultrascan 4000 CCD camera at a step size of 1.883
Å / pixel. The CCD images were saved in the Gatan DM3 format and later converted to the tiff
format. All images were recorded with an objective lens underfocus value of 1.25 to 3.0 µm. The
total electron dose on each image was ~24 to 28 e- / Å2.
Three-Dimensional Reconstruction of AAV:Fab Complexes
The software package RobEM (http://cryoEM.ucsd.edu/programs.shtm) was used to
extract individual particle images from collected micrographs or film. Preprocessing of the
selected images occurred as previously described, as well as estimation of the defocus levels of
each micrograph was as previously described (Baker et al., 1999). A random-model computation
procedure (Yan, Dryden, et al., 2007) was used to generate a starting model at ~30Å resolution
from 150 particle images. This map was then used to initiate full orientation and origin
determinations and refinement of the entire set of images using the current version of
64
AUTO3DEM (Yan, Sinkovits, et al, 2007). Corrections to compensate for the effects of phase
reversals in the contrast-transfer functions of the images were performed as previously described
(Bowman et al., 2002; Zhang et al., 2003), but amplitude corrections were not applied. Final 3D
maps, reconstructed from the selected particle images were estimated to be reliable to ~10.8-
22.8Å resolution (Table 4-2), using the 0.5 threshold of the Fourier Shell Correlation (FSC0.5)
criterion (data not shown). The handedness of the maps was determined by comparison of the
surface features with those of AAV structures determined by X-ray crystallography. Graphical
representations of the constructed maps were generated with the RobEM and Chimera
(http://www.cgl.ucsf.edu/CHIMERA; Peterson et al., 2004) visualization software packages.
Fitting of Atomic and Pseudo-Atomic Models into CryoEM Density Maps
The atomic coordinates for the capsid structures of AAV2 (PDB accession number 1LP3;
Xie et al., 2002) and AAV8 (Nam et al., 2007), AAV1 and 5 (unpublished data), and a model
built for AAV6 based upon AAV1, were used for capsid docking and scaling of the cryoEM
reconstruction density maps. When necessary, the capsid coordinates were rigid-body docked
into the cryo-reconstructed AAV-complexed maps using the COLORES subroutine in the SITUS
program, version 2.3 (Chacon and Wriggers, 2002). A visual inspection of the docked model in
the program COOT (Emsley and Cowtan, 2004) identified either several small surface loop
regions extending above the density envelope or into the capsid interior density depending on the
specific map, possibly indicating an incorrect initial map scaling factor. The absolute scales for
the maps were thus determined by creating electron density maps of the cryo-reconstructed
AAV:Fab structures at various Å/pixel values and comparing these maps to an electron density
map generated for structure factors calculated for the docked homology model at cell-dimensions
and grid sizes matching each of the scaled maps. The various cryo-reconstructed and model
maps were then normalized in the program MAPMAN (Kleywegt, 1996) and compared using the
65
program’s similarity function (http://xray.bmc.uu.se/usf/mapman_man.html). The highest
correlation coefficient was observed between the maps generated at 1.80 and 1.883Å/pixel (the
original scale), depending on the respective structure (Table 4-3) and was considered the
absolute map scale. Difference maps between the cryoEM reconstructed maps and the atomic
data (or AAV6 3D model) maps (as used above in scaling) were then generated in MAPMAN
using the operation function (OP). The excess density that was displayed in the difference maps
were identified as the bound Fab. Atomic coordinates for an unrelated (generic) Fab (PDB 2FBJ;
Suh et al., 1986) structure was manually docked into the Fab density, as web-antibody models
(WAM) models of the respective Fabs were only generated for the Fv arm of the antibody. Once
the unrelated Fab was docked into the respective density, indicated by the least amount of amino
acids seen jutting from the density, a 60-mer of the capsid and Fab model was generated in order
to mimic the fully bound structure of the AAV:Fab complex. Electron density maps generated
from structure factors calculated for these fully bound models (at cell-dimensions and grid sizes
matching each of the scaled maps) were then compared to their respective complex density map
using the similarity function in MAPMAN. Models with the highest correlation (Table 4-3), or
similarity index were then chosen as the properly docked representative Fab model. WAM
models generated from the sequenced Fabs were then superimposed onto the docked unrelated
Fab (2FBJ) in COOT (SSM Superpose; Krissinel and Henrick, 2004) in order to visualize the
variable loops on the AAV capsid surface that appear to interact with the canonical and non-
canonical loop regions on the Fab models. Visualization and analysis of these models in context
of their respective maps were completed in Chimera (Petteresen et al., 2004).
Mutagenesis of Proposed Antigenic Regions
Since a majority of the structures solved in the cryoEM studies were of AAV1,
mutagenesis studies were focused on this serotype. Specific regions of the capsid surface that
66
appeared to be important in capsid recognition were identified by the previous docking
procedures (Table 4-4) and primers were designed to generate antigenic escape mutants in these
regions by muatatons of the AAV1 cap gene. The recombinant AAV1 plasmid (pAAV1cap,
~8kb) was a gift from Dr. Jay Chiorini (NIH, Bethesda, MD) and contained the AAV1 cap gene
and the AAV2 rep gene. Forward (Fwd) and reverse (Rvs) primers were designed to mutate
multiple regions in the AAV1 cap gene to structurally equivalent regions in AAV2 (Table 4-4).
The mutations from AAV1 to AAV2 were based on the observation that AAV2 was not
recognized by anti-AAV1 antibodies (See Figure 3-1A and C). The method involved mutating 1-
2 amino acids at a time so several primers were designed for subsequent rounds of mutagenesis.
For example, mutant 1AB (Mut1AB) was created by mutating AQNK (residues (aa) 456-459) of
the AAV1 cap gene to TTQS (of AAV2) using, Fwd1a: 5’
GTCCGGAAGTACCCAACAGAAGGACTTGCTGTTTAGCCGTGG 3’, Rev1a: 5’
CCACGGCTAAACAGCAAGTCCTTCTGTTGGGTACTTCCGGAC 3’ and Fwd1b: 5’
GTCCGGAAGTACCACACAGAGCGACTTGCTGTTTAGCCGTGG 3’ , Rev1b: 5’
CCACGGCTAAACATGAAGTCAGTCTGTTGGGTACTTCCGGAC 3’, where 1a mutated
AQNK TQQK and 1b mutated TQQKTTQS. This method was repeated for all other
mutants. Mut2AB was mutated from TK (aa492-493) to SA using, Fwd2a: 5’
GCAGCGCGTTTCTAAAACAAGCACAGACAACAACAACAGC 3’, Rev2a: 5’
GCTGTTGTTGTTGTCTGTGCTTGTTTTAGAAACGCGCTGC 3’ and Fwd2b: 5’
GCAGCGCGTTTCTAAAACAAGCGCAGACAACAACAACAGC 3’, and Rev2b: 5’
GCTGTTGTTGTTGTCTGCGCTTGTTTTAGAAACGCGCTGC 3’. Finally, Mut3AB was
mutated from SSST (aa586-591) to RGNR using, Fwd3a: 5’
GCAGTCAATTTCCAGAGAGCAACACAGACCCTGCGACCGG 3’, Rev3b:5’
67
CCGGTCGCAGGGTCTGTGTTGCTTCTCTGGAAATTGACTGC 3’ and Fwd3b: 5’
GCAGTCAATTTCCAGAGAGGCAACCGAGACCCTGCGACCGG 3’ and Rev3b:5’
CCGGTCGCAGGGTCTCGGTTGCCTCTCTGGAAATTGACTGC 3’. These mutants were
then used in combination with the designed primers to generate AAV1 capsid with multiple
AAV2 region suvstitutions. For example, Mut1AB served as the template to generate
Mut1AB2AB, which served as the template to create a mutant with all three regions mutated,
Mut1AB2AB3AB. Mutations were made with the QuickChange LightningTM kit (Stratagene)
following recommended parameters, and primer and reagent volumes. Final products were run
out on a 0.8% agarose gel with ethidium bromide to check for purity and sent to the UF ICBR
core for Sanger sequencing.
Production and Purification of Antigenic Mutants
Plasmids with positive mutations, as well as plasmids containing Ad helper functions
(pXX6, ~18kb; Xiao et al., 1998) and green fluorescing protein (GFP) gene (UF11, ~7.5kb;
Klein et al., 1998), were amplified by inoculating 1L cultures of TB with glycerol stocks.
Cultures were shaken at ~225 rev min-1 with incubation at 37°C for ~24 hours and were
harvested by low-speed centrifugation (15 minutes, 5000 rev min-1, 4°C). 1L Cell pellets were
resuspended in 20 mls buffer ( 25mM Tris pH8.0, 10mM EDTA, 15% sucrose (w/v), 0.1mg/ml
RNAse) and incubated with 50mgs lysozyme for 5-10 minutes on ice. Cells were lysed with the
addition of 48 mls lysis buffer (1% SDS, 0.2 N NaOH) and gentle shaking. The solution was
neutralized with the addition of 36 mls 3M sodium acetate (C2H3NaO2) pH5.2, and gently
inverted several times. Chloroform (0.2 mls) was added and the solution was mixed by gentle
inversion. The lysed cells were then spun at 9000 rev min-1 for 20 minutes and the supernatant
was poured into a new centrifuge bottle through a layer of gauze to remove any debris. The
plasmid DNA was then precipitated by the addition of an equal volume of isopropanol and left
68
overnight at 4°C. The DNA was then pelleted at 9000 rev min-1 for 10 minutes at 4°C and
resuspended in 20 mls double distilled filtered water (ddH2O). After allowing the DNA to
dissolve, 20 mls 5.5 M lithium chloride (LiCl) was added to precipitate any proteinaceous
material. The solution was incubated for 30 minutes on ice and then spun at 9000 rev min-1 for
10 minutes at 4°C. To avoid disrupting the protein pellet, the supernatant was carefully decanted
into two 50 ml conical tubes and equal volumes of isopropanol were added to re-precipitate the
plasmid DNA. This was spun at 9000 rev min-1 for 10 minutes at 4°C and the supernatant was
discarded and the pellet was air-dried for several minutes. The pellet was allowed to gently
dissolve in 7 mls ddH2O and cesium chloride (CsCl) was added at 1 g/ml. The CsCl was allowed
to completely dissolve on ice and then 125uls of 10mg/ml ethidium bromide (EtBr) was added.
This solution was then split into two 4.9ml Optiseal tubes (Beckman) and sealed. The EtBr
gradient was transferred to an NVT90 rotor and allowd to run at 78000 rev min-1 for 4 hours at
15°C, or 55000 rev min-1 overnight (15°C). Bands were collected with an 18 gauge needle and
transferred to a 15 ml centrifuge tube. EtBr was removed through several iso-amyl alcohol
extractions and the DNA was re-precipitated by the addition of 2.5 volumes ddH2O and two-
times the combined volume of 95% ethanol. The DNA was then pelleted at 9000 rev min-1 for 15
minutes at 4°C. The supernatant was discarded and the DNA pellet was allowed to air-dry, and
then allowed to resuspended in 500 uls ddH2O. The sample was subjected to two rounds of
phenol/chloroform/iso-amyl extractions and one chloroform/iso-amyl extraction before a final
round of re-precipitation. Finally, 0.1 volumes of 3 M C2H3NaO2 pH5.2, followed by 2.5
volumes 95% ethanol were added to precipitate the DNA and the mixture was spun at 14000 rev
min-1 at room temp. The pellet was then washed with 1 ml 75% ethanol and allowed to resuspend
in 1 ml ddH2O. Purified DNA concentrations were obtained through optical density
69
measurements at OD260 and DNA isoforms were verified on agarose gels. The plasmid DNA was
also checked for purity and identified by restriction digests.
Human endothelial kidney cells (HEK 293) were grown to ~70% confluency in 15cm
plates with DMEM supplemented with 1% antibiotics and 10% serum (complete DMEM) at
37°C and 5% carbon dioxide (CO2). Cell media was aspirated and replaced with 12 mls of fresh
complete DMEM and were then triple transfected using polyethyleneimine (pEI) inoculums.
Transfection inoculums were made using 45ugs/plate pXX6, 15ugs/plate UF11, 15ugs/plate
mutant plasmid, 160uls/plate pEI, and 1000uls/plate of DMEM (no FBS). Transfected cells were
incubated for 24 hours and assayed for quality of transfection on a florescence scope by green
cell visualization. Virus infected cells were then harvested at ~60 hours post-infection by
scraping and low-speed centrifugation (1200 rev min-1, 15 minutes, 4°C). Cell pellets were
resuspended in 1 ml TD (1x PBS, 5mM MgCl2, 2.5mM KCl) per plate supplemented with
protease inhibitor cocktail (Roche) and cells were lysed by three rapid freeze/thaws with the
addition of benzonase (Novagen) at the second thaw, and subjected to a low-speed spin to pellet
cell debris (10000 rev min-1, 15 minutes, 4°C). Clarified cell lysates were then used in dot blot,
infectivity, and packaging assays.
For further purification, clarified cell lysates were loaded onto an iodixanol gradient and
fractions were collected (as previously explained). Ion-exchange chromatography was used to
further purify capsids and remove iodixanol as previously described in Zolotukhin et al., 2002.
Briefly, a 1 ml HiTrap Q-column (GE Healthcare) was equilibrated with 5 mls buffer A (20mM
Tris pH8.5, 15mM NaCl) followed by 5 mls buffer B (20mM Tris pH8.05, 500mM NaCl) and
another round of 5 mls buffer A on a Gilson peristaltic pump at 1 ml/minute. The iodixanol
fractions were then diluted 1:1 with buffer A and loaded onto the equilibrated column at 1
70
ml/minute and washed with 50 mls of buffer A. Bound virus was eluted with a gradient reaching
100% buffer B over a 5 minute time span at 1 ml/minute on a Pharmacia ATKA FPLC system.
One ml fractions were collected and assayed for protein concentrations by OD280 optical density
readings.
Western Blot Analysis
Fractions were assayed for VPs by western bolt analysis. Fractions were denatured by 10%
SDS-PAGE and transferred to PVDF membranes (Millipore) overnight at 27V. The membrane
was then blocked in 10% non-fat milk (BioRad) for 1 hour and probed with B1 (Wistuba et al.,
1997), a mouse monoclonal that recognizes denatured AAV capsids (except AAV serotype 4).
The blot was then treated with secondary antibody conjugated to horse radish peroxidase (HRP)
for another hour. The bound antibody-HRP was detected with Supersignal luminescent substrate
and exposed to X-ray film.
AAV1 Enzyme-Linked Immunosorbent Assay (ELISA)
Mutant AAV1 capsids were tittered using an AAV1 titration ELISA kit (PROGEN).
Briefly, standards were prepared by diluting the provided kit standard from 1:2 to 1:64 and cell
lysates were initially diluted from 1:250 to 1:100000. Prepared dilutions and a blank were added
to a microtiter strip and incubated at 37°C for 1 hour. The strips were washed in 1x wash buffer
provided by the kit and anti-AAV1 biotin conjugate was added to the wells for 1 hour at 37°C.
After washing, a streptavidin peroxidase conjugate was then added and the strips were incubated
again at 37°C for 1 hour. Strips were washed and then reacted with tetramethylbenzidine (TMB)
or substrate for 10 minutes. A blue color indicated bound capsid and the color reaction was
stopped with Stop Buffer, provided by the kit. The plate was then read on an ELx800 plate
reader (BioTek) at a wavelength of 450nm.
71
Dot Blot Analysis for Antigenic Reactivity
Recombinant AAV1 mutant capsids (rAAV1muts) were generated, expressed, and
purified as described above and further characterized for antibody recognition and cross-
reactivity against mAbs originally generated against the rAAV1 parental virus. Native dot blots
using cell lysates of the rAAV1muts were performed as described above.
Pixel Intensity Calculations from Native Dot Blots
Native dot blots were scanned at about 600 pixels per inch (or higher) and analyzed using
the freeware, imageJ (http://rsbweb.nih.gov/ij). Briefly, the stacked RGB image was divided into
the different gamma channels; red, green, and blue, and stacks containing the most information
were combined. Background and noise was removed from the images using a rolling ball
function at a threshold just below the actual data points so as not to smooth the actual data.
Finally, data points were chosen and analyzed using the “plot lanes” function. Pixels intensities
were given as integrated optical densities. Negative cell lysates values were subtracted from the
raw data and the final values were plotted in Excel against the respective mutant. Percent (%)
calculations were performed by setting the rAAV1 pixel intensities for each mAB to 100% and
subsequent data was calculated from this normalized value.
Infectivity Assays
Infectivity of the rAAV1muts was checked by green cell assays in HEK293 cells as
previously described (Wu et al., 2000). Briefly, HEK293 cells were plated at a 1:4 dilution in 96-
well dishes and were allowed to attach for ~24 hours. In a separate 96-well dish 20uls of cell
lysates were added to the first column and serial dilutions were made from 10-1 to 10-8 in DMEM
(no FBS) spiked with Ad 5 virus at a MOI of 10. To the cells, 100 uls of the dilutions were added
to corresponding rows and plates were incubated for 1 hour at 37°C with 5% CO2. The dilutions
were removed and wells were refilled with 100 uls complete DMEM. Cells were allowed to
72
grow for 48-60 hours and infectivity was checked by counting green cells (GFP production)
under a fluorescent microscope. The total number of green cells counted was multiplied by the
original dilution factor and the plate dilution to get a final infectious particle count. Negative
controls included wells with Ad 5 virus only and no virus or cell lysate.
DNA Extraction and Real-Time Polymerase Chain Reaction
Packaging efficiency was checked by determining the amount of vector genomes (vgs) in
the cell lysate virus. Briefly, 2 uls of benzonase and benzonase buffer were added to 10 uls of
cell lystate and incubated at 37°C for ~2 hours. To the sample, 2 uls proteinase K and 22 uls
proteinase K buffer was added for a final volume of 224 uls. This reaction was allowed to
incubate for 1 hour at 37°C. Samples were then subjected to two rounds of
phenol/chloroform/iso-amyl extractions and one round of plain chloroform extraction. After the
final extraction, 22.4 uls of 3 M C2H3NaO2 pH5.2, 2 uls glycogen (Novagen), and 670 uls of
100% ethanol were added to precipitate the DNA. Samples were left overnight at -20°C and then
spun at 14000 rev min-1 for 20 minutes. The pellet was then washed in 75% ethanol and spun
again at 14000 rev min-1 for 5 minutes. The DNA pellet was allowed to air-dry and was then
resuspended in 20 uls ddH2O. Concentrations were checked by OD260 readings and were then
diluted to be between 10 and 1 ng/ul for real-time polymerase chain reaction (qPCR) analysis.
A master mix containing iQ SYBR Green Supermix (Bio-Rad) and forward and reverse
primers for each sample, was made and then delivered to 96-well PCR plates (BioRad). To this
mix, extracted DNA samples or standards (ranging from 10 ngs to 0.0001 ngs) were added to
respective wells for a final reaction volume of 25 uls and qPCR capture was performed. PCR
parameters were set as suggested in the iQ SYBR Green Supermix handbook and analysis of the
data was performed on the My iQ 2.0 software (Bio-Rad).
73
Figure 2-1. Coomassiee stained SDS-PAGE and EM images of purified AAV VLPs for structural studies. Images for AAV serotypes illustrate stained bands for VP1, 2, and 3 at ~92, 78, and 67kDa, repectively and a corresponding EM image for each VLP depicting white, negatively stained intact VLPs against a black carbon background.
Figure 2-2. Images of the Fab WAM models. Complimentary determining regions are colored differently an the heavy and light chain portions are indicated in gray. Heavy and light chain hyper-variable regions are indicated as H1-3 and L1-3, respectively. Models were generated for A) AA5H7.D11 B) AA4E4.G7 C) BB3C5.F4
Figure 2-3. Production, purification, and characterization of mAb and Fabs. A) Superdex-75 column data showing the milli-absorbance units (mAU) at OD280 on the y-axis and peak fractions (vertical red dotted lines) and column volume on the x-axis. The Fab peak at fraction 11-12 corresponds to a molecular weight calculation of ~41.7kDa. B) An SDS-PAGE analysis of the Superdex-75 fraction 11 and the whole IgG before papain digestion indicating the loss of the Fc portion. Both images indicate by size and purity that Fabs have been obtained for the example IgG.
74
CHAPTER 3 CHARACTERIZATION OF ANTIBODIES AGAINST AAV1 AND AAV5 CAPSIDS,
CROSS-REACTIVITY AND NEUTRALIZATION CAPABILITIES
Introduction
Antibodies are unique structures that have been found in various tissues of the body and
are responsible for recognition and neutralization of foreign objects. There are five main classes,
some of which have several subtypes. These classes represent the effector function and heavy-
chain structure of each isotype, which are: Immunoglobulin G (IgG), IgM, IgD, IgE, and IgA.
The IgG and IgM molecules are the first of the secreted Igs to be present in early infection, while
the IgD molecule functions at the B-cell surface as a receptor. The IgE molecules are generally
associated with allergies and parasites, whereas the IgA has roles in preventing colonization at
the mucosal surface. Of these isotypes, IgG is the most abundant isotype in plasma and
represents the typical antibody structure. This structure is composed of four polypeptide chains;
two heavy chain molecules (Hc) and two light chain molecules (Lc) (Figure 3-1A). These chains
are broken down further into constant (C) and variable (V) regions. The variable regions VH and
VL make up the Fab and contain the complimentary determining region (CDR) (Figure 3-1A).
This region is important for interaction with antigens and contains the paratope specific for
antigen recognition. Both the heavy and light chains have three CDRs, generally referred to as
hypervariable regions 1-3 (Figure 2-2). Structures among the different isotypes are generally
similar for IgG, IgD, and IgE, where differences occur mainly in the hinge regions and
glycosylation. The IgM molecule exists as a pentameric structure (Figure 3-1C) with five IgE-
like molecules attached by a small peptide referred to as the J-chain. The IgA is also held
together by the J-chain, but exists in a dimeric form (Figure 3-1B). These structures can be
abundantly produced by activated B-cells, which has led to the creation of hybridomas. In this
methodology, hybridomas were made by fusing B-cells with myeloma cells (a B-cell cancer cell
75
line) (Alkan, 2004). Hybridomas specific for a certain antigen could then be separated and
propagated to generate mAbs. This technology won Cesar Milstein, Georges J. F. Köhler and
Niels Kaj Jerne the Nobel Peace Prize in Physiology or Medicine in 1984 (Alkan, 2004). Since a
virus can be neutralized in several different ways that are determined primarily by the specificity
of the reacting antibody, analysis of specific antibody-virus interactions is only possible using
these mAbs (Reading and Dimmock, 2007). Thus mAbs are used in various clinical and research
approaches, including diagnosis of disease and production of vaccines (Nabel, 2004).
The antibody response is a key immunity that develops against pathogens and foreign
objects alike, making viral vectors no different than their wild-type parents. Many people have
already been infected with some parvoviruses (Schneider et al., 2008; Lindner and Modrow,
2008; Candotti et al., 2004) and AAV serotypes (Gao et al., 2004; Calcedo et al., 2009) and have
developed antibodies which can interfere with the application of AAV vectors (Manno et al.,
2006). In addition, it is likely that therapeutic application of AAV vectors will induce immune
responses, which may interfere with re-administration. The primate and human AAV serotypes
are mostly antigenically distinct when tested with polyclonal antibodies, although much of the
sequences are similar, especially in structurally conserved domains (Gao et al., 2004; Agbandje-
Mckenna and Chapman, 2006). However, despite the important role of antibodies in the
recognition of AAV, the mechanisms of binding and neutralization or other effects are not well
understood. Information regarding antigenicity has been obtained for AAV2, as well as the
autonomous viruses; CPV, human B19 parvovirus (B19), and MVM (Wobus et al., 2000;
Hafenstein et al., 2009; Tolfvenstam et al., 2000). These studies suggest that although most
antibodies appear to bind on or near the protrusions at or surrounding the icosahedral three-fold
axes, there is a complex interaction between the capsid surface and the host antibody response,
76
and that the sites involved in antibody binding may be at least partially constrained by the capsid
surface. Studies of AAV2 examining a small number of mAbs, or polyclonal anti-capsid
antibodies, show that the antibody response is affected by changes in only a small number of
sites on the capsids, and that reactivity can be reduced by mutations of residues within those sites
(Lochrie et al., 2006; Huttner et al., 2003; Wu et al., 2000b). For AAV2 some epitopes have been
described by mutational analysis and peptide mapping. These show neutralization of infectivity
at either attachment or post-attachment steps (Wobus et al., 2000).
Here, a number of mAbs (both IgGs and IgMs) directed against the capsids of AAV1 and
AAV5 have been generated. By comparing their reactivities with the parental capsids and other
AAV serotypes we seek to define the general mechanisms of antibody attachment, the processes
of neutralization, and to compare the structures that allow these different serotypes to be
successful in nature. This study also aims to characterize the antigenic nature, mainly the B-cell
response, of the AAV serotypes and provide optimized tools for serotype recognition.
Results
Production of Anti-AAV Monoclonal Antibodies and Isotyping
A panel of mAbs against AAV1 and AAV5 capsids (Table 3-1) were prepared in the
Parrish laboratory at Cornell University. Two different immunization strategies were utilized.
The first involved immunization 2-3 months prior to the fusion of spleen cells, giving 4 IgG- and
4 IgM-secreting hybridomas against AAV1. Seeking to derive cross-reactive antibodies between
AAV1 and AAV5 capsids, mice were immunized with AAV1 capsids, then given an intravenous
immunization with AAV5 capsids 4 days prior to the fusion. This protocol resulted in the
isolation of several hybridomas producing 9 IgMs and 1 IgG against AAV5. Three antibodies
were chosen from each panel to continue on with the structural studies: 4 IgGs and 2 IgMs. The
anti-AAV1 monoclonals were AA4E4.G7, AA5H7.D11, both IgG2a subtypes, and the IgG1
77
subtype, AA9A8.B12 (Table 3-1). For the anti-AAV5 final panel; the two IgMs were
BB9F7.F12 and BB8F1.E8, with the last being BB3C5.F4, an IgG3 (Table 3-1).
Cross-Reactivity among Different AAV Serotypes
Cross-reactivity among the available serotypes was tested by dot blot analysis using the
anti-AAV panel antibodies previously chosen. At low dilutions of ~1:4 cross- reactivity was
observed with the anti-AAV5 mAbs against AAV1 (Figure 3-2A) however, at higher dilutions of
hybridoma sera, this phenomenon was completely abolished (Figure 3-2B). The slight
recognition of AAV1 by A20 (Figure 3-2A) has been previously reported at high concentrations
of the A20 monoclonal (Kuck et al., 2007). All antibodies showed specific reactions with their
homologous antigen, but no cross-reaction with the other capsids among AAV1, 2, 4, and 5-9
other than those previously mentioned (Figure 3-2C). All of the antibodies tested recognize
conformational epitopes on assembled capsids rather than linear epitopes in denatured capsid
proteins (Figure 1-1), as is seen for antibodies against other parvovirus capsids (Wobus et al.,
2000; Kuck et al., 2007, Hafenstein et al., 2009; Wistuba et al., 1997; Parrish and Carmichael,
1983).
Neutralization Ability of the Monoclonal Antibody Panel
Each of the antibodies was tested for its ability to neutralize virus infectivity by
determining the concentration of immunoglobulin required to neutralize AAV1 and AAV5
vectors carrying the GFP-gene. Most of the antibodies tested in this study neutralized AAV,
where some of the antibodies were non-neutralizing. In the anti-AAV5 study, both IgMs were
able to neutralize approximately 50% of cells at hybridoma supernatant dilutions of 1:200 and
1:400, BB9F7.F12 and BB8F1.E8, respectively (Figure 3-3). The only IgG in the anti-AAV5
panel, BB3C5.F4, was non-neutralizing, with the undiluted stock leading to only a 20%
reduction in gene expression (Figure 3-3). This observation may have been due to an aggregation
78
of viral particles due to excess of antibody and interstingly, the ability of non-neutralizing
antibodies to reduce gene transfer has been noted (Li and Ertl, 2009). This occurrence may be
due to the aggregation of viral particles and further characterization of these events may
eventually allow for useful generalizations in the future. Similar results were observed for the
anti-AAV1 panel. Both IgG2a antibodies, AA5H7.D11 and AA4E4.G7, neutralized cells at
approximately 50% at dilutions of ~1:1800 and ~1:1200, respectively (Figure 3-3), while the
IgG1, AA9A8.B12, was not able to neutralize AAV1-GFP vector (Figure 3-4). Studies were also
completed using available AAV1-like vectors, X1, X25, and AAV6 (Figure 3-5). Neutralization
of 50% was achieved for AAV6 with AA4E4.G7 at ~1:2000 (Figure 3-5A) and ~1:3200 for
AA5H7.D11 (Figure 3-4C), while AA9A8.B12 was not neutralizing (Figure 3-5B). The X25
vector was neutralized at ~50% at dilutions of ~1:3200 for both AA4E4.G7 and AA5H7.D11
(Figures 3-5A and 3-5C). AA9A8.B12 was also non-neutralizing for X25 (Figure 3-5B).
Interestingly, the X1 capsid was not neutralized with AA5H7.D11, whereas this antibody
neutralized the X25, AAV1, and AAV6 vectors (Figures 3-4 and 3-5C). It showed similar
profiles as seen before with AA4E4.G7, neutralizing 50% of vector at a dilution of 1:1200, while
AA9A8.B12 was again non-neutralizing (Figures 3-5A and 3-5B).
Discussion
The antigenic nature of viral capsids recognized by antibodies is crucial to the
development of protective immune responses. The AAVs provide an important model system to
examine the antigenic structure of theses simple capsids as there are several different genetically
distinct viruses, which have appeared to be antigenically distinct when analyzed in limited
serological studies. Antibodies are an important protective immunity against viral infection,
including the AAVs. Immunoglobulins can develop shortly after initial infection, and can
continue to provide protection through the duration of infection and by generation of a memory
79
response (Dörner and Radbruch, 2007). These responses are also effective against infection by
antigenically related viruses. The use of monoclonals in viral neutralization assays is beneficial
due to the existence of several epitopes on the capsid surface, which can lead to various
mechanisms of neutralization through different antibodies (Reading and Dimmock, 2007). Here
we developed a variety of new IgG and IgM mAbs against the capsids of two AAV serotypes to
examine the antigenic structures of the viruses, the cross-reactivity with the representative AAV
clade members, and also the effects of IgG and IgM antibodies. The AAV capsid was a potent
antigen; generating a panel of mAbs against AAV5 four days post immunization (Dr. Parrish,
Cornell University). This is consistent with the fact that multivalent viral capsids and particles
appear to be potent B-cell antigens and to give some immediate T-independent responses
(Reading and Dimmock, 2007; Janeway et al., 2005). For AAV5, although the mice had been
previously immunized with AAV1, most of the antibodies produced were IgMs directed against
AAV5, with only 1 IgG. The IgM response is one of the first responses to be seen in early
infection due to circulating low-affinity molecules (Janeway et al., 2005). It is possible that T-
helper cell epitopes shared between the AAV1 used to pre-immunize and the AAV5 capsids
given in the boost enhanced the responses to the second virus.
Characterization of antibodies produced after vector delivery has become increasingly
more important. In the data presented here Balb/c mice were used to generate a diverse response
of antibodies to AAV1 and 5 capsids. Most animal studies have found patterns in antibody
generation depending upon the animal strain and route of administration. Both IgG2a and IgG3
responses were seen in intramuscularly (i.m.) injected Balb/c mice, while C57B1/6 mice
exhibited IgG2a (Chirmule et al., 1999). The subcutaneous (s.q.) injections used by the Parrish
group produced IgG2a/b and IgG1 responses against AAV1, but AAV5 immunizations produced
80
an IgG3 with the majority of the response being IgMs (Table 3-1). Intravenous (i.v.)
administration in C57B1/6 mice also produced IgG2a and IgG2b responses (Zhang et al., 2005).
This issue becomes important due to the different responses seen against both vector and
transgene (Lu and Song, 2009; Lu et al., 2006; Vandenberghe et al., 2006; Lin and Ertl, 2009;
Hernandez et al., 1999). Few studies have fully characterized the antibody response in humans,
although it has been repeatedly reported that the human population has titerable antibodies to the
better characterized AAV serotypes 1-9 (Erles et al., 1999, Chirmule et al., 1999; Moskalenko et
al., 2000; Calcedo et al., 2009; Xiao et al., 1999; Mingozzi and High, 2007). One study sought to
examine IgG subsets in serum samples of normal human subjects exposed to wild-type AAV,
injected i.m. with AAV vectors, or subjects injected i.v. with AAV vectors (Murphy et al., 2009).
Although the ultimate findings were that a complex, non-uniform pattern of responses existed, it
was interesting to note that IgG1 and IgG3 responses appeared more prominent in those exposed
to AAV vector, while IgG2 and IgG4 appeared more robust in normal donors (Murphy et al.,
2009). Thus, the inclusion of different isotypes into antigenic studies, such as this study, will
further characterize the antigenic nature of the AAV vector and may help to isolate immunogenic
differences among vector delivery routes, concentrations, and target tissues. Another such study
identified that IgG1 responses were the highest among the naturally infected population and
suggests that this may be indicative of a cured subject or a chronic carrier of this virus (Sylvie et
al., 2010). Most importantly, it appears that the antibody response is subjective to many factors,
including immunocompetence of the patient and their pre-existing immunity. Further
characterization of these responses may one day enable us to predict the immune outcome based
on the patient, the disease, and the vector type and administration.
81
The new monoclonal antibodies described here react with the AAV1 and AAV5 capsids,
providing a variety of useful reagents for the study of AAVs in particular. Cross-reactivity was
checked by dot blot to determine if conserved regions on the AAV capsid serotypes were
recognized among the different antibodies. Of note, the anti-AAV1 antibodies recognized the
AAV6 capsid structure. This serotype was not used in the immunizations, but is 99% identical to
AAV1 (Xiao et al., 1999; Schmidt et al, 2006). This cross-recognition implies that the
differences between AAV1 and 6 are probably not involved in the antibody epitope, but different
affinities may be seen due to these differences. However, it is interesting to find that in a recent
seroprevalence study AAV6 IgG levels in human subjects was lower (46%) than those found for
AAV1, suggesting that the differences at the capsid surface level play an important role in
immunogenicity among humans (Sylvie et al., 2010). Although no cross-reactivity was seen
among other serotypes not used in the mouse immunizations, a weak cross-reactivity of the anti-
AAV5 antibodies with the AAV1 capsid was observed. This is likely a result of pre-
immunization with AAV1 before the AAV5 boost. However, the recognition was very weak, as
serial dilutions in hybridoma supernatant drastically lowered the recognition seen (Figure 3-2B).
Similarily, in an unrelated study looking at pre-existing immune response models, mice
immunized with Ad-AAV8cap vectors and then subsequently challenged with AAV2-transgene
vectors produced low-levels of non-neutralizing antibodies that bound AAV2 capsids (Li and
Ertl, 2009). Also of interest, these low-levels of non-neutralizing antibodies against AAV8,
found to be of IgG2a isotype, were able to disrupt gene transfer in AAV2 studies (Li and Ertl,
2009).
Preliminary neutralization studies on the anti-AAV1 and AAV5 mAbs (conducted by Dr.
J. Chiorini, NIH) identified neutralizing ability for most of the mAbs generated. Among the anti-
82
AAV5 mAbs, both of the IgMs selected for further study were neutralizing. Here the expression
of GFP was reduced by ~50% at hybridoma supernatant (sup) dilutions of ~1/300-1/150 for
BB9F7.F12 and ~1/400 for BB8F1.E8 (Figure 3-3). These data were only repeated once at this
point in the study and the jump seen for BB9F7.F12 between dilutions of 1/300 and 1/150 may
have to do with pipeting errors or the avidity of the IgM itself. Repetitions will be needed to
confirm the values for this anti-AAV5 IgM. The only IgG selected from the anti-AAV5 mAbs
was non-neutralizing. Neutralization assays did not identify a diltion of BB3C5.F4 sup that was
able to knock down GFP expression by 50%. In fact, at a 1/25 dilution, green cells were still
sustained at about 80% (Figure 3-3). However, neutralization assays for AAV5:BB3C5.F4 at a
sup dilution of 1/25 show almost no change in GFP expression patterns (Figure 3-3). Since
particles that escape from this aggregate become infectious once again, the on and off rates, or
kinetics of antibody binding may allow for enough viral particles to escape and cause infectious
events. Of note, these infections were done in the absence of helper virus and since few cells are
infected to begin with, a significant change may not been seen in the materials used here. A
similar trend was seen in anti-AAV1 neutralization assays. Two of the three mAbs chosen for
further study were also neutralizing while one was non-neutralizing. The two IgG2a mAbs,
AA4E4.G7 and AA5H7.D11, both neutralized ~50% of AAV1-GFP vectors at sup dilutions of
~1/1200 and >1/1600, respectively (Figure 3-4). Both of these mAbs are strongly neutralizing as
almost complete knockdown of GFP transduction was seen at dilutions of ~1/100. In comparison
to the non-neutralizing anti-AAV5 mAb, BB3C5.F4, gene transfer of the AAV1 capsids was
slightly more susceptible to the non-neutralizing anti-AAV1 mAb, AA9A8.B12, as ~50%
knockdown was seen at sup dilutions of ~1/50 (Figures 3-3 and 3-4). Again, in-house green cell
83
assays mimic this observation as green cells are reduced by almost half in capsid:mAb ratios of
1:4 (data not shown).
Neutralization assays were also done for the AAV1 mAbs against AAV1-like viruses, X1,
X25, and AAV6. These vectors differ from AAV1 by only a few amino acids at the capsid
surface. AAV6 differs from AAV1 at several sites on the VP1 protein, but only three of these are
exposed at the capsid surface and reside in the common C-terminus, VP3. These residues are
E418D, E531K, and E584L (Table 3-2). The X25 capsid differs from AAV1 at T162S, V198L,
which are buried in the capsid on VP2, and Q386K, which is surface exposed (Table 3-2). The
X1 capsid differs from AAV1 at N327S, D495G, R514H, and D590H (Table 3-2). All 4 of these
residues can be mapped to the capsid surface and fall in or around the three-fold axis (Figure 3-
6). These viruses gave results similar to wtAAV1 in that AA9A8.B12 was still non-neutralizing
while AA4E4.G7 was still able to neutralize (Figure 3-5). However, a change was noted in the
ability of AA5H7.D11 to neutralize against the X1 virus. A reduction of ~50% in green cells was
now seen at a sup dilution of ~1/200 (Figure 3-5C). The decline in sensitivity to this previously
neutralizing virus is most likely due to the differences seen at the aa level. Interestingly, single
mutations on the AAV1 capsid surface to mimic the X1 surface (made in the Chiorini lab, NIH),
resulted in no significant changes in neutralizing activity against the AA4E4.G7 and
AA5H7.D11 monoclonals (data not shown). This supports that idea that several amino acids are
involved in the epitope and that it may be conformational in nature. This data corresponds to the
mapping of the AAV2 mAb epitopes (Figure 1-1), supporting the idea that these symmetrical
axes are important in the antigenicity of the AAV capsids and also implies that some of these
residues are involved in the epitope itself, or at least plays a role in stabilization.
84
The mAbs presented here will aid in the further characterization of antigenic regions on the
AAV capsid surface in hopes of engineering AAV capsids for efficient immune response
evasion, whether it be pre-existing or an initial response. Conversely, this data will also aid in the
manipulation of these sites to generate a desired response, as in vaccine development. The next
step in these studies is to determine the epitopes for these mAbs using a structural approach as
detailed in the following chapter. These data may then define the consequence of manipulation in
these capsid regions and better define the antigenic nature of the capsid surface while adding to
the general knowledge of antigen-antibody interactions.
85
Table 3-1. Mouse monoclonal antibodies selected for further studies Monoclonal antibody (mAb) Parental AAV antigen Isotype AA4E4.G7 AAV1 IgG2a AA5H7.D11 AAV1 IgG2a AA9A8.B12 AAV1 IgG1 BB3C5.F4 AAV5 IgG3 BB9F7.F12 AAV5 IgM BB8F1.E8 AAV5 IgM Table 3-2. Amino acid differences between AAV1-like viruses and AAV1* 162 198 327 386 418 495 514 531 584 AAV6 T V N Q D D R K L X1 T V S Q E G H E F X25 S L N K E D R E F AAV1 T V N Q E D R E F *Red residues signify different amino acids
Figure 3-1. Basic immunoglobulin structure. A) General structure of a monomeric antibody representing the IgG, IgE, and IgD classes. The heavy chains (Hc) are colored in dark shades which are connected to the light chains (Lc, lighter shade) by disulfide bridges. The Hc and Lc portion of the Fab structure is further divided into a variable (V) and constant (C) region, VH and VL, and CH and CL. The variable region contains the CDRs housing the hypervariable domains responsible for the antibody paratope. B) The IgA class is represented by a dimer which is held together by a J-chain (small structure colored gray). C) The pentameric structure depicts an IgM molecule. These structures are also held together by a J-chain (gray). Different colored structures represent a separate functional Ig unit.
86
Figure 3-2. Monoclonal antibody characterization and cross-reactivity. A) Native blot using a mAb hybridoma supernatant dilution of 1:4. Recognition of native AAV1 by the anti-AAV5 monoclonals is evident at this dilution. B) Serial dilutions of the anti-AAV5 mAb hybridoma supernatant tested against native AAV1 and 5 capsids. A lack of recognition is already seen at a dilution of 1:25. C) Native dot blot analysis of AAV1, 2, and 4-9 against the anti-AAV1 and -AAV5 monoclonal antibodies. No cross-reactivity is seen, except for anti-AAV1 recognition of AAV6 capsids. The B1 antibody recognizes denatured capsids at a linear epitope near the two-fold axis and is used here to emphasize lack of native virion binding. The ADK4 monoclonal only recognizes native AAV4 capsids, as the B1 epitope does not exist in this serotype.
87
Figure 3-3. Preliminary neutralization assay data for the anti-AAV5 monoclonals. The non-
neutralizing effects of the IgG BB3C5.F4 (blue) can be seen from this assay, as GFP expression has shown only ~20% knockdown at dilutions of even 1:25. Both IgMs, BB8F1.E8 (magenta) and BB9F7.F12 (yellow) appear to be completely abolishing the ability of the vector to deliver GFP cDNA at dilutions of ~1:100 and ~1:50, respectively.
Figure 3-4. Preliminary neutralization assay data for the anti-AAV1 monoclonals. The non-
neutralizing monoclonal, AA9A8.B12 (A8.B12; green) among the AAV1 antibodies appears to have an effect on the vectors ability to transduce cells at dilutions below ~1:100. While the other monoclonal, AA4E4.G7 (E4.G7; red) and AA5H7.D11 (H7.D11; blue), appear to be able to completely abrogate transduction at dilutions of ~1:100 and ~1:200, respectively.
88
Figure 3-5. The ability of the anti-AAV1 monoclonals to neutralize the AAV1-like viruses: X1, X25, and AAV6. A) AA4E4.G7 data. The G7 monoclonal antibody appears to retain neutralizing ability, although compared to the AAV1 data, a slight increase in transduction capability is seen for all vectors, especially among the X1 virus. B) AA9A8.B12 data. B12 monoclonal appears to retain its overall non-neutralizing effect even against the AAV1-like viruses. C) AA5H7.D11 data. Neutralization data for D11 shows a significant reduction in neutralizing activity against the X1 virus, while X25 and AAV6 are still neutralized at levels similar to AAV1. AAV6 is shown in blue, X1 is red, and X25 is yellow.
89
Figure 3-6. Differnces between the AAV1 and X1 sequences are mapped to the atomic coordinates of AAV1 and cluster at the three-fold axis of symmetry. A 3D surface representation of the AAV1 capsid is shown in radial coloring from blue to red, where red represents the furthest distance from the capsid center. A three-fold axis is circled to illustrate the region of interest shown in more detail on the right. Here, three symmetry realted monomers, shown in green, red, and blue, make up the interactions of the there-fold spike. Residue differences netween AAV1 and X1 are highlighted in corresponding colors to depict the highly interactive nature of this region. It is also evident from this image how easily a conformational epitope could be generated from the capsid surface.
90
CHAPTER 4 DETERMINATION OF DOMINANT ANTIGENIC SITES ON THE CAPSID SURFACE OF
ADENO-ASOCIATED VIRUS SEROTYPES 1, 2, 5, 6, AND 8 THROUGH CRYO-ELECTRON MICROSCOPY STUDIES
Introduction
The use of structure approaches in biology has shed light on the mechanisms of many
interactions, especially macromolecular assemblies and their function in general (Johnson and
Chiu, 2000). It has also been equally important in virology, as it has given insight into the
structural basis of assembly, nucleic acid packaging, particle dynamics, and interactions with
cellular molecules (Johnson and Chiu, 2000). Structural characterizations of viruses are crucial to
the design of therapeutics and vaccines against them and has allowed for the elucidation of
mechanistic pathways. There are many well developed methods for structural determination
approaches available, and the one undertaken for a particular study is often dependent on the
resolution and type of information desired. For instance, both nuclear magnetic resonance
(NMR) and crystallography can give atomic resolution detail on bond interaction and backbone
placement, but NMR also provides dynamic (ensemble) information while crystallography
provides a “snapshot” and is often considered static. In cryoEM, macromolecules are frozen in
their native state allowing for discrete selection of dynamic states to be visualized, albielt at
lower resolution. The largest issue separating cryoEM from crystallography, besides size and the
limitations of crystal formation, is resolution. CryoEM has generally been considered a low-
resolution technique, giving reconstructions around 15-30Å, but with advances in sample
handling, instrumentation, image processing, and model building, near-atomic resolution can
now be achieved (Zhou, 2008). In reality, hybrid approaches, combining NMR, X-ray
crystallography and cryoEM, is often adopted, which provides a powerful means of filling gaps
that can arise in structural characterization of macromolecues. For example, in studies where
91
large complexes or macromolecular assemblages cannot be crystallized, subcomponents can be
crystallized to obtain high resolution information which can then be used to interprete the
complex structure at lower resolution obtained by cryoEM (Rossmann, 2000). An excellent
example of this approach is highlighted in a recent review on the influenza virus where several
crystallographic studies involving smaller portions of the viral polymerase were combined with a
large cryoEM complex to enable functional interpretation (Ruigrok et al., 2010). Insights into the
function of the replication machinery, like those of the viral polymerase, were deemed from
these studies. This information has become increasingly important due to rising concerns of
influenza pandemics, much like that reported for the 2009 H1N1 strain, and forms a basis for
virtual screening of drug compounds for clinical trials (Ruigrok et al., 2010).
A number of reported cryoEM virus complex structures offer insights into receptor or
antibody interactions, as well as conformational changes associated with replication and virion
assembly (Baker, 1999). CryoEM of immune complexes provides a method for visualizing
antigen–antibody interactions for viral capsids (Smith, 2003). These structures are often
interpreted by docking the known structures of the capsid and FAb or homologous 3D models
(from the RCSB protein data bank, (http://www.rcsb.org/pdb/home/home.do) into a cryoEM
density map of the Fab-decorated capsid (Harris et al., 2006). The resulting pseudo-atomic model
may then be used to define the FAb binding sites in an attempt to identify the peptide(s) that
make up the epitope (Harris et al., 2006). These studies better define the often dynamic surface
of viruses and also further characterize antibody-antigen interactions while offering insight into
their mechanisms of neutralization and potential therapeutic targets.
Numerous virus structures have been determined to date, big or small, pathogenic or non-
pathogenic. While these viruses have unique host characters and function, many similarities can
92
be found among them. Most of the structural studies on virus-Ab complexes began with the
influenza virus, which laid the foundations for thoughts on how viruses avoid immune
surveillance (Smith, 2003). However, the first study to visualize an intact virion complexed with
a neutralizing antibody was done with rotavirus in 1990 (Prasad et al., 1990). This study had
remarkable impact on the field at two levels. First, it established that protein-ligand interactions
could be visualized by cryoEM and second, it correlated structural and functional information for
the rotavirus capsid proteins (Smith, 2003). Later on, studies with Cowpea Mosaic Virus
(CPMV) would show that pseudo-atomic models could be generated from atomic models fitted
in the density maps, which were then used to identify important regions on the capsid surface
(Wang et al., 1992; Porta et al., 1994). Studies with the Human Rhinovirus 14 (HRV14) took it
one step further and demonstrated that pseudo-atomic models could be generated using all of the
structures in the complex and subsequently used to determine binding regions (Smith, 2003).
There are several structures of HRV14 bound to an antibody molecule (Smith, 2003), which all
seem to be localized in one region. Studies with HRV2 brought more insight into antibody
flexibility, with the observation that Fabs binding to the same site can have distinctly different
orientations and mode of binding, such as bivalency, as well as the identification of a second
antigenic site on the rhinovirus capsid (Smith, 2003). Peptide mapping and analysis of escape
mutants were approaches originally used to identify dominant neutralizing sites to CPV and FPV
capsids (Smith, 2003). CryoEM structures of virus-Ab complexes confirmed these sites (for
CPV) and further demonstrated the recognition of a common antigenic epiope by several
different mAbs generated against AAV the capsid as was observed for the HRVs (Hafenstein et
al., 2009; Wikoff et al., 1994). Many of these studies highlight the fact that the location of these
antigenic loops is a critical factor in determining how to engineer vectors and vaccines and the
93
notion that dominant antigenic regions exist on the capsid surfaces. All of these data further
support the techniques used here and the work presented in this chapter.
Here cryoEM and image reconstruction of capsid-Fab complexes was used to define the
specific structural contacts between 6 different mAbs (5 neutralizaing and one not) and the
capsids of AAV1, AAV2, AAV5, AAV6, and AAV8, to provide a comprehensive view of the
antigenic structure of the AAVs. These viruses are representative members of the AAV antigenic
clade and clonal isolate groups (Gao etal., 2004) and are ~55 to 99% identical in VP amino acid
sequence, with AAV1 and AAV6 being the most similar and belonging to the same clade group
A. Docking of atomic structures and models into the reconstructed density identified the AAV
surface loops which interact with the respective Fab.
Results
Analysis of the Fab-VLP CryoEM Complexes
In order to identify the binding sites of antibodies recognizing different AAV serotypes,
six different mAbs were chosen for analysis in complex with their respective capsid VLP (Table
4-1). These studies report seven different complexes with resolutions between ~10.8 -22.8Å
(Table 4-2). Two complexes involved the same Fab AA9A8.D11 in complex with AAV1 and
AAV6 serotypes, while the other three mAbs do not cross-react with other serotypes. One
complex, AAV5:F4 appeared to be non-neutralizing in preliminary studies, while the other mAbs
neutralize the virus that they recognize (Figures 3-2 and 3-3). For each complex the final map
was scaled by comparison with known atomic virus capsid coordinates, which were then docked
into the cryoEM reconstruction with fairly high confidence (ranging from a cc of 0.6-0.75).
Excess Fab density most likely accounted for correlations less than 100% (Table 4-3). Models of
the Fab Fv portions were derived from the sequencings of the heavy and light variable domains
of the antibodies which were submitted to the WAM antibody modeler server (Figure 2-2; see
94
chapter 2). These Fab models were then manually fit into the Fab density that extended from the
capsid surface. Once the atomic capsid coordinates and the Fab homology models were fit into
the cryoEM maps the exposed capsid surface loops that extended into the open Fab density could
be analyzed for possible contact areas. The docked Fab models served as guide points to
determine which loops were most likely involved in protein-protein contacts.
Fab G7 and AAV1
A final map of ~13Å resolution was derived from 2,870 particle images. The most notable
extensions of density occurred at the capsid three-fold axis of symmetry, and extended across the
two-fold region to overlap with the density extension from a symmetry related two-fold region
(Figure 4-1A). This extension of density was recognized as the binding site of the Fab on the
viral capsid surface. The Fab density (~1.0σ) at this region is not as strong as the capsid density
(2.30σ), indicating that steric hindrance may prevent two Fabs binding at the same time. Capsid
surface loops that play a role in the footprint are clustered in VRIV, V, and VIII (Figure 4-1D
and 4-2). These are loops from symmetry-related capsid VPs that must be forming a
conformational epitope. Residues that contact the fitted Fab in the model of a VP1 monomer
include residues 456-459 (AAV1 numbering), giving the short peptide AQNK in VRIV, and
residues 585-590 highlighting the QSSSTD peptide in VRVIII; while residues in a three-fold
related monomer include residues 492-498 - TKTDNNN in VRV (Table 4-4).
Fab D11 with AAV1 and AAV6 Capsids
Reconstructions were generated from 262 particles for AAV1:D11 and 2,527 particles for
AAV6:D11, at resolutions of 22.8Å and 15.3Å, respectively (Table 4-2). Docking of the atomic
coordinates and Fab models identified the binding region as the large protrusion from the center
of the three-fold symmetry axis of both complexes (Figure 4-1B and 4-1C). Density closer to the
three-fold protrusions was stronger (2.20σ) than that seen for the region above the capsid (0.60σ)
95
correlating to the constant portions of the D11 Fab. Indeed, fitting of a model in both density
portions, i.e. capsid and Fab density, revealed that the constant region of the Fab arm was angled
over the three-fold canyon, just above the center of the two remaining three-fold peaks. This
suggests that only one Fab at a time was bound, possibly due to steric hindrance. Capsid surface
loops in the Fab contact region were identified as VRIV, V, and VIII (Figure 4-1B and 4-1C,
Figure 4-2), which create the three-fold protrusion. AAV1 and AAV6 differ by only six amino
acids, with three of them, E418D, E531K, and F584L located on the capsid surface at the three-
fold axis (Table 3-2). As the Fab bound both viruses in the same position, those residues are
most likely not directly involved directly in the Fab binding, but they may play a role in stability
of the protein interactions. Residues mapped to one VP3 monomer include 456-459 (AAV1
numbering) with the sequence AQNK in VRIV; and 585-590 of QSSSTD in VRVIII (Table 4-4).
A three-fold related interaction places VRV between the VRIV and VIII fingers of the reference
monomer; notably, D495, N496, N497, and N498. The same sequences were identified in the
AAV6 complex due to the high degree of conservation among these two serotypes (Table 3-2).
Fab C37B and AAV2
A total of 4,392 particles were extracted from 149 CCD images to generate a final 3D map
with a resolution of ~10.8Å (Table 4-2). Density was present at each three-fold protrusion,
indicating that there is no steric hindrance for the binding of all three protrusions simultaneously
(Figure 4-1D). The latter region has been identified as being important for the binding of the
AAV2 primary receptor, HSPG (Opie et al., 2003; Kern et al., 2003; Levy et al., 2009;
O’Donnell et al., 2009), and incubation of mAb C37B with AAV2 capsids has shown a strong
reduction of infectivity, possibly due to blocking of receptor mediated entry (Wobus et al.,
2000). The C37B mAb is not cross-reactive with any of the serotypes studied here (Wobus et al.,
2000). Peptide scans identified two binding regions on the AAV2 capsid surface, including
96
residues 493 to 502 (SADNNNSEYSWT) which showed the highest binding, while residues 601
to 610 (LPGMVWQDRD) was involved to a lesser extent (Wobus et al., 2000). Those peptides
are found near the three-fold axis of symmetry with 493-502 located on the surface of the three-
fold protrusions in VRV, which is cradled by the fingers (VRIV and VIII) of a three-fold related
symmetry monomer. Residues 601-610 map to an interior region just below the central
depression of the three-fold protrusions which is not accessible from the surface and is therefore
most likely important for the structure of the three-fold axes. Currently there is no sequence data
available for this mAb, thus an unrelated Fab structure (PDB 2FBJ) was used for docking into
the Fab density (Figure 4-3). Close inspection of the docked models revealed Fab coverage at the
tips of the three-fold protrusions highlighting the variable regions (VR) within this region;
VRIV, V, and VIII (Figure 4-2). The most obvious residues involved include a portion of the
sequence identified as the major epitope by Wobus et al (2000), residues 496-500 (Table 4-4).
The peptide identified here, NNNSE, includes the two asparagines (underlined residues, N497
and N498) that when mutated in peptide competition experiments caused a partial reversal of
competition (Wobus et al., 2000). The other two proposed regions, in VRIV and VIII (peptides
RGNRQ and TTT, respectively) flank VRV as symmetry-related monomers that come together
to form the three-fold protrusions, and possibly only contribute stability to the main interaction
found at VRV. A surface exposed sequence identified to be important in receptor binding of
HSPG (residues 585-590), located in VRVIII was also identified here (Opie et al., 2003; Kern et
al., 2003), consistent with the reduction in infectivity of AAV2 capsids when incubated with
C37B mAbs due to receptor blocking (Wobus et al., 2000).
Fab F4 and AAV5 Capsids
The fnal density map of AAV5 with Fab BB3C5.F4 included 2,694 particles submitted for
3D rendering, producing a density map to 15.7Å (Table 4-2). These Fabs obscure the capsid
97
topology, and the 3D map shows a stronger portion of the Fab arm extending from the shoulder
of the three-fold protrusion towards the five-fold axis of symmetry (Figure 4-1E). Additional
analysis also reveals a weaker extension between the shoulders of the three-fold protrusions at
the base of the two-fold axis of symmetry (Figure 4-4). It may be possible that these two regions
are the binding sites for the heavy and light chain of the Fv portion, where residues at the end of
the chain may be stabilized in a non-specific manner. Docking of the AAV5 atomic coordinates
and analysis of the bound Fab structure were done at a high sigma level (~3.5-4.0σ) in order to
identify the capsid regions below the strongest density - the likely contact sites. These regions
were visually mapped to the same symmetry regions that make up the three-fold spike on the
capsid surface at VRIV, V, and VIII. Residues involved in this region were mapped to a
reference monomer at the three-fold as 442-448 (AAV5 VP1 numbering) giving the peptide
NNTGVQ at VRIV; and 572-578 of NNQSSTT at VRVIII. The three-fold related monomer
contributes peptide SGVNR (residues 479-483) at VRV (Table 4-4).
Fab ADK8 and AAV8 Capsids
A total of 1,123 particles were extracted from 78 CCD images to generate a final 3D map
with a resolution of ~18.4Å (Figure 4-5A). Density was present at each three-fold protrusion,
indicating that there was no steric hindrance for the binding of all three protrusions
simultaneously, as was seen for a previous reconstruction, AAV2:C37B (Figure 4-1D). An entire
capsid molecule, or 60-mer, was generated from the AAV8 atomic coordinates (PDB 2QA0;
Nam et al., 2007) and was docked into the respective capsid density with a correlation coefficient
(cc) of ~0.42. The cc of the fit was most likely low due to the presence of the uninterpretted
density belonging to the bound Fabs as well as the use of a C-α model (Figure 4-5B). Since the
ADK8 antibody has yet to be sequenced, a model has not been generated. The atomic
coordinates for a generic Fab model (PDB 2FBJ) were then docked into the identified Fab
98
density and a density map was generated from this new model (Figure 4-5C). A new cc was
calculated (see chapter 2) and was found to be ~0.68. The correlation was improved when the
excess density was filled, but this low correlation is most likely due to the use of C-α model, as
previously mentioned, in addition to the use of rigid-body and manual fitting techniques. In this
case, these approaches do not account for conformational differences in the docked structures.
This is evident from a comparison of the maps used in the determination of the cc between the
cryoEM reconstruction and the fitted models (Figure 4-5D).
Docking atomic models into the cryoEM map allowed for further characterization of the
ADK8 binding site on the AAV8 capsid (data presented in Appendix B). Visual inspection
identified the Fab binding to the spikes at the three-fold axis of symmetry, leaving the two- and
five-fold axes un-occluded. A more detailed inspection of the docked VP3 capsid monomers
identified VRIV, V, and VIII (Figure 4-2) of the three-fold axes to be important in the antibody
interaction. A lesser, but possible interaction was also noticed at VRVI corresponding to a
KDDEE peptide at aa530-534. These regions were also noted in the previously determined
reconstructions (Table 4-4). Sequence identification of this region proposes that the peptides
GTANTQ (VRIV; aa455-460), TTTGQNNNS (VRV; aa493-501), and LQQQNT (VRVIII;
aa586-591) could be involved in the epitope. It is interesting to note that portions of the peptides
from VRIV and V in AAV8 have been identified in studies as being immunogenic in cytotoxic
T-cell assays, while the peptide identified in VRVIII was found as an epitope in MHC II
molecules (Chen et al., 2006). Another interesting note includes the observation the sequences at
VRV and VIII both have triplicate repeats of hydrophilic, polar residues, i.e. QQQ (Table 4-4).
Fab ADK1a and AAV1 Capsids
The data collected for the ADK1 Fab and AAV1 yielded a reconstruction to ~10.8Å from
1,271 particles (Figure 4-6A). Data for this particular complex was taken from 42 film images at
99
a magnification of 50,000. The data was collected on the 200KV FEI Sphera, instead of the
300KV FEI Polara used for the other complexes, from 1.5-2.75µm underfocus. Scanning of the
films was done on a Nikon super coolscan 8000ED at a scan step size of 6.35μm/pixel, giving a
final image size of 1.27Å/pixel. Film data was most likely the deciding factor in the higher
resolution obtained for this structure.
Fab density can be seen at each individual three-fold protrusion and appears to bind at an
estimated angle ~45° to the capsid surface. The binding indicates that all axes are open for
interactions, and if the antibody is neutralizing, it most likely does so by steric interference. The
ADK1a monoclonal antibody has been shown to recognize assembled capsids (Kuck et al., 2009)
most likely implying a conformational epitope at the tip of the three-fold protrusion. Fitting of
atomic models into the cryoEM map (Figure 4-6B) has illustrated that the most probable region
of interaction is at VRIV and corresponds to the peptide 453SGSAQ457. These residues flank
the region previously found at VRIV in the other structures. It appears that the region also
previously identified at VRV, 496NNNSNF501 may help to stabilize the interaction. Because
this region is also found in AAV2 (495 NNNSN498) it is unlikely that it is involved in the
epitope itself. However, it is possible that the last two residues at AAV1, 500NF501 are
involved, since these differ from AAV2 at 499SE500. Mutagenesis will be able to confirm this
interaction. This antibody has also been found to recognize AAV6 capsids, which further
supports the region identified here due to the identity between AAV1 and AAV6 at these
regions.
Common Features of Antibody Binding Sites on the AAV Capsid
Mapping the antigenic regions of the different virus-antibody complexes shows that there
is significant overlap among these epitopes at the three-fold axis of symmetry (Figure 4-7). A
structure-based sequence alignment of AAV1, 2, 5, 6, and 8 (Figure 4-8) shows that the epitope
100
residues have stretchess of high variability, adopt different surface loop conformations, but form
similar structural features, such as the protrusions in the proposed regions of antibody binding.
Discussion
Here the antigenic structures of the AAV capsids that are recognized by a set of antibodies
have been characterized. This study also shows that there are common antibody recognition
properties that are shared by the capsids of these four distinct AAV serotypes. The two mAbs
tested that recognize the intact AAV1 capsid, AA4E4.G7 and AA5H7.D11, were produced after
extended immunization of mice and likely represent affinity mature IgGs, including the C37B,
ADK1a, and ADK8 IgGs. The BB3C5.F4 anti-AAV5 mAb was generated only 4 days after
intact AAV5 capsid immunization of a mouse which had been previously immunized with
AAV1and may represent an immature antibody.
Viruses as Antigens - Features Revealed
The antibodies tested here all bound on or near the three-fold protrusions on the capsids of
the four distinct serotypes. While the specific sites of interaction vary, with some binding nearer
the top and others on the side of these structures, there was a common overlapping area within
the footprints of all the antibodies on all capsids (Figure 4-7). All epitopes appeared
conformational in nature, as each antibody made multiple contacts with sequences from
symmetry-related monomers, which were displayed as variable loops in the assembled structure
of the capsid. The capsid surface contacted by the Fabs did not show any specific features, other
than being a raised structure with three-fold symmetry, and the distributions of amino acids in
the interacting regions were not significantly different from those seen in the remainder of the
capsid surfaces. However, it is interesting to note that some residues identified in these proposed
epitopes were highly repetitive. Indeed, further examination of the AAV1-9 sequences show that
there are several instances of identical triplet aa stretches, i.e. TTT, almost all occurring at the
101
capsid surface regions. However, the significance of this find is yet to be determined. The
regions identified as possible epitopes were also found to exhibit an overall hydrophilic
environment with predominantly polar side chains.
The most obvious feature of the structures bound by the antibodies was that they were
raised above the capsid surface. This feature is a structural property that has been revealed for
antigenic sites on the canine and feline parvoviruses, as well as other viral capsids (Kaufmann et
al., 2007; Hafenstein et al., 2009; Moskalenko et al., 2000; Smith, 2003). However, the other
raised region of the AAV capsid, the cylinder that surrounds the pore structure at the five-fold
axis of symmetry, did not bind the antibodies in this study or among the 8 antibodies examined
for the canine and feline parvovirus capsid (Hafenstein et al., 2009), and it may therefore require
more than just structural prominence to induce the formation and affinity maturation of
antibodies. The loops in that structure are flexible which may influence the attachment of
antibodies and the selection of high affinity IgGs (Agbandje-McKenna and Chapman, 2006).
Structures of Bound Antibodies and Possible Mechanisms of Neutralization or Effects on Capsid Functions
Studies of other viruses and antibodies have shown that there are several possible
mechanisms of neutralization by different antibody-virus combinations (Smith, 2003). The
neutralization mechanisms of many anti-AAV antibodies have not been investigated in detail,
but some examples have been reported. Although we have not yet examined the capsid-cell
infection processes of all the antibodies and Fabs tested here, the structural analysis provides a
number of predictions about the likely effects of antibody binding. At least for the AAV2:C37B
structure there is more conclusive evidence that the mechanism of neutralization is indeed
blocking by obscuring cellular receptor attachment as biochemical data (Wistuba et al., 1997)
and now structural data identify the HSPG receptor site as being involved.
102
Since the structures were determined in context with Fabs instead of the full mAb it is
difficult to conclude that bivalent binding had occurred for any of the antibodies studied here.
However, there are some generalizations that can be made that suggest bivalent binding for the
AA4E4.G7 mAb. In the case of the AA4E4.G7 Fab, there is strong density seen at the base of the
three-fold protrusion and hovereing above the two-fold axes in a cross-linking manner (Figure 4-
2A). However, it has been shown that bivalent binding does not have to occur over the two-fold
axis (Thouvenin et al., 1997). Bivalency on such a repetitive antigenic surface as a virl capsid
generally hints at an increased avidity causing both Fab arms to bind the capsid surface, resulting
in a marked decrease in capsid aggregation. There is no structure of the AA9A8.B12 Fab bound
to an AAV1 capsid, but preliminary complexing studies using negative staining techniques
identified a lack of aggregagtion at several dilutions of mAb (Figure 3-6). It was previously
noted in studies with a bivalent mAb against RHDV that small amounts of aggregation may be
indicative of bivalency (Thouvenin et al., 1997). There are two thoughts here, one being the
direct overlap of the antibody attachment site and the receptor binding site that would likely lead
to competition for the binding site, and the other would be the occlusion of these regions on the
capsid surface by the Fab structures, particularly where the Fabs bind at oblique angles so that
they would block an even larger region of the capsid than the direct footprint. The effectiveness
of competition may be lower for the AA5H7.D11 Fab, as that only occupies one of 3 possible
binding sites near the three-fold axis, meaning that forty sites would remain open even when the
capsid was saturated for antibody binding. However, other studies have noted that not all sites
need to be covered to induce neutralization (Thouvenin et al., 1997). Some antibodies can
stabilize the capsid against structural changes required for uncoating within the cell, which
would still allow cell binding and internalization (Smith, 2003). It appears that AA4E4.G7, if
103
binding across the two-fold as proposed, could lock the capsid in a position where uncoating may
not be possible, if the two-fold is involved in uncoating. Indeed, further studies are required to
isolate the step in the life cycle that is neutralized by these antibodies, but studies in reovirus
have identified large structural changes at the binding site (Nason et al., 2001). Conversly, it has
been proposed that antibodies do not neutralize by gross conformational changes (Smith, 2003).
Most likely due to the thought that for antibody-induced conformational changes to be important
for neutralization, clonal expansion of B-cells would need a mechanism to choose for B-cells
that would generate antibodies that induce conformational changes (Smith, 2003). Although
small structural changes were noted in a few virus-antibody reconstructions (Li et al., 1994;
Bothner et al., 1998, Lewis et al., 1998), it is more likely that capsid dynamics play a role here
instead of antibody-induced changes (Smith, 2003).
Collaborative studies with Drs. Jurgen Kleinschmidt and Christina Raupp included the
ADK8 data to further characterize this mAb and its interaction with the AAV capsid. In these
studies mutations were made in the proposed antigenic regions on the AAV8 capsid (Table 4-4)
generating similar regions corresponding to AAV2 and ultimately found that when the AAV8
peptide 586LQQQNT591 at VRVIII was changed to the AAV2 peptide 586LQNRGNR591
ADK8 recognition was abolished (Appendix B). This region was further confirmed when AAV2
capsids were mutated to the corresponding AAV8 antigenic peptide (Appendix B). Further
analysis included investigating neutralization efficiency, using in vivo and in vitro techniques,
and mechanisms of neutralization. Both in vivo and in vitro data confirmed neutralization activity
by the ADK8 monoclonal and preliminary mechanistic studies further identified that binding and
internalization of AAV8 in HepG2 cells was not perturbed (Appendix B).
104
Table 4-1. Summary of AAV-Fab complexes generated for cryoEM data studies Monoclonal antibody (mAb)
Parental Antigen
Antigen used in complex
Complex formed
AAV4E4.G7 AAV1 AAV1 AAV1:G7 AA5H7.D11 AAV1 AAV1 AAV1:D11 C37B AAV2 AAV2 AAV2:C37B BB3C5.F4 AAV5 AAV5 AAV5:F4 AA5H7.D11 AAV1 AAV6 AAV6:D11 ADK1a AAV1 AAV1 AAV1:ADK1a ADK8 AAV8 AAV8 AAV8:ADK8
Table 4-2. CryoEM data reconstruction statistics Antigen: antibody complex
Number of micrographs
Range of defocus
Includes focal pair data
Number of particles boxed
Number of particles used
Final resolution (~Å)
AAV1:G7 95 1.50-3.0 No 2,376 1,914 12.3 AAV1:D11 31 1.25-3.0 No 313 262 22.8 AAV2:C37B 149 1.50-3.0 Yes 4,731 4,392 10.8 AAV5:F4 122 1.25-3.0 No 2,806 2,552 15.7 AAV6:D11 49 1.00-3.0 Yes 5,265 2,527 15.3 AAV1:ADK1a 49/film 1.00-3.0 No 1,387 1,271 10.8 AAV8:ADK8 78 1.00-3.0 No 1,923 1,123 18.4
Table 4-3. Scaling and fitting statistics of cryoEM 3D reconstructions with atomic models 3D Reconstruction Scaling CC Final Å/pixel Fitted AAV + Fab
model CC AAV1:G7 0.80 1.863 0.82 AAV1:D11 0.81 1.850 0.65 AAV2:C37B 0.65 1.883 0.73 AAV5:F4 0.67 1.800 n/a AAV6:D11 0.77 1.850 0.63 AAV8:ADK8 0.60 1.900 0.68
Table 4-4. Proposed epitopes for each respective complex including residue numbers and
sequence data Complex VRIV* peptide: amino
acid sequence VRV* peptide: amino acid sequence
VRVIII* peptide: amino acid sequence
AAV1:G7 456-459:AQNK 492-498:TKTDNNN 585-590:QSSSTD AAV1:D11 456-459:AQNK 492-498:TKTDNNN 585-590:QSSSTD AAV1:ADK1a 453-456:SGSA -- -- AAV6:D11 456-459:AQNK 492-498:TKTDNNN 585-590:QSSSTD AAV2:C37B 454-456:TTT 496-500:NNNSE 585-590:RGNRQ AAV5:F4 442-448:NNTGVQ 479-483:SGVNR 572-578:NNQSSTT AAV8:ADK8 455-460:GTANTQ 493-501:TTTGQNNNS 586-591:LQQQNT
*residue numbering is respective to the serotype indicated in each complex
105
Figure 4-1. CryoEM 3D reconstructions shown with respective roadmap of epitope positions on the 3D capsid surface A) AAV1:G7 complex to ~12.3Å B) AAV1:D11 complex to ~22.8 Å C) AAV6:D11 complex to ~15.3 Å D) AAV2:C37B complex to ~10.8 Å E) AAV5:F4 complex to ~15.3 Å. All viral surfaces are colored respectively, AAV1-purple, AAV2-blue, AAV5-grey, AAV6-pink, while the Fab density is shown in light grey. Maps are viewed down the two-fold axis as indicated in the roadmap image.
106
Figure 4-2. Variable regions of the AAV-VP3 monomer A) Colored loops on the monomer represent the variable regions (VR) of the monomeric unit and are indicated with Roman numerals. The conserved core structure is shown in grey. Icosahedral symmetrical axes are depicted with a triangle and the number 3 for the three-fold axis, while the two-fold axis is depicted by an oval and the number 2 and the five-fold is shown as the number 5 set into a pentagon shape. B) The same monomer repeated 60-times generates an entire capsid structure and is viewed here in a surface representation. The VRs can be seen interacting with one another in context of the surface and it is evident that the capsid surface is highly variable. The VRs are colored the same as in part A and a triangle represents the asymmetric unit of the AAV capsid.
Figure 4-3. Docking of atomic models into cryoEM density A) The atomic coordinates of AAV2 (PDB 1LP3) were docked into the capsid density for the AAV2:C37B complex while atomic coordinates of a generic Fab (PDB 2FBJ) was used for the Fab density. Symmetry related capsid monomers at a three-fold axis are depicted in green, blue, and magenta and the Fab is shown in yellow. B) Zoomed in image of the broken line-boxed region shown in part A. Surface loops from the Fab model can be seen in close proximity to the capsid surface monomers. These regions were chosen as sights of possible interaction.
107
Figure 4-4. Difference map between the AAV5:F4 reconstruction and a similar scale AAV5-uncomplexed map. Difference map density can be seen as black mesh on the light grey capsid surface. The red-dash oval depicts the region covered by an entire Fab molecule. Within this region there appeared to be two sites contacting the capsid surface. The first can be seen at the regions between the five and three-fold axes, in a black-dash box and the second at the shoulders of the three-fold axis, in a black-solid box, spilling into the two-fold. Model docking appears to correlate better with Fab fitting to the region identified with the black-solid box. The 3D map is viewed down the two-fold axis at a 90 degree rotation from the maps in Figure 4-1.
108
Figure 4-5. Three-dimensional reconstruction and model fitting for the AAV8:ADK8 complex. A) Final density map for the AAV8:ADK8 structure to ~18.4Å. The AAV8 viral surface is colored green while the excess density in white was identified as the Fab density. B) Fitting of the AAV8 (PDB 2QA0) structure into the respective capsid density. Density map is shown in grey and the atomic coordinates are in green. The excess white space remaining for the Fabs could easily account for the low fitting statistics. C) Fitting of both Fab and AAV8 capsid models into the reconstructed density. One Fab portion, modeled in as PDB 2FBJ, is shown in red and is hovering above three symmetry related monomers, colored in green, blue, and magenta, at the three-fold axis of symmetry. D) Visual comparison of the maps used to generate the final correlation of the docked models (green) and the original cryoEM reconstruction (grey). An excess of green density from the fit model map can be seen extending beyond the cryoEM map, indicating regions that would decrease the final corrolation. Black arrows mark example regions of density extension.
109
Figure 4-6. AAV1:ADK1a reconstruction and model fitting. A) Three-dimensional rendering of the ~10.8Å density map. The virus is shown in purple with the Fabs depicted in white and the three-fold axis is marked by a black triangle. B) Atomic models were fit into the virus:Fab cryoEM map to isolate regions on the capsid that are interacting with capsid surface. The capsid monomers making up the highest point on one three-fold protrusion are colored in green and blue. It appears that one major loop, identified as VRIV, is most involved in the interaction. It is highly possible that other VRs close by are helping to stabilize or interact directly with the antibody.
Figure 4-7. Illustration of overlapping antigenic regions at the capsid surface level A) An AAV capsid surface is depicted in cartoon diagram illustrating the overlapping regions identified through structural epitope mapping. The DE and HI loops that make up the five-fold pore and surrounding floor, respectively, are offset in dark grey for orientation, while the overlapping antigenic residues are highlighted in black spheres. These regions sit at the shoulders of the three-fold protrusions. B) A roadmap of the 3D capsid surface maps the regions of interest in black to the three-fold axes. An assymetrical unit is shown on both surfaces, red triangle in A and a black triangle in B. A pentamer shape is representing the five-fold axes, while an oval depicts the two-fold, and filled triangles map the three-fold axes.
110
Figure 4-8. Structure based sequence alignment of AAV1, 2, 5, and 6. Variable regions (VR) are marked by roman numerals and specific sequences of interest are underlined in red. The numbering to the above the sequences follows AAV2 VP3 numbering.
111
CHAPTER 5 MUTATIONAL ANALYSIS OF PROPOSED DOMINANT ANTIGENIC SITES ON THE
CAPSID SURFACE OF ADENO-ASSOCIATED VIRUS SEROTYPE 1
Introduction
The AAV capsid surface is responsible for a multitude of interactions and functions within
the viral life-cycle including host cell receptor recognition, pathogenicity determination, viral
genome encapsidation, capsid assembly, nuclear import and export, endosomal escape,
maturation to infectious virions, and recognition by and evasion from the host immune response
(Agbandje-Mckenna and Chapman, 2006). Mutational analysis, including the use of insertions
and cross-packaging, of the AAV capsids have answered many questions about the role of the
VPs in assembly, tissue tropisms, antigenicity, and even gene transfer (Lochrie et al., 2006;
Hauck et al., 2006; Opie et al., 2003; DiPrimo et al., 2009; Wu et al., 2000; Xu et al., 2005; Shi
et al., 2006 Rabinowitz et al., 1999, 2002; Grifman; et al., 2001; Huttner et al., 2003). However,
these studies have also created more questions. For example, AAV2 mutants that lack HSPG
binding ability can no longer transduce the liver at comparable levels, but still retain a robust
tropism for heart tissue (Kern et al., 2003). Since the atomic structures of almost all of the known
AAV serotypes have been solved, examination of the capsid surface in a three-dimensional
context has aided in the identification of functional regions important in the viral lifecycle (Xie
et al., 2002; Nam et al., 2007; Govinda et al., 2006; unpublished data).
Almost all of the structural studies conferring functionality to domains on the AAV capsid
have involved AAV2. These studies include the identification of the primary cell-surface
receptor, HSPG (Summerford and Samulski, 1998) and the capsid residues that confer binding
(Opie et al., 2003; Kern et al., 2003) as well as the identification of potential co-receptors
necessary for internalization of AAV2 (Qing et al., 1999; Kashiwakura et al., 2005; Summerford
et a., 1999; Akache et al., 2006). Previous studies aimed at identifying antigenic regions have
112
also identified both T-cell and B-cell epitopes which generally map to the core structural
components and the capsid surface, respectively (Figure 1-1; Wobus et al., 2000; Mays et al.,
2009). Further studies in cellular uptake and trafficking have hinted that rearrangements at the
capsid surface level occur during receptor interaction that may allow for the exposure of residues
necessary for subsequent steps in the trafficking pathway (Levy et al., 2009; Lochrie et al.,
2006). This work has given new insight into the dynamics of these regions and the role that they
play in the capsid framework. In further support, there are several structures available which
incorporate receptors and antibodies against AAV, including the work presented here, to further
identify the multitude of interactions at the capsid surface level and beyond (Levy et al., 2009;
O’Donnell et al., 2009; unpublished data).
Increased interest in alternate serotypes has broadened many of these types of
investigations to include other serotypes. Several studies have already identified cell-surface
receptors for AAV1, 3, 4, 5, and 6 (Walters et al., 2001; Kaludov et al., 2001; Wu et al., 2006;
Di Pasquale et al., 2003). In this study we aim to further characterize the antigenic properties of
AAV1 through mutational analysis at variable regions previously identified to be involved in
antibody recognition (Chapter 4).
Results
Generation of AAV1 Capsid Mutants
Recombinant AAV1 mutants were generated in a plasmid encoding the AAV1 cap gene
and an AAV2 rep gene. Both the packaged genome, GFP, and the Ad virus helper functions
were supplied in trans on another plasmid, using UF11 with the pXX6, respectively. Mutational
primers were designed against the VRs IV, V, and VIII and targeted the previously determined
sequences (Table 4-4). The idea was to change the AAV1 sequence to that of the structurally
similar region in AAV2 because the AAV2 capsids were not recognized in native dot blot
113
analyses (Figures 3-2A and 3-2C). This would generate 7 mutants where the three regions could
be tested separately and in context with one another. The first ‘single-site’ mutant at VRIV was
changed from 456AQNK459 on AAV1 to 455TTQS458 on AAV2 and named mut1AB as all the
mutations were done in a two-step process (see methods chapter 2). The second mutant had
mutations at VRV, from 492KT493 on AAV1 to 493SA494 on AAV2 and was referred to as
mut2AB. The final single-site was at VRVIII and was changed from 586SSST589 on AAV1 to
585RGNR588 on AAV2, which has been associated with HSPG binding, and was referred to as
mut3AB. The four remaining mutants included all the possible combinations of the single-sites
to create ‘double-sites’, 1AB2AB, 1AB3AB, and 2AB3AB, and finally a mutant which harbors
AAV2 sequences at VRIV, V, and VIII, 1AB2AB3AB, which has yet to be generated (Table 5-
1).
Analysis of Mutant Protein Expression
The ability of the capsid mutants to be expressed in the 293 cell system was checked by
standard western blotting techniques using the B1 antibody that recognizes a common, linear
epitope at the C-terminus (see chapter 2). The monoclonal, 1F, was also used to detect that Rep
proteins were being expressed. Western blots on cell lysates of single-site mutants revealed that a
similar amount of the VPs were being expressed during transfections (Figure 5-1). This blot also
identified the expression of Rep proteins 52 and 42 (Figure 5-1).
Characterization of Mutant Capsid Production
The observation of VPs in a western blot does not indicate that intact capsids are being
generated during the transfection process. To check for the ability of the antigenic mutants to
form particles a native dot blot could not be done as the nature of the mutants was to evade
detection by capsid antibodies, and the lack of recognition could therefore not confirm antibody
escape, but possibly just be due to the lack of intact particles. Thus, in order to identify capsid
114
populations, cell lysates were further purified by iodixanol-gradients and ion-exchange
chromatography (see chapter 2 methods). Fractions from the iodixanol gradient were similar to
the rAAV1 vector having B1 positive fractions at the 40/60% and 40%, except for mut3AB
which showed VPs in the 40% and 40/25% (Figure 5-2A). Anion exchange chromatography also
showed identical elution fractions for all samples (Figure 5-2B). These purified samples were
then visually inspected by negative stain EM for the presence of intact particles. The single-site
mutants were all capable of forming intact virions as indicated by the presence of particles in the
EMs (Figure 5-3).
Quantification of Mutant Particle Titers, Infectivity, and Packaging Ability
In order to determine mutant capsid titers an ELISA was performed using the ADK1a
mAb. The binding region for this antibody has been mapped to regions just flanking the mutants
at VRIV (Table 4-4) and it was determined by native dot blot (Figure 5-4) that this antibody will
still bind the antigenic mutants and that any changes in titer seen can be directly correlated with a
defect in the mutant itself. Capsid titers were determined from separate analyses using cell
lysates from independent transfections performed at different times. Titers show that mut1AB
and mut2AB produced similar intact capsid amounts when compared to each other at 1.06 x 1014
and 1.72 x 1014 respectively (Table 5-2). However, there was at least a two log decrease in the
titers seen for mut3AB at 1.22 x 1012 (Table 5-2). rAAV1 controls produced higher amounts of
intact capsid than all three single-site mutants at an average titer (n=3) of 2.70 x 1014 (Table 5-2).
Preliminary particle assembly values for the double-site mutants confer a similar observation
where mutants involving aa swapps from AAV1 to AAV2 at both VRIV and VRV
(mut1AB2AB) have near control values (~2.34 x 1014 particles/ml), while those harboring
swapps from VRVIII show a slight loss when combined with VRV (2AB3AB; ~2.91 x 1013) and
a drastic drop when VRIV is included (mut1AB3AB; 1.55 x 1011) (Table 5-2).
115
Infectivity of the mutants was checked by infectious assays using the direct visualization of
green cells (see methods chapter 2). The mutants were all determined to be infectious at similar
values with mut1AB being lowest by less than one log (Table 5-2). Double-site mutants have yet
to be checked for infectivity.
Quantitative PCR (qPCR) was used to determine the packaging ability of each mutant in
context of rAAV1 as a control. However, results indicated that more genome was packaged than
capsids were made (Table 5-2). The error could easily be due to several instances, including
background interference of non-specific primer recognition, pipetting errors, technique, or is a
reflection of the sensitivity differences between ELISA and qPCR. Data for the double-site
mutations is currently under evaluation and is not available for inclusion in this report. Assays
will need to be repeated to ensure accurate data collection, possibly including optimization of
DNA extraction protocols. Acquisition of the data for the double-site mutations is currently in
progress.
Ability of Antigenic Mutants to Evade Monoclonal Antibody Recognition
Cell lysates were used in a native dot blot analysis to check for the ability of the parental
mAbs to recognize the mutants. Of the single-site mutants, it appears that the mutations in VRIV
and VRV reduce the ability of the AA4E4.G7 antibody to recognize the AAV1 capsid, seen as a
~27% and ~75% drop in pixel intensity, respectively, (Figure 5-5). Prelimianry analysis thus
indicates VRIV and VRV as important in AA4E4.G7 mAb recognition, with mut2AB (VRV)
having a greater role, but further analysis with combinatorial VR mutations in VRV and VRVIII
(mut2AB3AB) shows a drastic reduction in recognition (Figure 5-5B). Intersetingly, there is an
increase (~32%; Figure 5-5B) in AA4E4.G7 recognition of mut3AB, possibly indicating a
stabilization or increased affinity for these capsids. The AA5H7.D11 antibody was still able to
recognize all of the single-site mutants (Figure 5-5A), but a reduced pixel intensity was observed
116
for mut3AB (~56% reduction) as well as the double-site mutants mut1AB3AB and 2AB3AB (at
~61% and 62% respectively). The AA9A8.B12 antibody still recognizes mut2AB and mut1AB,
however there is definitely a loss in the AA9A8.B12 ability to recognize mut3AB (~86%
reduction; Figure 5-5B). This shows that the epitope for B12 exists in the AAV1 VRVIII. A
negative control of un-transfected cells was also used to isolate any unspecific recognition from
the cell lysates, as this background was subtracted in pixel intensity calculations.
Direct Effects of Mutations on Neutralization
An infectious assay for the singe-site mutants was repeated (as indicated for the
neutralization assays done by the Chiorini lab, NIH) and the AA4E4.G7 and AA5H7.D11
neutralizing mAbs (Nabs) were included at a ratio of one mAb per one capsid binding site (or
monomer). As seen in the native dot blot, both neutralizing antibodies knocked down rAAV1
infectivity to below visible detectable levels (Figure 5-6A-C), while infectious units were
drastically knocked down to ~10% of the original recombinant infectivity units (Figure 5-7A). In
conjunction with mut1AB un-treated levels (Figure 5-6D) the mutant’s ability to escape
recognition by the AA4E4.G7 Nabs (Figure 5-6F) was evident as comparable levels of infectious
units were accounted for in the green cell assays (Figure 5-7B). This data also confirms that
mut1AB has retained recognition by AA5H7.D11 as was previously seen in the native dot blot
(Figure 5-6E and Figure 5-5). Although, mut2AB has shown comparable results to rAAV1 titers,
infectivity, and packaged genomes (Figure 5-6A and Figure 5-6G; Table 2), as well as
recognition by the AA5H7.D11 NAb (Figure 5-6H), it has now acquired the ability to escape
recognition by Nab AA4E4.G7 (Figure 5-6I). The levels of infectious mutant virus treated with
AA4E4.G7 are near un-treated levels of mut2AB (Figure 5-7C). In accordance with the native
dot blot (Figure 5-5), mut3AB was still recognized by AA4E4.G7 and AA5H7.D11 (Figure 5-6J-
L). However, a slight drop in recognition by the AA5H7.D11 mAb for mut3AB was observed as
117
seen by the appearance of green cells in the infectious assay, which is also in accordance with the
pixel intensity calculations for this mutant (Figure 5-5B).
Discussion
Previous structural studies involving complexes of AAV capsids and anti-AAV Fabs
(Chapter 4) led to the observation that antigenic regions on the capsid surface cluster at and
around the three-fold axis. A majority of these structures involved AAV1 and a different
monoclonal antibody, including ADK1a, AA4E4.G7, and AA5H7.D11, hence the use of AAV1
as the backbone for mutational analyses. Further analysis of these structures allowed for the
selection of VRs within the capsid monomers (VP3) that appeared to have contact with the
antigen binding regions. The capsid sites were identified as VRIV, VRV, and VRVIII (Table 4-
4) and although there are several VRs that give rise to the topology seen at the three-fold, these
VRs specifically make up the three-fold protrusions. In order to confirm that these VRs were
involved in the antigenic epitope, the AAV1 cap gene was mutated in VRs IV, V, and VIII to
corresponding VRs in AAV2 and subsequently tested against the anti-AAV1 antibody panel.
Preliminary Analysis on Mutant AAV1 Viruses Reveals a Retained Ability to Assemble Infectious Capsids
Assembly studies and mutational analysis on the AAV2 VPs have identified various
regions that confer properties important for assembly, packaging, and infectivity which have
variable levels of tolerance for mutagenesis (Wu et al., 2000, DiPrimio et al., 2008; unpublished
data). In addition, combinatorial studies among several different serotypes also eluded that
certain VRs can be switched between the different AAVs in order to produce chimeric vectors
with different functional profiles (Rabinowitz et al., 2002). The mutations made here included
swapping amino acids in variable surface loop regions among two different serotypes, AAV1
and AAV2, which have almost 83% identity. For the singe-site mutations, particle assembly
118
appeared to be near recombinant AAV1 control values (Table 5-2) indicating that these regions
in the AAV1 capsid tolerate the corresponding functional domains from the AAV2 capsid.
However, inclusion of the AAV2 receptor binding domain (VRVIII) onto AAV1 (mut3AB)
appears to have decreased particle production by at least two logs (Table 5-2). This region is
known to carry a highly basic charge and may offer a challenge for the structural integrity of
these residues in context of the AAV1 background. Similar observations have been identified for
the double-site mutants where mut1AB3AB shows a drastic reduction (~3 logs) in particle
assembly compared to recombinant controls (Table 5-2). Although the addition of VRV from
AAV2 (mut2AB3AB) appears to rescue this defect as reasonable (near control values) capsid
tites were identified (Table 5-2). This phenomenon is likely due to the structural arrangement of
these VRs, as VRV from a symmetry related monomer sits between VRIV and VIII of another
three-fold related monomer. The addition of residues from AAV2 in VRV most likely act as a
‘buffer’ by helping to further stabilize the normal interactions of the AAV2 type residues in the
AAV1 background. Interestingly, VRV is fairly conserved between AAV1 and 2 (Figure 4-9) yet
the inclusion of the two residues that differ here (AAV1:KT AAV2:SA) possibly offer more
stability for capsid assembly at this region. This data may also support the importance of these
residues for AAV2 stability in this region.
Packaging ability was determined by qPCR analysis of extracted DNA and indicated that
all capsids created retained the ability to package. Since the AAV1 single-site mutant capsids
now harbor amino acids in the outermost extensions of the three-fold VRs IV, V, and VIII,
which were functional in the parental AAV2 capsid, it appears to be possible that no packaging
defect exists for these mutants. Previous studies in AAV2 indicated packaging defects for 4-5aa
linker insertion at residues 595 and 597(VP1 numbering) (Rabinowitz et al., 2000), however this
119
phenotype is not observed here as substitutions, and not insertions, were created. Also, previous
studies have not exploited these regions as being important for interactions with the Rep proteins
or packaging (Wu et al., 2002). However, since the values for PCR were higher than those values
obtained for capsid production (Table 5-2), it is hard to conclude that the packaging data is
accurate. In further analysis, this may be an artifact of the sensitivity difference between the two
techniques and will require more thorough confirmation, possibly with optimization of the DNA
extraction protocol.
Green cell assays revealed that two of the three single-site mutants retained near control
levels of infectivity, with the exception of mut1AB, which harbors VRIV residues from AAV2.
Since total capsid levels of this mutant were lower compared to the others, it is possible that this
drop is due to the difference in capsid production and also implies that infectivity assays should
be repeated holding capsid titers comparable during the assay. With the data available, it appears
that infectivity levels are within the same range suggesting that infectivity has not been altered
and that these regions are not interfering with any primary cell surface receptor recognition in the
cell type (HEK 293) tested. However, infectivity data will most likely be different depending on
the cell type used and studies will benefit greatly from taking these mutants into in vivo
infectivity assays. AAV1 capsids have been shown to very weakly bind heparin agarose beads
(Opie et al., 2003), and AAV1 mutants with VRVIII (mut3AB, mut1AB3AB, and 2AB3AB)
from AAV2 now contain the heparin-binding motif RGNR (Opie et al., 2003; Kern et al., 2003).
Subsequently, these mutants were tested on heparin-agarose columns and results concluded that
the more intense heparin-binding ability of AAV2 was also transferred to the AAV1 capsids
(Figure 5-8). Overall, these data show that the single-site mutant capsids were assembly
competent (Figure 5-3) and retained the ability to infect cells at levels comparable to the rAAV1
120
control (Table 5-2). The double-site mutants were only assumed to create capsids based on
recognition by the ADK1a ELISA and further studies are currently underway (Table 5-2).
Ability of Antigenic Escape is Limited to Specific Regions on the Capsid Surface
Single-site AAV1 mutants were tested for recognition by the anti-AAV1 panel of
antibodies that were used in the structure determination of this project (Chapter 4). Although
these studies identified the three VRs IV, V, and VIII as being involved in the epitope,
mutational analysis has made it increasingly more evident that antigen recognition is not a broad
activity. The site for AA4E4.G7 binding has been disrupted with the mutation of either VRIV
(mut1AB) or VRV (mut2AB), although the reduction appears greater for mut2AB (Figure 5-5B).
It is quite possible that either both sites are involved in a conformational epitope, or that steric
hindrance occurs when one site or the other is altered. Intriguingly, the double-site mutants with
either one or both of these regions mutated appears to have a reduction in recognition of the
capsid antigen, with ~65% and 56% reductions in VRIV+VRV (mut1AB2AB) and
VRIV+VRVIII (mut1AB3AB), with VRV+VRVIII (mut2AB3AB) having nearly abolished
recognition completely (Figure 5-5B). This drastic reduction indicates that the combination of
VR substitutions in mut2AB3AB (both VRV and VRVIII) contribute to the AA4E4.G7 epitope.
In order to confirm this, it is necessary to place these regions of AAV1 onto another serotype and
re-probe for recognition. The lack of recognition for mut3AB mutants by AA9A8.B12 indicates
that this region is involved in the antibody epitope (Figure 5-5). Double-site mutants
incorporating AAV2 VRVIII residues also confirm this observation as mut1AB3AB and
mut2AB3AB appear to have escaped recognition by the AA9A8.B12 antibody with an ~87% and
91% reduction respectively (Figure5-5B). Since there is no structure yet of the AAV1 capsid and
the AA9A8.B12 Fab, this find illustrates the strength of this technique in the elucidation of
common antigenic regions. Again, transferring these residues into the background of another
121
serotype will confer the binding region for this antibody. As an aside, it is intriguing to note that
this specific monoclonal is non-neutralizing, and following that its binding site is structurally
superimposable with the HSPG binding site of AAV2, the primary cell surface receptor-binding
region is most likely not within this region. On the other hand, it has been noted before that
parvoviruses have differing sites of receptor attachment (Agbandje-McKenna and Chapman,
2006). AA5H7.D11 still appears to recognize all double-site mutant capsids through native dot
blot analysis, although recognition reductions were seen in the pixel intensity calculations for
mut3AB, mut1AB3AB and 2AB3AB (Figure 5-5B). Infectivity assays conducted in the presence
of the AA5H7.D11 Nab indicate that mutations in VRVIII (mut3AB) may play a role in escape
from this mAB (Figure 5-6), as double-site mutants with VRVIII mutations also indicate a
reduced binding ability (Figure 5-5B). Overall, it appears that mutations in VRV and VRVIII
(mut2AB3AB) have the greatest affect on antibody recognition for the specific panel of
antibodies studied here as native dot blots revealed reduced recognition for AA4E4.G7 (near
100%), AA5H7.D11 (~62%), and AA9A8.B12 (~91%). In support, the AAV1-like virus X1 has
mutations D495G and D590H, which are also mutated in mut2AB3AB, and correlate to the drop
in AA5H7.D11 mAb recognition (Figure 5-5B). Infectivity assays for this double-site mutant
should confirm this observation.
122
Table 5-1. AAV1 mutational nomenclature and sequence differences AAV1 mutant AAV2 VRIV:
455TTQS458 AAV2 VRV: 493SA494
AAV2 VRVIII: 585RGNR588
mut1AB X - - mut2AB - X - mut3AB - - X mut1AB2AB X X - mut1AB3AB X - X mut2AB3AB - X X
X indicates by that column is contained within the AAV1 capsid sequence. A dash (-) indicates that it does not exist. Table 5-2. AAV1 single-site mutant characterization
Vector Total capsid/ml Infectious capsid/ml* Vector genomes/ml* rAAV1 2.70 x 1014 6.90 x 107 1.07 x 1015
mut1AB 1.06 x 1014 4.88 x 106 3.50 x 1014
mut2AB 1.72 x 1014 2.75 x 107 1.53 x 1014
mut3AB 1.22 x 1012 1.38 x 107 1.38 x 1014
mut1AB2AB 2.34 x 1014 Nd Nd mut1AB3AB 1.50 x 1011 Nd Nd mut2AB3AB 2.91 x 1013 Nd Nd
*(nd) not determined.
123
Figure 5-1. Western blot analysis on single-site mutant cell lysates. Cell lysates were probed with B1 antibody to detect VP1, 2, and 3 ratios and 1F to detect Rep proteins in the cell lysates. Both were used to ensure that all proteins from the rep/cap plasmid were being expressed. Mutant VP levels appear comparable to the rAAV1 control. Viral proteins are indicated with text (VP1, 2, and3) where VP3 is also indicated by a red arrow. Rep 52 and 42 are also indicated by text where rep52 is indicated below VP3 with a green arrow.
Figure 5-2. Purification of single-site antigenic escape mutants A) B1 western blot analysis on iodixanol fractions for each mutant and the rAAV1 control. Bands for VP1, 2, and 3 can be seen in the 60/40% and 40% fractions for the rAAV1 control as well as mut1AB and 2AB, while the VPs are seen in the 40% and 40/25% for mut3AB. The lower bands in the 40% and 40/25% fractions appear to be degredation products, possibly an indication of unstable capsids. B) An example elution (mut1AB) from the ion-exchange column indicating the peak that capsids were seen (Figure 5-3). Inset shows B1 positive western blot analysis for the ion-exchange elutin fractionfor each single-site mutant. VP3 is indicated by a red arrow.
124
Figure 5-3. Electron micrograph images of AAV1 mutant capsids. Ion-exchange fractions buffer
exchanged out of high salt and were spotted onto carbon-coated electron microscopy grids which were viewed at 47, 000 times magnification. White spherical images indicate negatively stained AAV capsids against a black carbon background. A) mut1AB capsids B) mut2AB capsids C) mut3AB capsids.
Figure 5–4. ADK1a native dot blot of mutant cell lysates. Mutant cell lystates were blotted against ADK1a to determine if capsids could be detected by the ADK1a monoclonal for use by an ADK1a ELISA. The AA5H7.D11 is used as a positive control as it has already shown to recognize the single-site mutants.
125
Figure 5-5. Analysis of a native dot blot of single and double-site mutants. A) Mutants were checked for recognition with anti-AAV1 mAbs AA4E4.G7, AA5H7.D11, and AA9A8.B12. The rAAV1 was used as a positive control and non-transduced cell lysates as a negative control. B) ImageJ analysis results indicating pixel intensities for the blots. Preliminary double-site mutant analysis revealed a reduced recognition by AA4E4.G7 against all double-site mutants; 1AB2AB, 1AB3AB, and 2AB3AB, indicating that VRIV and V are involved in the AA4E4.G7 epitope. A slightly reduced recognition for 3AB, 1AB3AB, and 2AB3AB by AA5H7.D11 is seen from the pixel intensity calculations indicating VRVIII in the D11 epitope. A similar observation is indicated with the AA9A8.B12 mAb.
126
Figure 5-6. Infectious green cell assay images of NAb treated single-site mutants. Mutant cell lysates and rAAV1 control were left un-treated or were saturated (1 mAb:1 VP) with the AA4E4.G7 and AA5H7.D11 Nabs and applied to 293 cells. Cells were imaged under UV light for the production of GFP. A) rAAV1 control untreated virus. B) rAAV1:D11 NAb. C) rAAV1:G7 NAb. Both AA5H7.D11 and AA4E4.G7 neutralize rAAV1 infections as indicated by the lack of green cells in (B) and (C). D) Mut1AB untreated virus. E) Mut1AB:D11. F) Mut1AB:G7. G) Mut2AB untreated virus. H) Mut2AB:D11. I) Mut2AB:G7. AA5H7.D11 treated virus shows retained recognition and neutralization of mut1AB and mut2AB, while AA4E4.G7 appears to have lost the ability to recognize and subsequently neutralize both of these capsids. J) Mut3AB untreated virus. K) Mut3AB:D11. L) Mut3AB:G7. rAAV1 capsids with VRVIII mutations (mut3AB) appears to have retained recognition by both the neutralizing Nabs, although quantification (Figure 5-7D) and visualization indicate a slight loss in recognition and neutralization of this particular mutant.
127
Figure 5-7. Histograms for in vitro Nab treated mutant virus. Group ANOVA statistics and
Tukey’s post hoc results are indicated separately for each mutant as their comparisons are internal. A) rAAV1 control virus with D11(AA5H7.D11) and G7 (AA4E4.G7) where F[2,3]=496.6, P= 0.0002 (***). B) mut1AB with D11 and G7 where F[2,3]=2217, P=0.0001 (***). C) mut2AB with D11 and G7 where F[2,3]=266.6, P=0.0004 (***). Here, a slightly significant value is noted as; *=P<0.05 D) mut3AB with D11 and G7, where F[2,3]=230.2, P=0.0005 (***).
Figure 5-8. Heparin-binding analysis. Single and double-site mutants with VRVIII (mutations including 3AB) from AAV2 were analyzed on a heparin-agarose column. Mutants 3AB, 1AB3AB, and 2AB3AB were loaded onto the column and eluted off with 1 M sodium chloride. Empty baculovirus expressed AAV2 capsids (bAAV2) were loaded as a positive control and non-transfected cell lysates were loaded as a negative control. The flow-through (FT) was collected during loading of the samples. bAAV2 blot samples were checked using mAB C37B, while AAV1 mutants were tested with mAb ADK1a.
128
CHAPTER 6 SUMMARY AND FUTURE DIRECTIONS
Summary
The AAVs have shown great promise as vectors for gene delivery in therapeutic medicine.
However, since the AAV capsid is a foreign antigen, patient immune responses are activated
causing significant decrease in the therapeutic potential of gene therapy. The future aims of the
field include generating vectors that escape immune surveilence, specifically pre-existing
humoral responses. Biochemical engineering of the capsid surface requires a working knowledge
of not only capsid regions that confer antigenicity, but also receptor binding and tissue tropism.
This study used a structural approach to determine antigenic regions on the capsid surface of
several different AAV serotypes using monoclonal antibodies directed against intact capsid
antigen. Although different serotypes (AAV1, 2, 5, 6, and 8) and monoclonal antibodies were
used, data identified that the monoclonals studied were all recognizing overlapping regions at the
three-fold axes of the capsid surface, regardless of the serotype examined (Figure 4-7). This led
to the conclusion that AAV capsids harbor common antigenic regions at the three-fold axes
including VRs IV, V, and VIII. A majority of the structures included AAV1 as the anti-AAV1
panel was largest at the time the study was initiated and thus this data was used for further
verification of structurally mapped antigenic sites. Based on the structural analyses, two to four
amino acid stretches in the AAV1 capsid VRs IV, V, and VIII were chosen for mutation (Table
4-4). These stretches were subsequently mutated to amino acid residues present in the AAV2 VP
at structurally equivalent positions since this serotype was observed to have no cross-reactivity
with the anti-AAV1 mAbs and is also ~83% similar to AAV1, possibly allowing for tolerable
mutations. This led to the creation of six AAV1 mutant capsids, three single-site mutants termed
mut1AB, mut2AB, and mut3AB, and three double-sites mutants termed mut1AB2AB,
129
mut1AB3AB, and mut2AB3AB. All mutants appeared to have comparable amounts of capsid
production, although mutants harboring mut3AB (AAV2 VRVIII) had a decrease in capsid
production (Table 5-2). It was proposed that these mutants will most likely not have packaging
defects as the regions mutated have not been previously identified in packaging or assembly
deficiencies of the recombinant virion. Preliminary vector genome evaluation could not be
trusted due to the error seen between the ELISA and qPCR values. Undeniably, PCR is a much
more sensitive technique and will most likely require optimization of the DNA extraction
protocol in order to reduce contaminating cellular DNA.
Infectivity assays with single-site mutant cell lysates revealed comparable levels of
infectivity, although mut1AB had a slightly lower infectivity among them (Table 5-2). However,
the lower value for mut1AB may be due to lower capsid production and/or packaged genome.
Interstingly, double-site mutant capsid production for mut1AB3AB and mut2AB3AB had
markedly less particles/ml (Table 5-2), possibly indicating a challenge in maintaining such a
basic patch (AAV2 VRVIII residues 585RGNR588) within the background of AAV1 VP3.
Interstingly, but not unexpected, mutants which contained the AAV2 heparin-binding motif in
VRVIII, were able to bind heparin (Figure 5-8). It is appealing to propose that receptor binding
regions are heavily supported in the context of the capsid structure through neighboring residues
and that these regions may offer dual functionality in governing receptor-mediated entry as well
as capsid stability. This observation will most likely cause issues in retaining desired
functionality during capsid engineering, but may also reveal more simplistic engineering as
changing one site to generate a desirable phenotype may remove the previous functionality that
was conferring therapeutic challenge.
130
Following preliminary mutant characterization, in vitro neutralization assays using a GFP
reporter mirrored recognition levels seen on a native dot blot. Here, capsids with mutations in
VRIV and V showed the ability to escape recognition by neutralizing antibody AA4E4.G7
(Figure 5-6A-C). Interstingly, green cell assays involving mut3AB illustrate a slight reduction in
infectivity when incubated with the neutralizing mAb AA4E4.G7 which contradicts the dot blot
analysis (Figure 5-5B). Here, an increase in recognition was observed by AA4E4.G7 with
mut3AB capsids. As the epitope for the non-neutralizing mAb AA9A8.B12 was found to include
VRVIII (mut3AB), it is likely that the increased recognition of the capsid has hindered its ability
to either bind a receptor or traffic properly. This observation supports the notion that the close
proximity of these variable regions confers a mixed functionality through indirect structural
support. Also of note, a native dot blot indicated that all capsids were still recognized by the Nab
AA5H7.D11, but pixel intensity calculations and green cell assays identified a slight drop in
recognition by mutants with VRVIII muations (Figure 5-5B) and an enhanced ability to infect in
the presence of the Nab (Figures 5-6D and 5-7K), respectively. It is likely that engineering of
unique vectors will also include the replacement of adjacent residues to support the desired
mutations/deletions in order to retain viability.
Taken as a whole, all three VRs identified through the structural analysis of complexed
AAV capsid revealed antigenic nature as mutations in these regions altered recognition to some
level. From the data presented for this antibody panel, it appears that VRVIII plays an important
role in the antigenic nature of the AAV1 capsid, including the surrounding regions of VRIV and
VRV. This study has at least developed a foundation for mutational analysis of antigenic regions
on the AAV capsid and offers a starting point for engineering antigenic escape vectors. Ideally,
the whole AAV capsid is antigenic in nature and current data shows that the majority of the
131
raised regions on the AAV capsid surface, other than the top of the five-fold cylindrical channel,
are succestible to antibody recognition. Thus it is anticipated that new antigenic regions will
arise following the use of escape mutant vectors and that new antibodies against the engineered
vectors will arise. Further analysis of these responses will lead to strategies for better vector
development because it will provide information on capsid regionswhere minimal mutations
create novel antigenic profiles. This information can then inform the engineering of new esacape
vectors which encompase all potential epitopes.
Future Directions
The studies initiated here offer a pathway for further elucidation and definition of the
common antigenic regions on the AAV capsid surface. Including more diverse serotypes, like
AAV4, and more similar serotypes, like AAV9, and expanding the number of antibodies tested,
will most likely offer more support for these common antigenic regions, but may also identify
other areas of the capsid that should be included in engineering antibody escape vectors.
Broadening the antibody panel to include different isotypes, i.e. IgA, may also offer a different
perspective among antibody binding and may also indicate preferential sites among isotypes.
These studies could also identify mechanistical differences among the isotypes and the types of
response mounted, i.e. IgM vs. IgG. Also, the use of serum from patients involved in AAV trials,
and those with natural exposure, may also better identify which antigenic sites are more
common, if such a thing exists. Structural studies including receptor binding regions will also
support the functionality of the variable surface loops and will help to better guide capsid
engineering.
Future studies involving characterization of the mechanisms of neutralization for the
antibodies proposed here may also link specific isotypes or binding regions with certain
mechanisms, allowing for the prediction of antigen-antibody interactions. More immediate goals
132
will be to swap the identified epitope regions from AAV1 onto other serotypes to confer their
specificity for the antibodies studied here. Also, testing of the intermediate mutants, i.e. 1A,
where only half of the site has been altered may confer more specific residue interactions at the
capsid surface level. Furthermore, a more thorough mutational analysis through this region will
likely better indicate residues involved in antigenicity and may further define those residues
important for the structural integrity of this region.
133
APPENDIX A HUMAN BOCAVIRUS CAPSID STRUTURE: INSIGHTS INTO THE STRUCTURAL
REPERTOIRE OF THE PARVOVIRIDAE
Introduction
Human Bocavirus (HBoV), a newly discovered member of the family Parvoviridae, was
originally isolated in randomly selected nasopharyngeal aspirates (Allander et al., 2005). Since
this initial discovery, HBoV has also been detected worldwide, predominantly in children under
the age of two with respiratory infections, in serum, urine, and fecal samples (Lindner and
Modrow, 2008). Symptomatic children commonly exhibit acute diseases of the upper and lower
respiratory tracts (Arnold et al., 2006; Kesebir et al., 2006; Monteny et al., 2007; Vicente et al.,
2007), and possibly, gastroenteritis (Kapoor et al., 2009; Vicente et al., 2007), though the link to
gastroenteritis outbreaks has been questioned (Campe et al., 2008). It is still unclear if HBoV is
the sole etiologic agent of respiratory disease as higher rates of co-infections with other
respiratory pathogens such as human rhinovirus and Streptococcus spp. are often observed
(Allander et al., 2007). However, Allander et al. recently reported (2007) that HBoV was found
in 19% of children with acute wheezing, thereby making it the fourth most common virus, after
rhinoviruses, enteroviruses, and Rous Sarcoma Virus, detected in children exhibiting this
symptom. These findings suggest that, at high viral load, HBoV could be an etiologic agent of
respiratory tract disease (Allander et al., 2007). HBoV infection is common in the first few years
of life and clinical research suggests it may follow the primary period for acquisition of Human
Parvovirus B19 (B19), though there is no antigenic cross-reactivity between B19 and HBoV
(Kahn et al., 2008; Kantola et al., 2008). By age five most people have circulating antibodies
against HBoV as is also true for other respiratory viruses such as rous sarcoma virus,
rhinoviruses, and Human Metapneumovirus (Chow and Esper, 2008). HBoV has also been
134
identified in adults, with ~63% of samples tested being seropositive, showing a positive
correlation with age and a slight positive bias towards women (Cecchini et al., 2009).
The Parvoviridae is a family of small, non-enveloped viruses that package a single-
stranded DNA (ssDNA) genome of ~5000 bases. These viruses are subdivided into two
subfamilies: Parvovirinae and Densovirinae (Table A-1). The Parvovirinae are further
subdivided into five genera, all of whose members infect vertebrates. The Densovirinae (four
genera) only infect invertebrates. Phylogenetic analysis places HBoV in the recently classified
Bocavirus genus (Table A-1). In addition to HBoV, numerous parvoviruses circulate among the
human population. Among these are several dependoviruses, Adeno-associated virus (AAV)
serotypes AAV1-3, AAV5, and AAV9, the Erythrovirus B19, and the newly discovered human
parvovirus genotypes 4 (Parv4) and 5 (Parv5) (Jones et al., 2006; Fryer et al., 2006; Schneider et
al., 2008). Of these, only B19 had been implicated in disease until the discovery of HBoV and
Parv4, which has been isolated from patients who present symptoms of acute HIV infection
(Schneider et al., 2008).
The HBoV genome, like that of all members of the Bocavirus genus, contains three open
reading frames (ORFs). The first ORF (ns), at the 5’ end, encodes for NS1, a nonstructural
protein. The next ORF, unique to the Bocaviruses, encodes for NP1, a second nonstructural
protein. The third ORF (cap), at the 3’ end, encodes the two structural capsid viral proteins
(VPs), VP1 and VP2. The HBoV VPs share 42% and 43% amino acid sequence identity with the
corresponding VPs of bovine parvovirus and canine minute virus, respectively (Allander et al.,
2005). More recently, two additional HBoV-like viruses, HBoV-2 and HBoV-3, were identified
in stool samples from children (Arthur et al., 2009; Kapoor et al., 2009). The genome
organization of these viruses is identical to that of HBoV, with the NS1, NP1, and VP proteins of
135
HBoV2/3 being ~80/90%, ~70/80%, and ~80/80% identical to the respective proteins in HBoV
(Arthur et al., 2009; Kapoor et al., 2009).
Parvovirus genomes are packaged into a T=1 icosahedral capsid that is assembled from 60
copies of a combination of up to six types of capsid VPs (VP1-VP6), all of which share a
common C-terminus. VP1 is always a minor component, typically comprising about five copies
per capsid, whereas the smallest VP is always the major component. The unique N-terminal
region of VP1 (VP1u) contains a conserved phospholipase A2 (PLA2) motif within the first 131
amino acids that is essential for infection (Qu et al., 2008; Zadori et al., 2001). Interestingly,
AMDV, the only member of the Amdovirus genus, is the only exception in that this motif is
absent, which suggests this virus employs a different mechanism to escape the endosome during
infection (Tijssen et al., 2006).
The X-ray crystal structures of several parvoviruses show that all VPs contain a conserved,
eight-stranded β–barrel motif (βB-I) that forms the core of the capsid (Chapman and Agbandje-
McKenna, 2006). There is also a conserved α-helix (α-A) observed in all parvovirus structures
determined to date. The bulk of the VP consists of elaborate loops between the strands that form
the surface of the capsid. For example, the GH loop between the βG and βH strands is ~230
residues. The composition and topology of these loops encode several important functions,
including tissue tropism, pathogenicity, and the antigenic response directed against each
parvovirus during infection (Agbandje-McKenna and Chapman, 2006).
A number of parvoviruses have been studied by cryo-electron microscopy (cryoEM) and
three-dimensional (3D) image reconstruction in concert with and complementary to X-ray
crystallographic studies (reviewed in Agbandje-McKenna and Chapman, 2006). Report here is
the 3D structure of a recombinant HBoV capsid solved to 7.9-Å resolution using cryoEM. The
136
capsid of HBoV was compared to representative members of Parvoviridae (Table A-1) with
known atomic structures (AAV2, MVM, B19, GmDNV) or pseudo-atomic models built into
cryoEM reconstructed density (AMDV) to identify similarities and differences. The capsid
topology of the newly emerging HBoV incorporates a combination of structural elements at the
surface seen in other members of the Parvovirinae. However, homology modeling, based on a
comparison of sequences and available parvovirus structures, places the HBoV VP2 topology
closest to B19, the only other structurally characterized parvovirus that is pathogenic to humans.
Materials and Methods
Expression and Purification
The putative HBoV VP2 gene was constructed by de novo synthesis using the published
HBoV ST1 sequence (NCBI Accession No. DQ000495; start codon, nt 3373), and cloned into a
pUC vector (Epoch Biolabs, Inc.). The vector was cut with EcoRV and HindIII to release the
insert, and recloned into a Fastbac1 plasmid (Invitrogen). The sequence and orientation of the
resulting plasmid were confirmed, and the plasmid was used to produce a recombinant
baculovirus using the standard Bac-to-Bac technology (Invitrogen).
Sf9 insect cells (ATCC) were maintained in Grace’s medium (Invitrogen) with 10% fetal
calf serum (FCS) and antibiotics, infected with recombinant baculovirus, harvested 4 or 7 days
post infection into the medium, and collected by centrifugation. Production of baculovirus was
confirmed by a combination of immunofluorescence of infected cells with mouse antibody
against baculovirus (eBioscience), and the typical cytopathic effect of baculovirus-infected cells.
For the immunofluorescence testing, slides were prepared by cytocentrifugation (200 x g for 8
min in a Shandon Cytospin 4) or by spotting a concentrated cell suspension onto each of 8 spots
of a fluorescent antibody slide (Bellco Glass, Inc). The slides were allowed to air dry and were
fixed in 1:1 acetone:methanol at -20°C for 10 min. The monoclonal antibody was diluted 1:50 in
137
PBS plus 10% FCS, and 25 µl of the solution was incubated with each cell spot (~200 cells) for
1 h at 37 °C. After washing with PBS, the cells were incubated with 1:100 anti-mouse-FITC
conjugated antibody in PBS plus 10% FCS and 1:200 Evans Blue (Sigma) for 1 h at 37°C. After
further washing, the cells were examined by UV microscopy.
Sf9 cells were grown in suspension at 27 °C in Sf-900 II SFM media (Gibco/Invitrogen
Corporation) and infected at a multiplicity of infection of five plaque-forming units per cell.
Resultant HBoV virus-like particles (VLPs) were released from infected cells by three cycles of
freeze–thaws in lysis buffer (50 mM Tris–HCl pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.2%
Triton X-100), with the addition of benzonase (Merck KGaA) after the second cycle. The sample
was clarified by centrifugation at 12,100 x g for 15 min at 4 °C. The cell lysate was pelleted
through a 20% (w/v) sucrose cushion (in 25 mM Tris–HCl pH 8.0, 100 mM NaCl, 0.2% Triton
X-100, 1 mM EDTA) by ultracentrifugation at 149,000 x g for 3 h at 4 °C. The pellet from the
cushion was resuspended in the same buffer overnight at 4 ˚C. The sample was subjected to
multiple low-speed spins at 10,000 x g in order to remove insoluble material. The clarified
sample was purified by separation within a 10-40 % (w/v) sucrose-step gradient spun at 151,000
x g for 3 h at 4 °C. Equilibrium dialysis was performed against 20 mM Tris–HCl pH 7.5, 150
mM NaCl, 2 mM MgCl2 at 4 °C. The final concentration of the sample was 10 mg mL-1
calculated from optical density measurements at 280 nm with an extinction coefficient of 1.7 M-1
cm-1. The purity and integrity of the viral capsids were monitored by SDS-PAGE and negative-
stain electron microscopy, respectively.
Negative Stain Transmission Electron Microscopy
Small (3.5 µL) aliquots of purified VLPs (~5 mg mL-1) were applied to a continuous
carbon support film that had been glow-discharged for 15 s in an Emitech K350 glow discharge
unit and subsequently stained with 1% aqueous uranyl acetate. Micrographs were recorded on a
138
2k x 2k Gatan CCD camera in an FEI Sphera microscope at a nominal magnification of 40,000 x
at an accelerating voltage of 200 keV.
CryoEM and Image Reconstruction
Small (3.5 µL) aliquots of purified VLPs (~10 mg mL-1) were vitrified via standard rapid
freeze-plunging procedures (Adrian et al., 1984; Dubochet et al., 1988). Samples were applied to
Quantifoil holey grids that had been glow-discharged for ~15 s in an Emitech K350 glow
discharge unit. Grids were then blotted for ~5 s, plunged into liquid ethane, and transferred into a
pre-cooled FEI Polara multi-specimen holder, which maintained the specimen at liquid nitrogen
temperature. Micrographs were recorded on Kodak SO-163 electron-image film at 200 keV in a
FEI Polara microscope under minimal-dose conditions (~9 e/Å2) at a nominal magnification of
39,000 x.
Sixteen micrographs exhibiting minimal astigmatism and specimen drift, recorded at
underfocus settings ranging between 0.95 and 3.25 µm, were digitized at 7 µm intervals
(representing 1.795 Å pixels) on a Zeiss SCAI scanner. The software package RobEM
(http://cryoEM.ucsd.edu/programs.shtm) was used to extract 3,754 individual particle images
(each 1,872 pixels), pre-process the images, and estimate the defocus level of each micrograph
(Baker et al., 1999). The random-model procedure (Yan, Dryden, et al., 2007) was used to
generate an initial reconstruction at ~ 25 Å resolutions from 150 particle images. This was then
used as a starting model to initiate full orientation and origin determinations and refinement of
the entire set of images using AUTO3DEM (Yan, Sinkovits, et al., 2007). Corrections to
compensate for the effects of phase reversals in the contrast-transfer functions of the images
were performed as previously described (Zhang et al., 2003; Bowman et al., 2002), but
amplitude corrections were not applied. A final 3D map at an estimated resolution limit of 7.9 Å,
based on the 0.5 threshold of the Fourier Shell Correlation (FSC0.5) criterion (van Heel and
139
Schatz, 2005) (data not shown), was reconstructed from all particle images. To verify the new
structure, we computed two, completely independent, ab initio reconstructions from separate sets
of micrographs consisting of 1770 and 1839 particle images. The resultant reconstructions
showed excellent correspondence in all coarse and fine features as evidenced by good agreement
(FSC0.5) at all spatial frequencies out to 1/9.8 Å-1. An inverse temperature factor of 1/100 Å2
was applied to the final 3D reconstruction (Havelka et al., 1995) to enhance fine details and to
aid in the analysis and interpretation of the capsid structure. A Gaussian function was applied to
attenuate the Fourier data smoothly to zero between 7.24 and 6.75 Å. The absolute handedness of
the reconstruction was set to be consistent with surface features seen in parvovirus crystal
structures (see below).
Generation of Density Maps for Representative Members of the Parvoviridae Genera
Density maps were generated from the crystal structure Cα coordinates of AAV2 (PDB
accession No. 1LP3 (Xie et al., 2002)), B19 (PDB accession No. 1S58 (Kaufmann et al., 2004)),
GmDNV (PDB accession No. 1DNV (Simpson et al., 1998)), and MVM (PDB accession No.
1Z1C (Kontou et al., 2005)), and from the Cα pseudo atomic model of AMDV (McKenna et al.,
1999). The coordinates for each VP2 subunit were used to generate a complete icosahedral
capsid using the Oligomer generator in VIPERdb (Carrillo-Tripp et al., 2009). The CCP4
software suite was used to calculate structure factors (7.9 Å resolution cutoff, B-factor =100 Å2)
and electron density maps, using the SFALL and FFT routines, respectively (The CCP4 suite:
programs for protein crystallography, 1994). The calculated maps were rendered with RobEM
(http://cryoEM.ucsd.edu/programs.shtm) and Chimera (Pettersen et al., 2004).
140
Sequence Analysis and Generation of Unrooted Phylogenic Tree
VP2 sequences of representative members of the six Parvoviridae genera (taken from the
PDB accession files listed above), AMDV (NCBI accession No. ABG20974), and HBoV (NCBI
accession number ABN46897) were uploaded to the Biology Workbench (Subramaniam, 1998).
A multiple sequence alignment was conducted using ClustalW (Thompson et al., 1994) with
default values and with B19 as the reference sequence. An unrooted phylogenetic tree was then
created using the PHYLIP algorithm (Felsenstein, 1989).
Generation and Docking of a Homology Model of HBoV VP2 into Cryo-Reconstructed Density
A homology model was built for the HBoV VP2 (amino acids 37-542) using the Swiss-
PDBviewer-Deep View 3.7 software program (Guex and Peitsch, 1997), using the VP2 sequence
(NCBI accession No. ABN46897) and the B19 VP2 coordinates (PDB accession No. 1S58)
supplied as a template. B19 VP2 was chosen as a template for building the HBoV VP2 pseudo-
atomic model because the comparison of density maps generated for the members of the
different Parvoviridae genera (see above) identified B19 as being the most structurally similar to
HBoV even though the AAV2 VP2 sequence had a slightly higher sequence identity.
The HBoV VP2 homology model was used to generate the coordinates for a complete, 60-
subunit capsid through icosahedral matrix multiplication in VIPERdb (Carrillo-Tripp et al.,
2009). The 60-mer was then visualized within the HBoV cryo-EM density map in COOT
(Emsley and Cowtan, 2004). The core β-barrel (β-strands B-I) and helix αA regions fitted into
density contoured at a 3.3-σ threshold without further adjustment. The loops between β-strands
D and E (DE loop) and between β-strands H and I (HI loop) of a reference VP2 monomer were
manually adjusted to fit into density contoured at a threshold of 1.0 σ guided by their distinct
structural features. The remaining surface loops were adjusted to be within the reconstructed
141
envelope at a density threshold of 0.5 σ. Further model building included addition of amino acids
at the N-terminus of the VP2 monomer (residues 28-36) to satisfy density (0.5-σ threshold)
inside the capsid under the icosahedral five-fold axes. Following these adjustments, the geometry
of the model was regularized under idealized geometric constraints in COOT. The NCS function
in COOT was used to generate the remaining 59 icosahedral symmetry related VP2 monomers to
verify their fit within the cryo-EM density map. The pseudo-atomic capsid model coordinates
were used to generate a density map for comparison with the HBoV cryo-EM density in
MAPMAN ( Kleywegt and Jones, 1996) using the program’s similarity function
(http://xray.bmc.uu.se/usf/mapman_man.html).
Comparison of Parvovirus VP2 Structures
The structures of VP2 from representative members of six Parvoviridae genera (Table A-
1) were aligned by superposition of the Cα positions of their atomic coordinates (AAV2, B19,
GmDNV, and MVM) or pseudo-atomic coordinates built into cryo-reconstructed densities
(AMDV and HBoV) using the MatcherMaker Tool (Meng et al., 2006) in Chimera.
Results and Discussion
HBoV Capsid Structure
Self-assembled VP2 VLPs expressed and purified from insect cells were checked for
composition and purity by SDS-PAGE (Figure A-1A), and for integrity by negative-stain
electron microscopy (Figure A-1B) prior to sample vitrification and cryoEM (Figure A-1C). A
3D reconstruction of the HBoV capsid, at an estimated resolution limit of 7.9 Å, was computed
from 3,754 individual particle images (Figure A-2) and compared to the capsid structures of
representative members of the Parvovirinae (B19, AAV2, AMDV, MVM) and Densovirinae
(GmDNV) subfamilies (Figure A-3).
142
The HBoV capsid exhibits three characteristic features common to other vertebrate
parvoviruses. The first of these is a dimple-like depression at each icosahedral two-fold axis
(“dimple”) (Figures A-2A and A-3). The second is a large, trimeric protrusion that surrounds
each three-fold axis or is located at the three-fold axis (“protrusion”) (Figures A-2A and A-3).
The third is a channel at each five-fold axis (“channel”), whose outermost opening is formed by
a small, pentameric structure encircled by a wide, canyon-like region (“canyon”) (Figures A-2A
and A-3). While the dimple is also observed among the invertebrate parvoviruses, these members
lack the three-fold protrusions and canyon around the five-fold channel (e.g. GmDNV in Figure
A-3).
The external diameter of the HBoV capsid ranges from ~215 Å at the lowest points of the
dimple and canyon to ~280 Å at the top of the protrusion. For comparison, the maximum
diameters of AMDV and GmDNV (the largest and smallest capsids of the representative
parvoviruses being compared) are 310 and 270 Å respectively, and the minimum diameter
calculated from the capsid density maps generated from the atomic or pseudo-atomic coordinates
of all the viruses being compared is ~220 Å (Figure A-3). The spherical, relatively smooth
topology of the invertebrate GmDNV clearly distinguishes it from the others, especially those
with more prominent surface features such as AMDV, MVM, and AAV2. HBoV appears to be
morphologically most similar to B19 since both display features intermediate to the above
extremes.
Radial density projections of the six representative Parvoviridae capsids reveal that
structural similarity among them is greatest at low radii and most divergent at the capsid surface
(Figure A-4). Of note, all the vertebrate parvoviruses have strikingly similar structures at radii <
143
98 Å, suggesting they share a common core. Again, GmDNV stands out among the parvoviruses
as having a structure that is most distinct from the others at all radii.
Pseudo-Atomic Model of HBoV VP2 Built Based on the B19 VP2 Crystal Structure
The HBoV capsid surface topology was observed to be most similar to B19 (Figure A-3),
even though the AAV2 VP2 protein sequence is closer to that of HBoV in percentage identity
(Table A-2). Preliminary docking of the crystal structure of the VP2 subunit for both AAV2 and
B19 into the HBoV cryo-reconstruction showed that the β-barrel and α-A helix in both viruses
fitted into the HBoV density (with a MAPMAN correlation coefficient of 0.6, but for AAV2, the
loop regions that form each protrusion extended outside the HBoV envelope (not shown)
consistent with HBOV’s smoother surface features (Figure A-3). Thus, the B19 crystal structure
coordinates (VP2 residues 19-554) were used to construct the initial homology model of HBoV
VP2 (residues 37-542) for docking into the cryo-EM density map (Figure A-5A and A-5B).
A map of HBoV at sub-nanometer resolution enabled precise, rigid-body fitting of the
conserved secondary structural elements (βA, β-barrel core, and α-A helix) as well as the DE
and HI loops, which required only small manual adjustments followed by bond geometry
refinement (Figure A-5B-D). The remaining inter-strand loops and the C-terminus of the VP2,
including variable regions (VRs as defined in (Govindasamy et al., 2006)), were adjusted to
conform to the HBoV density envelope (Fig. 5A and B). The initial homology model of HBoV
started at residue 37 which corresponds to B19 residue 19, the first N-terminal residue reported
in the crystal structure (Kaufmann et al., 2004). However, a clearly defined region of density at
the base of the channel adjacent to residue 37 of the initial HBoV model remained un-interpreted
following the fitting procedure. This density (denoted by an asterisk in Figure A-2C and A-2D)
was modeled as residues 28-36 contributed by five VP2 monomers (Figure A-5E). These nine
144
residues were modeled as a small loop that projects into the additional density and extends
towards the channel as seen in B19 and MVM (Agbandje-McKenna et al., 1998; Kaufmann et
al., 2008) (dark blue loop in Figure A-5A and A-5E). This interpretation positions residue 28 of
the HBoV model towards the remaining un-interpreted density that lies close to the entrance of
the channel and appears to form a “plug-like” structure inside the capsid (black arrow in Figure
A-2C and A-5E). The correlation coefficient between the map of the 60-subunit capsid
(generated from the final HBoV VP2 pseudo-atomic model, residues 28-542) and the cryo-EM
density map is 0.70.
A similar plug at the base of the five-fold channel was also observed in the cryo-
reconstruction of B19 VLPs comprised solely of VP2 (Kaufmann et al., 2008). This plug
appeared as negative difference density when the B19 VLP cryo-reconstruction was subtracted
from reconstructions of wild-type full (VP1, VP2, and DNA) and empty (VP1 and VP2) B19
capsid structures (Kaufmann et al., 2008). The location of the first ordered N-terminal residue in
the B19 VP2 VLP crystal structure (starting with residue 18) overlapped this negative density,
suggesting that the remaining, un-ordered residues lie inside the VLP capsid (Kaufmann et al.,
2008). In the B19 report no effort was made to model the missing residues into the negative
difference density. However, positive difference density was observed, in the B19 full and empty
capsid minus VLP density map subtraction, projecting into the channel towards the capsid
surface, emerging between the DE loop β-ribbons onto the canyon at the capsid surface. This
density was modeled as B19 VP2 residues 1-17 (Kaufmann et al., 2008). The differences in
locations of the N-terminal residues in B19 VLPs versus capsids with or without DNA and VP1
are likely dictated by the presence of the DNA and/or VP1. The VP1u motif was not observed in
B19 capsids (Kaufmann t al., 2008) most likely because only a few copies are present per
145
particle, and their contributions to the density map would be averaged out by the icosahedral
symmetry imposed during structure determination. The B19 channel appears to be closed or
tightly constricted at the capsid surface (Figure A-3) (Kaufmann et al., 2008; Kaufmann et al.,
2004). Thus, disposition of the N-terminal residues of the B19 capsid proteins on the outer
surface is consistent with previous reports that placed VP1u on the B19 surface (Anderson et al.,
1995; Kajigaya et al., 1991; Kawase et al., 1995), which differs from other parvoviruses in which
externalization is proposed to occur via the channel. In HBoV the channel is constricted at its
base but open at the top (Figure A-2C and A-D, and Figure A-3). Thus, it is difficult to assess
whether the HBoV VP2 and/or VP1 N-termini are also already located on the capsid surface. Our
current model of HBoV suggests the N-terminal residues (Figure A-5E) may protrude out
through the channel in virions. Structural studies of capsids that contain VP1 or both VP1 and
DNA are needed to verify this prediction.
HBoV VP2 Shares Low Sequence Identity but High Structural Similarity to Other Parvovirinae VP2s
Sequence alignments and an un-rooted phylogenetic tree were generated to quantitatively
compare the HBoV VP2 sequence with the corresponding VP sequences in five other parvovirus
genera (Table A-1) and to help guide homology modeling based on the HBoV cryo-EM density
map. The pairwise sequence identities ranged between 12 and 23%, with GmDNV, the sole
invertebrate member in the comparison, being the most divergent from HBoV (Table A-2).
Phylogenetic analysis based on a multiple sequence alignment of the six viruses being compared
placed them almost equidistant from each other in an un-rooted tree (data not shown) suggesting
high divergence.
Structural superimposition of the Cα positions of the atomic (B19, AAV2, MVM) and
pseudo-atomic (AMDV) coordinates of the Parvovirinae VP2s with the pseudo-atomic model
146
generated for HBoV (residues 28-542) showed that, despite the high (~50%) sequence
variability, the eight-stranded β-barrel (BIDG-CHEF) core, βA, and αA helix are conserved in all
the viruses (Figure A-6A). The most VRs in the VP2 structures of the Parvovirinae being
compared to HBoV are located within the loops between the β-strands and most occur at and
around the three-fold axis (Chapman and Agbandje-McKenna, 2006). In the HBoV pseudo-
atomic model, the VRs in the loops are also predicted to be located around this axis (Figure A-
5A). However, at the current resolution of the HBoV density map, the exact positions of the loop
structures are unknown and thus preclude a comparison with those of the other viruses. As with
the Parvovirinae members, the core β-strands are structurally very similar between HBoV and
GmDNV, though the exact position of αA differs slightly (Figure A-6A).
HBoV Capsid Surface Topology is Unique, but Shares Features Common to Other Vertebrate Parvoviruses
Based on the overall surface morphologies of their capsids, parvoviruses have been
structurally assigned to three groups (Padron et al., 2005). Groups I and III comprise members of
the Parvovirinae subfamily, all of which share three features as described earlier: dimples at
two-fold axes, protrusions at or surrounding three-fold axes, and channels surrounded by
canyons at the five-fold axes (Figure A-3). The primary difference between these two groups
occurs in the vicinity of each three-fold axis. Group I capsids have a single, relatively flat,
pinwheel-shaped protrusion, whereas group III capsids have three distinct protrusions. Current
structural data place members of the Parvovirus genus in group I and members of the
Dependovirus, Amdovirus, and Erythrovirus genera in group III (Table A-1, (Padron et al.,
2005)). A third topology (group II), seen in GmDNV and other members of the Densovirinae
subfamily (Bruemmer et al., 2005; Simpson et al., 1998), is characterized by capsids which are
relatively spherical and featureless compared to the vertebrate parvoviruses (Table 1, (Padron et
147
al., 2005)). This morphology, exemplified by GmDNV, results from the reduced length of the
surface loops compared to members of the other parvovirus genera (Simpson et al., 1998).
HBoV possesses features common to groups I and III and also others that are unique. One
distinguishing feature in HBoV arises from a difference in surface topology at the two-fold axis,
where the HBoV dimple is narrower and shallower than that in the other parvoviruses (Figure A-
3). However, like B19, the HBoV dimple is flanked by wider “walls” between the two- and five-
fold depressions compared to the other viruses which may be due a conformational loop
difference in modeled VR-IX located close to the two-fold axis (Figure A-5A). Residues that
form the walls of the parvovirus dimple have been implicated in receptor attachment for MVM
(Lόpez-Bueno et al., 2006) and as determinants of tissue tropism and pathogenicity for other
members of the parvovirus genus (reviewed in (Kontou et al., 2005)). There is no evidence
however that this region plays an analogous role in B19 or HBoV.
Other characteristic parvovirus features at the two-fold region are conserved in HBoV.
Helix αA, observed in all parvovirus structures determined to date (reviewed in (Chapman and
Agbandje-McKenna, 2006)), is also present in HBoV (Figure A-5C and A-6A). This helix forms
stabilizing interactions at the icosahedral two-fold interface (Govindasamy et al., 2006). Even
though there are fewer interactions at the two-fold interface compared to those at the three-fold
interface, they are proposed to play a crucial role in parvovirus capsid assembly (Govindasamy
et al., 2006; Wu et al., 2000). Conservation of this structural feature in the HBoV VP2 thus
suggests that it plays a similar, central role in capsid assembly.
HBoV topology at the three-fold axis is intermediate between that of groups I and III.
AAV2 and AMDV, group III viruses, have large protrusions that arise from the βEF and βGH
loops. The protrusions in HBoV are less pronounced than those of the group III parvoviruses,
148
which may be due to shorter sequences and/or conformational differences in its βEF and βGH
loops. This results in a less pronounced depression in this region (Figure A-3). MVM and
AMDV, group I viruses, also lack depressions at their three-fold axes (Figure A-3). Residues
near the three-fold axes in AAV2 and B19 have been implicated as receptor attachment sites
(Chipman et al., 1996; Kern et al., 2003; Levy et al., 2009; O'Donnell et al., 2009, Opie et al.,
2003). The protrusions surrounding this axis are responsible for the antigenic reactivity of
members of the Parvovirinae subfamily (reviewed in (Agbandje-McKenna and Chapman,
2006)). Given this antigenic role of the group I and III surface protrusions, the lack of similar
structures in group II viruses is consistent with the fact that invertebrate viruses are not subject to
adaptive pressures of the host immune response (Simpson et al., 1998).
As mentioned above, the capsid structures of members of the Parvovirinae have a
conserved channel at the icosahedral five-fold axis, surrounded by a canyon (Figure A-3). The
channel is a β-cylinder formed by five symmetry-related β-ribbons inserted between β-strands D
and E with a loop at the top (referred to as the DE loop). In HBoV the DE loops form finger-like
protrusions that surround the channel (Figures A-5A, A-5D, and A-6B) and project away from
the five-fold axis in a manner similar to the conformation observed in AAV2, resulting in an
open channel (Figure A-3). In contrast, for the other Parvovirinae structures, the DE loops
cluster nearer the five-fold axis and constrict the channel (Figure A-3 and A-6B). This channel in
parvoviruses is postulated to serve several functions such as the site of (a) VP2 externalization in
members of the parvovirus genus (e.g. MVM, in which VP3 is generated by a maturation
cleavage event), (b) VP1u externalization (except in B19 and GmDNV) for its PLA2 function,
and (c) genomic DNA packaging. However, because the diameter of the channel varies between
~5 and 25 Å at its narrowest and widest points, respectively, this is likely too small to
149
accommodate extrusion of the ~130aa in the PLA2 domain. It has thus been proposed that
flexibility and structural re-arrangement of the DE loop (and possibly the HI loop, see below) are
required to facilitate these functions. Conservation of this channel in HBoV suggests that it
performs analogous functions. Notably, HBoV appears to have the largest surface opening of the
parvovirus structures determined to date (Figure A-3).
The canyon depression in HBoV is less pronounced compared to that seen in all other
parvovirus structures except for GmDNV, which lacks that feature (Figure A-3). In HBoV the
HI loop lies in the canyon and is structurally most similar to the HI loop of AAV2, and although
shorter in HBoV, generates a similar surface topology for the two viruses (Figure A-3 (red loop
features on the canyon floor), Figure A-5D, and A-6B). The HI loop conformation differs
significantly in B19 because it projects radially from the capsid surface and lies very close to the
DE loop to form a continuous, raised surface in the canyon (Figure A-3). This B19 HI loop
conformation is similar to that proposed to occur in AAV2 following receptor attachment (Levy
et al., 2009). The HI loops in MVM and AMDV, which also lie on the canyon floor, are
structurally similar to each other but differ from HBoV (Figure A-6B). In GmDNV the HI loop
lies very close to the DE loop, and together they “fill in” the canyon floor (Fig. 3 and 6B). The
HI loop which has been proposed to undergo structural re-arrangements in AAV2 to facilitate
VP1 externalization upon receptor binding has also been reported to play a role in capsid
assembly (DiPrimio et al., 2008; Levy et al., 2009).
Summary
The structure of the HBoV capsid presented here represents the first for a member of the
newly established Bocavirus genus of the Parvoviridae. The sub-nanometer resolution obtained
in the cryo-reconstruction clearly reveals that the newly identified HBoV assembles a capsid that
retains the highly conserved structural core adopted by all other parvoviruses, and does so
150
despite low sequence similarity between its major capsid protein and those of the other viruses
which have existed for many years. Conservation of core structural features in the major coat
protein, including the eight-stranded β-barrel, the αA helix, the DE loop, and the HI loop, is
consistent with their contributing essential functions in capsid assembly and stability.
Conservation of the β-cylinder channel formed by the DE loop strongly suggests that this
channel is essential for productive HBoV infection as it likely serves as a portal for VP1u
externalization and DNA packaging as reported for other parvoviruses.
Overall, HBoV shares most of its capsid surface features with the two other human
parvoviruses (B19 and AAV2) in this comparative study, and these result in a unique capsid
morphology whose functional domains are yet to be verified. The resemblance of B19 and
AAV2 capsids at the three-fold axes suggests that this region confers similar, receptor
attachment capability. On the other hand, the structural difference of the HBoV five-fold axis
from B19 (reported to have its VP1/VP2 localized on the capsid surface) and its similarity to
AAV2 (proposed to externalize its VP1u for its PLA2 function via the channel) suggests that the
HBoV capsid is required to undergo similar conformational transitions as AAV2 during cellular
trafficking. Variable HBoV VP2 surface loops, as built in the pseudo-atomic model based on
known parvovirus structures (Chapman and Agbandje-McKenna, 2006) and constrained by the
reconstructed density envelope, likely reflect adaptations related to host cell recognition and
evasion of the adaptive immune response. Significantly, the structure of HBoV provides an
additional framework for the molecular annotation of parvovirus VP structures, which have
evolved to facilitate infectivity through the utilization of structural elements and capsid surface
features that fulfill myriad functions.
151
Table A-1. Selected properties of representative members of Parvoviridae Genus Members Number of VPs Group Major VP Subfamily Parvovirinae – vertebrate hosts Parvovirus
MVM*, CPV*, FPV*
3
I
VP2
Erythrovirus
B19*†, SPV 2 III VP2
Dependovirus AAV2*, AAV4*, GPV
3 III VP3
Amdovirus
AMDV† 2 III VP2
Bocavirus HBoV†, BPV, CnMV
2 III VP2
Subfamily Densovirinae - invertebrate hosts Densovirus
GmDNV*, JcDNV†
4
II
VP4
Iteravirus BmDNV† 4-6 II VP1-4
Brevidensovirus AaeDNV, AalDNV
2 or 3 n/a VP1 or 2/3
Pefudensovirus PfDNV 5 n/a VP1 *Denotes structures determined by X-ray crystallography, †Denotes structures determined by cryo-EM, (n/a) denotes structural group not assigned
Table A-2. Sequence alignment statistics for representative members with structural data Sequence % Identity* % Strong
similarity* % Weak similarity*
% Different*
AAV2 23 17 14 46 B19 19 17 14 50 MVM 16 17 16 51 AMDV 14 17 13 56 GmDNV 12 13 11 64
*Percentage identity, similarity, and difference are as defined in Thompson et al., 1994.
152
Figure A-1. Characterization of recombinant HBoV VLPs. A) 10% SDS-PAGE of HBoV VLPs
(right lane) and molecular mass marker (left lane), showing that VLPs consist solely of VP2 (~61 kDa). B) Micrograph of negatively stained VLPs shows that most capsids exhibit a circular profile and fill with stain, consistent with them being empty, spherical shells. C) Micrograph of unstained, vitrified VLP sample highlights that these HBoV particles are intact, empty, and spherical.
153
Figure A-2. HBoV cryo-reconstruction. A) Stereo, shaded-surface representation of the entire capsid viewed along a two-fold direction. The black equilateral triangle bounded by two three-fold (3) axes (separated by a two-fold (2) axis) and a five-fold (5) axis identifies one asymmetric unit of the icosahedral capsid. B) Stereo, half particle image of (A) sectioned at the center and generated as a perspective view. C) Central cross-section of the reconstructed map, with highest and lowest density features depicted in black and white, respectively. Arrows highlight the two- (2), three- (3), and five-fold (5) axes that lie in the plane of the cross-section in one quadrant. Density “plugging” the five-fold channel at low radii is indicated by the short, black arrow for one channel. D) Enlarged view of upper left quadrant from (C) with circles drawn at radii corresponding to the radial density projections shown in Figure A-4.
154
Figure A-3. Comparison of six representative parvovirus capsid structures. Stereo, radially color-cued, surface representation of HBoV cryo-reconstruction shown with corresponding two-fold views of B19, AAV2, AMDV, MVM, and GmDNV highlight similarities and differences in the outer surfaces of the capsids. Density maps of the latter five virus capsids were computed from atomic (B19, AAV2, MVM, and GmDNV) or pseudo-atomic (AMDV) coordinates and all rendered at the same resolution (7.9 Å)
155
and magnification. Vertical bar depicts color cueing as a function of particle radius in Å.
Figure A-4. Radial density projections of six representative parvoviruses. The virus labels are as defined in Table 1. Density distributions at different radii (see Figure A-2D) illustrate similarities (inside) and differences (towards the outside) among the virus capsids. Contrast same as that depicted in Figure A-2C.
156
Figure A-5. Pseudo-atomic model of HBoV VP2. A) A Cα-trace of the VP2 pseudo-atomic model. Labels highlight the β-strands regions (BIDG-CHEF) that form the core β-barrel, the α- helix A (α-A), the DE and HI loops, and the nine variable surface regions (VR-1 to VR-IX). The approximate two-, three- and five-fold axis are shown by the numbers 2, 3, and 5, respectively. The dark blue, N-terminal portion of the model represents nine residues that were not included in the initial HBoV homology model, but that can be fit into the HBoV cryo-reconstruction. B) An equatorial slab of pseudo-atomic HBoV VP2 monomers docked into the reconstructed density map. Several icosahedral symmetry axes are labeled as in Figure A-2B. Regions outlined in red are enlarged in subsequent panels to further emphasize the fit of the model. C) Stereo view of a small portion of the VP2 model (in red) showing the structurally conserved, β–barrel core (β-strands A, B, I, D and G) and α-A helix modeled into the reconstructed density (grey mesh, contoured at a threshold of 3.3σ). D) Same as (C) for a view of the region of the capsid near the five-fold axis, showing the fit of five, symmetry-related, DE loop β-ribbon models that form the cylindrical channel and the fit of five HI loops that form the petal-like structures on the canyon floor. The five VP2 monomers are distinguished by color. E, left) Same as (D) but for a view from inside the capsid that highlights the channel “plug”. Residues 28-36 in the HBoV VP2 model are colored blue. E, right) Same as (E, left) but viewed from the side to show the model of one monomer and the DE-loops of four symmetry-related monomers at the pore of the five-fold channel. N-terminal residues 28-36 (in blue) were modeled as extending up into the five-fold channel. Color scheme same as in (D).
157
Figure A-5. Continued
158
Figure A-6. Superposition of VP2 structures from six representative members of the Parvoviridae. The orientation of the VP monomer is relative to the exterior and interior of the capsid is indicated by labels. A) Enlarged view of the conserved β-barrel motif (BIDG and CHEF sheets), strand βA, and helix α-A. Structural elements are labeled as in Figure A-5A. B) Enlarged view of the HI loops
159
APPENDIX B MONOCLONAL ANTIBODY AGAINST THE ADENO-ASSOCIATED VIRUS TYPE 8 CAPSID AND THE IDENTIFICATION OF CAPSID DOMAINS INVOLVED IN THE
NEUTRALIZATION OF THE AAV8 INFECTION
Introduction
As previously mentioned, the AAVs have become a useful tool in gene transfer. Several
serotypes have been tested for use as vectors in gene therapy, with AAV2 being the best
characterized and most tested (Rabinowitz et al., 1999; Nicklin et al., 2001; Wu et al., 2000;
Bartlett et al., 1998; Fischer et al., 2003; Simonelli et al., 2009; Wobus et al., 2000; Opie et al.,
2003; Moskalenko et al., 2000; Flotte, 2005). AAV vectors based on other serotypes show
transduction efficiency in various tissues: AAV1 in muscle, pancreatic islets, heart, vascular
endothelium, brain and central nervous system (CNS), and liver; AAV3 in Cochlear inner hair
cells; AAV4 in brain; AAV5 in brain and CNS, lung, eye, arthritic joints, and liver; AAV6 in
muscle, heart, and airway epithelium; AAV7 in muscle; and AAV8 in muscle, pancreas, heart,
and liver (Jiang et al., 2006). Most vectors are broad in their tropism, but many have specific
points of localized infectivity that make them more efficient for one tissue over the other.
Interstingly, i.v. and intraperitoneal (i.p.) injection of AAV8-based vectors have been found to
cross the blood vessel barrier and deliver transgene efficiently to striated muscles where as
AAVs 1 and 6 have been shown to better deliver localized responses (Wang et al., 2003). Also,
in studies to identify the best vector for transgene expression in specific tissues the AAV8
serotype repeatedly out-transduces other serotype-based vectors in the mouse liver (Gao et al.,
2002; Jiang et al., 2006; Moscioni et al., 2006; Nakai et al., 2005; Davidoff et al., 2005; Sarker et
al., 2004; Conlon et al., 2005; Sun et al., 2005). Concordantly, the first study to successfully
correct hemophilia A in mice used an AAV8 vector and demonstrated the same phenomenon
independent of vector cassette or injection route (Sarkar et al., 2004). Studies comparing AAV2,
160
6, and 8 vectors in the liver have identified that AAV8 can efficiently transduce liver cells due to
rapid uncoating of the viral capsid, thus allowing for a constant presence of ssDNA available for
annealing and dsDNA production (Thomas et al., 2004). This phenomenon allows for more than
10% transduction of mouse liver cells compared to AAV2 vectors (Thomas et al., 2004).
Although AAV2/8 vectors have been successful in the correction or reduction of pathology in
many murine disease models, there are many differences noted in the efficiency of treatment and
gene expression. One study successfully transduced adult mouse livers resulting in a sustained
correction of a metabolic disease, but noted that juvenile livers, still growing, had markedly
decreased gene expression over time (Cunningham et al., 2008, 2009). A follow up study
identified that re-administration was necessary, but difficult due to an anti-capsid humoral
condition and further concludes that efficient vector design and re-administration techniques will
be necessary (Cunningham et al., 2009). Although, the fold increase of transformed-hepatic cells
with AAV8-based vectors seen in the mouse is yet to be repeated in larger animal models
(Murphy et al., 2009), recent studies in hemophilia dog models have demonstrated AAV-
mediated continuous expression of canine factor VIIa (cFVIIa) (Margaritis et al., 2009). These
results provide the first evidence for efficacy and safety of the gene-based FVIII/FIX bypassing
agent FVIIa for the treatment of hemophilia in a large animal model using rAAV8 vectors
(Margaritis et al., 2009). However, antibody responses to the capsid have been shown to
efficiently block gene transfer in large animal models (Wang et al., 2010; Davidoff et al., 2005).
The same has held true in human patients (Manno et al., 2006; Li et al., 2009). Indeed, lower
transgene expression was primarily linked to AAV-binding antibodies rather than to AAV
capsid–specific CD8+ T cells (Manno et al., 2006).
161
Few animal models exist where pre-existing antibody responses can be studied that will
relate to the human conditions. Many studies use passive immunity models where animals are
injected with human-derived immunoglobulin, often termed intravenous immunoglobulin (IVIG)
(Scallan et al., 2006; Wang et al., 2010). Pre-existing immune response studies in mice using
AAV2 and 8 vectors indeed support that even non-neutralizing antibodies can have a negative
effect on gene transfer stability, as well as an interesting find that mice immunized with
AAV8cap vector and subsequently challenged with an AAV2-gene vector induced cross-reactive
antibodies to AAV2, but the converse was not found in the opposite studies (Li et al., 2009).
However, when compared to rh32.33 vectors, AAV8 generated a lower immune response to the
viral vector allowing for longer sustainment of the transduced gene (Mays and Wilson, 2009).
The use of alternate serotypes to bypass pre-existing immune responses (Halberd et al., 2006; De
et al., 2006; Wang et al., 2005) has drawn much attention as the list of naturally and engineered
capsids continue to expand (Gao et al., 2004; Koerber et al., 2008; Excoffon., 2009; Choi et al.,
2005). These new vectors utilize variability among the surface loops as a general means of
antibody escape and retargeting. It is a well-known fact that the human population already has
high neutralizing antibody titers to the more common AAV, serotype 2 (Blacklow et al., 1968,
1971; Chirmule et al., 1999; Erles et al., 1999; Calcedo et al., 2009; Boutin et al., 2010). A more
comprehensive study looking at serum samples from around the world, implicated that titers to
the non-human primate isolates AAV7 and 8 were relatively low compared to AAV2 and 1,
while rh32.33, an engineered vector from primate tissue, was rarely detected (Calcedo et al.,
2009). Of note, a more recent study identified that AAV8 seroprevelence was the lowest, less
than 40%, when compared to AAV1,2,5,6, and 9 (Boutin et al., 2010). These vectors may
therefore be useful in clinical trials with patients who have already been exposed to other
162
serotypes (Calcedo et al., 2009; Boutin et al., 2010). However, given that AAV-reactive
antibodies are present in a large portion of the population (Moskalenko et al., 2000), combined
with the fact that a high degree of homology exists among the serotypes, it may become
increasing difficult to overcome pre-existing immunity (Lin and Ertl, 2009). Indeed, it has been
noted that patients who have titerable antibodies to AAV1 will generally have antibodies to the
rest of the serotypes currently being studied (Dr. Luk Vandenberghe, personal communication).
This underlines the increasing importance of understanding the antigenic nature of the AAV
vector in context with the patient response.
A better understanding of these regions will aid in engineering or choosing vectors that
have specific properties for use with a certain disease or special circumstances that arise among
the different patients. This study specifically focuses on identifying regions on the AAV8 capsid
surface that are recognized by a monoclonal antibody, ADK8. In order to identify regions
involved in the antigen-antibody interaction, 3D reconstructions were generated from cryoEM
data of the AAV8:ADK8 complex (see chapter 2). Possible sites of antibody recognition were
identified on the capsid surface and mutational analysis was then performed. These vectors were
tested in vivo and in vitro for their inherent ability to evade pre-existing responses while
retaining the transduction efficiency previously noted for the first generation AAV8 vector.
Materials and Methods
Animals, Cell Lines and Cell Culture
Six-to-nine weeks old, female NMRI mice were used for all experiments. Mice were
purchased from Charles River Wiga (Sulzfeld, Germany). Mice were kept according to the
guidelines of the German Cancer Research Center. 293T (Pear et al., 1993), HeLa cells
(laboratory stock) and HepG2 (Knowles et al., 1980) cells were maintained in Dulbecco’s
Modified Eagles Medium (DMEM) supplemented with 10% heat inactivated fetal-calf serum
163
and 100 U/ml penicillin and 100 µg streptomycin at 37°C in 5% CO2. For transfection, a
confluency of 70% of 293T cells is needed. For binding assays, a 90 % confluency of Hep G2
cells should be ensured. X63/Ag8 lymphoma cells (kindly provided by P.Krammer, DKFZ) were
cultured in RPMI 1640 supplemented with 10% heat-inactivated fetal-calf serum, 100 U/ml
penicillin, 100 µg/ml streptomycin, 2 mM L-Glutamin, 20 mM Hepes pH 7.2, at 37°C, 5% CO2.
Plasmids and Site-Directed Mutagenesis
Vector plasmid pTRUF2CMV-Luc is a recombinant AAV2 plasmid expressing a
luciferase reporter gene from a cytomegalovirus (CMV) promoter and is flanked by inverted
terminal repeats (Zolotukhin et al., 1996). Plasmid pDGΔVP, expresses all essential Ad helper
proteins. It is based on plasmid pDG but prevents expression of VPs (Grimm et al., 1998). The
p5E18-VD2/8 helper construct (Gao et al., 2002) supports the synthesis of AAV2 Rep proteins
and AAV8 VPs, pBSΔTR18 provides AAV2 Rep and Vp proteins (Weger et al., 1997). Both
plasmids served as templates for site-directed mutagenesis reactions. Mutagenesis was
performed using the QuickChange site-directed mutagenesis kit (Stratagene, Amsterdam, The
Netherlands) according to the manufacturer’s manual. For each mutant two complementary PCR
primers were designed containing the substitution, flanked by 10 to 15 homologous base pairs on
each site of the mutation. Three sets of primer were produced and inserted into the cap gene of
P5E18-VD2/8: forward primer 5´GGAGGCACGACAACTCAGTC GACCCTCTGGGC 3´,
reverse primer 5´GCCCAGAGTCGACTGAGTTGTCGTGCCTCC 3´
(455GTANTQ→GTTTQS460); forward primer 5´GCTATTGTTGTTGTCCGCGCTTGT
CTTTGAGACGCG 3´, reverse primer 5´CGCGTCTCAAAGACAAGCGCGGACAACAAC
AA TAGC 3´ (493TTTGQ→KTSAD497); forward primer 5´ GCAGATAACTTGCAGC
GGGGAAACAGGGCTCCTC 3´, reverse primer 5´GAGGAGCCCTGTTTCCCCGCTGCA
AGTTATCTGC 3´ (586LQQQNT→LQRGNR591). Two sets of primer were produced and
164
inserted into the cap gene of pBSΔTR18: forward primer 5́CCAAGTGG
AACCACCAATACGCAAACTCTTGGGTTTTCTCAGGC 3´, reverse primer 5`GCCTG
AGAAAACCCAAGAGTTTGCGTATTGGTGGTTCCACTTGG 3´ (457QSRLQ→TQ TLG
461); forward primer 5´CCACCTCCAGCAACAAAACACAGCGGCAGCTACCGC 3´, reverse
primer 5´GCGGTAGCTGCCGCTGTGTTTTGTTGCTGGAGGTTGG3´ (585RGNR
Q→QQNTA589). The complete fragments were sequenced to exclude additional mutations.
Plasmids and Peptide Insertions
To insert a peptide into the AAV8 VPs, a SFII binding site had to be constructed (Muller,
Kaul et al. 2003) and a fragment of 743 bp was synthesized (GeneArt, Regensburg, Germany). It
was cloned with XcmI 1277 - EcoRV 2020 into the plasmid P5E18-VD2/8. Four peptides were
chosen for insertion into the SFII binding site: PSVPRPP, VNSTRLP, GQHPRPG (Ying Y. et
al., paper under revision) and ASSLNIA (Yang et al., 2009). The fragments containing the
insertion sites were sequenced for verification.
Transfection of 293T Cells and Vector Particle Purification
A triple transfection was carried out by calcium phosphate precipitation. For each mutant,
twenty plates with 5E+6 cells each, were seeded out and transfected 24 h later. Per plate, 50 – 60
µg DNA was resuspended in 1.125ml sterile Braun H20, mixed with 125 µl CaCl2 (Sigma, St.
Louis, USA) and added slowly to 1.25 ml of 2 x HBSS (280 mM NaCl, 50 mM Hepes, 1.5mM
Na2HPO4, 10 mM KCL, 12 mM glucose, pH 7.05), shaking constantly. After 1 min of
incubation, the mix was added to 7.5 ml medium and append to the cells. After 48 h at 37°C, 5%
CO2, cells were harvested, washed with PBS, centrifuged at 200x g for 15min and stored at -
80°C until further purification steps took place.
165
Purification and Titration of Mutant AAV Stocks
15 ml of a lysis buffer (150 mM NaCl, 50 mM Tris-HCl, pH8.5) was added to the cell
pellets, followed by 5 rounds of freeze-thawing (N2 and 37°C), 30 min of Benzonase (50 U per
ml) treatment at 37°C and 15 min at 3000 x g . Via an iodixanol-step-gradient ( 40 ml Beckman
Quick Seal Tube), AAV particles were purified from the crude cell lysate (Zolotukhin et al.,
1999). Sealed gradients were centrifuged at 50,000 x g at 4°C for 2 h. Viral particles were
aspirated from the 40 % iodixanol phase and deep-frozen at 20°C. Quantitation of AAV genomic
particles was achieved by quantitative real-time PCR adapted from previously described methods
(Veldwijk et al., 2002). After the alkaline lysis of the AAV particles, genomes were subjected to
Taqman Universal Master mix including primer (for-5´TGCCCAGTACATGAC CTTATGG 3´,
rev- 5´GAAATCCCCGTGAGTCAAACC 3´), probe (6-fam-AGTCATCGC TATTACCATGG-
MGB) and analysed under standard QR-PCR conditions (Applied Biosystems Inc, Foster City,
USA).
Analysis of Viral Protein Expression
Western blot analysis was performed with equal amounts of genome containing particles
according to standard methods (Harlow and Lane, 1988). Monoclonal antibodies were applied as
previously described (Wistuba et al., 1995).
Production and Purification of Monoclonal Antibody ADK8
After fusion (Wobus et al., 2000) and screening of hybridoma supernatants (Kuck et al.,
2007), a monoclonal antibody was identified and its hybridoma cells were cultivated with
increasing medium amounts (maximum 1.5 ml) in expanded surface roller bottles (Sigma-
Aldrich, St. Louis, USA) for 12 days. Cells were centrifuged at 4000 x g, 10min, and supernatant
was collected. For preservation, 0.01% thimerosal was added to the supernatant. A monoclonal
antibody was purified from the hybridoma supernatant by affinity chromatography using a
166
protein G sepharose column (GE Healthcare Europe GmbH, Munich, Germany). Hybridoma
culture was filtered through a 0.45µm filter and applied onto the column at room temperature
overnight. Wash column with PBS and elute the antibody with 10 x 1ml Na azide (0.1 M)/NaCl
(0.15 M), pH 4.4. The eluted antibody was neutralized with 5% Tris pH 9.5. The total antibody
amount was determined (Nanodrop ND-1000, Peqlab Biotechnology, Erlangen, Germany) at an
OD of 280nm.
Determination of Antibody Isotypes
Subclasses of antibodies were determined by using the Amersham mouse antibody
isotyping kit (RPN29, Braunschweig, Germany) according to manufactuere’s guidelines.
AAV Capture ELISA
ELISA was carried out using ADK8 (50ng per well) coated flexible microtiter plates
(Becton Dickinson, Heidelberg, Germany). Particles of rAAV8 for the standard were determined
by particle calculation of negative staining from 10 randomly taken pictures on an electron
microscope. 1E+10 genome containing particles were applied and serial diluted. HRP-labelled
(ApD Serotech, Oxford, UK) ADK8 (1µg/ml) antibodies were applied and analysed by an
ELISA reader (Ascent FL, Thermo Labsystems, Egelsbach, Germany).
In Vitro and In Vivo Neutralization
HepG2 cells were seeded out on a 96-well plate (NuncTM, Rochester, USA) 24 h in
advance. Monoclonal antibodies were preincubated with vector particles at 37°C for 30 min.
Antibody particle mix was added to FCS-free medium of HepG2 cells. After 4 h, medium was
aspirated, cells were washed with PBS and DMEM+FCS was added. After 72 h medium was
aspirated, cells were washed with PBS once again and 1x RLB (Reporter lysis buffer, Promega
GmbH, Mannheim, Germany) was applied. Cell extracts were stored at -80°C until expression
analysis was performed.
167
Mice were injected intraperitoneal with 50 or 500 µg monoclonal antibodies 4 h prior to
intravenous injection of the viral particles. 1E+11 viral genome containing particles were
injected into each mouse. 14 days after injection animals were analysed by the IVIS Imaging
software ( Xenogen, Alameda, USA) for 5 min, 10 min after D-Luciferin (30 mg/ml) (Synchem
OHG, Altenburg, Germany) injection. Mice were sacrificed, liver and heart were extracted and a
luciferase expression analysis was performed. The amount of protein was determined with a
NanoOrange protein quantitation kit (Invitrogen, Karlsruhe, Germany).
Binding Analysis by DNA Dot Blot
For the DNA binding analysis, HeLa cells were seeded out on 6-well plates 24 h before
infection. Viral particles were pre-incubated with 50 µg or 250 µg of antibody at 37°C for 30
min. After incubation, virus antibody mix was shifted to 4°C before being applied onto the cells
for 30 min. Cells were washed with medium and either harvested directly or transferred to 37°C
for 2 h and harvested. Without a temperature shift, cells were centrifuged at 200 x g for 10 min
and medium aspirated. The latter were centrifuged, medium aspirated, 100 µl of pre-warmed
trypsin (0.05%) added for 5 min, centrifuged again and medium carefully removed. Pellets were
stored at -20°C. After Nuclease (1mg/ml) and Proteinase K (10 mg/ml) treatment (Roche,
Pensberg, Germany), phenol-chloroform extraction was performed, DNA was spotted onto a
nylon membrane (Gene ScreenTM, DuPont , Boston, USA)in a dot blot chamber, 10 min
denaturated (1.5 M NaCl, 0.5 M NaOH), 10 min neutralized (0.5 M Tris-HCL pH 7.0, 0.3 M
Tris-Sodium-Citrate, 3M NaCl) and by UV-cross-linking at 1200 Joule. A radioactive probe,
directed against CMV was produced with standard protocol of the DNA Labeling kit (Roche,
Pensberg, Germany) and 5µl [α-32P]dCTP (50 μCi). On the nylon membrane immobilized
template DNA was pre-hybridized in 15 ml hybridization buffer (125 mM Na2HPO4, 250 mM
NaCL, 1 mMEDTA, 45% formamide, 7% SDS), probe was added and DNA was labelled in a
168
hybridization oven with rotation at 42°C overnight. Membrane was washed 6 x 5 min with wash
buffer I (2 x SSC, 0.1% SDS) at 42°C and 3 x 20 min with wash buffer II (0.1 x SSC, 0.1% SDS)
at 68 °C. Membrane was exposed to an X-ray film (Kodak BioMax MS, Sigma, St. Louise,
USA).
Immune Fluorescence
HeLa cells were seeded out on cover slides in 24-well plates. Viral particles were pre-
incubated with 0.1 mg antibody ADK8 or IVA7 (own stock) at 37°C for 30 min then on ice for
10 min. Virus-antibody mix was added to pre-cooled cell slides and kept at 4°C for 30 min. Cells
were washed with PBS and directly fixed with ice-cold methanol for 10 min. After fixation,
ADK8 hybridoma culture was applied overnight, then washed 3 x 10 min with PBS, kept 1 h at
room temperature in the dark after secondary antibody (Dianova GmbH, Hamburg, Germany)
goat-anti-mouse Alexa-Fluor 488, diluted 1:700 in PBS 1% BSA, was applied. Finally cell slides
were washed twice with PBS and embedded in Permafluor mounting medium. Examination was
carried out with a fluorescence microscope (Leica DMRD).
Native Dot Blot Assay
Viral genome containing particles per mutant vector were incubated at 37°C for 30 min
either without or with the monoclonal antibody ADK8 (2 mg/ml), then 10 min on ice. Next,
samples were subjected to a temperature treatment, 37°C, 65°C, 75°C or 99°C for 5 min.
Samples were spotted on a nitrocellulose membrane (Schleicher&Schuell, Dassel, Germany) in
a dot blot chamber. Membrane was blocked in PBS/10% skimmed milk powder and incubated
with A1 antibody, which had been linked to HRP (Zenon labelling Kit, Invitrogen). Particle
visualization was performed using an enhanced chemiluminescence detection kit (Perkin-Elmer,
Boston, USA)
169
Epitope Mapping
See above section in Chapter 2.
Results and Discussion
Three-Dimensional Reconstruction and Model Docking of the AAV8:ADK8 Complex
A total of 982 particles were extracted from 78 CCD images to generate a final 3D map
with a resolution of ~18.4Å (Figure B-1A). Independent density was present at each three-fold
protrusion, indicating that there was no steric hindrance for the binding of all three protrusions
simultaneously, as was seen for a previous reconstruction, AAV2:C37B (Figure 4-2). An entire
capsid molecule, or 60-mer, was generated from the AAV8 atomic coordinates (PDB 2QA0;
Nam et al., 2007) and was docked into the respective capsid density with a correlation coefficient
(cc) of ~0.42. The cc of the fit was most likely low due to the presence of the excess density
belonging to the bound Fabs as well as the use of a C-α model (Figure B-1B). Since the ADK8
antibody has yet to be sequenced, a model has not been generated. The atomic coordinates for a
generic Fab model (PDB 2FBJ; Suh et al., 1986) were then docked into the identified Fab density
and a density map was generated from this new model (Figure B-1C). A new cc was calculated
(see chapter 2) and was found to be ~0.68. The correlation was improved when the excess
density was filled, but this low correlation is most likely due to the use of C-α model, as
previously mentioned, in addition to the use of rigid-body and manual fitting techniques. In this
case, these functions do not refine or take possible low-energy conformational movements into
account, causing portions of the atomic structures to be outside of the density in order to better fit
more important regions. This is evident from a comparison of the maps used in the determination
of the cc between the cryoEM reconstruction and the fitted models (Figure B-1D).
170
Analysis of the 3D Models Identifies Proposed Monoclonal Antibody ADK8 Binding Sites on the AAV8 Capsid
Docking atomic models into the cryoEM map allowed for further characterization of the
ADK8 binding site on the AAV8 capsid. First glance identifies the Fab binding to the protrusions
at the three-fold axis of symmetry, leaving the two- and five-fold axes un-occluded. A more
detailed inspection of the docked VP3 capsid monomers identifies VRIV, V, and VIII (Figure 4-
3) of the three-fold axes to be prominent in the antibody interaction. A lesser, but possible
interaction was also noticed at VRVI corresponding to a KDDEE peptide at aa530-534. The
symmetry related manner of this region brings VRV into a position to be ‘cradled’ by VRs IV
and VIII from another symmetry related monomer. These regions were also noted in the
previously determined reconstructions (Table 4-4). Sequence identification of this region
proposes that the peptides GTANTQ (VRIV; aa455-460), TTTGQNNNS (VRV; aa493-501),
and LQQQNT (VRVIII; aa586-591) could be involved in the epitope. It is also interesting to
note that portions of the peptides from VRIV and V have been identified in studies as being
immunogenic in cytotoxic T-cell assays, while the peptide identified in VRVIII was found as an
epitope in MHC II molecules (Chen et al., 2006). Indeed, these regions also overlap with the
sequence based structural alignment from previously determined reconstructions and show
similar patterns. Mainly, the sequences at VRV and VIII both have triplicate repeats of
hydrophilic, polar residues, i.e. QQQ (Table 4-4).
Characterization of rAAV8- and rAAV2- Capsid Mutants to Elucidate the ADK8 Binding Epitope
Many monoclonal antibodies, such as B1, A1, A69 and A20, have been well studied and
characterized in the past (Wistuba et al., 1995; Wobus et al., 2000). The collection of generated
monoclonal antibodies has been rapidly increasing and has greatly contributed to a detailed
analysis of AAV2 (Grimm et al., 1999; Kern et al., 2003), the best studied serotype. The
171
monoclonal antibody ADK8 is of importance as it could help to further analyze the rhesus
monkey derived AAV8, which is of particular interest to gene therapy (Gao et al., 2002; Wang et
al., 2005).
To investigate possible binding epitopes of ADK8, we chose to construct mutants with
amino acid exchanges of AAV8 towards AAV2 and vice versa. Mutations in the cap gene were
chosen according to the aa identified in the previously described antibody mapping. However, a
homology alignment confirmed that only three of the four optional binding sites are localized in
non-homologous region between AAV 2 and AAV8 (Figure B-2). Therefore, binding site
KDDEE (aa sequence underlined in green) was excluded in the next set of experiments. A
reduced VP expression combined with an impaired capsid assembly has been observed for
several mutants in previous studies (Bleker et al., 2005). On that account, quantitative real-time
PCR and western blot analysis was assayed. In two independent productions, rAAV2 and
rAAV8 as well as the five mutants had been produced and purified. The results of the
quantitative real-time PCR indicate that all mutants, except mutant rAAV8→rAAV2
586LQRGNR591 showed no decrease in capsid assembly (Figure B-3A). The latter did show a
20-fold decrease, but VP expression was not impaired for any mutant (Figure B-3B). It is
interesting to note an extra band for the rAAV2rAAV8 493KTSAD497 mutant and a change
in the banding position for all three rAAV2rAAV8 vectors. It is quite possible that these small
changes at the capsid surface cause differences in glycosylation events or a change in the break-
down pattern of the capsid during denaturing. These results permitted an ADK8-ELISA to parse
mutant vectors for their monoclonal antibody binding ability. We decided to use 1E+10 viral
genome containing particles to check if the assay could be used to detect the amount of mutant
capsids. In case of the first set of produced mutants (aa exchanged from rAAV8→ rAAV2), the
172
capsids of two mutant vectors, rAAV8→ rAAV2 455GTTTQS460 and rAAV8→ rAAV2
493KTSAD497, could be detected in range of rAAV8. Ten-fold more capsids were detected than
viral genome containing particles had been applied (Figure B-4A). This finding is conform with
the literature, a 10- to 100- fold higher amount of capsids is normally found after the purification
with a single iodixanol step-gradient (Grimm et al., 1999). In contrast to the previously described
mutants, the third mutant vector, rAAV8→ rAAV2 586LQNRGNR591, could not be detected
with the ADK8-ELISA. It indicates that the binding site 586-591 LQQQNT is of utmost
importance in ADK8 binding to rAAV8. This finding was further substantiated by a second set
of produced mutants (aa exchange from rAAV2→ rAAV8). Whereas the mutant vector
rAAV2→ rAAV8 457TQTLG461 could not be detected by the ADK-ELISA, just like rAAV2,
the mutant vector rAAV2→ rAAV8 585QQNTA589, was detected (Figure B-4B). If a seven aa
peptide was inserted at position 590 of rAAV8, ADK8 ELISA was not able to detect any capsids
(Figure B-4C). The motifs had been chosen either from an AAV2 in vitro and in vivo heart tissue
selection (Ying et al. paper under revision) and a muscle-targeting peptide (Yu et al., 2009).
Taken together, the experiments could demonstrate that of all possible binding sites, the position
586-591 LQQQNT is directly involved in ADK8 binding to AAV8. Due to a mutation from
AAV8 to AAV2 at this position, ADK8 was not able to recognize the AAV8 mutant anymore.
Additionally, AAV2 mutated to AAV8 at the given locus, allowed mutant vector capsid
detection in the ADK8-ELISA. Insertion of a 7 aa peptide motif in the binding site at aa 590
caused a complete loss of capsid recognition in the ADK8-ELISA.
In Vitro and In Vivo Neutralization of AAV8 by ADK8
It has been revealed that the seroprevalence of AAV8 is lower than AAV1 and AAV2 in
ten different countries on four different continents (Calcedo et al., 2009). Due to the fact that
AAV2 directed antibodies dominate the global seroprevalence, the AAV8 serotype is more
173
feasible for the avoidance of antibody-mediated AAV vector neutralization in clinical
application.
To gain insight into the basis of neutralization, it is of importance to characterize
neutralization competency of the derived monoclonal antibody ADK8. We decided to investigate
if ADK8 impacts transduction efficiency in vitro and in vivo. At first, three different vectors,
rAAV2, rAAV8 and mutant rAAV2→ rAAV8 585QQNTA589 were pre-incubated either
without, with 250 ng or 500 ng of monoclonal antibodies (ADK8 or A20). ADK8 was able to
decrease transduction efficiency of AAV8 as well as of mutant rAAV2→ rAAV8
585QQNTA589 by more than 50-fold compared to vector transduction without antibody addition
(Figure B-5A). Furthermore, neutralization was already found if less antibody (250 ng) ADK8
had been applied. However, ADK8 was not able to neutralize AAV2 as this sequence is not
identical in AAV2. As a second control, vectors were pre-incubated with A20, a monoclonal
antibody for AAV2. AAV2 transduction was almost completely neutralized after A20 pre-
incubation. Conversely, A20 did not have any influence on AAV8 transduction, indicating that
the sequence responsible for A20 binding to AAV2 is not found in AAV8. This further implies
that the A20 epitope is most likely in a variable region where AAV2 and 8 differ. Transduction
efficiency for mutant rAAV2→ rAAV8 585QQNTA589 was slightly reduced. Moreover, ADK8
neutralization was also analyzed in vivo. Four groups of 4 (6-9 week old, female NMRI) mice
were chosen for this experiment. As a control, mice were injected with rAAV8 without any
antibodies. The other three groups were i.p. injected with ADK8 (50 µg and 250 µg) or with
ADK4 (250 µg) and 4 h later with rAAV8. Two weeks later, luciferase expression was measured
with an IVIS imaging system. The imaging ascertained that already the amount of 50 µg ADK8
was enough to achieve a complete neutralization of rAAV8 vector transduction (Figure B-5B).
174
Without i.p. antibody injection, transduction efficiency had a typical imaging pattern of a
hepatotropic AAV vector (Sarkar et al., 2004). The results of the fourth group of mice which had
been injected with ADK4, a monoclonal antibody against AAV4 only, clearly demonstrated that
a monoclonal antibody has to bind rAAV8 for neutralization. Antibodies present in the mouse
system which are not able to bind AAV8, are rapidly cleared from the system and do not impact
reporter expression. Results were verified by reporter expression analysis of liver and muscle
after dissection of the animals (Figure B-5C). As already observed in the imaging, luciferase
reporter expression was reduced more than 100-fold in liver and heart after the 50 µg ADK8 i.p.
injections. After a 250 µg ADK8 i.p. injection, RLU were even below the detection limit.
Injection of ADK4 did not have any impact on luciferase reporter expression. Similar
transduction efficiency was confirmed as for rAAV8 transduction without antibody injection.
The obtained results are in agreement with the fact that ADK8 is a monoclonal antibody
which neutralizes rAAV8 vector and mutant rAAV2 (with inserted binding epitope of ADK8,
585QQNTA589) transduction in vitro and in vivo. Therefore, this monoclonal antibody can be
used to elucidate the basis of neutralization of rAAV8.
Neutralization Analysis of ADK8
Clear attempts have been made to understand the molecular mechanisms of novel
hepatotropic AAVs (Thomas et al., 2004). However, the transduction pathway itself has not
been analyzed as a whole yet. ADK8 is known to neutralize AAV8 in vitro and in vivo, it is of
interest to ask if neutralization occurs at cell-surface binding, endosomal release, or uncoating.
To identify the point of ADK 8 neutralization, a series of experiments was designed.
Binding of AAV8 in presence of ADK8 was analysed by immune fluorescence and DNA dot
blot. In the past it has been demonstrated that monoclonal antibodies such as C37B or C24-B
(Wobus et al., 2000) prevent AAV2 binding. This does not appear to be the case for ADK8.
175
Neither in presence of IVA7 nor in presence of ADK8, could a reduction in rAAV8 vector
binding to HeLa cells be demonstrated (Figure B-6A), which was also consistent with the DNA
dot blot analysis (Figure B-6B). The same amounts of vector genomes of bound particles could
be identified with and without ADK8 pre-incubation. Interestingly, the addition of ADK4
improved rAAV8 vector binding to the HepG2 cells, but not for rAAV2. To investigate if all
bound particles can enter, infected cells were incubated in 0.05% trypsin to deplete defective or
un-internalized particles from cell surfaces. For rAAV8 and mutant rAAV2 (with inserted
binding epitope of ADK8, 585QQNTA589), less particles could be detected after trypsin
treatment, but not rAAV2. Additionally, pre-incubation of ADK8 showed a reduction in detected
vector genomes after trypsin treatment. However, spots are very low in contrast and needed
confirmation by the analysis of the spots with the plot lane analysis software ImageJ (Figure B-
6C and D). According to the pixel intensity, AAV2 binding and internalization is not hampered
but is even improved by the addition of ADK8 (about 15% increase in pixel intensity). In case of
rAAV8 and mutant rAAV2, ADK8 does not impact vector binding, but viral entry. About 15%
less viral genomes for AAV8 and the AAV2 mutant could be detected after trypsin treatment.
ADK8 pre-incubation caused an additional decrease of detectable viral genomes, the pixel
intensity was 20% reduced in these spots compared to vector genomes without previous antibody
treatment.
It has been observed previously that the A20 antibody has a direct impact on N-terminus
externalization (data not shown). The exposure of the VP1 N termini was conducted by using a
monoclonal antibody A1. A1 is known to bind a defined epitope close to the PLA2 domain on
the N terminus of VP1 (Wobus et al., 2000). Figure B-7 shows that after the incubation of virions
for 5 min at 37°C, no reaction upon A1 supernatant application to the nitrocellulose membrane
176
could be detected. Detection was found after 5 min treatment at 65°C for rAAV2 and mutant
rAAV2, but not for rAAV8. N termini became accessible at this point. At 71°C and 99°C, signal
detection was shown for all analysed vectors after A1 supernatant was applied. Due to the fact
that the secondary antibody GαM-HRP recognizes A1 supernatant and ADK8 which is used in
the preincubation step, the experiment had to be repeated with purified monoclonal A1 directly
coupled to HR-peroxidase. The results were similar, although the N-terminus was already
partially accessible for rAAV8 at 65°C. If rAAV8 particles were pre-incubated with ADK8, the
N-terminus became even better accessible at 65°C. Reduction of detection signal due to ADK8
pre-treatment was not observed.
Taken together, these results indicate, that ADK8 does not influence AAV8 binding to the
cell but has a slight impact on endocytosis. Additionally, there is no direct or indirect influence
on the N-terminus externalization. This may lead to the hypothesis that ADK8 prevents
uncoating and neutralizes reporter expression of vectors below detection limit if antibody
binding site is present on the capsids.
Impact on rAAV8 vector generation
Pre-existing immune reponses have proven to be a challenge in the field of gene therapy.
Current studies involved in circumventing these responses include shuffling of the cap gene to
produce chimeric vectors and generating wild-type capsids under immunoselective strain, as well
as identifying natural AAVs from other animal species. These methods create highly diversified
AAVs with capsids that incorporate different VRs that often mirror those seen in already
identified genotypes, yet also yield radically different phenotypes. Preliminary charcterization of
many of these vectors have identified poor capsid production, as well as reduced packaging and
altered infectivity profiles, which generally leads to a limited number of viable vectors. Studies
aimed at determining functional regions on the capsid surface have identified VRs that are
177
necessary for structural integrity, receptor binding, infectivity, and antigenicity. All of these
types of studies add to the basic biology and further characterization of the capsid surface as
human engineering of the capsid component is only fruitful when applied appropriately in a
manner that allows for the generation of vectors that mimic or exceed wt virus production and
infectivity values. The knowledge provided in the study presented here has identified VRs on the
AAV8 capsid surface that are specifically important in antigenicity and thus add to the basic
biology of the capsid surface. The VRs identified here will provide a basis where vector
development using mutational analysis in these regions may help generate antigenic escape
vectors.
178
Figure B-1. Three-dimensional reconstruction and model fitting for the AAV8:ADK8 complex. A) Final density map for the AAV8:ADK8 structure to ~21Å. The AAV8 viral surface is colored green while the excess density in white was identified as the Fab density. B) Fitting of the AAV8 (PDB 2QA0) structure into the respective capsid density. Density map is shown in grey and the atomic coordinates are in green. The excess white space remaining for the Fabs could easily account for the low fitting statistics. C) Fitting of both Fab and AAV8 capsid models into the reconstructed density. One Fab portion, modeled in as PDB 2FBJ, is shown in red and is hovering above three symmetry related monomers, colored in green, blue, and magenta, at the three-fold axis of symmetry. D) Visual comparison of the maps used to generate the final correlation of the docked models (green) and the original cryoEM reconstruction (grey). An excess of green density from the fit model map can be seen extending beyond the cryoEM map, indicating regions that would decrease the final corrolation. Black arrows mark example regions of density extension.
179
Figure B-2. Homology alignment of a portion of the G-H loop between AAV2 and AAV8. Amino acid differences are highlighted in blue and sequences of interest are underlined in black. The sequence underlined in green was found to have a possible interaction in the structure analysis, but is obviously conserved among the two viruses, indicating little or no involvement in the epitope.
Figure B-3. Analysis of rAAV2 and rAAV8 mutant vector production. A) A quantitative real-time analysis demonstrated viral genome containing particles for all produced mutants above the detection level (detection limit 5E+7Vg/ml). Compared to AAV2 and AAV8, none of the produced mutants showed a completely diminished packaging capacity. Production of viral genome containing particles was performed twice for each mutant and an average is shown. B) In a western blot analysis, VP proteins are analyzed with a monoclonal antibody (B1). VP1, 2 and 3 are marked with arrows, the stoichiometry of ~1:1:10 is evident.
180
Figure B-4. Binding ability of rAAV2 and 8 mutant capsids by the ADK8 monoclonal. A) Shows the reaction of the ADK8 antibody with capsids containing 1E+10 viral genomes in an AAV8 capsid ELISA. The effects on the mAb ADK8 detection due to the exchanges of aa 455 - 460 (GTANTQ → GTTTQS), aa 493 - 497 (TTTGQ → KTSAD), aa 586 - 591 (LQQQNT → LQRGNR) from rAAV8 to rAAV2 are demonstrated. B) Detection of AAV2 capsid mutants derived from the exchange of aa 457- 461 (QSRLG → TQTLG) and aa 585 - 589 (RGNRQ → QQNTA) in the AAV 2 capsid. C) Seven aa peptide insertion at position 590 into the rAAV8 capsid and disrupt sequence QNTA influence the mAb ADK8 detection. Mean standard deviations from four independent experiments are shown, an asterik indicates mutants for which capsids could not be detected in the ADK8 ELISA (detection limit, 1E+8 capsids/ml).
181
Figure B-5. in vitro and in vivo neutralization data A) Prior toHepG2 cell infection (MOI 5E+4), viral vectors were pre-incubated with different amounts (250 ng and 500 ng) of antibody for 30 min at 37°C. After 6 h, viral vector-antibody mix was washed off and cells were harvested two days after infection for the determination of reporter gene expression. B) Four hours after i.p. injection of ADK8 (50 µg and 250 µg) or ADK4, respectively, NMRI mice (female, 9 weeks old, n= 4) were i.v. injected with 1E+11 viral genome containing rAAV8 vectors. Two weeks after injection, mice were injected with D-luciferin, imaged for 5 min (10 min after D-luciferin injection) and sacrificed. C) Transgene expression per mg protein was determined for heart and liver of sacrificed mice with a luciferase reporter assay and NanoOrange. Dashed line indicates the detection limit of the luciferase reporter assay
182
Figure B-6. Neutralization analysis of the ADK8 mAb. A) Hela cells which had been seeded out on glas slides 24 h earlier, were infected with rAAV8 (MOI 1E+6). After viral vector preincubation with or without ADK8 (0.1 mg) for 30 min at 37°C,. infected Hela cells were kept at 4°C for 30 min. Next, glas slides were washed with PBS, then cells were fixed with ice cold methanol for 10 min and washed with PBS once again. As a negative control, ADK8 was added to the cells or no virus without antibody preincubation at all was applied. As a positive control, rAAV8 infected Hela cells without ADK8 preincubation or with IVA7 preincubation. After fixation, ADK8 hybridoma supernatant was applied for 24 h. Three washes with PBS were applied before GαM-A488 was added (1:700). After a last washing step, glas slides were embedded in Permafluor. B).Binding analysis was performed by DNA dot blot analysis after infection of HepG2 cells with rAAV8, rAAV2 and rAAV2 with aa exchanges at position (585RGNRQ → QQNTA589). At first, genomes containing particles (MOI5E+4) were preincubation with ADK8 or ADK4 for 30 min. Then, cells were infected for 30 min and washed with PBS. Cells were either directly harvested or kept at 37°C for 2 h before harvest and trypsin treatment. For all samples, DNA was extracted and transfered to a membrane for radioactive labelling. Positive control is the viral load without cell infection. The dot blot was performed three times. C) Analysis of spots by a plot lane analysis was performed using ImageJ software. Pixel intensity was determined for all spots containing viral genomes of bound vector particles to the cells. D) shows pixel intensities of viral genome containing particles which entered the cells at 37°C.
183
Figure B-6. Continued.
Figure B-7. ADK8 and its impact on the N terminus externalization in the post-entry process. AAV8 genome containing particles (1E+9) were incubated with or without ADK8 (1 μg /spot) for 30 min at 37°C. Thereafter, they were kept at either 37°C, 65°C, 71°C or 95°C for 5 min. Particles or particle-antibody mixes were spotted onto a nitrocellulose membrane and reacted with A1 hybridoma supernatant or HR peroxidase coupled A1 antibodies.
184
LIST OF REFERENCES
Aasted, B., Alexandersen, S., and Christensen, J. (1998). Vaccination with Aleutian mink disease parvovirus (AMDV) capsid proteins enhances disease, while vaccination with the major non-structural AMDV protein causes partial protection from disease. Vaccine 16, 1158-1165.
Aasted, B., Tierney, G. S., and Bloom, M. E. (1984). Analysis of the quantity of antiviral
antibodies from mink infected with different Aleutian disease virus strains. Scand. J. Immunol 19, 395-402.
Adrian, M., Dubochet, J., Lepault, J., and McDowall, A. W. (1984). Cryo-electron microscopy of
viruses. Nature 308, 32-36. Agbandje, M., McKenna, R., Rossmann, M. G., Strassheim, M. L., and Parrish, C. R. (1993).
Structure determination of feline panleukopenia virus empty particles. Proteins 16, 155-171.
Agbandje, M., Parrish, C. R., and Rossmann, M. G. (1995). The structure of parvoviruses. Semin
Virol 6, 299-309. Agbandje-McKenna, M., Llamas-Saiz, A. L., Wang, F., Tattersall, P., and Rossmann, M. G.
(1998). Functional implications of the structure of the murine parvovirus, minute virus of mice. Structure 6, 1369-1381.
Akache, B., Grimm, D., Pandey, K., Yant, S. R., Xu, H., and Kay, M. A. (2006). The 37/67-
kilodalton laminin receptor is a receptor for adeno-associated virus serotypes 8, 2, 3, and 9. J. Virol 80, 9831-9836.
Alkan, S. S. (2004). Monoclonal antibodies: the story of a discovery that revolutionized science
and medicine. Nat Rev Immunol 4, 153-156. Allander, T., Jartti, T., Gupta, S., Niesters, H. G. M., Lehtinen, P., Osterback, R., Vuorinen, T.,
Waris, M., Bjerkner, A., Tiveljung-Lindell, A., et al. (2007). Human bocavirus and acute wheezing in children. Clin. Infect. Dis 44, 904-910.
Allander, T., Tammi, M. T., Eriksson, M., Bjerkner, A., Tiveljung-Lindell, A., and Andersson,
B. (2005). Cloning of a human parvovirus by molecular screening of respiratory tract samples. Proc. Natl. Acad. Sci. U.S.A 102, 12891-12896.
Anderson, S., Momoeda, M., Kawase, M., Kajigaya, S., and Young, N. S. (1995). Peptides
derived from the unique region of B19 parvovirus minor capsid protein elicit neutralizing antibodies in rabbits. Virology 206, 626-632.
185
Arbetman, A. E., Lochrie, M., Zhou, S., Wellman, J., Scallan, C., Doroudchi, M. M., Randlev, B., Patarroyo-White, S., Liu, T., Smith, P., et al. (2005). Novel caprine adeno-associated virus (AAV) capsid (AAV-Go.1) is closely related to the primate AAV-5 and has unique tropism and neutralization properties. J. Virol 79, 15238-15245.
Arnold, J. C., Singh, K. K., Spector, S. A., and Sawyer, M. H. (2006). Human bocavirus:
prevalence and clinical spectrum at a children's hospital. Clin. Infect. Dis 43, 283-288. Arthur, J. L., Higgins, G. D., Davidson, G. P., Givney, R. C., and Ratcliff, R. M. (2009). A novel
bocavirus associated with acute gastroenteritis in Australian children. PLoS Pathog 5, e1000391.
Atchison, R. W., Casto, B. C., and Hammon, W. M. (1965). Adenovirus-Associated Defective
Virus Particles. Science 149, 754-755. Auricchio, A., Rivera, V. M., Clackson, T., O'Connor, E. E., Maguire, A. M., Tolentino, M. J.,
Bennett, J., and Wilson, J. M. (2002). Pharmacological regulation of protein expression from adeno-associated viral vectors in the eye. Mol. Ther 6, 238-242.
Azzouz, M., and Mazarakis, N. (2004). Non-primate EIAV-based lentiviral vectors as gene
delivery system for motor neuron diseases. Curr Gene Ther 4, 277-286. Baac, H., Hajós, J. P., Lee, J., Kim, D., Kim, S. J., and Shuler, M. L. (2006). Antibody-based
surface plasmon resonance detection of intact viral pathogen. Biotechnol. Bioeng 94, 815-819.
Bachmann, M. F., and Zinkernagel, R. M. (1996). The influence of virus structure on antibody
responses and virus serotype formation. Immunol. Today 17, 553-558. Bainbridge, J. W. B., Smith, A. J., Barker, S. S., Robbie, S., Henderson, R., Balaggan, K.,
Viswanathan, A., Holder, G. E., Stockman, A., Tyler, N., et al. (2008). Effect of gene therapy on visual function in Leber's congenital amaurosis. N. Engl. J. Med 358, 2231-2239.
Baker, T. S., Olson, N. H., and Fuller, S. D. (1999). Adding the third dimension to virus life
cycles: three-dimensional reconstruction of icosahedral viruses from cryo-electron micrographs. Microbiol. Mol. Biol. Rev 63, 862-922, table of contents.
Barker, D. D., and Berk, A. J. (1987). Adenovirus proteins from both E1B reading frames are
required for transformation of rodent cells by viral infection and DNA transfection. Virology 156, 107-121.
Bartlett, J. S., Samulski, R. J., and McCown, T. J. (1998). Selective and rapid uptake of adeno-
associated virus type 2 in brain. Hum. Gene Ther 9, 1181-1186.
186
Bartlett, J. S., Wilcher, R., and Samulski, R. J. (2000). Infectious entry pathway of adeno-associated virus and adeno-associated virus vectors. J. Virol 74, 2777-2785.
Basak, S., and Turner, H. (1992). Infectious entry pathway for canine parvovirus. Virology 186,
368-376. Bekeredjian, R., Chen, S., Frenkel, P. A., Grayburn, P. A., and Shohet, R. V. (2003). Ultrasound-
Targeted Microbubble Destruction Can Repeatedly Direct Highly Specific Plasmid Expression to the Heart. Circulation 108, 1022-1026.
Bessis, N., GarciaCozar, F. J., and Boissier, M. (2004). Immune responses to gene therapy
vectors: influence on vector function and effector mechanisms. Gene Ther 11 Suppl 1, S10-17.
Blacklow, N. R., Hoggan, M. D., and Rowe, W. P. (1968). Serologic evidence for human
infection with adenovirus-associated viruses. J. Natl. Cancer Inst 40, 319-327. Blacklow, N. R., Hoggan, M. D., Sereno, M. S., Brandt, C. D., Kim, H. W., Parrott, R. H., and
Chanock, R. M. (1971). A seroepidemiologic study of adenovirus-associated virus infection in infants and children. Am. J. Epidemiol 94, 359-366.
Bleker, S., Sonntag, F., and Kleinschmidt, J. A. (2005). Mutational analysis of narrow pores at
the fivefold symmetry axes of adeno-associated virus type 2 capsids reveals a dual role in genome packaging and activation of phospholipase A2 activity. J. Virol 79, 2528-2540.
Bloom, M. E., Best, S. M., Hayes, S. F., Wells, R. D., Wolfinbarger, J. B., McKenna, R., and
Agbandje-McKenna, M. (2001). Identification of aleutian mink disease parvovirus capsid sequences mediating antibody-dependent enhancement of infection, virus neutralization, and immune complex formation. J. Virol 75, 11116-11127.
Bloom, M. E., Kanno, H., Mori, S., and Wolfinbarger, J. B. (1994). Aleutian mink disease:
puzzles and paradigms. Infect Agents Dis 3, 279-301. Bloom, M. E., Martin, D. A., Oie, K. L., Huhtanen, M. E., Costello, F., Wolfinbarger, J. B.,
Hayes, S. F., and Agbandje-McKenna, M. (1997). Expression of Aleutian mink disease parvovirus capsid proteins in defined segments: localization of immunoreactive sites and neutralizing epitopes to specific regions. J. Virol 71, 705-714.
Bothner, B., Dong, X. F., Bibbs, L., Johnson, J. E., and Siuzdak, G. (1998). Evidence of viral
capsid dynamics using limited proteolysis and mass spectrometry. J. Biol. Chem 273, 673-676.
Boutin, S., Monteilhet, V., Veron, P., Leborgne, C., Benveniste, O., Montus, M., and Masurier,
C. (2010). Prevalence of serum IgG and neutralizing factors against adeno-associated virus types 1, 2, 5, 6, 8 and 9 in the healthy population: implications for gene therapy using AAV vectors. Human Gene Therapy, 100122110120098.
187
Bowles, D., J.E.R, and Samulski, R. J. (2006). The genus Dependovirus. In Parvoviruses, J. R.
Kerr, M. E. Bloom, R. Linden, and C. Parrish, eds. (New York: Oxford University Press Inc.), pp. 15-23.
Boxall, A. B. (2004). The environmental side effects of medication. EMBO Rep 5, 1110-1116. Boyce, N. (2001). Trial halted after gene shows up in semen. Nature 414, 677. Bramson, J. L., Graham, F. L., and Gauldie, J. (1995). The use of adenoviral vectors for gene
therapy and gene transfer in vivo. Current Opinion in Biotechnology 6, 590-595. Brown, C. S., Jensen, T., Meloen, R. H., Puijk, W., Sugamura, K., Sato, H., and Spaan, W. J.
(1992). Localization of an immunodominant domain on baculovirus-produced parvovirus B19 capsids: correlation to a major surface region on the native virus particle. J. Virol 66, 6989-6996.
Bruemmer, A., Scholari, F., Lopez-Ferber, M., Conway, J. F., and Hewat, E. A. (2005). Structure
of an insect parvovirus (Junonia coenia Densovirus) determined by cryo-electron microscopy. J. Mol. Biol 347, 791-801.
Buchschacher, G. L. (2001). Introduction to retroviruses and retroviral vectors. Somat. Cell Mol.
Genet 26, 1-11. Budker, V., Zhang, G., Danko, I., Williams, P., and Wolff, J. (1998). The efficient expression of
intravascularly delivered DNA in rat muscle. Gene Ther 5, 272-276. Burton, D. R., Saphire, E. O., and Parren, P. W. (2001). A model for neutralization of viruses
based on antibody coating of the virion surface. Curr. Top. Microbiol. Immunol 260, 109-143.
Burton, D. R. (2002). Antibodies, viruses and vaccines. Nat Rev Immunol 2, 706-713. Calcedo, R., Vandenberghe, L. H., Gao, G., Lin, J., and Wilson, J. M. (2009). Worldwide
epidemiology of neutralizing antibodies to adeno-associated viruses. J. Infect. Dis 199, 381-390.
Campe, H., Hartberger, C., and Sing, A. (2008). Role of Human Bocavirus infections in
outbreaks of gastroenteritis. J. Clin. Virol 43, 340-342. Candotti, D., Etiz, N., Parsyan, A., and Allain, J. (2004). Identification and characterization of
persistent human erythrovirus infection in blood donor samples. J. Virol 78, 12169-12178.
188
Carrillo-Tripp, M., Shepherd, C. M., Borelli, I. A., Venkataraman, S., Lander, G., Natarajan, P., Johnson, J. E., Brooks, C. L., and Reddy, V. S. (2009). VIPERdb2: an enhanced and web API enabled relational database for structural virology. Nucleic Acids Res 37, D436-442.
Carter, B. J. (2005). Adeno-associated virus vectors in clinical trials. Hum. Gene Ther 16, 541-
550. Cecchini, S., Negrete, A., Virag, T., Graham, B. S., Cohen, J. I., and Kotin, R. M. (2009).
Evidence of prior exposure to human bocavirus as determined by a retrospective serological study of 404 serum samples from adults in the United States. Clin. Vaccine Immunol 16, 597-604.
Chacón, P., and Wriggers, W. (2002). Multi-resolution contour-based fitting of macromolecular
structures. J. Mol. Biol 317, 375-384. Chang, L. (2009). Immune responses to lentiviral vectors. In Gene Therapy Immunology (New
Jersey: Wiley-Blackwell), pp. 85-96. Chang, S. F., Sgro, J. Y., and Parrish, C. R. (1992). Multiple amino acids in the capsid structure
of canine parvovirus coordinately determine the canine host range and specific antigenic and hemagglutination properties. J. Virol 66, 6858-6867.
Chao, H., Liu, Y., Rabinowitz, J., Li, C., Samulski, R. J., and Walsh, C. E. (2000). Several log
increase in therapeutic transgene delivery by distinct adeno-associated viral serotype vectors. Mol. Ther 2, 619-623.
Chapman, M. S., and Agbandje-McKenna, M. (2006). Atomic structures of viral particles. In
Parvoviruses, J. R. Kerr, S. F. Cotmore, M. E. Bloom, Linden, R L, and Parrish, C R, eds. (New York, NY: Edward Arnold,Ltd.), pp. 107-123.
Chen, C., Jensen, R. L., Schnepp, B. C., Connell, M. J., Shell, R., Sferra, T. J., Bartlett, J. S.,
Clark, K. R., and Johnson, P. R. (2005). Molecular characterization of adeno-associated viruses infecting children. J. Virol 79, 14781-14792.
Chen, J., Wu, Q., Yang, P., Hsu, H., and Mountz, J. D. (2006). Determination of specific CD4
and CD8 T cell epitopes after AAV2- and AAV8-hF.IX gene therapy. Mol. Ther 13, 260-269.
Chiorini, J. A., Kim, F., Yang, L., and Kotin, R. M. (1999). Cloning and characterization of
adeno-associated virus type 5. J. Virol 73, 1309-1319. Chiorini, J. A., Yang, L., Liu, Y., Safer, B., and Kotin, R. M. (1997). Cloning of adeno-
associated virus type 4 (AAV4) and generation of recombinant AAV4 particles. J. Virol 71, 6823-6833.
189
Chipman, P. R., Agbandje-McKenna, M., Kajigaya, S., Brown, K. E., Young, N. S., Baker, T. S., and Rossmann, M. G. (1996). Cryo-electron microscopy studies of empty capsids of human parvovirus B19 complexed with its cellular receptor. Proc. Natl. Acad. Sci. U.S.A 93, 7502-7506.
Chirmule, N., Propert, K., Magosin, S., Qian, Y., Qian, R., and Wilson, J. (1999). Immune
responses to adenovirus and adeno-associated virus in humans. Gene Ther 6, 1574-1583. Choi, V. W., McCarty, D. M., and Samulski, R. J. (2005). AAV hybrid serotypes: improved
vectors for gene delivery. Curr Gene Ther 5, 299-310. Chow, B. D. W., and Esper, F. P. (2009). The human bocaviruses: a review and discussion of
their role in infection. Clin. Lab. Med 29, 695-713. Cideciyan, A. V., Hauswirth, W. W., Aleman, T. S., Kaushal, S., Schwartz, S. B., Boye, S. L.,
Windsor, E. A. M., Conlon, T. J., Sumaroka, A., Roman, A. J., et al. (2009). Vision 1 year after gene therapy for Leber's congenital amaurosis. N. Engl. J. Med 361, 725-727.
Clarke, J. K., McFerran, J. B., McKillop, E. R., and Curran, W. L. (1979). Isolation of an adeno
associated virus from sheep. Archives of Virology 60, 171-176. Cockrell, A., and Kafri, T. (2007). Gene delivery by lentivirus vectors. Molecular Biotechnology
36, 184-204. Conlon, T. J., Cossette, T., Erger, K., Choi, Y., Clarke, T., Scott-Jorgensen, M., Song, S.,
Campbell-Thompson, M., Crawford, J., and Flotte, T. R. (2005). Efficient hepatic delivery and expression from a recombinant adeno-associated virus 8 pseudotyped alpha1-antitrypsin vector. Mol. Ther 12, 867-875.
Cotmore, S. F., D'abramo, A. M., Ticknor, C. M., and Tattersall, P. (1999). Controlled
conformational transitions in the MVM virion expose the VP1 N-terminus and viral genome without particle disassembly. Virology 254, 169-181.
Crick, F. (1970). Central dogma of molecular biology. Nature 227, 561-563. Cunningham, S. C., Dane, A. P., Spinoulas, A., Logan, G. J., and Alexander, I. E. (2008). Gene
delivery to the juvenile mouse liver using AAV2/8 vectors. Mol. Ther 16, 1081-1088. Cunningham, S. C., Spinoulas, A., Carpenter, K. H., Wilcken, B., Kuchel, P. W., and Alexander,
I. E. (2009). AAV2/8-mediated correction of OTC deficiency is robust in adult but not neonatal Spf(ash) mice. Mol. Ther 17, 1340-1346.
Davidoff, A. M., Gray, J. T., Ng, C. Y. C., Zhang, Y., Zhou, J., Spence, Y., Bakar, Y., and
Nathwani, A. C. (2005). Comparison of the ability of adeno-associated viral vectors pseudotyped with serotype 2, 5, and 8 capsid proteins to mediate efficient transduction of the liver in murine and nonhuman primate models. Mol. Ther 11, 875-888.
190
Davidson, B. L., Stein, C. S., Heth, J. A., Martins, I., Kotin, R. M., Derksen, T. A., Zabner, J., Ghodsi, A., and Chiorini, J. A. (2000). Recombinant adeno-associated virus type 2, 4, and 5 vectors: transduction of variant cell types and regions in the mammalian central nervous system. Proc. Natl. Acad. Sci. U.S.A 97, 3428-3432.
Davies, D. R., and Cohen, G. H. (1996). Interactions of protein antigens with antibodies. Proc.
Natl. Acad. Sci. U.S.A 93, 7-12. Davis, N. L., Willis, L. V., Smith, J. F., and Johnston, R. E. (1989). In vitro synthesis of
infectious venezuelan equine encephalitis virus RNA from a cDNA clone: analysis of a viable deletion mutant. Virology 171, 189-204.
Daya, S., Cortez, N., and Berns, K. I. (2009). Adeno-associated virus site-specific integration is
mediated by proteins of the nonhomologous end-joining pathway. J. Virol 83, 11655-11664.
De Smedt, S. C., Demeester, J., and Hennink, W. E. (2000). Cationic polymer based gene
delivery systems. Pharm. Res 17, 113-126. De, B. P., Heguy, A., Hackett, N. R., Ferris, B., Leopold, P. L., Lee, J., Pierre, L., Gao, G.,
Wilson, J. M., and Crystal, R. G. (2006). High levels of persistent expression of alpha1-antitrypsin mediated by the nonhuman primate serotype rh.10 adeno-associated virus despite preexisting immunity to common human adeno-associated viruses. Mol. Ther 13, 67-76.
Delalande, A., Bureau, M., Midoux, P., Bouakaz, A., and Pichon, C. (2010). Ultrasound-assisted
microbubbles gene transfer in tendons for gene therapy. Ultrasonics 50, 269-272. Denby, L., Nicklin, S. A., and Baker, A. H. (2005). Adeno-associated virus (AAV)-7 and -8
poorly transduce vascular endothelial cells and are sensitive to proteasomal degradation. Gene Ther 12, 1534-1538.
Di Paolo, N., and Shayakhmetov, D. (2009). Immune responses to adenoviral vectors. In Gene
Therapy Immunology, R. W. Herzog, ed. (New Jersey: Wiley-Blackwell), pp. 57-84. Di Pasquale, G., Davidson, B. L., Stein, C. S., Martins, I., Scudiero, D., Monks, A., and Chiorini,
J. A. (2003). Identification of PDGFR as a receptor for AAV-5 transduction. Nat. Med 9, 1306-1312.
Dimmock, N. J. (1984). Mechanisms of neutralization of animal viruses. J. Gen. Virol 65 ( Pt 6),
1015-1022. Dimmock, N. J., and Hardy, S. A. (2004). Valency of antibody binding to virions and its
determination by surface plasmon resonance. Rev. Med. Virol 14, 123-135.
191
Dintzis, R. Z., Middleton, M. H., and Dintzis, H. M. (1983). Studies on the immunogenicity and tolerogenicity of T-independent antigens. J. Immunol 131, 2196-2203.
DiPrimio, N., Asokan, A., Govindasamy, L., Agbandje-McKenna, M., and Samulski, R. J.
(2008). Surface loop dynamics in adeno-associated virus capsid assembly. J. Virol 82, 5178-5189.
Dörner, T., and Radbruch, A. (2007). Antibodies and B cell memory in viral immunity.
Immunity 27, 384-392. Duan, D., Li, Q., Kao, A. W., Yue, Y., Pessin, J. E., and Engelhardt, J. F. (1999). Dynamin is
required for recombinant adeno-associated virus type 2 infection. J. Virol 73, 10371-10376.
Dubochet, J., Adrian, M., Chang, J. J., Homo, J. C., Lepault, J., McDowall, A. W., and Schultz,
P. (1988). Cryo-electron microscopy of vitrified specimens. Q. Rev. Biophys 21, 129-228.
Dubuisson, J., and Rice, C. M. (1993). Sindbis virus attachment: isolation and characterization of
mutants with impaired binding to vertebrate cells. J. Virol 67, 3363-3374. Dupuy, L. C., and Schmaljohn, C. S. (2009). DNA vaccines for biodefense. Expert Rev Vaccines
8, 1739-1754. Ehrengruber, M. U., and Lundstrom, K. (2007). Alphaviruses: Semliki Forest virus and Sindbis
virus vectors for gene transfer into neurons. Curr Protoc Neurosci Chapter 4, Unit 4.22. Emsley, P., and Cowtan, K. (2004). Coot: model-building tools for molecular graphics. Acta
Crystallogr. D Biol. Crystallogr 60, 2126-2132. Erles, K., Sebökovà, P., and Schlehofer, J. R. (1999). Update on the prevalence of serum
antibodies (IgG and IgM) to adeno-associated virus (AAV). J. Med. Virol 59, 406-411. Excoffon, K. J. D. A., Koerber, J. T., Dickey, D. D., Murtha, M., Keshavjee, S., Kaspar, B. K.,
Zabner, J., and Schaffer, D. V. (2009). Directed evolution of adeno-associated virus to an infectious respiratory virus. Proc. Natl. Acad. Sci. U.S.A 106, 3865-3870.
Farkas, S. L., Zadori, Z., Benko, M., Essbauer, S., Harrach, B., and Tijssen, P. (2004). A
parvovirus isolated from royal python (Python regius) is a member of the genus Dependovirus. J Gen Virol 85, 555-561.
Felsenstein, J. (1989). PHYLIP—Phylogeny inference package (version 3.2). Cladistics 5, 164-
166. Feschotte, C. (2010). Virology: Bornavirus enters the genome. Nature 463, 39-40.
192
Fischer, A. C., Beck, S. E., Smith, C. I., Laube, B. L., Askin, F. B., Guggino, S. E., Adams, R. J., Flotte, T. R., and Guggino, W. B. (2003). Successful transgene expression with serial doses of aerosolized rAAV2 vectors in rhesus macaques. Mol. Ther 8, 918-926.
Flotte, T. R. (2005). Adeno-associated virus-based gene therapy for inherited disorders. Pediatr.
Res 58, 1143-1147. Flotte, T. R. (2007). Gene therapy: The first two decades and the current state-of-the-art. J. Cell.
Physiol. 213, 301-305. Foust, K. D., Nurre, E., Montgomery, C. L., Hernandez, A., Chan, C. M., and Kaspar, B. K.
(2009). Intravascular AAV9 preferentially targets neonatal neurons and adult astrocytes. Nat. Biotechnol 27, 59-65.
Fryer, J. F., Kapoor, A., Minor, P. D., Delwart, E., and Baylis, S. A. (2006). Novel parvovirus
and related variant in human plasma. Emerging Infect. Dis 12, 151-154. Gao, G., Alvira, M. R., Somanathan, S., Lu, Y., Vandenberghe, L. H., Rux, J. J., Calcedo, R.,
Sanmiguel, J., Abbas, Z., and Wilson, J. M. (2003). Adeno-associated viruses undergo substantial evolution in primates during natural infections. Proc. Natl. Acad. Sci. U.S.A 100, 6081-6086.
Gao, G., Vandenberghe, L. H., Alvira, M. R., Lu, Y., Calcedo, R., Zhou, X., and Wilson, J. M.
(2004). Clades of Adeno-associated viruses are widely disseminated in human tissues. J. Virol 78, 6381-6388.
Gao, G., Vandenberghe, L. H., and Wilson, J. M. (2005). New recombinant serotypes of AAV
vectors. Curr Gene Ther 5, 285-297. Gao, G., Alvira, M. R., Wang, L., Calcedo, R., Johnston, J., and Wilson, J. M. (2002). Novel
adeno-associated viruses from rhesus monkeys as vectors for human gene therapy. Proc. Natl. Acad. Sci. U.S.A 99, 11854-11859.
Gao, X., Kim, K., and Liu, D. (2007). Nonviral gene delivery: What we know and what is next.
AAPS J 9, E92-E104. Girod, A., Wobus, C. E., Zádori, Z., Ried, M., Leike, K., Tijssen, P., Kleinschmidt, J. A., and
Hallek, M. (2002). The VP1 capsid protein of adeno-associated virus type 2 is carrying a phospholipase A2 domain required for virus infectivity. J. Gen. Virol 83, 973-978.
Govindasamy, L., Hueffer, K., Parrish, C. R., and Agbandje-McKenna, M. (2003). Structures of
host range-controlling regions of the capsids of canine and feline parvoviruses and mutants. J. Virol 77, 12211-12221.
193
Govindasamy, L., Padron, E., McKenna, R., Muzyczka, N., Kaludov, N., Chiorini, J. A., and Agbandje-McKenna, M. (2006). Structurally mapping the diverse phenotype of adeno-associated virus serotype 4. J. Virol 80, 11556-11570.
Grieger, J. C., Johnson, J. S., Gurda-Whitaker, B., Agbandje-McKenna, M., and Samulski, R. J.
(2007). Surface-exposed adeno-associated virus Vp1-NLS capsid fusion protein rescues infectivity of noninfectious wild-type Vp2/Vp3 and Vp3-only capsids but not that of fivefold pore mutant virions. J. Virol 81, 7833-7843.
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.
Grimm, D., Kern, A., Pawlita, M., Ferrari, F., Samulski, R., and Kleinschmidt, J. (1999).
Titration of AAV-2 particles via a novel capsid ELISA: packaging of genomes can limit production of recombinant AAV-2. Gene Ther 6, 1322-1330.
Grimm, D., Kern, A., Rittner, K., and Kleinschmidt, J. A. (1998). Novel tools for production and
purification of recombinant adenoassociated virus vectors. Hum. Gene Ther 9, 2745-2760.
Guex, N., and Peitsch, M. C. (1997). SWISS-MODEL and the Swiss-PdbViewer: an
environment for comparative protein modeling. Electrophoresis 18, 2714-2723. Hacein-Bey-Abina, S., von Kalle, C., Schmidt, M., Le Deist, F., Wulffraat, N., McIntyre, E.,
Radford, I., Villeval, J., Fraser, C. C., Cavazzana-Calvo, M., et al. (2003). A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency. N. Engl. J. Med 348, 255-256.
Hacein-Bey-Abina, S., Le Deist, F., Carlier, F., Bouneaud, C., Hue, C., De Villartay, J.,
Thrasher, A. J., Wulffraat, N., Sorensen, R., Dupuis-Girod, S., et al. (2002). Sustained Correction of X-Linked Severe Combined Immunodeficiency by ex Vivo Gene Therapy. N Engl J Med 346, 1185-1193.
Hafenstein, S., Bowman, V. D., Sun, T., Nelson, C. D. S., Palermo, L. M., Chipman, P. R.,
Battisti, A. J., Parrish, C. R., and Rossmann, M. G. (2009). Structural comparison of different antibodies interacting with parvovirus capsids. J. Virol 83, 5556-5566.
Hafenstein, S., Palermo, L. M., Kostyuchenko, V. A., Xiao, C., Morais, M. C., Nelson, C. D. S.,
Bowman, V. D., Battisti, A. J., Chipman, P. R., Parrish, C. R., et al. (2007). Asymmetric binding of transferrin receptor to parvovirus capsids. Proc. Natl. Acad. Sci. U.S.A 104, 6585-6589.
194
Halbert, C. L., Miller, A. D., McNamara, S., Emerson, J., Gibson, R. L., Ramsey, B., and Aitken, M. L. (2006). Prevalence of neutralizing antibodies against adeno-associated virus (AAV) types 2, 5, and 6 in cystic fibrosis and normal populations: Implications for gene therapy using AAV vectors. Hum. Gene Ther 17, 440-447.
Hardy, S. A., and Dimmock, N. J. (2003). Valency of antibody binding to enveloped virus
particles as determined by surface plasmon resonance. J. Virol 77, 1649-1652. Harris, A., Belnap, D., Watts, N., Conway, J., Cheng, N., Stahl, S., Vethanayagam, J., Wingfield,
P., and Steven, A. (2006). Epitope Diversity of Hepatitis B Virus Capsids: Quasi-equivalent Variations in Spike Epitopes and Binding of Different Antibodies to the same Epitope. Journal of Molecular Biology 355, 562-576.
Hauck, B., Xu, R. R., Xie, J., Wu, W., Ding, Q., Sipler, M., Wang, H., Chen, L., Wright, J. F.,
and Xiao, W. (2006). Efficient AAV1-AAV2 hybrid vector for gene therapy of hemophilia. Hum. Gene Ther 17, 46-54.
Hauswirth, W. W., Aleman, T. S., Kaushal, S., Cideciyan, A. V., Schwartz, S. B., Wang, L.,
Conlon, T. J., Boye, S. L., Flotte, T. R., Byrne, B. J., et al. (2008). Treatment of leber congenital amaurosis due to RPE65 mutations by ocular subretinal injection of adeno-associated virus gene vector: short-term results of a phase I trial. Hum. Gene Ther 19, 979-990.
Havelka, W. A., Henderson, R., and Oesterhelt, D. (1995). Three-dimensional structure of
halorhodopsin at 7 A resolution. J. Mol. Biol 247, 726-738. van Heel, M., and Schatz, M. (2005). Fourier shell correlation threshold criteria. J. Struct. Biol
151, 250-262. Heller, R., Gilbert, R., and Jaroszeski, M. J. (2000). Clinical Trials for Solid Tumors Using
Electrochemotherapy Available at: http://dx.doi.org.lp.hscl.ufl.edu/10.1385/1-59259-080-2:137 [Accessed January 7, 2010].
Horie, M., Honda, T., Suzuki, Y., Kobayashi, Y., Daito, T., Oshida, T., Ikuta, K., Jern, P.,
Gojobori, T., Coffin, J. M., et al. (2010). Endogenous non-retroviral RNA virus elements in mammalian genomes. Nature 463, 84-87.
Huang, X., and Yang, Y. (2009). Innate immune recognition of viruses and viral vectors. Hum.
Gene Ther 20, 293-301. Huber, M., and Trkola, A. (2007). Humoral immunity to HIV-1: neutralization and beyond. J.
Intern. Med 262, 5-25. Hueffer, K., Govindasamy, L., Agbandje-McKenna, M., and Parrish, C. R. (2003). Combinations
of two capsid regions controlling canine host range determine canine transferrin receptor binding by canine and feline parvoviruses. J. Virol 77, 10099-10105.
195
Huttner, N. A., Girod, A., Perabo, L., Edbauer, D., Kleinschmidt, J. A., Büning, H., and Hallek,
M. (2003). Genetic modifications of the adeno-associated virus type 2 capsid reduce the affinity and the neutralizing effects of human serum antibodies. Gene Ther 10, 2139-2147.
Inagaki, K., Piao, C., Kotchey, N. M., Wu, X., and Nakai, H. (2008). Frequency and spectrum of
genomic integration of recombinant adeno-associated virus serotype 8 vector in neonatal mouse liver. J. Virol 82, 9513-9524.
Ivics, Z., and Izsvák, Z. (2006). Transposons for gene therapy! Curr Gene Ther 6, 593-607. Izsvák, Z., Chuah, M. K. L., Vandendriessche, T., and Ivics, Z. (2009). Efficient stable gene
transfer into human cells by the Sleeping Beauty transposon vectors. Methods 49, 287-297.
Janeway, C., Travers, P., Walport, M., and Shlomchik, M. (2005). Immunobiology: the immune
system in health and disease (New York: Garland Science Publishing). Jiang, H., Lillicrap, D., Patarroyo-White, S., Liu, T., Qian, X., Scallan, C. D., Powell, S., Keller,
T., McMurray, M., Labelle, A., et al. (2006). Multiyear therapeutic benefit of AAV serotypes 2, 6, and 8 delivering factor VIII to hemophilia A mice and dogs. Blood 108, 107-115.
Johnson, J. E., and Chiu, W. (2000). Structures of virus and virus-like particles. Curr. Opin.
Struct. Biol 10, 229-235. Jooss, K., and Chirmule, N. (2003). Immunity to adenovirus and adeno-associated viral vectors:
implications for gene therapy. Gene Ther 10, 955-963. Kahn, J. S., Kesebir, D., Cotmore, S. F., D'Abramo, A., Cosby, C., Weibel, C., and Tattersall, P.
(2008). Seroepidemiology of human bocavirus defined using recombinant virus-like particles. J. Infect. Dis 198, 41-50.
Kajigaya, S., Fujii, H., Field, A., Anderson, S., Rosenfeld, S., Anderson, L. J., Shimada, T., and
Young, N. S. (1991). Self-assembled B19 parvovirus capsids, produced in a baculovirus system, are antigenically and immunogenically similar to native virions. Proc. Natl. Acad. Sci. U.S.A 88, 4646-4650.
Kaludov, N., Brown, K. E., Walters, R. W., Zabner, J., and Chiorini, J. A. (2001). Adeno-
associated virus serotype 4 (AAV4) and AAV5 both require sialic acid binding for hemagglutination and efficient transduction but differ in sialic acid linkage specificity. J. Virol 75, 6884-6893.
196
Kantola, K., Hedman, L., Allander, T., Jartti, T., Lehtinen, P., Ruuskanen, O., Hedman, K., and Söderlund-Venermo, M. (2008). Serodiagnosis of human bocavirus infection. Clin. Infect. Dis 46, 540-546.
Kapoor, A., Slikas, E., Simmonds, P., Chieochansin, T., Naeem, A., Shaukat, S., Alam, M. M.,
Sharif, S., Angez, M., Zaidi, S., et al. (2009). A newly identified bocavirus species in human stool. J. Infect. Dis 199, 196-200.
Kashiwakura, Y., Tamayose, K., Iwabuchi, K., Hirai, Y., Shimada, T., Matsumoto, K.,
Nakamura, T., Watanabe, M., Oshimi, K., and Daida, H. (2005). Hepatocyte growth factor receptor is a coreceptor for adeno-associated virus type 2 infection. J. Virol 79, 609-614.
Kaufmann, B., Chipman, P. R., Kostyuchenko, V. A., Modrow, S., and Rossmann, M. G. (2008).
Visualization of the externalized VP2 N termini of infectious human parvovirus B19. J. Virol 82, 7306-7312.
Kaufmann, B., López-Bueno, A., Mateu, M. G., Chipman, P. R., Nelson, C. D. S., Parrish, C. R.,
Almendral, J. M., and Rossmann, M. G. (2007). Minute virus of mice, a parvovirus, in complex with the Fab fragment of a neutralizing monoclonal antibody. J. Virol 81, 9851-9858.
Kaufmann, B., Simpson, A. A., and Rossmann, M. G. (2004). The structure of human parvovirus
B19. Proc. Natl. Acad. Sci. U.S.A 101, 11628-11633. Kawase, M., Momoeda, M., Young, N. S., and Kajigaya, S. (1995). Most of the VP1 unique
region of B19 parvovirus is on the capsid surface. Virology 211, 359-366. Kern, A., Schmidt, K., Leder, C., Müller, O. J., Wobus, C. E., Bettinger, K., Von der Lieth, C.
W., King, J. A., and Kleinschmidt, J. A. (2003). Identification of a heparin-binding motif on adeno-associated virus type 2 capsids. J. Virol 77, 11072-11081.
Kesebir, D., Vazquez, M., Weibel, C., Shapiro, E. D., Ferguson, D., Landry, M. L., and Kahn, J.
S. (2006). Human bocavirus infection in young children in the United States: molecular epidemiological profile and clinical characteristics of a newly emerging respiratory virus. J. Infect. Dis 194, 1276-1282.
Khuri, F. R., Nemunaitis, J., Ganly, I., Arseneau, J., Tannock, I. F., Romel, L., Gore, M.,
Ironside, J., MacDougall, R. H., Heise, C., et al. (2000). a controlled trial of intratumoral ONYX-015, a selectively-replicating adenovirus, in combination with cisplatin and 5-fluorouracil in patients with recurrent head and neck cancer. Nat. Med 6, 879-885.
Kidd, K. A., Blanchfield, P. J., Mills, K. H., Palace, V. P., Evans, R. E., Lazorchak, J. M., and
Flick, R. W. (2007). Collapse of a fish population after exposure to a synthetic estrogen. Proc. Natl. Acad. Sci. U.S.A 104, 8897-8901.
197
Klasse, P. J., and Sattentau, Q. J. (2002). Occupancy and mechanism in antibody-mediated neutralization of animal viruses. J. Gen. Virol 83, 2091-2108.
Kleywegt, G. J., and Jones, T. A. (1996). xdlMAPMAN and xdlDATAMAN - programs for
reformatting, analysis and manipulation of biomacromolecular electron-density maps and reflection data sets. Acta Crystallogr. D Biol. Crystallogr 52, 826-828.
Knowles, B. B., Howe, C. C., and Aden, D. P. (1980). Human hepatocellular carcinoma cell
lines secrete the major plasma proteins and hepatitis B surface antigen. Science 209, 497-499.
Koerber, J. T., Jang, J., and Schaffer, D. V. (2008). DNA shuffling of adeno-associated virus
yields functionally diverse viral progeny. Mol. Ther 16, 1703-1709. Kontou, M., Govindasamy, L., Nam, H., Bryant, N., Llamas-Saiz, A. L., Foces-Foces, C.,
Hernando, E., Rubio, M., McKenna, R., Almendral, J. M., et al. (2005). Structural determinants of tissue tropism and in vivo pathogenicity for the parvovirus minute virus of mice. J. Virol 79, 10931-10943.
Kotin, R. M., Siniscalco, M., Samulski, R. J., Zhu, X. D., Hunter, L., Laughlin, C. A.,
McLaughlin, S., Muzyczka, N., Rocchi, M., and Berns, K. I. (1990). Site-specific integration by adeno-associated virus. Proceedings of the National Academy of Sciences of the United States of America 87, 2211-2215.
Krissinel, E., and Henrick, K. (2004). Secondary-structure matching (SSM), a new tool for fast
protein structure alignment in three dimensions. Acta Crystallogr. D Biol. Crystallogr 60, 2256-2268.
Kronenberg, S., Böttcher, B., von der Lieth, C. W., Bleker, S., and Kleinschmidt, J. A. (2005). A
conformational change in the adeno-associated virus type 2 capsid leads to the exposure of hidden VP1 N termini. J. Virol 79, 5296-5303.
Kuck, D., Kern, A., and Kleinschmidt, J. A. (2007). Development of AAV serotype-specific
ELISAs using novel monoclonal antibodies. J. Virol. Methods 140, 17-24. Kühler, T. C., Andersson, M., Carlin, G., Johnsson, A., and Akerblom, L. (2009). Do biological
medicinal products pose a risk to the environment?: a current view on ecopharmacovigilance. Drug Saf 32, 995-1000.
Lai, Y., Yue, Y., and Duan, D. (2010). Evidence for the failure of adeno-associated virus
serotype 5 to package a viral genome > or = 8.2 kb. Mol. Ther 18, 75-79. Langeveld, J. P., Casal, J. I., Vela, C., Dalsgaard, K., Smale, S. H., Puijk, W. C., and Meloen, R.
H. (1993). B-cell epitopes of canine parvovirus: distribution on the primary structure and exposure on the viral surface. J. Virol 67, 765-772.
198
Lee, J. S., Hadjipanayis, A. G., and Parker, M. D. (2005). Viral vectors for use in the development of biodefense vaccines. Adv. Drug Deliv. Rev 57, 1293-1314.
Levy, H. C., Bowman, V. D., Govindasamy, L., McKenna, R., Nash, K., Warrington, K., Chen,
W., Muzyczka, N., Yan, X., Baker, T. S., et al. (2009). Heparin binding induces conformational changes in Adeno-associated virus serotype 2. J. Struct. Biol 165, 146-156.
Lewis, J. K., Bothner, B., Smith, T. J., and Siuzdak, G. (1998). Antiviral agent blocks breathing
of the common cold virus. Proc. Natl. Acad. Sci. U.S.A 95, 6774-6778. Li, H., Lin, S., Giles-Davis, W., Li, Y., Zhou, D., Xiang, Z. Q., High, K. A., and Ertl, H. C. J.
(2009). A preclinical animal model to assess the effect of pre-existing immunity on AAV-mediated gene transfer. Mol. Ther 17, 1215-1224.
Li, Q., Yafal, A. G., Lee, Y. M., Hogle, J., and Chow, M. (1994). Poliovirus neutralization by
antibodies to internal epitopes of VP4 and VP1 results from reversible exposure of these sequences at physiological temperature. J. Virol 68, 3965-3970.
Li, S., and Ma, Z. (2001). Nonviral gene therapy. Curr Gene Ther 1, 201-226. Li, Z., Dullmann, J., Schiedlmeier, B., Schmidt, M., von Kalle, C., Meyer, J., Forster, M.,
Stocking, C., Wahlers, A., Frank, O., et al. (2002). Murine Leukemia Induced by Retroviral Gene Marking. Science 296, 497.
Liljeström, P., and Garoff, H. (1991). A new generation of animal cell expression vectors based
on the Semliki Forest virus replicon. Biotechnology (N.Y.) 9, 1356-1361. Lin, S., and Ertl, H. C. (2009). Immune responses to Adeno-associated virus vectors. In Gene
Therapy Immunology, Roland Herzog, ed. (Hoboken, New Jersey: John Wiley & Sons, Inc.), pp. 97-124.
Linden, R. M., Ward, P., Giraud, C., Winocour, E., and Berns, K. I. (1996). Site-specific
integration by adeno-associated virus. Proceedings of the National Academy of Sciences of the United States of America 93, 11288-11294.
Lindner, J., and Modrow, S. (2008). Human bocavirus--a novel parvovirus to infect humans.
Intervirology 51, 116-122. Liu, G., Martins, I. H., Chiorini, J. A., and Davidson, B. L. (2005). Adeno-associated virus type
4 (AAV4) targets ependyma and astrocytes in the subventricular zone and RMS. Gene Ther 12, 1503-1508.
Llamas-Saiz, A. L., Agbandje-McKenna, M., Parker, J. S., Wahid, A. T., Parrish, C. R., and
Rossmann, M. G. (1996). Structural analysis of a mutation in canine parvovirus which controls antigenicity and host range. Virology 225, 65-71.
199
Lochrie, M. A., Tatsuno, G. P., Christie, B., McDonnell, J. W., Zhou, S., Surosky, R., Pierce, G.
F., and Colosi, P. (2006). Mutations on the external surfaces of adeno-associated virus type 2 capsids that affect transduction and neutralization. J. Virol 80, 821-834.
López de Turiso, J. A., Cortés, E., Ranz, A., García, J., Sanz, A., Vela, C., and Casal, J. I. (1991).
Fine mapping of canine parvovirus B cell epitopes. J. Gen. Virol 72 ( Pt 10), 2445-2456. López-Bueno, A., Mateu, M. G., and Almendral, J. M. (2003). High mutant frequency in
populations of a DNA virus allows evasion from antibody therapy in an immunodeficient host. J. Virol 77, 2701-2708.
Lu, Y., and Song, S. (2009). Distinct immune responses to transgene products from rAAV1 and
rAAV8 vectors. Proc. Natl. Acad. Sci. U.S.A 106, 17158-17162. Luchsinger, E., Strobbe, R., Wellemans, G., Dekegel, D., and Sprececher-Goldbeeger, S. (1970).
Haemagglutinating adeno-associated virus (AAV) in association with bovine adenovirus type 1. Archives of Virology 31, 390-392.
Lufino, M. M. P., Edser, P. A. H., and Wade-Martins, R. (2008). Advances in high-capacity
extrachromosomal vector technology: episomal maintenance, vector delivery, and transgene expression. Mol. Ther 16, 1525-1538.
Lundberg, C., Björklund, T., Carlsson, T., Jakobsson, J., Hantraye, P., Déglon, N., and Kirik, D.
(2008). Applications of lentiviral vectors for biology and gene therapy of neurological disorders. Curr Gene Ther 8, 461-473.
Lundstrom, K. (2004). Gene therapy applications of viral vectors. Technol. Cancer Res. Treat 3,
467-477. Lundstrom, K. (2003). Latest development in viral vectors for gene therapy. Trends Biotechnol
21, 117-122. M, A., and Chapman, M S (2006). Corrleating structure with function in the viral capsid. In
Parvoviruses, Kerr, J R, Cotmore, S F, Bloom, M E, Linden, R M, and Parrish, C R, eds. (New York, NY: Edward Arnold,Ltd.), pp. 125-139.
Macnab, S., and Whitehouse, A. (2009). Progress and prospects: human artificial chromosomes.
Gene Ther 16, 1180-1188. Maguire, A. M., Simonelli, F., Pierce, E. A., Pugh, E. N., Mingozzi, F., Bennicelli, J., Banfi, S.,
Marshall, K. A., Testa, F., Surace, E. M., et al. (2008). Safety and efficacy of gene transfer for Leber's congenital amaurosis. N. Engl. J. Med 358, 2240-2248.
200
Manno, C. S., Pierce, G. F., Arruda, V. R., Glader, B., Ragni, M., Rasko, J. J., Rasko, J., Ozelo, M. C., Hoots, K., Blatt, P., et al. (2006). Successful transduction of liver in hemophilia by AAV-Factor IX and limitations imposed by the host immune response. Nat. Med 12, 342-347.
Marconi, P., Argnani, R., Epstein, A. L., and Manservigi, R. (2009). HSV as a Vector in Vaccine
Development and Gene Therapy. Adv. Exp. Med. Biol 655, 118-144. Margaritis, P., Roy, E., Aljamali, M. N., Downey, H. D., Giger, U., Zhou, S., Merricks, E.,
Dillow, A., Ezban, M., Nichols, T. C., et al. (2009). Successful treatment of canine hemophilia by continuous expression of canine FVIIa. Blood 113, 3682-3689.
Martin, A. C., Cheetham, J. C., and Rees, A. R. (1989). Modeling antibody hypervariable loops:
a combined algorithm. Proc. Natl. Acad. Sci. U.S.A 86, 9268-9272. Matthews, K. S., Alvarez, R. D., and Curiel, D. T. (2009). Advancements in adenoviral based
virotherapy for ovarian cancer. Advanced Drug Delivery Reviews 61, 836-841. Mays, L. E., and Wilson, J. M. (2009). Identification of the murine AAVrh32.33 capsid-specific
CD8+ T cell epitopes. J Gene Med 11, 1095-1102. McCarty, D. M., Young, S. M., and Samulski, R. J. (2004). INTEGRATION OF ADENO-
ASSOCIATED VIRUS (AAV) AND RECOMBINANT AAV VECTORS. Annu. Rev. Genet. 38, 819-845.
McCullough, K. C., De Simone, F., Brocchi, E., Capucci, L., Crowther, J. R., and Kihm, U.
(1992). Protective immune response against foot-and-mouth disease. J. Virol 66, 1835-1840.
McKenna, R., Olson, N. H., Chipman, P. R., Baker, T. S., Booth, T. F., Christensen, J., Aasted,
B., Fox, J. M., Bloom, M. E., Wolfinbarger, J. B., et al. (1999). Three-dimensional structure of Aleutian mink disease parvovirus: implications for disease pathogenicity. J. Virol 73, 6882-6891.
Meng, E. C., Pettersen, E. F., Couch, G. S., Huang, C. C., and Ferrin, T. E. (2006). Tools for
integrated sequence-structure analysis with UCSF Chimera. BMC Bioinformatics 7, 339. Meyer, K. B., Thompson, M. M., Levy, M. Y., Barron, L. G., and Szoka, F. C. (1995).
Intratracheal gene delivery to the mouse airway: characterization of plasmid DNA expression and pharmacokinetics. Gene Ther 2, 450-460.
Miao, C. H., Ye, P., Thompson, A. R., Rawlings, D. J., and Ochs, H. D. (2006).
Immunomodulation of transgene responses following naked DNA transfer of human factor VIII into hemophilia A mice. Blood 108, 19-27.
201
Midoux, P., Breuzard, G., Gomez, J. P., and Pichon, C. (2008). Polymer-based gene delivery: a current review on the uptake and intracellular trafficking of polyplexes. Curr Gene Ther 8, 335-352.
Midoux, P., Pichon, C., Yaouanc, J., and Jaffrès, P. (2009). Chemical vectors for gene delivery: a
current review on polymers, peptides and lipids containing histidine or imidazole as nucleic acids carriers. British Journal of Pharmacology 157, 166-178.
Miller, E. B., Gurda-Whitaker, B., Govindasamy, L., McKenna, R., Zolotukhin, S., Muzyczka,
N., and Agbandje-McKenna, M. (2006). Production, purification and preliminary X-ray crystallographic studies of adeno-associated virus serotype 1. Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun 62, 1271-1274.
Milich, D. R., and McLachlan, A. (1986). The nucleocapsid of hepatitis B virus is both a T-cell-
independent and a T-cell-dependent antigen. Science 234, 1398-1401. Mingozzi, F., Maus, M. V., Hui, D. J., Sabatino, D. E., Murphy, S. L., Rasko, J. E. J., Ragni, M.
V., Manno, C. S., Sommer, J., Jiang, H., et al. (2007). CD8(+) T-cell responses to adeno-associated virus capsid in humans. Nat. Med 13, 419-422.
Mislick, K. A., and Baldeschwieler, J. D. (1996). Evidence for the role of proteoglycans in
cation-mediated gene transfer. Proceedings of the National Academy of Sciences of the United States of America 93, 12349-12354.
Monteny, M., Niesters, H. G. M., Moll, H. A., and Berger, M. Y. (2007). Human bocavirus in
febrile children, The Netherlands. Emerging Infect. Dis 13, 180-182. Mori, S., Wang, L., Takeuchi, T., and Kanda, T. (2004). Two novel adeno-associated viruses
from cynomolgus monkey: pseudotyping characterization of capsid protein. Virology 330, 375-383.
Moscioni, D., Morizono, H., McCarter, R. J., Stern, A., Cabrera-Luque, J., Hoang, A.,
Sanmiguel, J., Wu, D., Bell, P., Gao, G., et al. (2006). Long-term correction of ammonia metabolism and prolonged survival in ornithine transcarbamylase-deficient mice following liver-directed treatment with adeno-associated viral vectors. Mol. Ther 14, 25-33.
Moskalenko, M., Chen, L., van Roey, M., Donahue, B. A., Snyder, R. O., McArthur, J. G., and
Patel, S. D. (2000). Epitope mapping of human anti-adeno-associated virus type 2 neutralizing antibodies: implications for gene therapy and virus structure. J. Virol 74, 1761-1766.
Muzyczka, N. (1994). Adeno-associated virus (AAV) vectors: will they work? J. Clin. Invest 94,
1351.
202
Muzyczka, N., and Berns, K. I. (2001). In Fields Virology, D. Knipe and P. Howley, eds. (New York: Lippincott, Williams and Wilkins), pp. 2327-60.
Muzyczka, N., and Warrington, K. H. (2005). Custom adeno-associated virus capsids: the next
generation of recombinant vectors with novel tropism. Hum. Gene Ther 16, 408-416. Nabel, G. J. (2004). Genetic, cellular and immune approaches to disease therapy: past and future.
Nat Med 10, 135-141. Nakai, H., Fuess, S., Storm, T. A., Muramatsu, S., Nara, Y., and Kay, M. A. (2005). Unrestricted
hepatocyte transduction with adeno-associated virus serotype 8 vectors in mice. J. Virol 79, 214-224.
Nam, H., Lane, M. D., Padron, E., Gurda, B., McKenna, R., Kohlbrenner, E., Aslanidi, G.,
Byrne, B., Muzyczka, N., Zolotukhin, S., et al. (2007). Structure of adeno-associated virus serotype 8, a gene therapy vector. J. Virol 81, 12260-12271.
Nanda, K., Vancini, R., Ribeiro, M., Brown, D. T., and Hernandez, R. (2009). A high capacity
Alphavirus heterologous gene delivery system. Virology 390, 368-373. Nason, E. L., Wetzel, J. D., Mukherjee, S. K., Barton, E. S., Prasad, B. V., and Dermody, T. S.
(2001). A monoclonal antibody specific for reovirus outer-capsid protein sigma3 inhibits sigma1-mediated hemagglutination by steric hindrance. J. Virol 75, 6625-6634.
Neumann, E., Schaefer-Ridder, M., Wang, Y., and Hofschneider, P. H. (1982). Gene transfer
into mouse lyoma cells by electroporation in high electric fields. EMBO J 1, 841-845. Nicklin, S. A., Buening, H., Dishart, K. L., de Alwis, M., Girod, A., Hacker, U., Thrasher, A. J.,
Ali, R. R., Hallek, M., and Baker, A. H. (2001). Efficient and selective AAV2-mediated gene transfer directed to human vascular endothelial cells. Mol. Ther 4, 174-181.
O'Donnell, J., Taylor, K. A., and Chapman, M. S. (2009). Adeno-associated virus-2 and its
primary cellular receptor--Cryo-EM structure of a heparin complex. Virology 385, 434-443.
Oliphant, T., and Diamond, M. S. (2007). The molecular basis of antibody-mediated
neutralization of West Nile virus. Expert Opin Biol Ther 7, 885-892. Olson, E. J., Haskell, S. R. R., Frank, R. K., Lehmkuhl, H. D., Hobbs, L. A., Warg, J. V.,
Landgraf, J. G., and Wünschmann, A. (2004). Isolation of an adenovirus and an adeno-associated virus from goat kids with enteritis. J. Vet. Diagn. Invest 16, 461-464.
Opie, S. R., Warrington, K. H., Agbandje-McKenna, M., Zolotukhin, S., and Muzyczka, N.
(2003). Identification of amino acid residues in the capsid proteins of adeno-associated virus type 2 that contribute to heparan sulfate proteoglycan binding. J. Virol 77, 6995-7006.
203
Pacak, C. A., Mah, C. S., Thattaliyath, B. D., Conlon, T. J., Lewis, M. A., Cloutier, D. E.,
Zolotukhin, I., Tarantal, A. F., and Byrne, B. J. (2006). Recombinant adeno-associated virus serotype 9 leads to preferential cardiac transduction in vivo. Circ. Res 99, e3-9.
Padron, E., Bowman, V., Kaludov, N., Govindasamy, L., Levy, H., Nick, P., McKenna, R.,
Muzyczka, N., Chiorini, J. A., Baker, T. S., et al. (2005). Structure of adeno-associated virus type 4. J. Virol 79, 5047-5058.
Parker, A. L., Newman, C., Briggs, S., Seymour, L., and Sheridan, P. J. (2003). Nonviral gene
delivery: techniques and implications for molecular medicine. Expert Rev Mol Med 5, 1-15.
Parker, J. S., Murphy, W. J., Wang, D., O'Brien, S. J., and Parrish, C. R. (2001). Canine and
feline parvoviruses can use human or feline transferrin receptors to bind, enter, and infect cells. J. Virol 75, 3896-3902.
Parker, J. S., and Parrish, C. R. (2000). Cellular uptake and infection by canine parvovirus
involves rapid dynamin-regulated clathrin-mediated endocytosis, followed by slower intracellular trafficking. J. Virol 74, 1919-1930.
Parks, R., and Graham, F. (1997). A helper-dependent system for adenovirus vector production
helps define a lower limit for efficient DNA packaging. J. Virol. 71, 3293-3298. Parrish, C. R., Aquadro, C. F., Strassheim, M. L., Evermann, J. F., Sgro, J. Y., and Mohammed,
H. O. (1991). Rapid antigenic-type replacement and DNA sequence evolution of canine parvovirus. J. Virol 65, 6544-6552.
Parrish, C. R., O'Connell, P. H., Evermann, J. F., and Carmichael, L. E. (1985). Natural variation
of canine parvovirus. Science 230, 1046-1048. Parrish, C. R. (2006). Pathogenesis of feline panleukopenia virus and canine parvovirus. In
Parvoviruses, J. R. Kerr, M. E. Bloom, R. Linden, and C. Parrish, eds. (New York: Oxford University Press Inc.), pp. 429-434.
Parrish, C., and Carmichael, L. (1983). Antigenic structure and variation of canine parvovirus
type-2, feline panleukopenia virus, and mink enteritis virus. Virology 129, 401-414. Patil, S. D., Rhodes, D. G., and Burgess, D. J. (2005). DNA-based therapeutics and DNA
delivery systems: A comprehensive review. AAPS J 7, E61-E77. Pear, W. S., Nolan, G. P., Scott, M. L., and Baltimore, D. (1993). Production of high-titer helper-
free retroviruses by transient transfection. Proc. Natl. Acad. Sci. U.S.A 90, 8392-8396. Peng, Z. (2005). Current status of gendicine in China: recombinant human Ad-p53 agent for
treatment of cancers. Hum. Gene Ther 16, 1016-1027.
204
Pettersen, E. F., Goddard, T. D., Huang, C. C., Couch, G. S., Greenblatt, D. M., Meng, E. C., and
Ferrin, T. E. (2004). UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem 25, 1605-1612.
Piechaczek, C., Fetzer, C., Baiker, A., Bode, J., and Lipps, H. J. (1999). A vector based on the
SV40 origin of replication and chromosomal S/MARs replicates episomally in CHO cells. Nucleic Acids Res 27, 426-428.
Porta, C., Wang, G., Cheng, H., Chen, Z., Baker, T. S., and Johnson, J. E. (1994). Direct Imaging
of Interactions between an Icosahedral Virus and Conjugate Fab Fragments by Cryoelectron Microscopy and X-Ray Crystallography. Virology 204, 777-788.
Prasad, B. V., Burns, J. W., Marietta, E., Estes, M. K., and Chiu, W. (1990). Localization of VP4
neutralization sites in rotavirus by three-dimensional cryo-electron microscopy. Nature 343, 476-479.
Pushko, P., Parker, M., Ludwig, G. V., Davis, N. L., Johnston, R. E., and Smith, J. F. (1997).
Replicon-helper systems from attenuated Venezuelan equine encephalitis virus: expression of heterologous genes in vitro and immunization against heterologous pathogens in vivo. Virology 239, 389-401.
Qing, K., Mah, C., Hansen, J., Zhou, S., Dwarki, V., and Srivastava, A. (1999). Human
fibroblast growth factor receptor 1 is a co-receptor for infection by adeno-associated virus 2. Nat. Med 5, 71-77.
Qiu, J., and Brown, K. E. (1999). Integrin alphaVbeta5 is not involved in adeno-associated virus
type 2 (AAV2) infection. Virology 264, 436-440. Qu, X., Liu, W., Qi, Z., Duan, Z., Zheng, L., Kuang, Z., Zhang, W., and Hou, Y. (2008).
Phospholipase A2-like activity of human bocavirus VP1 unique region. Biochem. Biophys. Res. Commun 365, 158-163.
Rabinowitz, J. E., Xiao, W., and Samulski, R. J. (1999). Insertional mutagenesis of AAV2 capsid
and the production of recombinant virus. Virology 265, 274-285. Rabinowitz, J. E., Rolling, F., Li, C., Conrath, H., Xiao, W., Xiao, X., and Samulski, R. J.
(2002). Cross-packaging of a single adeno-associated virus (AAV) type 2 vector genome into multiple AAV serotypes enables transduction with broad specificity. J. Virol 76, 791-801.
Rapberger, R., Lukas, A., and Mayer, B. (2007). Identification of discontinuous antigenic
determinants on proteins based on shape complementarities. J. Mol. Recognit 20, 113-121.
205
Reading, S. A., and Dimmock, N. J. (2007). Neutralization of animal virus infectivity by antibody. Arch. Virol 152, 1047-1059.
Reaves, P. Y., Gelband, C. H., Wang, H., Yang, H., Lu, D., Berecek, K. H., Katovich, M. J., and
Raizada, M. K. (1999). Permanent Cardiovascular Protection From Hypertension by the AT1 Receptor Antisense Gene Therapy in Hypertensive Rat Offspring. Circ Res 85, e44-50.
Rini, J. M., Schulze-Gahmen, U., and Wilson, I. A. (1992). Structural evidence for induced fit as
a mechanism for antibody-antigen recognition. Science 255, 959-965. Ros, C., Burckhardt, C. J., and Kempf, C. (2002). Cytoplasmic trafficking of minute virus of
mice: low-pH requirement, routing to late endosomes, and proteasome interaction. J. Virol 76, 12634-12645.
Ros, C., and Kempf, C. (2004). The ubiquitin-proteasome machinery is essential for nuclear
translocation of incoming minute virus of mice. Virology 324, 350-360. Rosenfeld, S. J., Yoshimoto, K., Kajigaya, S., Anderson, S., Young, N. S., Field, A., Warrener,
P., Bansal, G., and Collett, M. S. (1992). Unique region of the minor capsid protein of human parvovirus B19 is exposed on the virion surface. J. Clin. Invest 89, 2023-2029.
Rossmann, M. G. (2000). Fitting atomic models into electron-microscopy maps. Acta
Crystallogr. D Biol. Crystallogr 56, 1341-1349. Roy, R., Jerry, D. J., and Thayumanavan, S. (2009). Virus-Inspired Approach to Nonviral Gene
Delivery Vehicles. Biomacromolecules 10, 2189-2193. Ruigrok, R. W., Crépin, T., Hart, D. J., and Cusack, S. (2010). Towards an atomic resolution
understanding of the influenza virus replication machinery. Curr Opin Struct Biol. Available at: http://www.ncbi.nlm.nih.gov.lp.hscl.ufl.edu/pubmed/20061134 [Accessed January 28, 2010].
Russell, D. W., and Miller, A. D. (1996). Foamy virus vectors. J. Virol 70, 217-222. Samulski, R. J., Zhu, X., Xiao, X., Brook, J. D., Housman, D. E., Epstein, N., and Hunter, L. A.
(1991). Targeted integration of adeno-associated virus (AAV) into human chromosome 19. EMBO J 10, 3941-3950.
Sanders, D. A. (2002). No false start for novel pseudotyped vectors. Curr. Opin. Biotechnol 13,
437-442. Sarkar, R., Tetreault, R., Gao, G., Wang, L., Bell, P., Chandler, R., Wilson, J. M., and Kazazian,
H. H. (2004). Total correction of hemophilia A mice with canine FVIII using an AAV 8 serotype. Blood 103, 1253-1260.
206
Sarkis, C., Philippe, S., Mallet, J., and Serguera, C. (2008). Non-integrating lentiviral vectors. Curr Gene Ther 8, 430-437.
Sato, H., Hirata, J., Kuroda, N., Shiraki, H., Maeda, Y., and Okochi, K. (1991). Identification
and mapping of neutralizing epitopes of human parvovirus B19 by using human antibodies. J. Virol 65, 5485-5490.
Scallan, C. D., Jiang, H., Liu, T., Patarroyo-White, S., Sommer, J. M., Zhou, S., Couto, L. B.,
and Pierce, G. F. (2006). Human immunoglobulin inhibits liver transduction by AAV vectors at low AAV2 neutralizing titers in SCID mice. Blood 107, 1810-1817.
Schiedner, G., Morral, N., Parks, R. J., Wu, Y., Koopmans, S. C., Langston, C., Graham, F. L.,
Beaudet, A. L., and Kochanek, S. (1998). Genomic DNA transfer with a high-capacity adenovirus vector results in improved in vivo gene expression and decreased toxicity. Nat. Genet 18, 180-183.
Schlesinger, S. (2001). Alphavirus vectors: development and potential therapeutic applications.
Expert Opin Biol Ther 1, 177-191. Schmaljohn, A. L., Johnson, E. D., Dalrymple, J. M., and Cole, G. A. (1982). Non-neutralizing
monoclonal antibodies can prevent lethal alphavirus encephalitis. Nature 297, 70-72. Schmidt, M., Grot, E., Cervenka, P., Wainer, S., Buck, C., and Chiorini, J. A. (2006).
Identification and characterization of novel adeno-associated virus isolates in ATCC virus stocks. J. Virol 80, 5082-5085.
Schmidt, M., Voutetakis, A., Afione, S., Zheng, C., Mandikian, D., and Chiorini, J. A. (2008).
Adeno-associated virus type 12 (AAV12): a novel AAV serotype with sialic acid- and heparan sulfate proteoglycan-independent transduction activity. J. Virol 82, 1399-1406.
Schneider, B., Fryer, J. F., Oldenburg, J., Brackmann, H., Baylis, S. A., and Eis-Hübinger, A. M.
(2008). Frequency of contamination of coagulation factor concentrates with novel human parvovirus PARV4. Haemophilia 14, 978-986.
Schoenbach, K. H., Beebe, S. J., and Buescher, E. S. (2001). Intracellular effect of ultrashort
electrical pulses. Bioelectromagnetics 22, 440-448. Seiler, M. P., Miller, A. D., Zabner, J., and Halbert, C. L. (2006). Adeno-associated virus types 5
and 6 use distinct receptors for cell entry. Hum. Gene Ther 17, 10-19. Shi, J., and Zheng, D. (2009). An update on gene therapy in China. Curr. Opin. Mol. Ther 11,
547-553.
207
Shi, X., Fang, G., Shi, W., and Bartlett, J. S. (2006). Insertional mutagenesis at positions 520 and 584 of adeno-associated virus type 2 (AAV2) capsid gene and generation of AAV2 vectors with eliminated heparin- binding ability and introduced novel tropism. Hum. Gene Ther 17, 353-361.
de Silva, S., and Bowers, W. J. (2009). Herpes Virus Amplicon Vectors. Viruses 1, 594-629. Simonelli, F., Maguire, A. M., Testa, F., Pierce, E. A., Mingozzi, F., Bennicelli, J. L., Rossi, S.,
Marshall, K., Banfi, S., Surace, E. M., et al. (2009). Gene Therapy for Leber's Congenital Amaurosis is Safe and Effective Through 1.5 Years After Vector Administration. Mol. Ther. Available at: http://www.ncbi.nlm.nih.gov.lp.hscl.ufl.edu/pubmed/19953081 [Accessed January 19, 2010].
Simpson, A. A., Chipman, P. R., Baker, T. S., Tijssen, P., and Rossmann, M. G. (1998). The
structure of an insect parvovirus (Galleria mellonella densovirus) at 3.7 A resolution. Structure 6, 1355-1367.
Simpson, A. A., Hébert, B., Sullivan, G. M., Parrish, C. R., Zádori, Z., Tijssen, P., and
Rossmann, M. G. (2002). The structure of porcine parvovirus: comparison with related viruses. J. Mol. Biol 315, 1189-1198.
Singer, O., and Verma, I. M. (2008). Applications of lentiviral vectors for shRNA delivery and
transgenesis. Curr Gene Ther 8, 483-488. Sinn, P. L., Sauter, S. L., and McCray, P. B. (2005). Gene therapy progress and prospects:
development of improved lentiviral and retroviral vectors--design, biosafety, and production. Gene Ther 12, 1089-1098.
Smith, T. J. (2001). Antibody interactions with rhinovirus: lessons for mechanisms of
neutralization and the role of immunity in viral evolution. Curr. Top. Microbiol. Immunol 260, 1-28.
Smith, T. J. (2003). Structural studies on antibody-virus complexes. Adv. Protein Chem 64, 409-
453. Strassheim, M. L., Gruenberg, A., Veijalainen, P., Sgro, J. Y., and Parrish, C. R. (1994). Two
dominant neutralizing antigenic determinants of canine parvovirus are found on the threefold spike of the virus capsid. Virology 198, 175-184.
Subramaniam, S. (1998). The Biology Workbench--a seamless database and analysis
environment for the biologist. Proteins 32, 1-2. Suh, S. W., Bhat, T. N., Navia, M. A., Cohen, G. H., Rao, D. N., Rudikoff, S., and Davies, D. R.
(1986). The galactan-binding immunoglobulin Fab J539: an X-ray diffraction study at 2.6-A resolution. Proteins 1, 74-80.
208
Suikkanen, S., Antila, M., Jaatinen, A., Vihinen-Ranta, M., and Vuento, M. (2003). Release of canine parvovirus from endocytic vesicles. Virology 316, 267-280.
Summerford, C., Bartlett, J. S., and Samulski, R. J. (1999). AlphaVbeta5 integrin: a co-receptor
for adeno-associated virus type 2 infection. Nat. Med 5, 78-82. Sun, B., Zhang, H., Franco, L. M., Young, S. P., Schneider, A., Bird, A., Amalfitano, A., Chen,
Y., and Koeberl, D. D. (2005). Efficacy of an adeno-associated virus 8-pseudotyped vector in glycogen storage disease type II. Mol. Ther 11, 57-65.
Sundararajan, R. (2009). Nanosecond electroporation: another look. Mol. Biotechnol 41, 69-82. Surosky, R. T., Urabe, M., Godwin, S. G., McQuiston, S. A., Kurtzman, G. J., Ozawa, K., and
Natsoulis, G. (1997). Adeno-associated virus Rep proteins target DNA sequences to a unique locus in the human genome. J Virol 71, 7951-7959.
Tachibana, M., Sparman, M., Sritanaudomchai, H., Ma, H., Clepper, L., Woodward, J., Li, Y.,
Ramsey, C., Kolotushkina, O., and Mitalipov, S. (2009). Mitochondrial gene replacement in primate offspring and embryonic stem cells. Nature 461, 367-372.
Taniyama, Y., Tachibana, K., Hiraoka, K., Aoki, M., Yamamoto, S., Matsumoto, K., Nakamura,
T., Ogihara, T., Kaneda, Y., and Morishita, R. (2002). Development of safe and efficient novel nonviral gene transfer using ultrasound: enhancement of transfection efficiency of naked plasmid DNA in skeletal muscle. Gene Ther 9, 372-380.
Tattersall, P. (2006). The evolution of parvovirus taxonomy. In Parvoviruses, J. R. Kerr, S.
Cotmore, M. E. Bloom, R. Linden, and C. R. Parrish, eds. (Hodder Arnold), pp. 5-14. Terrosi, C., Fabbiani, M., Cellesi, C., and Cusi, M. G. (2007). Human bocavirus detection in an
atopic child affected by pneumonia associated with wheezing. J. Clin. Virol 40, 43-45. The CCP4 suite: programs for protein crystallography (1994). Acta Crystallogr. D Biol.
Crystallogr 50, 760-763. The News Staff (2009). The Runners-Up. Science 326, 1600-1607. Thomas, C. E., Storm, T. A., Huang, Z., and Kay, M. A. (2004). Rapid uncoating of vector
genomes is the key to efficient liver transduction with pseudotyped adeno-associated virus vectors. J. Virol 78, 3110-3122.
Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994). CLUSTAL W: improving the
sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673-4680.
209
Tijssen, P., Szelei, J., and Zádori, Z. (2006). Phospholipase A2 domains in structural proteins of parvoviruses. In Parvoviruses, J. Kerr, S. Cotmore, M. E. Bloom, R. Linden, and C. R. Parrish, eds. (New York, NY: Edward Arnold,Ltd.), pp. 95-105.
Tolfvenstam, T., Lundqvist, A., Levi, M., Wahren, B., and Broliden, K. (2000). Mapping of B-
cell epitopes on human Parvovirus B19 non-structural and structural proteins. Vaccine 19, 758-763.
Trobridge, G. D. (2009). Foamy virus vectors for gene transfer. Expert Opin Biol Ther 9, 1427-
1436. Valori, C. F., Ning, K., Wyles, M., and Azzouz, M. (2008). Development and applications of
non-HIV-based lentiviral vectors in neurological disorders. Curr Gene Ther 8, 406-418. Vandenberghe, L. H., Wilson, J. M., and Gao, G. (2009). Tailoring the AAV vector capsid for
gene therapy. Gene Ther 16, 311-319. Vandenberghe, L. H., and Wilson, J. M. (2007). AAV as an immunogen. Curr Gene Ther 7, 325-
333. Veldwijk, M. R., Topaly, J., Laufs, S., Hengge, U. R., Wenz, F., Zeller, W. J., and Fruehauf, S.
(2002). Development and optimization of a real-time quantitative PCR-based method for the titration of AAV-2 vector stocks. Mol. Ther 6, 272-278.
Vicente, D., Cilla, G., Montes, M., Pérez-Yarza, E. G., and Pérez-Trallero, E. (2007). Human
bocavirus, a respiratory and enteric virus. Emerging Infect. Dis 13, 636-637. Vihinen-Ranta, M., Wang, D., Weichert, W. S., and Parrish, C. R. (2002). The VP1 N-terminal
sequence of canine parvovirus affects nuclear transport of capsids and efficient cell infection. J. Virol 76, 1884-1891.
Voigt, K., Izsvák, Z., and Ivics, Z. (2008). Targeted gene insertion for molecular medicine. J.
Mol. Med 86, 1205-1219. Vorburger, S. A., and Hunt, K. K. (2002). Adenoviral Gene Therapy. Oncologist 7, 46-59. Walters, R. W., Yi, S. M., Keshavjee, S., Brown, K. E., Welsh, M. J., Chiorini, J. A., and Zabner,
J. (2001). Binding of adeno-associated virus type 5 to 2,3-linked sialic acid is required for gene transfer. J. Biol. Chem 276, 20610-20616.
Walters, R. W., Agbandje-McKenna, M., Bowman, V. D., Moninger, T. O., Olson, N. H., Seiler,
M., Chiorini, J. A., Baker, T. S., and Zabner, J. (2004). Structure of adeno-associated virus serotype 5. J. Virol 78, 3361-3371.
Walther, W., and Stein, U. (2000). Viral vectors for gene transfer: a review of their use in the
treatment of human diseases. Drugs 60, 249-271.
210
Wang, G., Porta, C., Chen, Z., Baker, T. S., and Johnson, J. E. (1992). Identification of a Fab
interaction footprint site on an icosahedral virus by cryoelectron microscopy and X-ray crystallography. Nature 355, 275-278. Wang, L., Calcedo, R., Nichols, T. C., Bellinger, D. A., Dillow, A., Verma, I. M., and Wilson, J.
M. (2005). Sustained correction of disease in naive and AAV2-pretreated hemophilia B dogs: AAV2/8-mediated, liver-directed gene therapy. Blood 105, 3079-3086.
Wang, L., Calcedo, R., Wang, H., Bell, P., Grant, R., Vandenberghe, L. H., Sanmiguel, J.,
Morizono, H., Batshaw, M. L., and Wilson, J. M. (2010). The pleiotropic effects of natural AAV infections on liver-directed gene transfer in macaques. Mol. Ther 18, 126-134.
Wang, L., Figueredo, J., Calcedo, R., Lin, J., and Wilson, J. M. (2007). Cross-presentation of
adeno-associated virus serotype 2 capsids activates cytotoxic T cells but does not render hepatocytes effective cytolytic targets. Hum. Gene Ther 18, 185-194.
Wang, Z., Zhu, T., Qiao, C., Zhou, L., Wang, B., Zhang, J., Chen, C., Li, J., and Xiao, X. (2005).
Adeno-associated virus serotype 8 efficiently delivers genes to muscle and heart. Nat. Biotechnol 23, 321-328.
Warrington, K. H., and Herzog, R. W. (2006). Treatment of human disease by adeno-associated
viral gene transfer. Hum Genet 119, 571-603. Waters, B., and Lillicrap, D. (2009). The immunology of gene transfer: an overview. In Gene
Therapy Immunology, R. W. Herzog, ed. (New Jersey: Wiley-Blackwell), pp. 1-18. Weber, M., Rabinowitz, J., Provost, N., Conrath, H., Folliot, S., Briot, D., Chérel, Y., Chenuaud,
P., Samulski, J., Moullier, P., et al. (2003). Recombinant adeno-associated virus serotype 4 mediates unique and exclusive long-term transduction of retinal pigmented epithelium in rat, dog, and nonhuman primate after subretinal delivery. Mol. Ther 7, 774-781.
Weger, S., Wistuba, A., Grimm, D., and Kleinschmidt, J. A. (1997). Control of adeno-associated
virus type 2 cap gene expression: relative influence of helper virus, terminal repeats, and Rep proteins. J. Virol 71, 8437-8447.
Weitzman, M. (2006). The parvovirus life cycle: an introduction to molecular interactions
important for infection. In Parvoviruses, J. R. Kerr, M. E. Bloom, R. Linden, and C. R. Parrish, eds. (New York: Oxford University Press Inc.), pp. 143-56.
Wells, D. J. (2004). Gene Therapy Progress and Prospects: Electroporation and other physical
methods. Gene Ther 11, 1363-1369.
211
Wikoff, W. R., Wang, G., Parrish, C. R., Cheng, R. H., Strassheim, M. L., Baker, T. S., and Rossmann, M. G. (1994b). The structure of a neutralized virus: canine parvovirus complexed with neutralizing antibody fragment. Structure 2, 595-607.
Wilson, J. M. (1996). Adenoviruses as gene-delivery vehicles. N. Engl. J. Med 334, 1185-1187. Wilson, J. M. (2009). Progress in the Commercial-Scale Production of Adeno-Associated Viral
Vectors. Human Gene Therapy 20, 695-695. Wistuba, A., Kern, A., Weger, S., Grimm, D., and Kleinschmidt, J. A. (1997). Subcellular
compartmentalization of adeno-associated virus type 2 assembly. J. Virol 71, 1341-1352. Wistuba, A., Weger, S., Kern, A., and Kleinschmidt, J. A. (1995). Intermediates of adeno-
associated virus type 2 assembly: identification of soluble complexes containing Rep and Cap proteins. J. Virol 69, 5311-5319.
Wolff, J. A., Malone, R. W., Williams, P., Chong, W., Acsadi, G., Jani, A., and Felgner, P. L.
(1990). Direct gene transfer into mouse muscle in vivo. Science 247, 1465-1468. Wu, H., Ceccarelli, D. F., and Frappier, L. (2000). The DNA segregation mechanism of Epstein-
Barr virus nuclear antigen 1. EMBO Rep 1, 140-144. Wu, P., Xiao, W., Conlon, T., Hughes, J., Agbandje-McKenna, M., Ferkol, T., Flotte, T., and
Muzyczka, N. (2000b). Mutational Analysis of the Adeno-Associated Virus Type 2 (AAV2) Capsid Gene and Construction of AAV2 Vectors with Altered Tropism. J Virol 74, 8635-8647.
Wu, Z., Asokan, A., and Samulski, R. J. (2006). Adeno-associated virus serotypes: vector toolkit
for human gene therapy. Mol. Ther 14, 316-327. Wu, Z., Miller, E., Agbandje-McKenna, M., and Samulski, R. J. (2006). Alpha2,3 and alpha2,6
N-linked sialic acids facilitate efficient binding and transduction by adeno-associated virus types 1 and 6. J. Virol 80, 9093-9103.
Xenariou, S., Liang, H., Griesenbach, U., Zhu, J., Farley, R., Somerton, L., Singh, C., Jeffery, P.
K., Scheule, R. K., Cheng, S. H., et al. (2010). Low-frequency ultrasound increases non-viral gene transfer to the mouse lung. Acta Biochim. Biophys. Sin. (Shanghai) 42, 45-51.
Xiao, W., Chirmule, N., Berta, S. C., McCullough, B., Gao, G., and Wilson, J. M. (1999). Gene
therapy vectors based on adeno-associated virus type 1. J. Virol 73, 3994-4003. Xie, Q., Bu, W., Bhatia, S., Hare, J., Somasundaram, T., Azzi, A., and Chapman, M. S. (2002).
The atomic structure of adeno-associated virus (AAV-2), a vector for human gene therapy. Proc. Natl. Acad. Sci. U.S.A 99, 10405-10410.
212
Xiong, C., Levis, R., Shen, P., Schlesinger, S., Rice, C. M., and Huang, H. V. (1989). Sindbis virus: an efficient, broad host range vector for gene expression in animal cells. Science 243, 1188-1191.
Xu, J., Ma, C., Bass, C., and Terwilliger, E. F. (2005). A combination of mutations enhances the
neurotropism of AAV-2. Virology 341, 203-214. Yan, X., Dryden, K. A., Tang, J., and Baker, T. S. (2007). Ab initio random model method
facilitates 3D reconstruction of icosahedral particles. J. Struct. Biol 157, 211-225. Yan, X., Sinkovits, R. S., and Baker, T. S. (2007). AUTO3DEM--an automated and high
throughput program for image reconstruction of icosahedral particles. J. Struct. Biol 157, 73-82.
Yan, Z., Zak, R., Luxton, G. W. G., Ritchie, T. C., Bantel-Schaal, U., and Engelhardt, J. F.
(2002). Ubiquitination of both adeno-associated virus type 2 and 5 capsid proteins affects the transduction efficiency of recombinant vectors. J. Virol 76, 2043-2053.
Yan, Z., Zak, R., Zhang, Y., Ding, W., Godwin, S., Munson, K., Peluso, R., and Engelhardt, J. F.
(2004). Distinct classes of proteasome-modulating agents cooperatively augment recombinant adeno-associated virus type 2 and type 5-mediated transduction from the apical surfaces of human airway epithelia. J. Virol 78, 2863-2874.
Yang, L., Jiang, J., Drouin, L. M., Agbandje-McKenna, M., Chen, C., Qiao, C., Pu, D., Hu, X.,
Wang, D., Li, J., et al. (2009). A myocardium tropic adeno-associated virus (AAV) evolved by DNA shuffling and in vivo selection. Proc. Natl. Acad. Sci. U.S.A 106, 3946-3951.
Yang, Y., Li, Q., Ertl, H. C., and Wilson, J. M. (1995). Cellular and humoral immune responses
to viral antigens create barriers to lung-directed gene therapy with recombinant adenoviruses. J. Virol 69, 2004-2015.
Yates, F., Malassis-Séris, M., Stockholm, D., Bouneaud, C., Larousserie, F., Noguiez-Hellin, P.,
Danos, O., Kohn, D. B., Fischer, A., de Villartay, J., et al. (2002). Gene therapy of RAG-2-/- mice: sustained correction of the immunodeficiency. Blood 100, 3942-3949.
Yates, V. J., el-Mishad, A. M., McCormick, K. J., and Trentin, J. J. (1973). Isolation and
characterization of an Avian adenovirus-associated virus. Infect. Immun 7, 973-980. Yoshimoto, K., Rosenfeld, S., Frickhofen, N., Kennedy, D., Hills, R., Kajigaya, S., and Young,
N. S. (1991). A second neutralizing epitope of B19 parvovirus implicates the spike region in the immune response. J. Virol 65, 7056-7060.
Yu, C., Yuan, Z., Cao, Z., Wang, B., Qiao, C., Li, J., and Xiao, X. (2009). A muscle-targeting
peptide displayed on AAV2 improves muscle tropism on systemic delivery. Gene Ther 16, 953-962.
213
Zádori, Z., Szelei, J., Lacoste, M. C., Li, Y., Gariépy, S., Raymond, P., Allaire, M., Nabi, I. R.,
and Tijssen, P. (2001). A viral phospholipase A2 is required for parvovirus infectivity. Dev. Cell 1, 291-302.
Zaiss, A. K., and Muruve, D. A. (2008). Immunity to adeno-associated virus vectors in animals
and humans: a continued challenge. Gene Ther 15, 808-816. Zaiss, A. K., and Muruve, D. A. (2005). Immune responses to adeno-associated virus vectors.
Curr Gene Ther 5, 323-331. Zanta, M. A., Boussif, O., Adib, A., and Behr, J. P. (1997). In vitro gene delivery to hepatocytes
with galactosylated polyethylenimine. Bioconjug. Chem 8, 839-844. Zhang, H., High, K. A., Wu, Q., Yang, P., Schlachterman, A., Yu, S., Yi, N., Hsu, H., and
Mountz, J. D. (2005). Genetic analysis of the antibody response to AAV2 and factor IX. Mol. Ther 11, 866-874.
Zhang, X., Walker, S. B., Chipman, P. R., Nibert, M. L., and Baker, T. S. (2003). Reovirus
polymerase lambda 3 localized by cryo-electron microscopy of virions at a resolution of 7.6 A. Nat. Struct. Biol 10, 1011-1018.
Zhou, Z. H. (2008). Towards atomic resolution structural determination by single-particle cryo-
electron microscopy. Current Opinion in Structural Biology 18, 218-228. Zolotukhin, S., Byrne, B. J., Mason, E., Zolotukhin, I., Potter, M., Chesnut, K., Summerford, C.,
Samulski, R. J., and Muzyczka, N. (1999). Recombinant adeno-associated virus purification using novel methods improves infectious titer and yield. Gene Ther 6, 973-985.
Zolotukhin, S., Potter, M., Hauswirth, W. W., Guy, J., and Muzyczka, N. (1996). A "humanized"
green fluorescent protein cDNA adapted for high-level expression in mammalian cells. J. Virol 70, 4646-4654.
Zolotukhin, S., Potter, M., Zolotukhin, I., Sakai, Y., Loiler, S., Fraites, T. J., Chiodo, V. A.,
Phillipsberg, T., Muzyczka, N., Hauswirth, W. W., et al. (2002). Production and purification of serotype 1, 2, and 5 recombinant adeno-associated viral vectors. Methods 28, 158-167.
214
BIOGRAPHICAL SKETCH
Brittney Lynn Gurda was born in Hinsdale, Illinois to Thomas and Christine Gurda. Her
family moved from Chicago to Orlando, Florida in the spring of 1985. She has two siblings.
Senior to her is CaryBeth Bevan, who is married to Nathan Bevan. She has one nephew, Corbin
Bailey Stevens, and two nieces, Eden and Rowan Bevan. Her brother, Dean Thomas, is younger
only by 25 months. Like most kids with a love for her pets, Brittney’s original goal was to
become a veterinarian. She graduated from Dr. Philips High School in May 1996 and
immediately moved to New Smyrna Beach to begin work at Beachwood Animal Hospital under
Dr. Donald Needham, D.V.M. While acquiring experience in the veterinary setting she went to
school full time at Daytona Beach Community College. She returned to Orlando, Fl in the
summer of 1997 and continued her undergraduate education at Valencia Community College
while working part-time to fund her schooling. She eventually obtained her Associate in Arts
from VCC in 1999 and began the necessary requirements for a Bachelor in Science in
microbiology and molecular biology at the University of Central Florida. Wanting to eventually
transfer to the University of Florida for vet school, Brittney moved to Gainesville, Florida in
2003. While attending UF, Brittney began undergraduate research under the guidance of Dr.
Mavis Agbandje-McKenna, Ph.D. in the department of Biochemistry and Molecular Biology.
This experience opened her up to a whole new realm of medicine and after graduating in the
spring of 2004 from UF with a BS in microbiology and cell science she acquired a full-time
position with Dr. Agbandje-McKenna as a laboratory technician. This role eventually led to her
acceptance into the Interdisciplinary Program in Biomedical Sciences in the College of Medicine
in the fall of 2005 where she continued her research as a graduate assistant in the Agbandje-
McKenna laboratory.