1 ENDOSOMAL PH MEDIATED STRUCTURAL TRANSITIONS IN ADENO ASSOCIATED VIRUSES By BALASUBRAMANIAN VENKATAKRISHNAN 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 2012
2 2012 Balasubramanian Venkatakrishnan
3 To my mother Sujatha Venkatakrishnan, my father Venkatakrishnan Subramanian, my sister Padmaja Venkatakrishnan and our p recious Scoobie
4 ACKNOWLEDGMENTS I thank my mentors, Dr. Robert McKenna and Dr. Mavis McKenna for their guidance, patience and support. The time I have spent in their lab has gone a long way in molding me from a raw enthusiastic student to a mature res earcher. They have been very involved in my professional and personal growth over the last five years. They have given me a lot of opportuni ty to learn and prove my abilities in their laboratory as well as in the general scientific community I thank them for being gentle teachers and for guiding me in developing my philosophy on science and the world in general. It is my privilege to have worked with them. I thank the past and present members of the McKenna lab for their help and collaboration. They have b een of invaluable help to me in the last five years and I hope to have future collaborators and long term scientific acquaintances in them. I thank my undergraduate student Joseph Yarbrough for his help in this study. I thank my committee members Dr. Joann a Long, Dr. Brian Cain, Dr. Linda Bloom and Dr. Nicholas Muzyczka for their guidance, critique and inputs. I could not have made it this far without the help of the administrative staff of the Biochemistry and Molecular Biology department and the IDP and I thank them for it. I thank my collaborators Dr. Shuo Qian, Dr. Lilin He ( Oak Ridge National Laboratory ) Dr. Richard Gillilan ( Cornell University ) Dr. Brian Bothner, Vamsee Rayaprolu ( Montana State University ) Dr. Olga Kozyreva ( University of North Caro lina ), and Dr. Barry Byrne, Dr. Nicholas Muzyczka and Max Salganik ( University of Florida ) for their assistance and involvement in this study
5 I thank my friends Matt Watts, Dennis Neeld, Lawrence Tartaglia and Mukundh Narayan for their support and counse l during tough times. They have been my family in Gainesville and I am grateful to them for making it a home away from home. I thank my parents Venkatakrishnan Subramanian and Sujatha Venkatakrishnan, my sister Padmaja Venkatakrishnan and my dog Scoobie fo r their love and support. I am grateful for their confidence in me, their support and life lessons that have had a big role in my professional and personal development.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 15 CHAPTER 1 BACKGROUND AND INTRODUCTION ................................ ................................ ..... 17 Adeno Associated Viruses ................................ ................................ ...................... 17 AAV Serotypes and their Tropisms ................................ ................................ .. 18 The Parvovirus Capsid Structure ................................ ................................ ...... 19 The Infectious Pathway of AAVs ................................ ................................ ............. 21 Intracellular Trafficking ................................ ................................ ............................ 23 VP1u Externalization ................................ ................................ ........................ 23 Capsid Stability and Uncoating ................................ ................................ ......... 27 Ov erall Goals ................................ ................................ ................................ .......... 29 2 MATERIALS AND METHODS ................................ ................................ ................... 38 Production and Purification of AAV VLPs and Virions ................................ ............. 38 AAV Capsid VLP Production in Sf9 Cells ................................ ......................... 38 AAV Mutant and DNA packaged Capsid Production in HEK 293 Cells ............ 39 Purification ................................ ................................ ................................ ........ 39 Validation of Purified Sample Quality ................................ ............................... 40 Transmission Electron Microscopy (TEM) ................................ ........................ 41 Circular Dichroism ................................ ................................ ................................ .. 41 Data Collection ................................ ................................ ................................ 41 Data Processing and Refinement ................................ ................................ ..... 42 Secondary Structure Determination Algorithm ................................ ................. 42 Differential Scanning Calorimetry ................................ ................................ ........... 43 Data Collection ................................ ................................ ................................ 43 Data Processing and Refinement ................................ ................................ ..... 44 Small angle Scattering ................................ ................................ ............................ 44 X ray Scat tering Data Collection ................................ ................................ ...... 44 Small angle Neutron Scattering Data Collection ................................ .............. 45 SAXS and SANS Data Processing and Refinement ................................ ......... 46 X ray Crystallography Data Collection and Structure Determination .................... 46 Protease Assays ................................ ................................ ................................ ..... 47
7 Computational Analysis ................................ ................................ .......................... 48 Secondary Structure Prediction and Modeling ................................ ................. 48 pI Calculations ................................ ................................ ................................ .. 49 Intrinsic Disord er Prediction ................................ ................................ .............. 49 3 GLOBAL CHANGES IN THE SOLUTION STRUCTURE OF AAV CAPSIDS AT DIFFERENT ENDOSOMAL PHS ................................ ................................ .............. 54 Crystal Structure of AAV8 at Different pHs ................................ ............................. 54 Small angle Scattering ................................ ................................ ............................ 56 Results ................................ ................................ ................................ .................... 57 Discussion ................................ ................................ ................................ .............. 59 4 STRUCTURAL AND BIOPHYSICAL ANALYSIS OF THE VP1 UNIQUE REGION ... 68 Background ................................ ................................ ................................ ............. 68 Results ................................ ................................ ................................ .................... 69 Ordered Structural S tate of the VP1u ................................ ............................... 71 helicity with Increase in Temperature ................................ ............... 71 helicity with Decrease in pH ................................ .............................. 72 Restoration of Secondary Structure ................................ ................................ 73 AAV5 and AAV8 ................................ ................................ ............................... 73 Discussion ................................ ................................ ................................ .............. 74 5 THE ROLE OF STRUCTURAL STABILITY IN THE INFECTIVE PATHWAY OF ADENO ASSOCIATED VIRUSES ................................ ................................ .............. 92 Background ................................ ................................ ................................ ............. 92 Results ................................ ................................ ................................ .................... 93 Discussion ................................ ................................ ................................ .............. 95 6 NOVEL PROTEOLYTIC ACTIVITY IN ADENO ASSOCIATED VIRUSES ............... 104 Background ................................ ................................ ................................ ........... 104 Results ................................ ................................ ................................ .................. 104 Discussion ................................ ................................ ................................ ............ 106 7 SUMMARY AND FUTURE DIRECTIONS ................................ ................................ 114 LIST OF REFERENCES ................................ ................................ ............................. 119 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 132
8 LIST OF TABLES Table page 1 1 Sequence similarities of the AAV serotype VP1s in comparison to AAV2. .............. 31 3 1 Rg values ( ) of AAV8 obtained from SANS data. All Rg values are expressed in Angstrom units. ................................ ................................ ................................ ... 62 3 2 Rg values ( ) of empty VLPs from SAXS data. ................................ ...................... 63 4 1 Thermal VP1u structural transition temperatures (in C) for AAV1, AAV5 and AAV8*. The numbers in parenthesis are the errors associated with the values base d on triplicate experiments. ................................ ................................ ............. 78 4 2 Percentage helicity from deconvolution of CD spectra of AAV1, AAV5 and AAV8 at different pHs. ................................ ................................ ............................ 79 5 1 Association energies and buried surface areas for AAV8 at pHs 7.5 and 4.0. This data was calculated using the VIPERdb server ................................ ............. 99 5 2 DSC melting temperatures (in C). ................................ ................................ ....... 100 6 1 Amino acid conservation in the catalytic triad region. ................................ ............ 108
9 LIST OF FIGURES Figure page 1 1 Multiple sequence alignment of the VP1 capsid protein in AAV serotypes used in this study. ................................ ................................ ................................ ............ 32 1 2 AAV capsid VPs and their organization. ................................ ................................ .. 33 1 3 Crystal structure of AAV1 capsid VP monomer (PDB ID: 3NG9). ........................... 34 1 4 Interfaces of symmetry on an icosahedral AAV1 (PDB ID: 3NG9) capsid. .............. 35 1 5 Schematic of the AAV life cycle. ................................ ................................ .............. 36 1 6 Equatorial slices of wild type AAV2 3D image reconstructions after incubation at RT and 65 C. ................................ ................................ ................................ ......... 37 2 1 Schematic of the production, purification and characterization of AAV empty capsid VLPs, DNA packaged capsids and capsid mutants for further biophysical analysis ................................ ................................ ................................ 51 2 sheet (red) and random coil (green). ................................ ................................ ................................ ................... 52 2 3 Contrast matching in Small angle Neutron Scattering. ................................ ............ 53 3 1 Conformational changes in pH quartet region with pH. ................................ ........... 64 3 2 Changes in AAV8 DNA propensity at A) pH 7.5 (blue) and B) pH 4.0 (purple). ...... 65 3 3 Pairwise density distr ibution function for AAV8 VLPs at different pHs. ................... 66 3 4 Pairwise distribution function of DNA containing AAV8 capsids. ............................. 67 4 1 Cartoon rendition of superimposition of VP1u models on PLA 2 crystal structure. ... 80 4 2 Surface rendition of superimposition of VP1u on to the capsid monomer from the crystal structure. ................................ ................................ ............................... 81 4 3 PONDR FIT plot showing intrinsic disorder in VP1/2 common region. .................... 82 4 4 Histogram showing pI values across all 12 AAV serotypes. ................................ .... 83 4 5 CD spectrum of AAV VP1u. ................................ ................................ .................... 84 4 6 CD spectrum of AAV1 empty capsid VLPs at different temperatures.. .................... 85 4 7 Electron micrographs of the AAV1 capsid VLPs. ................................ .................... 86
10 4 8 Plot of ellipticity values at 212 nm Vs temperature from CD experiments. .............. 87 4 9 CD spectrum of AAV1 empty capsid VLPs at different pHs. ................................ ... 88 4 10 CD spectrum of empty AAV8 capsids at different pHs. ................................ ......... 89 4 11 CD spectrum of empty AAV5 capsids at di fferent pHs.. ................................ ........ 90 4 12 AAV1 VP1u externalization model. ................................ ................................ ....... 91 5 1 DSC curves for AAV1, AAV5 and AAV8. ................................ .............................. 101 5 2 Electron micrographs of AAVs 1, 5 and 8 at different pHs.. ................................ .. 102 5 3 Electron micrographs of AAVs 1,5 and 8 at different temperatures. ..................... 103 6 1 Western Blot of AAV2 wild type and mutants showing degradation products. ...... 109 6 2 Superimposition of hypothesized aspartic protease sites (AAV2) onto catalytic aspartates of an aspartic protease (HIV1 protease). ................................ ............ 110 6 3 AAV1 capsid structure.. ................................ ................................ ......................... 111 6 4 Protease activity for AAV1, AAV2, AAV5, and AAV8 at pH 7.5 and 5.5. ............... 112 6 5 Protease activity assay plots for AAV1, AAV2, AAV5 and AAV8 in presence of inhibitors. ................................ ................................ ................................ .............. 113
11 LIST OF ABBREVIATION S Angstrom Units AAV Adeno Associated Virus ATP Adenosine Triphosphate Bp base pair C Degree Celsius Ca 2+ Calcium ion CaCl 2 Calcium Chloride CASP Critical Assessment of Techniques for Protein Structure Prediction CD Circular Dichroism CHESS Cornell High Energy Synchrotron Source Cl Chloride ion CO 2 Carbon Dioxide CPV Canine Parvovirus CsCl Cesium Chloride dAMP Deoxy Adenosine Monophosphate dCMP Deoxy Cytidine Monophosphate DEER Double Electron Electron Resonance DMEM DNA Deoxy ribonucleic acid DSC Differential Scanning Calorimetry EDTA ethylenediaminetetraacetic acid EM Electron Microscopy ER Endoplasmic Reticulum FBS Fetal Bovine Serum
12 FHV Flock House Virus FPLC Fast Protein Liquid Chromatography FRET g Rotational unit in terms of gravity GE General Electric Company GFP Green Fluorescent Protein H + Proton HCl Hydrochloric Acid HEK 293 Human Embryonic Kidney 293 cell line HFIR High Flux Isotope Reactor ITR Inverted Terminal Repeats K Kelvin Kb kilobase kDa kilodalton log logarithm M Molar m meter Mg mi lligram MgCl 2 Magnesium Chloride microliter ml milliliter mm millimeter mM MilliMolar mRNA messenger ribonucleic acid MVM Minute Virus in Mice
13 MW Molecular Weight N Normal (unit of concentration) Na 2 HPO 4 Sodium Phosphate (dibasic) NaCl Sodium Chloride NH 4 Cl Ammonium Chloride NLS Nuclear Localization Sequence Nm Nanometer nM nanoMolar N V Nudaurelia capensis omega virus ORF Open Reading Frame ORNL Oak Ridge National Laboratory PAGE Polyacrylamide Gel Electrophoresis PBS Phosphate buffer Saline pH negativ e log of hydrogen ion concentration pI Isoelectric point pKa negative log of acid dissociation constant PLA 2 Phospholipase A 2 rAAV Recombinant Adeno Associated Virus Rep AAV Replication protein gene Rg Radius of gyration RNA Ribonucleic acid Rpm Rotations Per Minute SANS Small angle Neutron Scattering SAXS Small angle X ray Scattering SDS Sodium Dodecyl Sulfate
14 Sf9 Spodoptera Frugiperda (Fall Armyworm) clonal isolate cell line SFM Serum Free Media ssDNA Single stranded Deoxy ribonucleic acid T Triangulation number TM Trademark Tris tris(hydroxymethyl)aminomethane Triton X 100 polyoxyethylene octyl phenyl ether a nonionic surfactant UV Ultraviolet Vis Visible light VLP Virus Like Particle VP Viral protein VP1u VP1 unique N terminal region v/v Volume per Vo lume wt wild type w/v Weight per Volume X Magnification
15 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 ENDOSOMAL PH MEDIATED STRUCTURAL TRANSITIONS IN ADENO ASSOCIATED VIRUSES By Balasubramanian Venkatakrishnan May 2012 Chair: Robert McKenna Major: Medical Sciences Biochemistry and Molecular Biology Adeno Associated Viruses (AAVs) are single stranded DNA viruses that bel ong to the Par v ovi ridae family. The AAV infective pathway involves the binding of the virus to a cell surface receptor and subsequent clathrin mediated endocytosis Inside the endosome, the virus capsid undergoes structural changes that permit the virus t o escape the endosome and traffick to the nucleus. There are s ome clues, to what structural changes may occur in the capsid during virus trafficking The VP1 capsid protein contains a unique (VP1u) phospholipase A 2 domain (PLA 2 ) and two nuclear localizati on signals. The phospholipase domain is required to remodel the endosom e to allow escape and the nuclear localization signal is required for nuclear entry. While the VP1u and the VP1/2 common region are not seen in the crystal structures of the viruses, an tibody assays confirm that the VP1u is not exposed on the surface of the capsid A ntibody assays have identified that heating the capsid to ~ 65 C expose s the VP1u. The externalization event has been shown to be crucial because capsids that are directly injected into the cytoplasm or the nucleus are ineffective in infection.
16 T he cellular trigger for AAV VP1u externalization in the endosome is unknown. T his study demonstrates that the VP1u is structurally ordered in solution in a predom helical. This secondary structural state is transition ed by temperature a nd/or pH where a gradual loss of secondary structural propensity signal was seen with increasing temperature and /or decreasing pH. When the pH was restored to 7.5 the second ary structural signal was also restored. E lectron microscopy confirmed that the capsids were intact when the pH was decreased or temperature, increased Di fferential Scanning Calorimetry and Small angle scattering analysis show mild structural and stabili ty changes in the capsid in response to pH. Small angle Neutron Scattering analysis confirmed that there was a genomic rearrangement event that accompanied the structural changes seen in the capsid. A novel autoproteolytic activity was identified in the AA V capsids that could have a role in capsid disa ssembly This data, taken as a whole, is the first physical evidence of the structural state of the VP1u during the effect of endosomal pH states
17 CHAPTER 1 BACKGROUND AND INTRO DUCTION Viruses are ubiquitous and infect everything from bacteria to humans. In order to be able to infect such a large diverse group of organisms, they are known to have evolved several strategies. These strategies are employed at every stage of the virus life cycle from host cell recognition and intracellular trafficking, to replication assembly and release of new progeny. These strategies have been optimized to effectiv ely evade host defens e mechanisms and use the host machinery for viral infectivity and replication. Viral infectivity is achieved despite having a relatively simple organization of a genetic material packaged by a capsid /lipid coat. The genetic material co uld be either DNA or RNA in single stranded or double stranded form. Based on the presence of a lipid membrane in the capsid coat, viruses are classified into either envelope viruses (lipid coat present) or non envelope viruses (lipid coat absent) (39) Nat urally, these differences in genetic material and coat composition lead to different strategies in host recognition, trafficking, replication, assembly and release. Since the genetic material is primarily involved in the replication process, the capsid is an important determinant if infectivity in the virus life cycle (39) Adeno Associated Viruses Adeno Associated Viruses (AAVs) are ssDNA viruses of the parvoviridae family (64) They are non envelope viruses that package a ~ 4.7 kb nucleotide genome (64) The y belong to the Dependovirus genus and infect many different vertebrates including humans. They are nonpathogenic and replicate productively only in the presence of a helper virus, such as adenovirus or herpes virus, which supplies a number of early functions for AAV that are required for gene expression and replication (10) Like the
18 related a utonomous parvoviruses, AAV capsids are icosahedral (approximately 260 in diameter) (64) The viral genome has two open reading frames (orf s ), rep and cap (64) The rep orf codes for four overlapping proteins required for replication and DNA packaging (6 4) Three capsid proteins (VPs) are made from two alternately spliced mRNAs from the cap orf (Figure 1 1A) One of these messages contains the entire capsid orf and encodes VP1. The other mRNA encodes for VP2, from an alternate start codon (ACG), and VP 3 from a conventional downstream ATG. VP3 is 61 kDa and (87 kDa) and VP2 (73 kDa), share the same C terminal amino acid sequence with VP3 but have additional N terminal sequences (64) A total of 60 copies of the three viral structural proteins, VP1, VP2 and VP3, in a predicted ratio of 1:1:8/10, form the 60 subunit T=1 icosahedral viral particle (Figure 1 2 B) (2 1) The only essential cis active sequences in AAV are the 145 bp terminal repeats (ITRs) which function as origins for DNA replication, packaging sequ ences and integration sites (79) Recombinant AAV vectors (rAAV) are generated by retaining the ITRs and replacing the internal wild type (wt) AAV coding sequences with therapeutic genes (56, 79) rAAVs infect both dividing and non dividing cells and can establish latency for the life of the an imal in non dividing cells (42) AAV S erotypes and their T ropisms Twelve distinct human and non human primate AAV serotypes (AAV1 12) have been sequenced to date, and numerous rAAVs h ave been identified in primate and human tissues through PCR studies (43, 84, 105) Their sequence similarities are described in Table 1 1. The viruses are not associated with any disease, are non toxic, and can package and deliver foreign genes to target cells. Considerable interest has
19 been generated in their development as gene delivery vectors and numerous studies show that each virus h as unique cellular transduction characteristics (16, 25, 43, 62, 84, 99, 121) Mo st of the gene therapy applications to date have been with AAV2, the most intensively studied serotype, but other serotypes have now been shown to have more promise for gene delivery to certain tissues. AAV1, for example, can transduce rodent skeletal musc le as much as a thousand fold more efficiently than AAV2 (20 ) and is now in a human clinical trial for muscle delivery. AAV5 has more diverse targeting, in addition to being more efficient at transducing neuronal and lung tissues (16, 25) For 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 f or transducing the liver (43, 44) and also show significant promise for muscle transduction (58, 91) Gene and protein expression by the AAVs have been shown to essentially last for the life span of the animal in numerous rodent studies, and seve ral years in large animals (85, 101) The recent success of AAV2 mediated gene transfer for the treatment of blindness (75) highlights the potential of these vectors and generated a considerable amount of national and international media attention along with public interest in the use of AAV vectors. The P arvovirus C apsid S tructure The parvovirus VPs are capable of perform ing a wide variety of structural and biological functions during the viral life cycle. These functions include host cell surface receptor recognition, endosomal entry and trafficking, viral genomic encapsidation, self assembly into capsids, maturation of v irions to produce infectious virus progeny, nuclear import (for assembly) and export (after assembly) and host immune response detection and evasion (53) The relatively small parvoviral genome (~ 4.7 kb) has allowed the use
20 of genetic manipulation to identify functional domains/regions of the VPs/capsid. Towards correlating these functions with structural features of the capsid in the AAV serotypes structures have been determined for AAV1, AAV2, AAV3B, AAV4, AAV5, AAV6, AAV7, AAV8 and AAV9 (47, 72, 82, 83, 87, 88, 98, 120, 131) A combination of sixty parvoviral VP1, 2, 3 ( and 4 for the densoviruses) build up the T=1 icosahedral capsid (Figure 1 2 B) in a 1:1:10 ratio (64) Only the common C terminal VP3 region (~ 53 0 a mino acids ) is obse rved in crystal structures (Figure 1 3 ) The unique VP1 N terminal region (VP1u) the N terminal ~ 40 60 amino acids of VP2 that overlaps with VP1 (VP1/2 common region) and the first ~ 15 24 residues of VP3, that are located inside the assembled capsid (13, 24, 36, 119, 122) have been disordered in the crystal structur es Th e f ocus of this dissertation is mainly on the N terminal VP1u region. This l ack of N terminal VP ordering has been attributed to the low copy numbers of VP1 and VP2 in mature capsids and to the possibility that the N termini of VP1, VP2 and VP3 adopt different conformations in the capsid. These conformations c ould be averaged out from the icosahedr al symmetry assumed during structure determination. Interestingly, cryo EM studies of heat shocked AAV2 capsids located inside the capsid at the icosahedral 2 fold axis that have been interpreted as the N term inal regions of VP1 and VP2 (69) but the structural topology of these regions remain to be elucidated. The structural topology of the common VP region is highly conserved, even for parvovirus members that are ~ 20% or less identical at the amino acid sequ ence level The V Ps consist of a core eight stranded anti barrel ( forms the contiguous capsid shell, with loop insertions between the strands forming the
21 surface (Figure 1 3 ) strand structure and var iable regions seen in all parvoviruses The major capsid surface features include depress ions at the icosahedral 2 fold symmetry axis and surrounding the 5 fold axis (blue surface regions) and protrusions at or surr ounding the 3 fold axes ( red surface regions). The floor of the 2 fold depression is only one polypeptide chain thick, making this the thinnest regions of the capsid (Figure 1 4 A) A the wall of the 2 fold depression The 3 fold protrusions are formed from the intertwining of loops from 3 fold symmetry related VP3 monomers and are the most variable regions within the parvovirus capsid with respect to sequence and structure (Figure 1 4 B) Two small stretches of a radial ribbon, which clustered at 5 fold icosahedral symmetry related VPs (Figure 1 4 C) This forms a conserved cylindrical channel that connects the interior to the exterior of the capsid A structurally conserved loop loop) for ms the most extensive 5 fold related VP contacts and lies o n the 5 fold surface. The I nfectious P athway of AAVs Structural and mutational analyses have demonstrated that parvovirus host tropism, transduction properties of AAV vectors, and antigenic differe nces arise from variations in surfa ce amino acids (53) Cell transduction phenotypes for the AAVs are dictated by the utilization of different cell surface glycans for cell binding and entry (5 3) Figure 1 5 is a simple schematic of the life cycle of the AAVs. Heparan sulfate proteoglycan, the first receptor identified for an AAV virus (116) appears to function primarily in attachment of AAV2 and AAV3 serotypes to the cell surface (89) whe reas AAV1, which is ~ 2 2 6 N linked sialic acid but not heparin (25, 99, 129) AAV4 and AAV5, which are ~ 55% identical to each other
22 2 3 O li nked sialic acids, 2 3 N linked sialic acids (62, 121) Platelet derived growth factor receptor is the protein mediating AAV5 infection (93) AAV6, a recombinant of AAV1 and AAV2, binds both heparin sulfate and sialic acid (88, 129) The receptors utilized by AAV7 12 for cellular transduction have yet to be determined (106) Studies aimed at deciphering AAV tissue tropism and transduction properties have utilized mutagenesis, cell binding, and transduction assays, mainly on AAV2, to identify individual capsid protein amino acid sequences that play a central role in cellular re cognition and transduction. Mutagenic studies have identified the amino acids R588, with R585 and R588 being the most critical for this interaction (63, 89, 129) The availability of the high resolution 3D structure for AAV2 (131) enabled the mapping of the AAV2 heparin binding site. Residues are contributed from symmetry related VP monomers at the wall of the protrusions surrounding the icosahedral 3 fold axes and in the valley that runs from the 2 to 3 fold axis (63, 89, 129) C ryo EM and image reconstruction studies of an AAV2 heparan sulfate oligosaccharide complex have confirm ed the heparin binding site at the icosahedral 3 fold axis (73) X ray crys tallographic studies of AAV5 complexed with sialic acid, glycan array analysis, and mutagenesis to map the receptor binding site for this serotype also point to the icosahedral 3 binding region (unpubl ished) Interestingly, the receptor binding site for these two viruses, which represent two of the most distantly related human AAV serotypes, is located adjacent to common variable surface loop regions on the caps id surface Mutational analysis of the
23 AAV 1, AAV2 and AAV8 capsids also identifies common variable regions as being involved in the determination of transduction efficiency (54, 99, 109) These observations suggest that analogous AAV capsid regions have evolved to perform similar functions In this study, we use AAV1, AAV2, AAV5, AAV6 and AAV8 to represent the different clades in the AAVs to understand fundamental processes in the AAV life cycle. Intracellular T rafficking The best characteri zed AAV with respect to cellular entry and trafficking is AAV2 (53) Once the AAV2 capsid binds heparin sulfate, entry occurs via dynamin dependent clathrin mediated endocytosis following interaction with co receptors (108, 114) The acidic environment of the endosome is essential for virus infection, as inhibitors of the vacuolar H + ATPases such as bafilomycin A1 or trea tment of cells with NH 4 Cl, both inhibit AAV transduction and reduce trafficking to the nucleus (7, 30, 32, 52, 105) VP1u E xternalization S hortly after entering the early endosome, the N termini of the minor capsid proteins VP1 and VP2 become externalized on the capsid surface while the capsid remains assembled (111) This serves two purposes; it exposes a phospholipase A 2 (PLA 2 ) activity present in VP1 and nuclear localization sequences (NLSs) that are present i n both VP1 and VP2 N termini on the capsid surface. Genetic studies have shown that both the PLA 2 and NLS are necessary for efficient infection (12, 13, 35, 36, 77) Mutation of either region results in a 1 to 3 log reduction in infection o r transduction (35, 105, 109, 122, 130) This PLA 2 domain has been identified in in VP1u sequence of ~ 30 different parvoviruses (17) Its sequence similarity with more potent PL A 2 s is very weak and is mainly restricted to the catalytic site hist idine and aspartate residues and
24 the G X G calcium binding motif The parvo viral PL A 2 motifs lack cysteines, unlike all other previously characterized PL A 2 s In contrast, multiple disulfide bonds are a hallmark of all nonparvoviral PL A 2 s and are used as the basis for their classification. Moreover, viral PL A 2 helices that contain the act ive site residues of classical PL A 2 s The parvoviral PL A 2 s have display specific activities that are 100 1,000 f old lower than the most active PLA 2 s toward phospholipid vesicles (17, 110) The reasons for the relatively low activity of a subset of these PLA2s is not obvious from a comparison of their amino acid sequences of these enzymes as well as the x ray structures of a subset of them. It is possible that the physiological substrates for these low activity enzymes have structures distinct from those of standard phospholipids. Non parvoviral PLA 2 s are known as key enzymes in lipid membrane metabolism, signal transduction pathways, inflammation, acute hypersensitivity, and degenerative diseases (4, 27, 67) They hydrolyze phos pholipid substrates at the 2 acyl ester (sn 2) position to release lysopho spholipids and free fatty acids (134) Parvoviral PLA 2 shows the same enzymatic activity when exp ressed as a recombinant protein (17) The sequence similarity of non parvoviral and parvoviral sPLA2s is largely confined to the catalytic HDXXY domain and to the calcium binding GXG moti f (17) Recently, Farr et al reported that polyethyleneimine (PEI) induced endosomal rupture or co infection with endosomolytically active adenovir uses partially rescued the inf ectivity of a catalytic center PLA 2 mutant (H42R) of the autonomously replicating parvovirus Minute Virus of Mice (MVM). They concluded that the PLA 2 activity plays a role in breaching the endosomal membrane to facilitate endo somal escape of incoming MVM particles (37)
25 Treatment of particles in vitro with the acidic pHs of the early endosome has been shown to induce the externalization of the VP1/2 N terminus in autonomous canine parvovirus (CPV ) and MVM (76, 111) However, acidic pHs do not induce VP1/2 N termini extrusion in AAV2, suggesting that some combination of factors, perhaps receptor b inding and pH, is necessary to cause the required capsid structural changes (111) It is possible that antibody binding is affected by change in pH in AAV2 (but not in CPV) and therefore the process of externalization needs to be studied more biophysical ly. A significant amount of the capsid is released into the cytoplasm: one current model is that the function of the PLA 2 activity is to rupture the endosome so that capsids are released (37, 111, 115) and then the NLS sequences facilitate vi rus trafficking to the nucleus (49, 74, 111) However, not all parvovirus PLA 2 activities appear to rupture endosomes in a way that would easily release virus particles. Studies with CPV showed that PLA 2 activity enabled the release of dextran molecules of 3kDa but not 10kDa, well below the capsid MW of ~ 4000 kDa (115) Comparative studies have shown a 3 log difference in PLA 2 enzymatic activity between porcine parvovirus and AAV2 (17, 115) What seems clear is that a structural change in the virus capsid, either during or early after entry, is necessary for efficient infection by promoting endosome escape, nu clear entry, and/or DNA uncoating. This structural change is at least in part responsible for the extrusion of the VP1/2 N termini. Although the extrusion of the VP1 N terminus is a feature common to all parvoviruses and appears to be essential for efficie nt infectivity, the biological mechanism that triggers this event is unclear. The channel at the 5 fold axis of the parvovirus capsid is postulated as the site of externalization of the PLA 2 activity and the N termini of VP2 in capsids (134)
26 Mutational analysis also supports the proposal that this channel is the pore through which the viral DNA is threaded into the capsid during particle assembly (13) a nd release during infection (36) Although it is appreciated that capsid protein dynamics play a centra l role in parvovirus cell binding, entry, trafficking, DNA release, and egress following assembly, few studies have addressed this question. Although it is generally believed that AAV is released into the cytoplasm from the endosome, a substantial portion of the capsids remain present in an endosomal compartm ent for 8 20 h ou rs post infection (108, 111) The delayed onset of maximum gene expression after transduction in vivo (2 6 weeks) is believed to be due, at least in part, to slow trafficking to the nucleus following infection, as well as to the fact that the viral genome must synthesize a second strand prior to transcr iption (38, 40, 55, 99, 117) Evidence for this comes from the recent observation by the Kay group that AAV8 produced a much faster onset of expression than AAV2 in spite of both types of capsids having the same AAV2 derived recombinant g enome ( 117) Their work and that of others suggests that AAV8 is significantly more efficient in delivering its payload to the nucleus than AAV2 and that intact capsids of AAV2 persist in infected cells for extended periods post infection prior to DNA release or degradation (55, 99, 117) During infection virus is found in the early endosome, recycling endosome, late endosome, and lysosomal fractions (7, 31, 32, 51, 52, 105, 130) One sero type, AAV5, has been observed in th e Golgi (6) and is capable of transcytosis (92) The question of which virus fraction is critical for infection, the cytoplasmic or endosomal fraction, has not been resolved. A recent report suggests that the recycling endosome compartment, which becomes po pulated only at higher MOI, is the most efficient route for cell transduction
27 (31) The proteasome pathway also impacts viral infectivity. Proteasome inhibitors generally increase AAV transduction, albeit to different levels in different cell types (32, 133) This is in contrast to CPV, where proteasome inhibitors reduce infectivity (103) Another unresolved issue in par vovirus infection is the location of viral DNA uncoating. AAV accumulates around the nucleus and some intact virus is found within the nucleus (38, 111) It remains unclear if uncoating occurs in the cytoplasm or whether intact virus enters the nucleus and then uncoats; evidence for both possibilities has been presented (7, 52, 55, 76, 103, 108, 130) Capsid S tability and U ncoating The AAV capsid is more thermally stable than most proteins Particles remain intact at pH 4.0 and can tolerate temperatures of 65 C for up to 60 min utes with only a m odest reduction in infectivity (1 3) Calorimetry experiments demonstrate a unique transition that occurs at 70 80 C similar to that seen with MVM (24, 1 02) Moreover, computer analyses of buried surface and association energies at the icosahedral 2 3 and 5 fold symmetry related VP interfaces for the crystal structures of AAV2 and AAV8 revealed that the two viruses have similar values (87) Thus, th e difference in onset of gene expression in vivo between AAV2 and AAV8 (24, 102) is likely to be due to specific variable surface features of the capsid and their cellular interactions rather than the overall stability of the capsids. This implies that in addition to the early structural changes that result in VP1/VP2 extrusion, there may be additional caps id associated stru ctural changes and events that are required for uncoating. Given that an endosomal acidic environment is essential for AAV infection (111) the hypothesi s is that the trafficking events that occur prior to gene expression are due to subtle pH mediated transitions of the capsid which facilitate the interaction of the virus with cellular proteins
28 that lead to more efficient trafficking, ubiquitination, nuclear entry, or uncoating. This study f ocus es on identifying via biophysical analysis, the p hysiological transitions in the AAV capsids The transition conditions include isolated capsids, receptor interaction at physiological pH, and som e conditions encountered in the endocytic pathway. The externalization of the N terminus of VP1 for its PLA 2 function is important for parvovirus infectivity (24) yet as discussed above with the large number of structures available for these viruses, there is no structural information on this capsid region, other terminal region of VP1 in low resolution cryo EM density of heat treated AAV2 (69) (Figure 1 6 ) All the parvovirus structures determined so far other than for B19, suggest that N termini of VP1 and VP2 reside inside the parvovirus virion, supporting a proposal that capsid dynamics and rearrangements must be an essential component of parvovirus trafficking to enable exposure of these domains for endosomal escape. The crystal and cryo EM capsid structures available for these viruses represent low energy c onformations since it is known that in solution both reversible fluctuations in protein domains and large scale subunit rearrangements during maturation and infection occur for several others viruses (126) As an example, for poliovirus, exposure of the N terminus (assayed by trypsin and Staphylococcus V8 protease sensitivity) upon receptor binding i s essential for infectivity (41) and the VP4 polypeptide found on the inside surface of the virion is released early in infection while the particle is still intact. For many viruses, internalization and subsequent acidification of the endosome results in the rearrangement of capsid proteins (14, 33, 70, 94) These solution based experiments show that domains of the subunit polypeptide that are clearly internal to the virion in the
29 crystal structure become intermittently exposed to the outer surface. Similar large scale re arrangements are involved in the externalization of the parvovirus VP1 and VP2 N termini. It is hypothesi zed that capsid protein dynamics occurs during AAV endocytosis in response to receptor/pH/environmental changes without disrupting capsid integrity. The externalized VP2 N terminal domain of MVM is crucial for nuclear exit of DNA filled mature particles (76) and MVM VP2 cleavage by cellular proteases during entry removes approximately 20 N terminal amino acids (24) in a maturation process. This rearra ngement of MVM VP2 can be mi micked by heating the virus (57) For AAV2 heat is also able to expose VP1u but unlike the autonomous parvoviruses, pH alone is unable to trigger this rearrange ment as previously mentioned (57, 111) This observation suggests that there is a fundamental difference in the trafficking requirements for viruses that do not require helper function for infectivity versus those that do. Overall Goals This study focuses on the structural transition s that occur in the AAV capsid in endosomal pH conditions. These structural changes include the externalization of the VP1u domain. Biophysical tools including small angle scattering, Circular Dichroism (CD) and Differential Scanning Calorimetry (DSC) are used to characterize the steps in the externalization process along with the structural transitions that occur in other regions of the capsid. This study help s in the gaps in our understanding of the intracellular trafficking process in the AAVs. Additionally, it provides tools to biophysically characterize regions of the capsid that are not seen in the crystal structure. It is important to look at the stability of the capsid in synergy with these structural transitions because the capsid would have to disassemble at some point prior
30 to viral DNA replication. This study uses thermal stability as a parameter for structural stability of the capsid under different pH and receptor conditions.
31 Table 1 1 Sequence similarities of the AAV serotype VP1s in comparison to AAV2. Serotype Similarity AAV1 83.70% AAV3 89.16% AAV4 57.30% AAV5 57.06% AAV6 83.65% AAV7 82.39% AAV8 82.20% AAV9 81.21%
32 Figure 1 1 Multiple sequence alignment of the VP1 capsid protein in AAV serotypes used in this study. A sequence alignment of AAV1, AAV2, AAV5, AAV6 and AAV8 VP1 protein sequence show regions with identity (yellow) and similarity (blue). A consensus sequence is also shown below the alignment. The sequence alignment was done using ClustalW (118)
33 Figure 1 2 AAV ca psid VPs and their organization. A) AAV capsid VPs are coded for by the same orf ( cap ) in green The VP1 contains entire VP2 sequence and the VP1u. The VP2 contains the entire VP3 sequence and the VP1/2 common region. The other gene, rep (blue) and the flanking Inverted Terminal Repeat regions (ITRs) make the AAV genome. This figure has been adapted from Blechacz and Russell (2004) (11) B) The AAV1 capsid is a T = 1 icosahedral capsid made of 60 VP mon omers. The 3 fold, 2 fold and 5 fold symmetrical elements are shown with surface peaks shown in red and valleys in blue. A) B)
34 Figure 1 3 Crystal structure of AAV1 capsid VP monomer (PDB ID: 3NG9) strands helices in red and loops i n orange. The dotted lines show the relative positions of the 5 fold, 3 fold and 2 fold interfaces of symmetry from the center of the capsid. A core 8 barrel forms the fundamental unit of the monomer flanked by variable loop regions and other co nserved regions.
35 Figure 1 4 Interfaces of symmetry on an icosahedral AAV 1 (PDB ID: 3NG9) capsid. A) The 2 fold interface B) The 3 fold interface C) The 5 fold interface 3 A) B) ) C)
36 Figure 1 5 Schematic of the AAV life cycle. The AAV capsid (red and blue) trafficks to the cell surface after which it is internalized by endocytosis. It escapes the endosome and trafficks to the nucleus. Inside the nucleus the AAV genome is replicated (with helper genes from a co infected autonomous virus) and transcribed to create more capsids that are then packaged with viral DNA inside the nucleus. Lipids are shown in yellow with cell s urface receptors in green.
37 Figure 1 6 Equatorial slices of wild type AAV 2 3D image reconstructions af ter incubation at RT and 65 C. From left to right, the image shows empty capsids and capsids with increasing sizes of packaged DNA. This image has been adapted from Kronenberg et al. (69) Empty Full
38 CHAPTER 2 MATERIALS AND METHOD S This C hapter describes common experimental methods and reagents utilized in C hapters 3, 4, 5 and 6 of this dissertation The first section of the C hapter describes the methods used to purify and produce large quantities of AAV VLPs and virions for subsequent experiments. The later sections describe the different experiments and assays done with the purified AAV VLPs and virions. The final section des cribes computational techniques and programs used in the study. Production and P urification of AAV VLPs and V irions Depending on the quantity required and the type of sample needed, the AAV capsids were produced in either Sf9 insect cells or HEK 293 cells. The cells were subsequently lysed and the VLPs or virions were purified from them. AAV C apsid VLP P roduction in Sf9 C ells Sf9 cells were used to produce AAV capsid VLPs that did not encapsida te DNA R ecombinant baculoviral vector s for AAV1, AAV2, AAV5, AAV6 and AAV8 were used to create the respective capsid VLPs. These vectors were transposed into S f9 cells and harvested 72 hours post transfection .The recombinant baculoviral vectors were a gift from the lab of Dr. Sergei Zolotukin (University of F lorida) The cells were infected at a multiplicity of infection of 5.0 plaque forming units per cell. The cells were grown in Erlenmeyer flasks at 300 K using Sf 900 II SFM media (Gibco/Invitrogen Corporation). The culture was then centrifuged at low speed (3000 rpm) to harvest and pellet the cells and the pellet was frozen before lysis and purification.
39 AAV Mutant and DNA packaged Capsid Production in HEK 293 Cells HEK 293 cells were maintained in DMEM supplemented with penicillin and streptomycin at 100 U/ml and 10% FBS and they were maintained in 15 ml petri dishes at 37 C and 5% CO 2 40 such plates (one cell factory) were used to produce enough quantity of the AAV mutant VLP. The cells were passaged to achieve 75% confluence th e next day. Each plate was then transfected with three different plasmids pIM45 (18 g), pXX6 (54g) and UF11 (18 g). pIM45 was used to code for the capsid VP proteins. pXX6, contained the Ad helper genes needed for successful infection. UF11 contained th e GFP gene driven by the CMV promoter and AAV terminal repeats. The UF11 was included to produce AAV capsids with packaged DNA coding for GFP. The plasmids were CsCl purified and used to transfect the cells by calcium phosphate precipitation and incubated for 48 hours at 37 C. The cells were then harvested by centrifugation at 1140g for 20 minutes and the pellet resuspended in 1 ml lysis buffer (1xTD with protease inhibitor cocktail). The sample was then freeze thawed 3 times, treated with Benzonase for 30 minutes at 37 C and the cell lysate clarified by centrifugation at 3700xg for 20 minutes. P urification A complete schematic representation of the purification process is shown in F igure 2 1. The cells were lysed by three freeze thaw cycles in lysis buff er (50 mM Tris HCl pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.2% Triton X 100 ) with B enzonase added after the second cycle to remove the nucleic acid content The sample was then centrifuged at 12100g in a Beckmann J2 centrifuge for 15 minutes at 4 C to separate t he virus capsids from the c ell debris. The virus capsids were then pelleted by sucrose cushion (20% w/v sucrose in 50 mM Tris, 100 mM NaCl, 1 mM EDTA and 0.2% Triton
40 X 100) centrifugation at 149,000 g for 3 hours at 4C in vacuum using a 70Ti rotor The pellet was further purified by sucrose gradient (10 40% w/v su crose in 25 mM Tris, 100 mM NaCl, 0.2% Triton X 100 and 2 mM MgCl 2 ) centrifugation at 151,00 0g for 3 hours at 4 C in vacuum using a SW41Ti rotor For AAV2, the supernatant was then loaded onto a 15 60% iodixanol step gradient. The iodixanol gradient was then centrifuged at 55000rpm for 2 hours. For some of the samples requiring additional purification, a HiTrap TM Q column (GE Healthcare) chromatography step was used in an FPLC setup. 1 ml fractions of the elu a te were collected from the column and chromatograms were used to select the fractions that contained the purified VLPs or virions. Validation of Purified Sample Quality The purified samples were validated by UV/Vis Spectrometry (260/2 80 /310 nm) in quartz cuvettes with 1 cm path length to check for protein content using the equation Where A is the measured absorbance, c is the concentration, l is the path length of the cuvette used and is the molar ext inction coefficient ( 1 .7 for empty AAV capsid VLPs for concentration in mg/ml) 10% SDS PAGE stained with Coomassie blue stain was used to check for protein purity E lectron microscopy (50,000X) was used to check for capsid integrity. The purified samples were then exchanged into desired buffers These protocols have been described previously (135) The samples were then buffer exchanged by either dialysis or filter centrifugation into Phosphate Citrate buffers (with 150 mM NaCl) using specific ratios of 0.2 M Na 2 HPO 4 and 0.1 M Citrate to final pHs of 7.5, 6.0, 5.5 and 4.0. 1 00,000 kDa cutoff Apollo centrifugal concentrators (Orbital Biosciences) were used for the filter centrifugation.
41 Transmission Electron Microscopy ( TEM) After buffer exchanging, 5 L of sample was loaded onto a carbon coated copper grid and allowed to settle for 5 minutes. The drop was then drained off the grid using a Wattmann filter paper and the grid was washed with 6 8 drops of ddH 2 O. The grid was then negatively stained with 5 L Nano W (Nanoprobes) stain for 1 minute after which the excess stain was drained using a Wattmann filter paper. The grid w as then examined in a JOEL 1200 EX transmission electron microscope at 50,000X magnification to check for capsid integrity. Circular Dichroism CD is a spectro photo metric technique used to determine the secondary structural state of a protein. It is based on the differential absorption of left and right circularly polarized light by a protein. Bas ed on the secondary structural state, different proteins absorb circularly polarized light differently in the near UV region (180 260 nm). The absorption spectrum is then deconvoluted mathematically to determine the amino acid fractional content of seconda ry structural elements in the protein. Data C ollection All the CD experiments were done on an Aviv model 410 Circular Dichroism Spectrometer. All the data was collected between the wavelengths of 200 a nd 260 nm with sample concentrations of 0.4 mg/ml in t riplicates with 50 scans per experiment taken at 30 C. For experiments that measured the thermal transition temperatures of the AAVs, CD spectra were collected between 200 and 260 nm between 30 C to 90 C with three degree intervals (21 spectra). 1 scan was collected for every wavelength (61 total scans) and this scan was averaged over a 5 second exposure time. Quartz lengths were used. I t was observed that
42 data collected below 200 nm was very noisy and hence the lower wavelength limit was restricted to 200 nm. In house programs were used to deconvolute the data and estimate the degree of secondary structure in the samples. Data P rocessing and R efinement The data collected in triplicate was averaged across the 50 scans and the triplicate experiments to improve the signal to noise. The averaged data was then scaled to molar ellipticity values to account for concentration and to compare with other CD spectra. Secondary S tructure D etermination A lgorithm CD data collection below 200 nm wavelength drastically overloaded the dynode and created large variation in the data accompanied by large error values. Therefore all the data was collected between 260 200 nm. Standard CD deconvolution programs require data collected to at least 190 nm and therefore these programs could not be sheet and random coil structures of poly lysine were taken from data published by Greenfield and Fasman (Figure 2 2) (4 8) and used to generate an array of CD spectra corresponding to sheet and random coil between 0 100% with 10% intervals using the following equation: Where and are the percentage (by residue) of helix, sheet and random coil content in the sample respectively and calculated resultant molar ellipticity. molar sheet and random coi l respectively. This gave sheet and random coil compositions
43 For each experimental data set, the theoretical CD data was first scaled (by individual wavelength) to the experimental data The experimental data was the n fitted to each of the theoretical data sets and the best fit was determined by the least squares technique using the following equation: Where f is the least squares fit value, Where is the scale factor corresponding to the wavelength is the experimental molar ellipticity and 66 possible combinations of helix, sheet and random coil propensities. Differential Scanning Calorimetry DSC is a calorimetry technique that measures the amount of heat required to raise the temperature of the sample by one degree. This value is directly compared to a reference buffer and the difference in heat required is plotted against temperature. This technique is useful in identifying thermal transition points and transition temperatures Data C ollection The DSC experiments were carried out on a MicroCal VP DSC machine between temperatures of 10 and 100 C with sample concentrations of 0.7 mg/ml in triplicates. The sample and the buffer were loaded in the two different chambers in the machine. The data collected was plotted and analyzed using the Origin software sui te (OriginLab )
44 Data P rocessing and R efinement The triplicate DSC data was averaged to improve the signal to noise ratio. The baseli ne was to remove background noise. A veraged data was used to calculate standard deviation s which provided error value s Small a ngle S cattering Small angle scattering is a solution based scattering technique that involves the deflect ion of light by the mo lecules in the solution by small angles (>10 ) Based on the observed scatter, the size, shape and the orientation of the particles in solution can be determined at low resolution. It is used to observe and analyze multimeric assembli es of macromolecules and track global structural changes that occur in them. X ray Scattering D ata C ollection Some control Small angle X ray Scattering ( SAXS ) data was collected at ORNL to verify some of the S mall angle Neutron Scattering (S ANS ) results. The remainder of the SAXS data was collected at the Cornell High Energy Synchrotron Source (CHESS) using the Macromolecular Diffraction at CHESS (MacCHESS) facility on the G 1 beam line 2 mg/ml sample concentration s w ere used olume in plastic cuvettes The wavelength used was 1.296 with a detector distance of 1210 mm. Exposure times of 1 and 5 seconds were used with 10X and 20X attenuation levels to achieve optimum signal to noise ratio while at the same time having minimal r adiation damage to the sample. The data was collected between q of 0.001 and 0.275 The images were processed for intensity and q values using DATASQUEEZE ( University of Pennsylvania ).
4 5 Small angle Neutron Scattering D ata C ollection All the SANS data was collected in two separate data collection trips at the BioSANS beam line of the High Flux Isotope Reactor (HFIR) source at Oak Ridge National Labs (ORNL). for data collectio n. The data was collected at a wavelength of 6 with a wavelength spread of 0.14. Sample to detector distances were between 6.8 to 1 m to achieve a q range of 0.0065 0.35 1 Samples were concentrated to 2 mg/ml concentration prior to the data collection. Since neutron scattering is non destructive to the sample, it is possible to collect data over large time periods to compensate for the low flux of neutron beam lines However the data collection times in t hese experiments were limited by the amount of user beam line time available. In order to distinguish between the protein and the nucleic acid SANS signal in the sample, a technique called contrast variation was employed. When the D 2 O concentration in the sample buffer was increased the sample scattering length density increased due to the increased coherent scattering by the deuterium in buffer. This is illustrated in Figure 2 3. At ~ 12% D 2 O concentration, the buffer signal would match the signal from th e lipid content in the sample. At ~ 42% D 2 O, the buffer signal would match the signal from the protein concentration in the sample. This is a result of the difference in the amount of exchangeable hydrogen atoms available in lipids, proteins and nucleic ac ids. Proteins have more exchangeable hydrogen atoms than lipids and nucleic acids have more exchangeable hydrogen atoms than proteins. This leads to differences In their scattering ability (Figure 2 3). The samples and buffers used for contrast variation c ontained 42% D 2 O Subtracting the buffer signal from the sample signal would therefore retain SANS signals that arise only from the DNA content in the sample.
46 SAXS and SANS Data P rocessing and R efinement The ATSAS suite of programs was used to further proc ess the data after raw image processing (136) GNUPLOT was used for data plotting and radius of gyration (Rg) calculations. Pairwise distribution functions and radius of gyration values were calculated using the GNOM program from the ATSAS suite. DAMMIN was used to generate the 3 dimensional ab initio model. Ten DAMMIN simulations were averaged using the DAMAVER program to generate a final model. This model was converted to a SITUS volume map for dock ing purposes and the models were docked manually into the map u sing the CHIMERA program (96) Theoretical SAXS curves from the crystal structures of the monomer and the dimer were com puted using the CRYSOL. GNOM, DAMMIN, DAMA VER, and CRYSOL are all l ocated within the ATSAS suite. X ray C rystallography D ata C ollection and Structure Determination Structure determination by X ray crystallography would provide the highest resolution view of the structural features of the viral capsid. Crystallization conditions were already available for t he AAV capsids and several structures have been previously determined as described in C hapter 1. This study included the structure determination and analysis of AAV5 capsids at the different endosomal pHs. C rystal screens were set up using the hanging drop vapor diffusion method (81) with VDX 24 well plates and siliconized cover slips (Hampton Research, Laguna Niguel, CA, USA). The crystallization drops contained 2 l sample solution (at ~ 10 mg/ ml ) and 2 l precipitant solution equilibrated against 1 ml precipitant solution. AAV5 crystallization conditions were screened against precipitant solutions containing varying polyethylene glycol (PEG) 800 0 (0.5 2.5%), NaCl (250 and 350 m M ) and MgCl 2 (5 20 m M ) concen trations and a pH range (pH 6.0 8.5) a t room tem perature (RT) and 277 K.
47 A buffer concentration of 20 m M for both Bis Tris (pH 6.0 and 6.5) and Tris HCl (pH 7.0 8.5) was used for the pH screens. All the d iffraction data was collected at the CHESS F1 beam line The images were collected at 0.3 oscillat ion angles with ~ 30 images collected per crystal used with exposure times of 30 45 seconds. The data was collected at a wavelength of 0.97 with a 400 mm crystal to detector distance. Multiple crystals were used to collect the data and were scaled togeth er after image processing, in order to get enough completeness (at least 60%) The collected diffraction images were processed using HKL2000 (90) SCALEPACK was used to scale different data sets together and generate the structure factor files (90) All structures were at a resolution of 3.5 while the completeness values after scaling were 65% (pH 6.0) 63.2% (pH 5.5) and 45.2% (pH 4.0). The corresponding R symm values were 15.3%, 17.2% and 12.5% respectively. The CCP4 suite was used for molecular rep lacement (137) with the previously solved AAV5 structure as the model (unpublished) All subsequent model refinements were done using the CNS suite of programs (15) AAV5 has non crystallographic symmetry (NCS) due to the ic osahedral nature of the capsid In order to improve the quality of the data, the refinement process using CNS involved non crystallographic averaging as well. The data was refined to final R factor values of 23.1% (pH 6.0), 24.5% (pH 5.5) and 26.3% (pH 4.0). COOT was used to visualize and manually refine the models (34) PyMOL (DeLano Scientific) and CHIMERA (96) were used to visualize the models. Protease A ssays The protease activity of the VLPs, GFP coding DNA packaged AAV8, and controls (the serine protease Trypsin (as the posit ive control) and buffer and enzyme
48 (Protease Determine Quick) Protease Assay ( Athena enzyme systems ) acti vity in aqueous samples. The proprietary substrate is a casein derivative entrapped in a cross linked matrix with a dye conjugate that responds to a wide range of proteases including serine, metallo aspartate and cysteine proteases. e VLPs samples (0.1 mg/ml, 30 nM) and molar equivalent amounts of Trypsin and carbonic anhydrase were used for each measurement. Each sample was loaded into the assay vial containing an immobilized proteolytic substrate, at pH 7.5 and 5.5. After specific incubation (1, 2, 3, 4 and 5 h ou rs) times (80) at 37 C, supernatant solution from the vial was then centrifuged and measured for absorbance on a Beckmann UV Vis Spectrometer at 450 nm. These experime nts were then repeated with protease inhibition, using a cocktail of protease inhibitors (Halt TM Protease inhibitor Single Use Cocktail from Thermo Scientific) reaction vial. Each experiment was repeated 3 times and th e average measured value was used. Errors bars used were the standard deviations of the measurements. Computational A nalysis Secondary S tructure P rediction and M odeling Predicted models of the VP1/2 N terminal region were generated using the ROBETTA full c hain protein structure prediction server (65) ROBETTA uses an automated sequence based structure prediction tool. It uses sequence homology to build homology models for known domains and then uses a de novo structure prediction method for domains with no known homology. PyM OL (26) and Coot (34) were used to
49 structurally superimpose, visualize and generate images of the models. PROCHECK (7 1) was used to resolve the secondary structural propensity of the models at an amino acid level. These values were used to theoretically estimate the percentage secondary structure to compare with experimental CD data. pI C alculations The ExPASy Comput e pI/Mw tool (45) was used to calculate pI values for individual capsid domains and in house algorithms were used for subsequent calculations for the whole capsid. A pI value of 5.0 was used for DNA to calculate net pI values for capsids with genomic or G FP coding DNA (22) Capsid monomer copy number s (VP1:VP2:VP3 1:1:10) w ere used to scale these values Intrinsic Disorder P rediction The PONDR Fit algorithm was used to calculate the intrinsic disorder disposition in the capsid proteins (132) PONDR Fit is a meta predictor because the prediction results from a collection of predictors is used as the input for the PONDR Fit algorithm. This approach would improve its prediction accuracy as the different intrinsic disorder predictors that it uses for its inp ut data base their predictions on different sequence features and different training models. Intrinsically disordered protein sequences are typically characterized by a low number of bulky hydrophobic residues with several stretches of polar or charged am ino acid residues. The intrinsic disorder results from an inability to fold in such a manner that the hydrophobic regions of the protein for the buried core of the structure. These sequence signatures are used by the PONDR Fit algorithm to calculate a num erical intrinsic disorder disposition value. This value was plotted against residue number for different AAVs. The plot was scaled by sequence alignment to the AAV1 VP1 sequence
50 to better compare the different regions on the capsid monomer. Known proteins like Carbonic Anhydrase (no intrinsic disorder negative control) and calcineurin (intrinsically disordered positive control) were used to check if the predictions from the algorithm were valid.
51 Figure 2 1 Schematic of the production, purification and characterization of AAV empty capsid VLPs, DNA packaged capsids and capsid mutants for further biophysical analysis Biophysical analysis Buffer exchanging to other pHs Validation by Spectrometry, SDS PAGE and EM Ion exchange column chromatography for further purification Sucrose cushion and sucrose gradient centrifugation ( Iodixenol gradient centrifugation for AAV2) Nuclease treatment and centrifugation Cell lysis by multiple rounds of freeze thawing I nfection of cells with constructs of the virus capsid
52 Figure 2 2 sheet (red) and random coil (green). Values for these curves are taken from Greenfield and Fasman (1969) (48) 50000 30000 10000 10000 30000 50000 70000 190 200 210 220 230 240 250 Molar Ellipticity (deg cm 2 dmol 1 Wavelength (nm)
53 Figure 2 3 Contrast matching in Small angle Neutron Scattering. A plot of scattering length density Vs percentage concentration of D 2 O in the sample shows that the b uffers prepared with 42% D 2 O concentration would match the scattering length density of protein in the sample.
54 CHAPTER 3 GLOBAL CHANGES IN TH E SOLUTION STRUCTURE OF AAV CAPSIDS AT DIFFERENT ENDOSOMAL PHS The Parvovirus capsids are known to undergo structural transitions at different endosomal pHs (53, 85, 109) While it is known that pH alone does not induce these transitions in AAV capsids based on antibody binding assays (111) there may be some (but not all) transitions that do occur in the capsids in response to end o somal acidification Crystal S tructure of AAV8 at D ifferent pHs Crystal s tructures of AAV8 VLPs and rAAV8 GFP capsids have been previously determined at pH values that mimi c the conditions (7.5, 6.0, 5.5 and 4.0) encountered by AAVs during trafficking through the endocytic pathway and the low acidic pH of the lysosome (86) These structures did not show any global changes (corresponding to VP1u externalization) in capsid an d retained the space group and unit cell parameters of th e structure at pH 7.5. This could be an effect of crystal lattice constraints and steric hindrance. However there were some changes observed in side chain conformation at the different pHs (86) Cha nges were observed at surface residues R392, Y707, E566, and H529 and were termed (Figure 3 1) (86) At pH 7.5, the side chain of E566 is in position to form two hydrogen bonds with R392 from a 3 fo ld related monomer and a hydro gen bo nd with the hydrox yl group of Y707 from a 2 fold related monomer interface However, as the pH wa s lowered to 6.0, 5.5, and 4. 0, the side chain of E566 was seen to adopt an alternative conformation that facilitates the formation of a hydrogen bond with the imidazole si de chain of H529 This lower pH induced interaction is most likely due to a protonation of H529, with a pKa of 6.2, which is now able to donate a hydrogen to th e interaction.
55 The conformational change of E566 at the lower pH results in the loss of interactions with the 3 fold related R392 and th e 2 fold related Y707 and as a consequence, reduces the inter subunit contacts between the VP monomers. Concomitant with the change at E566, the side chain of Y707 becomes ori ented toward the 2 fold axis Finally, i n the crystal structure of AAV8 where the pH was decreased to 4.0 and then increased back to 7.5 the transitions are reversed to the conformations observed at pH 7.5 (86) Density consistent with a single dAMP nuc leotide was ordered in the interior of the AAV8 VLP structure determined at pH 7.5 (87) For the rAAV8 GFP structures, densities consistent with two connected nucleotides were ordered ins ide the capsid at pH 7.5 and 6.0, which could be modeled as a dAMP a nd dCMP dinucleotide (86) The DNA densities in both the VLP and rAAV8 GFP structures became less ordered as the crystal condition was decreased to pH 5.5 and were lost at pH 4.0 (Figure 3 2) The dAMP density was restored when the pH was restored to 7.5, suggesting that the lack of ordering at pH 4.0 was likely due to disrupted interactions with VP3 amino acids with the decreasing pH. The disappearance of the densities for the dAMP and dinucleotide in the structures of AAV8 VLPs and rAAV8 GFP, respectively, with decreasing pH values was concomitant with the H632 side chain shift described above. These observations taken together would suggest that pH was disrupting the interactions between the capsi d and its encapsidated DNA. It is possible that the loss of density for DNA at low pH may be part of a structural rearrangement of the genomic DNA to prepare it for capsid disassembly and DNA ejection /release
56 Previous studies suggest, based on antibody a ssays that pH alone does not induce the exposure of the VP1u region of the capsid (69, 111) However, while the antibodi es may bind to the VP1u at physiological pH, the binding of antibodies to the epitopes on the capsid can be affected by pH change. For example, the epitope for the A1 antibody directed towards the VP1u region has a linear epitope KRVLEPLGL (residues 123 13 1 by AAV1 numbering) that contains charged residues that could affect the binding of the antibody pH levels lower than physiological pH. Therefore, the lack of VP1u detection by antibodies at lower pH may be a result of ineffective antibody binding rather than a lack of exposure of the VP1u region. In this chapter Small angle X ray Scattering and Small angle Neutron scattering are used to probe radial changes in the mass distribution of AAV1, AAV5 and AAV8 capsids to confirm if there are changes in arrangem ent of DNA packaged inside the capsid and if there is a structural change in the VP1u region in response to decrease in pH from 7.5 (physiological) to 4.0 (lysosome). There have been previous studies that have used Small angle X ray scattering to observe p H dependent structural changes in viral capsids. Capsid maturation in Nudaurelia capensis omega virus (N V) involved a ~ 60 shrinkage of the T = 4 capsid in response to decrease in pH from 7.0 to 4.5 (7 8) The large conformational changes occurring in the capsid in response to pH change were tracked using Small angle X ray Scattering. Small angle S cattering The method of contrast variation has been described in the SANS section of C hapter 2. Briefly, c ontrast variation can be used in SANS to observe changes in DNA independent of the protein content in the sample. This permits the elimination of the capsid signal in the SANS data and allow s for direct detection of genomic DNA
57 r earrangement in the capsi d in response to pH changes S mall angle X ray scattering was used to track other global changes in the virus capsid as well as for control experiments. Very little SANS data i s available on the parvoviruses. A study on the Kilham Rat Virus (127) determi ned the radius of gyration value for the virus to be 105 in D 2 O and 104 in H 2 O They used SANS to propose that the virus had a triangulation number of 1 (T=1) and also proposed that the capsid had an inner and outer shell with all three capsid proteins contributing to the inner shell. This may have been the first small angle scattering documentation of the VP1u region in a parvovirus. V ery little structural information exists on the ssDNA structure inside the capsid. Results X ray crystallographic analysis (method described in detail in Chapter 2) of AAV5 capsids at different endosomal pHs (7.5, 6.0, 5.5 and 4.0) showed no significant changes in the secondary structural state of any amino acid side chains (including residues c orresponding to R392, E566 and Y707 that showed conformational changes in AAV8 (86) ) The experimental details for the SAXS and SANS experiments have been described in C hapter 2. One way of measur ing the mass distribution of the VLPs would be to use the Radius of Gyration (Rg) values of the particles in the sample as measure by small angle scattering. The Rg value can be determined by the following equation: Where I is the measured intensity and s is the corresponding angular velocity. This equation is referred to as the guinier approximation and is valid for intensities at low s values. Both SAXS and SANS analysis of the capsid showed Rg values of around 113
58 for the empty capsid VLPs at pH 7.5 (Table 3 1) These values were comparable to the Rg values reported for the parvovirus Kilham Rat Virus (~ 104 ) in previous studies (127) At lower pHs this value showed a consistent though small, decrease by ~ 3 This was confirmed i ndependent ly in both SAXS and SANS experiments. The Rg values of GFP coding DNA containing capsids showed a significant ~ 15 decrease in Rg value. This indicated the mass contribution of the DNA contained inside the capsid that would affec t the rotational state of the capsid and thereby the Rg value. Previous cryo EM studies have shown that conformational changes are induced in the capsid on complexing with cell surface primary receptor glycans (73) For this study, Neu5Ac 2,3 GalNAc 1,4 GlcNAc was used as a receptor molecule for AAV1 (88) Neu5Ac 2,3 Gal 1,4 GlcNAc was used as a receptor molecule for AAV5 and AAV8. S mall angle X ray scattering analysis of receptor complexed AAV1, AAV5 and AAV8 showed very small changes in Rg values that were within the experimental error (Table 3 2) The p airwis e density distribution function for empty AAV 8 capsid VLPs (based on SANS data) at different pHs (Figure 3 3) show s that the minor density changes seen in the capsid occur in the inte rior of the capsid. There was a 0.2 electrons/ 3 decrease in radial electron density between ~ 100 150 There are also accompanying changes seen at radial distances closer to the surface of the capsid (>220 ). The regions on the surface show similar val ues and therefore do not differ much in their density distribution. Contrast variation analysis of AAV8 capsids containing GFP coding DNA was preceded by controls experiments with the empty capsid VLPs at 42% D 2 O concentration. T he buffer subtracted scattering intensity values were large ly negative
59 indicating a match between the intensity values of the VLP proteins and the buffer, confirming that the contrast variation process was working at 42% D 2 O concentration in the sample and the buffer. The negative intensity values seen are a result of overcompensation in D 2 O concentration. Buffer subtracted intensity values for the GFP DNA containing capsids would correspond only to the DNA contained within the capsids. SANS data was colle cted for AAV8 with GFP coding DNA packaged at pHs 7. 5 and 5.5. The pairwise distribution functi on for the GFP coding D NA containing capsids showed a famili ar 3 decrease in Rg value from capsids at pH 7.5 (54.2 0.2 ) to 5.5 (51.1 0.2 ) just as seen in the case of the empty capsids (Figure 3 4 ) This would indicate that there is minor condensation in the arrangement of the DNA inside of the capsid in response to decrease in pH. Discussion No significant residue level changes were observed in the crystal structures of AAV5 determined at the endosomal pHs (7.5, 6.0, 5.5, 4.0). Conformational changes seen in residues R392, E566 and Y707 for AAV8 (86) were not seen in corresponding residues R38 2 E552 and Y6 88 in AAV5. Th is study hypothesized two structural changes in the AAV8 capsid in response to pH; 1) the externalization of the VP1u and 2) the reorganization of the encapsidated DNA. The small 3 decrease in the Rg value of the AAV8 empty capsid VLPs wit h decrease in pH from 7.5 to 4.0 would not necessarily confirm the VP1u ext ernalization process. However the corresponding pairwise density distribution function (Figure 3 3) did show changes internal to the capsid as well as on the surface. Accompanying c hanges were also seen at radial distances closer to the surface of the capsid. This
60 could be a process where the VP1u is undergoing structural changes in response to decrease in endosomal pH levels. Using SANS analysis, a similar 3 decrease wa s seen in the GFP coding DNA packaged in AAV8 with decrease in pH from 7.5 to 5.5 Differences were also observed in the pairwise distribution function from the data. This change would confirm that the DNA condenses in response to decrease in endosomal pHs to prepar e the capsid for DNA release (Figure 3 4 ) P revious studies on receptor complexed AAV2 have shown small, residue level changes in the capsid by cryo EM (73) Observing these changes is beyond the scope of the resolution and experimental error in the SAXS data in this study T he role of the receptor in induc ing the process of externalization of the VP1u where pH alone is inadequate remains unconfirmed as no major changes we re seen specific to receptor complexing. Unlike some RNA viruses (104) DNA insid e the AAV capsid is not icosahedrally ordered and therefore is not seen in the crystal structures of the capsid. The crystal structures of AAV8 at different pHs however showed density for two nucleotides at pH 7.5 that was no longer ordered at lower pHs bu t was regained again at pH 7.5. This prompted the investigation of the potential structural rearrangement of the DNA inside of the capsid. This may be involved in the process of readying the DNA for ejection /release from the capsid or facilitating other st ructural changes in the capsid that prepare it for disassembly or trafficking to the nucleus. We have shown by SAXS and SANS that there are changes in the structural organization of the DNA inside the AAV8 capsid in response to decrease in pH Empty
61 capsid VLPs also show that the reorganization occurs on the interior of the AAV capsid with decrease in pH. Th is is in good agreement with structural changes seen on the capsid crystal structures that accomp any these changes in the DNA. Taken as a whole, t his data suggests that the genomic DNA inside the capsid condenses in preparation for release in synergy with structural changes in the capsid in response to pH change.
62 Table 3 1 Rg values ( ) of AAV8 obtained from SANS data All Rg values are expressed in Angstrom units. Rg () Buffer SANS SAXS pH 7.5 113.2 0.9 114.8 0.4 Empty Capsids pH 6 112.8 0.9 pH 5.5 112.9 0.9 pH 4 110.8 1.0 111.3 1.3
63 Table 3 2 Rg values ( ) of empty VLPs from SAXS data. pH AAV1 AAV5 AAV8 7.5 114.3 0.1 109.7 0.1 112.5 0.1 6 113.4 0.1 109.7 0.1 111.1 0.1 5.5 111.5 0.1 110.9 0.1 110.5 0.1 4 111.2 0.1 111.5 0.1 109.1 0.1
64 Figure 3 1 Conformational changes in pH quartet region with pH. The residues involved at shown in A) blue at pH 7.5 and in B) purple at pH 4.0 for AAV8 and AAV1. pH 7.5 pH 4.0 A B
65 Figure 3 2 Changes in AAV8 DNA propensity at A) pH 7.5 (blue) and B) pH 4.0 (purple). The electron density seen for two nucleotides at pH 7.5 is no longer ordered at pH 4.0. H632 that shows association with the nucleotides at pH 7.5 changes conformation at pH 4.0. A B
66 Figure 3 3 Pairwise density distribution function for AAV8 VLPs at different pHs. Through pH 7.5 (purple), 6.0 (green), 5.0 (red) and pH 4.0 (black), changes seen in the radial electron density ar e in the interior of the capsid. R() (R)
67 Figure 3 4 Pairwise distribution function of DNA containing AAV8 capsids. The DNA at pH 7.5 is represented in blue while the DNA at pH 5.5 is represented in crimson. A 3 change in the Rg value is seen between the two samples in response to pH. 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0 50 100 150 Rg Value: pH 7 .5 > 54.2 0. 2 pH 5.5 > 51.1 0. 2 R ( ) (R)
68 CHAPTER 4 STRUCTURAL AND BIOPH YSICAL ANALYSIS OF T HE VP1 UNIQUE REGION Background Structural changes occurring in the VP1u region are crucial to the process of infectivity in AAVs and relate d parvoviruses (53) However, as described in the introduction C hapter (Chapter 1) the biological mechanism that leads to these structural changes has not yet been identified and t he exact nature of these structural changes wa s also not fully understood This C hapter describes stud ies that attempt to identify these structural changes and also the causative factors in the endosome that may lead to such conformational change s The AAV VP1u contains a PLA 2 phospholipase domain and two nuclear localization signals, one of which is essential for nuclear entry during an infection. While the AAV PLA 2 does not have as much activity as other known PLA 2 s, mutational analysis have confirmed that altering the HDXXY active site motif causes a decrease in infectivity ( 113) In this study, homology models of the cap sid N terminal regions were generated to compare to and interpret experimental results on the VP1u region To identify the role of intrinsic disorder in the dynamics of the N terminal regions t he capsid mo nomer sequence of AAV1, AAV2, AAV5 and AAV8 was analyzed for intrinsic disorder in the capsid and the electrostatic state of different regions of the capsid monomers w ere calculated for AAVs 1 12 C ircular Dichroism was used to demonstrate that the VP1/2 N terminal region in AAV1 has a orde helical secondary structure. In house algorithms were used to interpret the CD data and determine the secondary structural helical secondary structure was gradually lost with
69 decrease in pH from 7.5 to 4 and was regained when the pH was increased back to 7.5 demonstrating a reversible mechanism This study would be an important step in physically documenting conformational changes in the VP1u region and establishing the order of events in the escape of the capsid from the endosome. Results ROBETTA server (65) models (Figure 4 1 ) of the first 20 9 amino acids of the AAV1 VP1 sequence (not seen in the crystal structure) showed a c onsensus helix motif connected by loop regions. The VP1 / 2 common region did not show any helices and instead was indicated to comprise of coiled secondary structure Sup erimposition of two of these models on a crystal structure of bovine pancreatic phospholipase A 2 ( PDB ID: 1BP2) (Resolution: 1.7) showed that the helices containing the active site and the calcium binding site showed a high degree of conservation in struc ture (28) The RMSDs of these superimpositions were 3.3 and 2.8 respectively. Based on the Critical Assessment of Techniques for Protein Structure Prediction (CASP) assessment of ROBETTA predicted models (23) these values were quite reasonable. The PLA 2 active site (residues H48, Y52, Y73 and D99 on the bovine pancrea tic PLA 2 sequence) helices (Figure 4 1) The PLA 2 calcium binding site (residues Y28, G30, G32 and D49 on the bovine pancreatic PLA 2 sequence) was also structurally conserved in the VP1u models superimposed on the helices w hile the other three residues were located on the adjacent two sheet (Fig ure 4 1). T hese models were also superimposed onto the crystal structure of the AAV1 ( PDB ID: 3NG9, Resolution: 2.5 ) with residue 218 (the first residue seen on the AAV1 crystal structure monomer) overlapping between the VP1 N terminal model and the crystal structure. When the
70 VP1u was oriented directly beneath the 2 fold interface (based on previous cryo EM data from Kronenberg et al .) (69) the VP1 / 2 common loop region was in a position primed to deliver the VP1u through the 5 fol d pore (Figure 4 2). The intrinsic disorder propensity for the AAV VP1 s equence was calculated using PONDR Fit ( 132) described in Chapter 2. A plot of the predicted intrinsic disorder disposition versus the capsid protein residue number is shown in F ig ure 4 3 The PONDR Fit algorithm showed an intrinsic disorder stretch in the VP1/2 common region (residues ~ 140 202). The intrinsic disorder region continued into the VP3 sequence to ~ residue 220. This is in good agreement with the lack of ordered density for the first 10 15 residues of VP3 in all the solved crystal structures for the AAVs to date. It was also seen that the capsid variable regions (VRs) also demonstrate variability in intrinsic disorder as well while the structurally conserved regions show a consensus low intrinsic disorder disposition. A detailed description of the pI value calculations is given i n Chapter 2. A histogram based on the calculated pI values for the VPs of AAVs 1 12 is given in Fig ure 4 4 The VP3 and the intact empty capsid s show an average pI value of ~ 6.3 through all the AAVs (1 12) because of VP3 dominance in copy number (1:1:10 VP1:VP2:VP3) The addition of genomic DNA (pI ~ 5.0) to the capsid would effectively decrease the pI by ~ 0.4 to an average pI value of ~ 5.9. Interestingly, the VP1u exhibit s a consistent acidic pI of ~ 4.9 while the VP1/2 common region has a more basic pI value of ~ 7.3. There is a greater variability in the pI value of the VP1/2 common region however with pI values ranging from 5 10.
71 Ordered S tructural S tate of the VP1u Intact empty particles of AAV1 (purification described in C hapter 2) were used in the CD experiments. The CD spectrum of AAV1 at pH 7.5 and 25 C when deconvoluted (by algorithm described in C hapter 2) show ed a n helical secondary structure (Fig ure 4 5 ). Of the 518 residues seen in the crystal structure, only helix is observed, that contributes less than 1% of the total secondary structure of the capsid monomer whereas deconvolution of the AAV1 CD spectrum (details d escribed in Chapter 2) showed a 30 helical s ignal. Therefore, this helical component of the capsid must come from the part of the capsid protein that is not observed in the crystal structure the VP1/2 N terminal region. It is interesting to note t hat this effect is seen despite the fact that the VP1 and VP2 together form only ~ 16% of the total amino acid composition of the capsid based on copy number (1:1:10 VP1:VP2:VP3) and amino acid content When compared with a VP1 deletion construct of AAV6 (Fig ure 4 5 ), which differs from AAV1 at only 6 residues helical signal is indeed from the VP1u region. The AAV6 VP1 deletion construct showed a lower helical propensity as observe d in the crystal structure (88) It also suggests that the VP1/2 common region does helical propensity which is in agreement with the intrinsic disorder disposition of the domain and the homology models (Figure 4 1). L helicity with I ncrease in T emperature CD was used to experimentally probe structural changes in the VP1u. Electron micrographs were used to ensure that the structural changes observed were not a result of capsid disassembly or breakage. A detailed description of the me thodology involved
72 is provided in Chapter 2. Molar ellipticity values were used to generate normalized and scaled CD spectra that could be compared across experiments. As the temperature was increased from 25 C to 90 C, the CD spectra for AAV1 (between the wavelengths of 200 and 260 nm) showed a loss of secondary structure at ~ 70 C (Fig ure 4 6 ) while electron micrographs showed that the capsids were still intact ( Figure 4 7 ). This would suggest that the changes seen in the CD signal were primarily a local denaturation event of the VPs rather than whole capsid degradation. helicity with D ecrease in pH When the ellipticity values at 208 nm for AAV1 at different pHs (7.5, 6, 5.5 and 4) were plotted against temperature (Fig ure 4 8 ) it was clear that the transition temperature decreased with decrease in pH. In other words, this indicate s that a decrease in pH may serve as a mechanism to destabilize the caps id or make i t more flexible. CD experiments at the different pHs (at physiological temperature) of AAV1 capsids showed spectra that would imply different secondary structural characteristics (Fig ure 4 9 ) (a t pH 6, the AAV 1 CD spectrum showed a decrease in the degree o helicity by ~ 10% when compared to the spectrum at pH 7.5) This trend continued a t pHs 5.5 and 4 where almost no helical signals (<10% as determined by the in house deconvolution algorithm) were detectable (Table 4 2) In spite of this e lectron mic roscopy images of the capsids at these different pHs showed that the capsids were not losing their structural integrity in the process. Previous studies on AAV2 have shown that heating the capsid to a temperature of ~ 65 C results in the irreversible exte rnalization of the VP1u region (111) The unfolding of the VP1u region could have a role in the externalization process, as similar loss es of helical signal a re seen at high temperature (~ 70 C) (Figure 4 6) and low p H (5.5 and below) This unfolding was
73 found to be irreversible; decreasing the temperature to 25 C did not restore the secondary structural state signal. T his data would suggest that the loss of signal might be the un fol ding of the secondary structure of the VP1 N terminal region alone. Restoration of S econdary S tructure To see if this conformational change was reversible, AAV1 capsid sample s were treated at pH 4.0 and then w ere buffer exchanged back to pH 7.5. Using CD experiments i t was observed that the helical signal was restored (Fig ure 4 9 ). Electron micrographs were used to confirm that these capsids were also still intact. This suggests that the structural transitions that occur with a decreas e in pH from 7.5 to 4 .0 were reversible and increasing the pH again to 7 .5 at least partially restores the helical conformation of the VP1u region. AAV5 and AAV8 The CD spectrum of empty AAV8 VLPs at 25 C and pH 7.5 show ed a similar helical propensity (~30 40%) to AAV1 as would be expected (Figure 4 10 ). This helical propensity wa s absent (<10%) (Table 4 2) at the lower pHs (6.0, 5.5 and 4.0). Unlike a gradual unfolding in the VP1u secondary s tructure as seen for AAV1, a much more rapid unfolding process was observed for AAV8 with decrease in pH. The CD spectrum of AA V5 at 25 C and pH 7.5 was more similar to the AAV6 VP1u construct under the same conditions (Figure 4 1 1 A). SDS PAGE analysis of the AAV5 baculoviral constructs show ed little or no VP1 presence (Figure 4 11 B). This would explain the absence of an helic al signal for the AAV5 capsids. While the reason for the lack of VP1 packaging in AAV5 capsids is not known, it verifies that the helical signal seen for AAV1 and AAV8 is most likely do to the VP1u secondary structural elements.
74 Discussion The complete r eason for the absence of electron density for the VP1/2 N terminal region in crystal structure of the AAV1 capsid is still unknown. It could be a result of heterogeneity in the presence of VP1 and VP2 in capsids where icosahedral averaging may have lead to disordered features that are attributed to them. However these CD studies show that there is a d etectable order ed s econdary structure of the VP1/2 N terminal region. P revious studies have shown p hospholipase A 2 protein structure s from snake (124) and bee venom (107) helical. Since the VP1u (residues 1 13 7 ) also contains a PLA 2 domain it is not surprising that a helical propensity is seen for it, from predicted model s and from the CD data. There have been previous CD studies on a related parvovirus MVM and the CD spectrum was similar to the ones seen for the AAVs but the spectrum was not deconvoluted (18) The VP1/2 common region (residues 13 8 202) in the predicted models does not exhibit helical strand secondar y structural elements Whe n the sequence of VP1 from AAV1, AAV2, AAV5 and AAV8 w ere analyzed for potential disordered or unstructured regions using the PONDR Fit algorithm ( 132) it was observed that the VP1/2 common region and the first ~ 20 results of VP3 were predicted to be highly disordered in comparison to other regions of the capsid. This observation ma y be another reason for why s ome of the N terminal region of VP1 (first ~220 residues), VP2 (first ~60 residues) and VP3 (first ~20 residues) are not ordered in the crystal structure besides just being artifacts of averaging Previous studies have observed that the VP1u region is exposed when the capsid is subjected t o high temperatures (~ 65 C) at physiological pH (111) While high temperature may not be the biological cause for VP1u exposure during the infective
75 pathway of the virus, it can be inferred from this study that decrease in endosomal pH results in simi lar structural changes in the VP1u as seen from i ncrease in temperature; b oth involve the unfolding of the VP1u region. This may explain how high temperatures are able to induce exposure of the VP1u ; by first inducing the unfolding the VP1u region and then the externalization This study shows that even though decrease in pH (to 4.0) ha s not been sh own to induce exposure of VP1u (111) there is a definite pH induced structu ral change that takes place in the capsid. Th e loss of secondary structure with d ecrease in pH can complicate the question of VP1u externalization during the process of endosomal trafficking of the AAV capsid. The actual biological cause for the exposure of VP1u remains to be identified but helicity may have a role in the process of exposure of VP1u If so, then it may be true that pH would be an important factor in the mechanism of VP1u exposure even though it is not entirely induced by decrease in pH alone. It is possible that including certain other factors may help indu ce the exposure of VP1u in vitro An analysis of the calculated isoelectric point (pI) values of different capsid regio ns for AAV serotypes 1 12 show ed that the VP1u has an acidic character with the pI values for all the AAVs of ~ 4.8. The VP1/2 common re gion show ed more variety with value ranges from 5 to 10 across the 12 serotypes but the mean value (~ 7.3) was more basic than the VP3 sequence. The VP3 (pI = ~ 6.3) is the dominant capsid protein and therefore influenced the pI of the capsids the most. T he mean pI for the net capsid across the 12 AAV serotypes was ~ 6.3.
76 The accepted pI value for DNA (nucleotides) from literature is ~ 5.0 (61) The VP1u therefore is more acidic in nature and th is could complicate the picture of both the genomic DNA an d the VP1u co existing in the interior of the c apsid It is possible that charge charge repulsion between the VP1u and the DNA may help it get or iented in a specific manner to prepare it for release from the c apsid The VP3 and net capsid pI values would i ndicate that the VP3 sequence and the capsid would have a zwitterionic state at early to late endosomal pHs (~ 5.5 6.5) but the VP1u would still be charged. This may be a possible explanation for how the VP1u exhibits specific structural changes in the e ndosome while the rest of the capsid does not. These differences in pI may also be important in determining the copy numbers of individual capsid monomers in assembling the whole capsid. While steric hindrance may have a role in the localization of VP1 mon omer on the capsid (for example, steric clashes would not allow for adjacent monomers on the 5 fold to be VP1), it is possible that the charged state of the different capsid regions determines the distribution and copy number of the specific monomer (VP1, VP2 or VP3) in the capsid. Since the VP1u is actually ordered in solution while located inside the capsid, it becomes more difficult to visualize the process of externalization while the capsid stays intact. It is possible that the un fol ding of the VP1u wi th pH aids in this mechanism A study by Levy et al on AAV2 has identified structural changes at the 5 fold pore in the presence of a cell surface glycan receptor (heparin sulfate) (73 ) The 5 fold pore seems to open in an iris like rotation of the ring of residues leading to the widening of the top of the 5 fold channel in the process and this may be in synergy with the un fol ding of the VP1u. The un folde d VP1u may easily slide out t he expanded 5 fold pore. The
77 potentially unstructured nature of the VP1/2 common region may have a role in this. The flexibility imparted by the intrinsic disorder in the VP1/2 common region would be useful in permitting large motion of the VP1u to the 5 f old pore. Since this un fol ding process has been shown to be reversible (Figure 4 9) the VP1u could adopt its native structure when externalized, possibly when in contact with the endosomal membrane (Fig ure 4 12 ). This study shows that the VP1u in AAV1 wa s structurally ordered in an helical conformation inside the capsid. This secondary structural state wa s affected by decrease in pH from 7.5 to 4.0 and negligible secondary structural elements were seen at the lower pH. Increasing the pH again to 7.5 res tore d the helical conformation in the VP1u. The VP1/2 common region has a n intrinsic ally disordered nature that could impart the flexibility required to permit the externalization of the VP1u through the 5 fold pore Differences in the charged state of t he VP1u and the rest of the capsid may service to delay the proce ss of externalization until a late endosomal pH (~ 6.0) is achieved. This study provides major insights into the potential mechanism of capsid rearrangements in preparation for endosomal escape of the AAV capsid.
78 Table 4 1 Thermal VP1u structural transition temperatures (in C) for AAV1, AAV5 and AAV8 The numbers in parenthesis are the errors associated with the values based on triplicate experiments AAV serotype pH 7.5 6 5.5 4 1 72 (3) 63 (2) ID ID 5 84 (2) 75 (2) 75 (3) 75 (2) 8 72 (3) ID ID ID *ID Indeterminate due to very low signal intensity
79 Table 4 2 Percentage helicity from d econvolution of CD spectra of AAV1, AAV5 and AAV8 at different pHs. Serotype AAV1 AAV5 AAV8 pH 7.5 35 45% 5 15% 35 45% pH 6.0 25 35% 5 15% <10% pH 5.5 <10% 5 15% <10% pH 4.0 <10% 5 15% <10%
80 Figure 4 1 Cartoon rendition of s uperimposition of VP1u models on PLA 2 crystal structure. Two of the VP1u models (cyan and pink) were structurally superimposed on Bovine Pancreatic PLA2 (green) (PDB ID: 1BP2). The superimposition RMSD values were 3.3 and 2.8 respectively. The helical active site is well conserved in the models. This i mage was generated using PyMOL.
81 Figure 4 2 S urface rendition of s uperimposition of VP1u on to the capsid monomer from the crystal structure. The VP1u monomer (red) was structur ally superimposed on to overlapping regions (residues 2 18 25 5 ) of the VP3 capsid monomer (purple) from the AAV1 crystal structure (PDB ID: 3NG9) This image were generated using UCSF CHIMERA (97)
82 Figure 4 3 PONDR FIT plot showing intrinsic disorder in VP1/2 common region. Variable regions are marked in red while highly conserved regions of the capsid are in blue, green and orange in the basal markings. AAV1 (purple), AAV2 (blue), AAV5 (gray) and AAV8 (green) show similar disorder propensities with the exception of the vari able regions
83 Figure 4 4 Histogram showing pI values across all 12 AAV serotypes. The VP1u (blue) pI values are 1 unit lower than the average pI values of VP3 (green) and the whole capsid (purple). The VP1/2 common region (red) shows a variable basic pI through the serotypes. 0 1 2 3 4 5 6 7 8 9 10 Isoelectric point
84 Figure 4 5 CD spectrum of AAV VP1u. AAV1 empty capsid VLPs (purple) show helical propensity while AAV6 VP1 constructs (pink) show lack of helical propensity confirming that the VP1u is structurally ordered in solution and is helical. 7000 6000 5000 4000 3000 2000 1000 0 1000 200 210 220 230 240 250 260 cm 2 dmol 1 Wavelength (nm) AAV1 AAV6 VP1u
85 Figure 4 6 CD spectrum of AAV1 empty capsid VLPs at different temperatures. Through temperatures 30 C (blue), 50 C (green), 60 C (red), 65 C (orange), 70 C (brown) a nd 75 C (black) the VP1u signal is lost with increasing temperature. 7000 6000 5000 4000 3000 2000 1000 0 1000 2000 200 210 220 230 240 250 260 cm 2 dmol 1 Wavelength (nm) 30 50 60 65 70 75
86 Figure 4 7 Electron micrographs of the AAV1 capsid VLPs. At different temperatures and different pHs, the capsids are still intact. Only on heating to 95 C, the capsids showed complete denaturation
87 Figure 4 8 Plot of ellipticity values at 212 nm Vs temperature from CD experiments. At different pHs 7.5 (blue), 6.0 (orange), 5.5 (yellow) and 4.0 (green) the ellipticity values decrease with decreasing pH and show tran sitions at ~ 70 C at the higher pHs 18 16 14 12 10 8 6 4 2 0 2 0 20 40 60 80 100 Ellipticity (mdeg) Temperature ( C)
88 Figure 4 9 CD spectrum of AAV1 empty capsid VLPs at different pHs. At pHs 7.5 (navy blue), 6.0 (orange), 5.5 (yellow) and 4.0 (green) the AAV1 capsid VLPs show loss of secondary structural signal with decreasing pH. This signal is restored at least in part when the pH is restored to 7.5 (light blue) 7000 6000 5000 4000 3000 2000 1000 0 1000 2000 200 210 220 230 240 250 260 cm 2 dmol 1 Wavelength (nm) pH 7.5 pH 6.0 pH 5.5 pH 4.0 back to pH 7.5
89 Figure 4 10 CD spectrum of empty AAV8 capsids at different pHs. The helical signal seen at pH 7.5 (blue) is no longer seen at the lower pHs 6.0 (red), 5.5 (gre en) and 4.0 (purple). 6000 5000 4000 3000 2000 1000 0 1000 200 210 220 230 240 250 260 cm 2 dmol 1 Wavelength (nm) pH 7.5 pH 6.0 pH 5.5 pH 4.0
90 Figure 4 1 1 CD spectrum of empty AAV5 capsids at different pHs. A) A distinct helical signal is not seen at any pH but there is a single transition from pHs 7.5 (blue) and pH 6.0 (green) to pH 5.5 (red) and pH 4.0 (purple). B) SDS PAGE gel showing lack of VP1 presence in AAV5 capsids. 4000 3000 2000 1000 0 1000 2000 200 210 220 230 240 250 260 cm 2 dm ol 1 Wavelength (nm) pH 7.5 pH 6.0 pH 5.5 pH 4.0
91 Figure 4 12 AAV1 VP1u externalization model. The VP1u (red) when structur ally superimposed on to the AAV1 VP3 crystal structure (1NG9) monomer, appears directly beneath the 5 fold and 2 f old interface. When the VP1u is unfolded with decrease in pH it can be threaded out through the 5 fold pore and when the right conditions are available on the outside of the capsid, it can refold back to a native functional state. The images were generated using PyMOL.
92 CHAPTER 5 THE ROLE OF STRUCTUR AL STABILITY IN THE INFECTIVE PATHWAY OF ADENO ASSOCIATED VIRUSES Background While structural changes in the AAV capsid associated with endosomal trafficking have been identified, the st ructur e of the cap sid when it arrive s to the nucleus is still not known. The capsid would either have to eject the DNA inside the nucleus or disassemble to expose the DNA prior to replication. Endosomal cathepsins B and L have been identified as uncoating factors for AAV2 a nd AAV8 (1) However, the stability of the capsid structure would also have an important role in the capsid disassembly process. Previous studies have determined the c rystal structure of AAV8 at different pHs (7.5, 6.0, 5.5 and 4.0) (8 6) The PDBe PISA tool (68) and VIPERdb (19) were used to calculate the buried surface area and associ ation energies at the capsid symmetry interfaces (86) Based on this analysis, the 2 fold interface was shown to decrease in buried surface area and a ssociation energy with decrease in pH (Table 5 1) This may be part of the structural changes seen in the capsid with response to pH but it is also possible that an embrittlement of the capsid is occurring. It is also possible that st ructural changes are o ccurring in the capsid in response to decrease in pH that allow for better flexibility in the capsid (which would serve to help in the process of VP1u externalization) or promote capsid disassembly. It is therefore important to assess the stability of the capsid in response to pH. This can be done in a number of different ways. Here, thermal stability was chosen, as c hanges in the thermal denaturation temperature of the capsid would be a good gauge of the structural stability of the viral capsid at differe nt pH conditions
93 Previous thermal denaturation studies on related viruses have shown melting temperatures at ~ 70 85 C using DSC and fluorescence experiments (3, 18, 100) For a related parvovirus, MVM, thermal denaturation temperatures were de termined to be at ~ 77 C (18) with thermal inactivation of infectivity at ~ 70 C (100) It was also observed that in presence of 1.5 M Guanidium HCl, the thermal denaturation temperature decreased to ~ 62 C (18) indicating that the structure of the capsid had a role in determining capsid stability. CD experiments on MVM showed no change in the CD spectrum between 200 and 250 nm wavelength even at temperatures as high as 95 C (18) This was concluded as the stabilization of a denatured conformation of the MVM capsid (18) Since loss of secondary structural state was observed usin g CD in the VP1u in response to increase in temperature ( C hapter 4 Figure 4 8 ), it wa s important to verify if this loss is not a result of capsid denaturation. Thermal denaturation studies would be a good way of determining if multiple steps are involved in capsid disassembly This study uses DSC and EM to analyze the thermal stability of the capsid. The methodology involved is described in detail in Chapter 2. Results The Origin program suite (OriginLab, Northampton Massachusetts) was used to visualize, set baselines and measure melting temperatures. The data from the DSC experiments was plotted on a graph with Temperature on the X axis and enthalpy value on Y axis (Figure 5 1). AAV1 and AAV5 showed a single transition point (Figure 5 1) at the thermal me lting temperature (Tm) for the AAV capsids as opposed to multiple process of the virus ( f or example, from capsids to pentamers to monomers to unfolded
94 monomers). Single poi nt transitions have been observed in previous studies on a related virus MVM (18) and therefore it was not surprising to observe the same in the case of AAV1 and AAV8. Therm al denaturation temperatures measured by DSC were interest ing, in that they were di fferent from the conformation al transition temperatures seen from the CD experiments (Chapter 4 : Table 4 1 Table 5 2 ). For example, the conformational transition temperature for AAV1 at pH 7.5 was ~ 70 C while the denaturation tempe rature as measured by DSC was ~85 C. This is understandable when compared to the electron microscopy results (Chapter 4 :Table 4 1 Figure 5 3 ) that demonstrated that capsids remained intact at temperatures ab ov e the conformational transition temperatures observed in the CD spe ctra This verifies that the thermal melting seen in the CD corresponds to only VP1u unfolding and is not a result of whole capsid disassembly The different AAV serotypes analyzed show different changes in melting temperature in response to decrease in p H. AAV1 show ed a small detectable increase in stability with decrease in pH from 7.5 (Tm = 84.8 0.4 ) to 5.5 (Tm = 87.5 0.5 ) and then at pH 4.0 (Tm = 81.70.4 ) it show ed a decrease in stability in comparison to pH 7.5 (Table 4 2) AAV5 shows a gradual d ecrease in melting temperature with decrease in pH from 7.5 (Tm = 92.4 0.3 ) to 4.0 (Tm = 80.3 0.4 ) indicating thermal destabilization with decrease in pH (Table 4 2) Unlike AAV1, AAV5 does not show any increase in stability at any of the lower pHs tes ted. AAV8 at pH 7.5 show ed a unique double peak transition (Figure 5 1) ( 74.7 0.4,79.1 0.3 ) that could potentially signify two different states in the disassembly process Th is double peak is not seen at the lower pHs of 6.0,
95 5.5 and 4.0 (Table 4 2) No transition peaks are observed at the AAV1 and AAV2 (111) VP1u transition temperature (~ 70 C) for the AAVs studied. The presence of cell surface receptor glycans (at 2 receptor molecules : 1 capsid monomer concentration) does not significant ly aff ect the me lting temperatures of the AAVs. Neu5Ac 2,3 GalNAc 1,4 GlcNAc was used as a receptor molecule for AAV1 (88) Neu5Ac 2,3 Gal 1,4 GlcNAc was used as a receptor molecule for AAV5 and AAV8. However t he presence of CaCl 2 (2 mM) showed drastic c apsid destabilization for AAV1 (Tm = 66.5 0.5 ) and AAV5 (Tm = 74.3 0.7) at pH 7.5 This was a difference of ~ 20 C when compared to the AAV capsids without CaCl 2 presence. AAV8 did not show destabilization (Tm = 76.3 0.8) with CaCl 2 presence, but a single peak transition replaced the double peak transition. AAV1, AAV5 and AAV8 were observed for capsid disassembly and morphological changes by negative stain (Nano W, Nanoprobes) EM (Figure 5 2). At the four different endosomal pHs tes ted (7.5, 6.0, 5.5 and 4.0), there was no significant capsid disassembly observed. However, AAV1, AAV5 and AAV8 capsids did show differential stain penetration at the lower pHs when compared to pHs 7.5 and 6.0. The thermal denaturation of the capsids was c onfirmed by electron microscopy that showed no intact capsids above the melting temperatures seen from the DSC studies (Figure 5 3 ) Di scussion The fact that the melting temperature of the capsid differs from the se transition temperature seen for the VP1u transitions (from CD) is a major indicator of the differential structural behavior of the VP1u in comparison to the rest of the capsid. It was also interesting to note that there were no peaks seen corresponding to the thermal
96 unfolding temperature of the AAV1 or AAV8 VP1u as determined by CD (Chapter 4, Table 4 1). It is possible that the thermal melting process of the capsid follows a sequence of events for different regions of the capsid. First the VP1u region unfolds followed by the denaturation of th e entire capsid The interact ion s that hold the capsid monomers together are expected to be of lesser energy than the internal energy of the capsid. However, since a single peak transition was observed from the DSC data for the melting of the capsid, one o f two events could be occurring. Either the monomer wa s unfold ing before the capsid inter monomer interactions were lost (which is very unlikely) or the inter monomer interactions in the capsid we re lost and the unfold ing of the monomer was instantaneous The latter is more likely as the capsid is observed (by electron micrography) to be denature d at the melting temperature s identified by DSC. The small variation s seen in the melting temperatures of the capsids at different pHs we re reproducible and distinct. The studied AAV serotypes (AAV1, AAV5 and AAV8) showed different denaturation temperature changes in response to changes in pH. However, the biological effect of these small differences may be debatable as these conditions are While pH induced destabilization of the capsid would promote disassembly of the capsid post endosomal transport, the consequence of the difference in observed melting temperatures might not be as much as would be req uired to promote disassembly. In an attempt to include more endosome like conditions in the DSC experiments, 2mM CaCl 2 was used in the DSC experiments. The late endosome is known to have
97 about 2 3 mM Ca 2+ concentration (46) Ca 2+ has been known to affect capsid stability in other non envelope viruses like Flock House Virus (FHV) (5) No major morphological differences are observed in the electron micrographs of the capsid at different pHs (7.5, 6.0, 5.5 and 4.0) (Figure 5 2 ) and temperatures (25 C, 55 C, 75 C ) for the AAVs (Figure 5 3 ) Above the thermal denaturation temperatures for the AAVs (based on DSC data), intact capsids were not observed. While electron micrography may be a very low resolution method to identify gross morphological changes in the capsid in response to pH, it still is an efficient way of confirming capsid integrity. There was differential stain penetration observed in the AAV1, AAV5 and AAV8 capsids at pHs 5.5 and 4.0 when compared to the higher pHs of 7.5 and 6.0. This can be attr ibuted to structural changes in the capsid that could affect the permeability and staining properties of the capsid. While thermal stability would be one way of analyzing the inherent stability of the capsid, other methods to analyze structural stability would involve the unfolding/denaturation process of the capsid with increasing chaotropic salt (for example, Guanidium HCl) concentration. Previous studies on MVM identified that the presence of 1.5 M Guanidium HCl decreased the melting temperature by ~15 C (18) indicating that the folded state of the capsid had a role in determining the thermal stability of the capsid. Th is would be a way of testing if, for the AAVs, changes in stability seen in the capsids were a result of increased unfolded states of t he capsid or individual regions of the capsid. This study determines the melting temperatures for the AAV serotypes at different pHs. There are small changes in the melting temperatures at different pHs. These
98 melting temperatures are different from the t ransition temperatures seen from CD experiments. These differences show that DSC and CD are measuring two different independent properties of the capsid. This study would impact our understanding the process of AAV capsid disassembly and post endosomal tra fficking.
99 Table 5 1 Association energies and buried surface areas for AAV8 at pHs 7.5 and 4.0. This data was calculated using the VIPERdb server. pHs Association Energy (kcal/mol) Buried Surface Area (2) I 2 I 3 I 5 I 2 I 3 I 5 pH 7.5 67.4 213.1 102.1 3235 10373 5058 pH 4 63.8 213.7 101.6 3082 10411 5033
100 Table 5 2 DSC melting temperatures (in C). AAV1 AAV5 AAV8 pH 7.5 84.8 0.4 92.4 0.3 74.7 0.4,79.1 0.3 pH 6.0 86.2 0.7 91.9 0.5 79.1 0.6 pH 5.5 87.5 0.5 83.5 0.9 78.6 0.5 pH 4.0 81.70.4 80.3 0.4 73.2 0.8 Back to pH 7.5 85.1 0.3 92.6 0.7 77.5 0.8 Presence of receptor pH 7.5 86.5 0.6 91.5 0.8 75.6 0.5 Presence of receptor pH 5.5 86.2 0.3 84.1 0.6 75.3 0.7 Calcium presence pH 7.5 66.5 0.5 74.3 0.7 76.3 0.8
101 Figure 5 1. DSC curves for AAV1, AAV5 and AAV8. This plot of Temperature versus enthalpy value (Cp) shows the thermal denaturation temperatures for AAV1 (green), AAV5 (gray) and AAV8 (red) as signified by the peaks. The data was plotted using Origin (OriginLab, Northampton, MA) AAV1 AAV5 AAV8
102 Figure 5 2 Electron micrographs of AAVs 1, 5 and 8 at different pHs. It can be seen that the capsids stay intact even at low pHs. The capsids were stained with Nano W and the micrographs were collected at 50,000X magnification. pH 7.5 pH 6.0 pH 5.5 pH 4.0 AAV1 AAV5 AAV8
103 Figure 5 3 Electron micrographs of AAVs 1,5 and 8 at different temperatures. Intact capsids were observed at 25 C, 55 C and 75 C. At 95 C the capsids were found to be denatured. The capsids were stained with Nano W and the micrographs were collected at 50,000X magnificatio n. 25 C 55 C 75 C 95 C AAV1 AAV5 AAV8
104 CHAPTER 6 NOVEL PROTEOLYTIC AC TIVITY IN ADENO ASSOCIATED VIRUSES Background Several virus families have previously been shown to exhibit protease activity, essential to their infectious pathway, these include: Aspartic proteases found in human immunodeficiency virus nodavirus and tetravirus (59, 66) ; Cysteine proteases found in poliovirus protease 3C, hepatitis C virus, and foot and mouth disease virus L peptidase (9, 50, 123) ; and Serine proteases found in a strovir us, S indbis virus and t ellina virus among others (2, 8, 95) Previous studies using two hybrid screens showed that the AAVs were susceptible to cathepsins B and L and that these cathepsins are essential for AAV2 and AAV8 mediated transduction of mammalian cells (1) This data suggest ed that cleavage events in the endosome, could prime the AAV capsid for nuclear uncoating. Of note, was the observation that long term stored AAV capsids often exhibited in vitro degradation of VP1 3 (as indicated on SDS gels, Figure 6 1 ), which implies th at a self p rotease activity was possible. Upon clo se inspection of the AAV 2 crystal structure (131) using the interactive modeling program COOT (34) a surface exposed region, near the three fold axis ( E 53 1 D 56 1 E563 and H527) was identified that had structural characteristics of a possible a spartic protease catalytic triad (Figure 6 2, Figure 6 3 ). Furthermore this site was semi conserved between all the known AAV structures in the VP1, VP2 and VP3 sequence and impli ed in the other known sequences (Table 6 1). Results Degradation products were observed in B1 antibody ( 125) western blot analysis of wild type and mutant AAV2 capsids. The B1 antibody was directed towards a linear
105 epitope in the C terminal region of VP1, VP2 and VP3 capsid proteins (125) This degradation increased at pH 5.0 in comparison to pH 7.5. O n e consistent degradation product was observed to be ~ 15 kDa in size. Mass spectrometry analysis (not shown) suggested that multiple potential cleavage sites on the AAV2 capsid sequence. Mutants at these cleavage sites (N469A, S498A, N582A, R585A and D219 A) however were unable to abolish protease activity entirely and the degradation products were still observed using western blots. Protease activity assays were also used to determine if the AAVs tested (AAV1, AAV2, AAV5 and AAV8) had protease activity (Protocol described in C hapter 2). The protease activity assay s clearly showed positive for the AAVs tested at pH 7.5 but not 5.5, when compared to the positive ( equal molar concentration of trypsin) and negative (buffer and equal molar concentration of ca rbonic anhydrase) controls (Figure 6 4 ), with a measured cleavage activity of ~ 1/3 that of trypsin (as seen fr om the optical density values). As the viruses were concentrated for the assay, the sample flow through (Citrate Phosphate buffer with 150mM NaCl ) was collected and used as a negative control to ensure that the observed protease activity was not an artifact of a contaminant protease in the buffer or the purification process. A test against an uninfected sf9 lysate supernatant confirmed that the act ivity seen was not from an impurity from the cells. The observed activities were comparable to protease activities previously reported using this kit (29, 80) Within the experimental uncertainty of the experiment no conclusions could be made to the relative rates of protease activ ity between the AAVs tested. To confirm that the observations were a proteolytic reaction the pH 7.5 experiment s w ere repeated in the presence of a cocktail of protease
106 inhibitors (HALT TM Protease inhibitor Single Use Cocktail from Thermo Scientific) (Figure 6 5 ). This stopped the protease activity as the measured ODs were reduced to the negative control values, as was the control trypsin. Protease activity was also abolished in an AAV2 mutant E563A (preparation and purification described in C hapter 2) (Figure 6 5). To test i f DNA packaging in the capsid had an effect on the OD values in GFP coding DNA packaged AAV 8 were assayed. An increase in the protease activity was seen (Figure 6 4) indicating a role for DNA in enhancing protease activity. Discussion This study has ident ified autoproteolytic activity in the capsids of AAV1, AAV2, AAV5 and AAV8. Western blot analysis (Figure 6 1) confirm that lower molecular weight degradation products are seen for the AAVs and the intensity of these products on the blot increases with dec rease in pH from 7.0 to 5.0, indicating that pH has a role in the AAV protease activity. Mutagenesis experiments (N469A, S498A, N582A, R585A and D219A) followed by mass spectrometry and western blot analysis (Figure 6 1) were used to identify the cleavage site but data so far has been inconclusive. Based on the conserved VP3 structure of AAVs a comparative analysis of the other known AAV structures (AAVs 1 9) showed a putative as partic catalytic triad (in AAVs 1 9 with the exception of AAV5) for ( Table 6 1), which involves the coordination of a water molecule between the two aspartate residues. This may also be the case where one of the aspartates is replaced with a glutamate, in the case of AAV 2, AAV 4, and AAV 9, (Glutamic proteases, Table 6 1). In additio n, the other AAVs have amino clusters that could be potential Serine or other yet unclassified proteases An AAV2 mutant E563A showed absence of protease activity when tested (Figure 6 5 ). This may indicate a role for this site in the proposed protease ac tivity.
107 The observation that AAV 1, AAV 2, AAV 5, and AAV 8, exhibit protease activity at pH 7.5 but not 5.5 would imply this enzymatic process is require d either pre or post endosomal entry for efficient infection during virus trafficking through the cytop lasm. Previous studies have identified endosomal cathepsins B and L as uncoating factors for AAV2 and AAV8 (1) The autoproteolytic activity seen in the AAVs is also susceptible to decrease in endosome like pH levels. It is possible that these activities seen are related or synergistic. The effect of mutating surface accessible amino acids (that are postulated as involved in the protease activity) is currently in progress as is the identification of wh at type of protease the AAVs are. This can be done by assaying the AAVs for protease activity in presence of inhibitors specific to each type of protease separately. The role for proteolytic activity in the infective pathway of the AAVs is unknown but sinc e this activity is affected by change in pH, it is possible that the proteolytic activity may be seen when the AAVs are trafficked through the endosome.
108 Table 6 1 Amino acid conservation in the catalytic triad region. AAV Amino acid number and type 1 532Asp, 562Asp, 527His, 564Glu 2 531Glu, 561Asp, 526His, 563Glu 3 532Asp, 562Asp, 527His, 564Glu 4 530Asp, 560Ser, 527Gly 5 518Ser, 551Ser, 514Asn 6 532Asp, 562Asp, 527His, 564Glu 7 514Asn, 518Ser, 551Ser 8 534Glu, 564Ser, 529His 9 532Asp, 562Asn, 527His, 564Glu
109 Figure 6 1 Western Blot of AAV2 wild type and mutants showing degradation products. A set of degradation products are seen in below the VP3 band. There is an increase of degradation products at pH 5.0 when compared to pH 7.5. The bands were probed with a B1 antibody directed towards the C terminal domain of the VP1, VP2 and VP3 capsid proteins. This image was taken from Max Salganik.
110 Figure 6 2 Superimposition of hypothesized aspartic protease sites (AAV2) onto catalytic aspartates of an aspartic protease (HIV1 protease). The AAV2 residues (blue) are show similar orientation and proximity as the aspartic protease (green) catalytic aspartates This figure was made using PyMOL (DeLano Scientific). E563 E531
111 Capsid Exterior Capsid Interior barrel Figure 6 3 AAV 1 capsid structure. A) Ribbon diagram of the VP3 structure. The positions of icosahedral 5 3 and 2 fold symmetry axes are as labeled. B) Close up view of the putative catalytic active site, amino acids are as labeled. The yellow sphere represents an ordered solvent mole cule observed in the crystal structure. The figure was generated in PyM OL H526 E 5 31 D 5 61 E563 2 3 5
112 Figure 6 4 P rotease activity for AAV 1 AAV 2, AAV 5, and AAV 8 at pH 7.5 and 5.5. Plotted are AAV1 (purple), AAV 2 (light blue), AAV 5 (dark grey), AAV 8 (green) (at pH 7.5 and 5.5) and AAV8 with GFP coding DNA packaged (lime green). Also shown are Trypsin (red positive control) carbonic anhydrase (orange negative control) and buffer (light grey negative control). The OD readings are given in arbitr ary units. Each data point was measured in triplicate. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 1 2 3 4 5 O.D. Time (hours) pH 7.5 pH 5.5
113 Figure 6 5 Protease activity assay plots for AAV1, AAV2, AAV5 and AAV8 in presence of inhibitors. Plotted are AAV1 (purple), AAV2 (blue), AAV5 (dark grey) and AAV8 (green). With inhibitor presenc e AAV1 (pink), AAV2 (light blue), AAV5 (light grey), AAV8 (light green) and trypsin (orange) show lack of activity. The AAV2 mutant E563A (black) and the mock cell extract (crimson) also show lack of activity. 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0 1 2 3 4 5 6 O.D. Time (hours)
114 CHAPTER 7 SUMMARY AND FUTURE D IRECTION S The crucial role of the VP1u region and structural changes associated with it has been well documented by this study and others (13, 69, 111) But the cellular factors that trigger these changes and the nature of these changes ha ve not yet bee n established. The primary focus of this study is to establish the nature of the structural changes in the VP1u changes and determine the biological trigger to structural changes in the capsid and the VP1u region. It was determined by CD based assays that the VP1u externalization event that previously was seen in response to high temperature (111) involves the unfolding of the VP1u region. This unfolding occurs in response to temperature as determined previously and pH as well (Chapter 4) The r esponse to pH would be expected, as pH changes are one the primary features of the endosome. However, while un fol ding of the VP1u region has been established, the second step of the mechanism that would involve the process of externalization of the unfolded VP1u ha s not yet been identified. It can be hypothesized that the unfolded VP1u can be threaded out through the 5 fold pore (based on work by Levy et al. (73) ) and the ionic state of the VP1u at late endosomal pH could be a trigger to cause the VP1u to thread out When the pH of the environment is decreased from 7.5 to 4.0 and then restored to 7.5 C D studies show ed re versal of the secondary structural state of the VP1u. This was in agreement with the reversible structural changes seen in the AAV8 crystal structu re solved at different pHs (7.5, 6.0, 5.5, 4.0 and at 7.5 restored) (86) A restoration of physiological pH may or may not be the biological trigger for VP1u refolding. This study still proves that the VP1u has the intrinsic propensity to refold to a native functional
115 state. It is possible that once the VP1u is in contact with the lipid membrane of the endosome, it refolds to regenerate the PLA 2 active site to carry out its function of lysing the endosomal membrane for capsid escape. An analysis of the AAV VP1u sequence indicates a putative C2 lipid binding domain (112) acti vity, which would be expected for most phospholipases. Future experiments would involve the testing of interactions between the AAV capsid and an in vitro reconstitution of the lipid membrane. A number of different techniques can be used to analyze the int eractions but one way of biophysically analyzing the interactions would be to label the lipid layer and the capsid with spin labels or fluorescent labels to test by Double Electron Electron Resonance er (FRET) respectively. These methods can be optimized to analyze multiple aspects of the interaction between the lipid layer and the PLA 2 domain like binding, conformational change and even activity. While it has been established that the VP1u contains a PLA 2 activity, the level of activity is still very low in comparison to more potent PLA 2 s like bee venom (107) or snake venom phospholipase (124) Due to this, there h ave been some concerns on whether the AAV PLA 2 is indeed a true phospholipase. Previo us studies do, however, indicate that mutations in the PLA 2 active site motif on AAV2 decrease infectivity (113) Since the VP1u has been known to externalize at high temperature for the AAVs (111) this method has routinely been used to test the AAVs for phospholipase activity. This however can be a destructive method for phospholipase activity determination. As it has been identified in this study, the VP1u domain irreversibly unfolds at the same temperatures used to externalize the VP1u domain. Ther efore, it is possible that the low activity seen for the AAV PLA 2 is due to the loss of structure and a resultant los s of
116 function of the VP1u region. Determining the biological trigger for externalization would be a huge boost to understanding the enzymat ic nature of AAV PLA 2 domains. A construct of the VP1u alone would also be very useful in studying its functions and verifying its structural properties. Some of the major challenges to studying the VP1u region include the unstructured state of the VP1/2 common region and the low copy number of the VP1u region. While the pI differences between the VP1u and the other regions of the capsid can effect the unique behavior of the VP1u, it may also have a role in determining copy number during the process of cap sid assembly. It is very possible that mutations in specific polar or charged residues can alter the electronic configuration of the whole capsid. It would be interesting to see how copy numbers in the capsid are altered by mutagenesis in charged residues that are not located at the inter monomeric interfaces. While this study has employed a reductionist approach to in vitro simulation of endosomal conditions by studying just pH alone using other endosomal factors could be key to identifying the actual bi ological mechanism of endosomal escape. Along with changes in pH the endosome also involves large changes in ionic strength. It may be difficult to a ccount for all the different ions present in the endosome partly because the concentration of a lot of thes e ions at different endosomal states is still not known properly. One of the known factors is that there is a large decrease in endosomal Ca 2+ (from ~ 30 mM early endosome to ~ 3 mM late endosome) ion concentration (46) but an increase in Cl ion conc entration. CaCl 2 was identified to destabilize the capsid for AAV1, AAV5 and AAV8. While the implication of this in the endosomal trafficking process is unclear, there is still a major effect of ionic presence on the capsid structure.
117 The presence of the g lycan receptor moieties has not shown any major structural or stability changes to the capsid. It is possible that the glycans will have to be in glycoprotein form to effect structural changes in the capsid, if indeed th ey have a role in the process. It wo uld be good to test if the presence of co recept ors could have an effect as well There seems to be a lot of diversity in the intracellular trafficking process for the AAVs after endosomal escape (53) This study does not shine much light on this area bu t the altered state of the capsid is important to focus on when looking at post endosomal trafficking and changes in the capsid. There is some recent evidence suggesting that the AAVs interact with an endoplasmic reticulum (ER) chaperone protein GRP78 (60) This could suggest that the AAVs could exploit the intracellular retrograde transport system and move from the Golgi to the ER to get closer to the nucleus. It would be useful to biophysically analyze the capsid structure under more oxidizing ER conditi ons. Most previous studies have employed antibody assays as the primary means of determining the structural changes occurring in the endosome (69, 111) While these assays have been very useful, they are limited by a number of factors including the indirect nature of the assays and the difference in binding ability of the antibodies with changes in conditions. For example, at lower pH (< 7.5) the A1 antibody (128) (directed towards the VP1u region) does not bind well (data not shown) This could be because of th e presence of a Glutamate residue in the linear epito p e for A1 (128) It is also difficult to check if the binding of A1 is seen as a result of capsid breakage. While there are structural antibodies that bind to intact capsids, they have been seen to sti ll bind to
118 disassembled capsid products. The design of more robust antibodies would help experiments be a lot more reproducible. A good antibody assay system would be a good compliment to biophysical analysis of capsid structural transitions. A novel protease activity in the AAVs has been identified in this study. While this study is very preliminary in characterizing the role of protease activity in the trafficking and infective pathway of the virus, it is possible that the activity seen in the virus could have a role in the slow disassembly of the virus. This autocatalytic nature of the virus adds a new dimension to the understanding of the viral life cycle. Further studies would involve the identification of the enzymatic nature of the protease activ ity and the potential substrate and the effect of mutagenesis of protease activity residues on the infectivity of the virus. This study adds to our understanding of the structural basis of intracellular viral trafficking in a substantial way. While the dir ect use of this data in clinical gene therapy may be limited, a thorough understanding of the viral life cycle would go a long way in developing better gene therapy strategies and add to our fundamental understanding of non envelope viral life cycles.
119 LIST OF REFERENCES 1. Akache B., D. Grimm, X. Shen, S. Fuess, S. R. Yant, D. S. Glazer, J. Park, and M. A. Kay 2007. A two hybrid screen identifies cathepsins B and L as uncoating factors for adeno associated viru s 2 and 8. Mol. Ther. 15 :330 339. 2. Aliperti G., and M. J. Schlesinger 1978. Evidence for an autoprotease activity of sindbis virus capsid protein. Virology 90 :366 369. 3. Ausar S. F., T. R. Foubert, M. H. Hudson, T. S. Vedvick, and C. R. Middaugh 200 6. Conformational stability and disassembly of Norwalk virus like particles. Effect of pH and temperature. J. Biol. Chem. 281 :19478 19488. 4. Balsinde J., M. A. Balboa, P. A. Insel, and E. A. Dennis 1999. Regulation and inhibition of phospholipase A2. An nu. Rev. Pharmacol. Toxicol. 39 :175 189. 5. Banerjee M., J. A. Speir, M. H. Kwan, R. Huang, P. P. Aryanpur, B. Bothner, and J. E. Johnson 2010. Structure and function of a genetically engineered mimic of a nonenveloped virus entry intermediate. J. Virol. 84 :4737 4746. 6. Bantel Schaal U., B. Hub, and J. Kartenbeck 2002. Endocytosis of adeno associated virus type 5 leads to accumulation of virus particles in the Golgi compartment. J. Virol. 76 :2340 2349. 7. Bartlett J. S., R. Wilcher, and R. J. Samulski 2000. Infectious entry pathway of adeno associated virus and adeno associated virus vectors. J. Virol. 74 :2777 2785. 8. Bass D. M., and S. Qiu 2000. Proteolytic Processing of the Astrovirus Capsid. J. Virol. 74 :1810 1814. 9. Belsham G. J., G. M. McIne rney, and N. Ross Smith 2000. Foot and Mouth Disease Virus 3C Protease Induces Cleavage of Translation Initiation Factors eIF4A and eIF4G within Infected Cells. J. Virol. 74 :272 280. 10. Berns K. I. 1990. Parvovirus replication. Microbiol. Rev. 54 :316 32 9. 11. Blechacz B., and S. J. Russell 2004. Parvovirus vectors: use and optimisation in cancer gene therapy. Expert Rev Mol Med 6 :1 24. 12. Bleker S., M. Pawlita, and J. A. Kleinschmidt 2006. Impact of capsid conformation and Rep capsid interactions on adeno associated virus type 2 genome packaging. J. Virol. 80 :810 820.
120 13. Bleker S., F. Sonntag, and J. A. Kleinschmidt 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. 14. Brabec M., D. Schober, E. Wagner, N. Bayer, R. F. Murphy, D. Blaas, and R. Fuchs 2005. Opening of size selective pores in endosomes during human rhinovirus serot ype 2 in vivo uncoating monitored by single organelle flow analysis. J. Virol. 79 :1008 1016. 15. Brnger A. T., P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros, R. W. Grosse Kunstleve, J. S. Jiang, J. Kuszewski, M. Nilges, N. S. Pannu, R. J. Read, L. M. R ice, T. Simonson, and G. L. Warren 1998. Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54 :905 921. 16. Burger C., O. S. Gorbatyuk, M. J. Velardo, C. S. Peden, P. Will iams, S. Zolotukhin, P. J. Reier, R. J. Mandel, and N. Muzyczka 2004. Recombinant AAV Viral Vectors Pseudotyped with Viral Capsids from Serotypes 1, 2, and 5 Display Differential Efficiency and Cell Tropism after Delivery to Different Regions of the Centr al Nervous System. Mol Ther 10 :302 317. 17. Canaan S., Z. Zdori, F. Ghomashchi, J. Bollinger, M. Sadilek, M. E. Moreau, P. Tijssen, and M. H. Gelb 2004. Interfacial enzymology of parvovirus phospholipases A2. J. Biol. Chem. 279 :14502 14508. 18. Carreira A., M. Menndez, J. Reguera, J. M. Almendral, and M. G. Mateu 2004. In vitro disassembly of a parvovirus capsid and effect on capsid stability of heterologous peptide insertions in surface loops. J. Biol. Chem. 279 :6517 6525. 19. Carrillo Tripp M., C. M. Shepherd, I. A. Borelli, S. Venkataraman, G. Lander, P. Natarajan, J. E. Johnson, C. L. Brooks 3rd, and V. S. Reddy 2009. VIPERdb2: an enhanced and web API enabled relational database for structural virology. Nucleic Acids Res. 37 :D436 442. 20. Chao H., Y. Liu, J. Rabinowitz, C. Li, R. J. Samulski, and C. E. Walsh 2000. Several Log Increase in Therapeutic Transgene Delivery by Distinct Adeno Associated Viral Serotype Vectors. Mol Ther 2 :619 623. 21. Chapman M. S., and M. Agbandje McKenna Atom ic structure of viral particles, p. 107 123. In Parvoviruses. Edward Arnold Ltd., New York. 22. Chargaff E., and J. N. Davidson 1960. The nucleic acids: chemistry and biology. Academic Press.
121 23. Chivian D., D. E. Kim, L. Malmstrm, P. Bradley, T. Rober tson, P. Murphy, C. E. M. Strauss, R. Bonneau, C. A. Rohl, and D. Baker 2003. Automated prediction of CASP 5 structures using the Robetta server. Proteins 53 Suppl 6 :524 533. 24. 1999. C ontrolled conformational transitions in the MVM virion expose the VP1 N terminus and viral genome without particle disassembly. Virology 254 :169 181. 25. Davidson B. L., C. S. Stein, J. A. Heth, I. Martins, R. M. Kotin, T. A. Derksen, J. Zabner, A. Ghodsi and J. A. Chiorini 2000. Recombinant adeno associated virus type 2, 4, and 5 vectors: Transduction of variant cell types and regions in the mammalian central nervous system. Proceedings of the National Academy of Sciences 97 :3428 3432. 26. Delano W. 2 002. The PyMOL Molecular Graphics System. 27. Dennis E. A. 1997. The growing phospholipase A2 superfamily of signal transduction enzymes. Trends Biochem. Sci. 22 :1 2. 28. Dijkstra B. W., K. H. Kalk, W. G. Hol, and J. Drenth 1981. Structure of bovine pan creatic phospholipase A2 at 1.7A resolution. J. Mol. Biol. 147 :97 123. 29. Dommisch H., W. O. Chung, M. G. Rohani, D. Williams, M. Rangarajan, M. A. Curtis, and B. A. Dale 2007. Protease activated receptor 2 mediates human beta defensin 2 and CC chemokine ligand 20 mRNA expression in response to proteases secreted by Porphyromonas gingivalis. Infect. Immun. 75 :4326 4333. 30. Douar A. M., K. Poulard, D. Stockholm, and O. D anos 2001. Intracellular trafficking of adeno associated virus vectors: routing to the late endosomal compartment and proteasome degradation. J. Virol. 75 :1824 1833. 31. Duan D., Q. Li, A. W. Kao, Y. Yue, J. E. Pessin, and J. F. Engelhardt 1999. Dynamin is required for recombinant adeno associated virus type 2 infection. J. Virol. 73 :10371 10376. 32. Duan D., Y. Yue, Z. Yan, J. Yang, and J. F. Engelhardt 2000. Endosomal processing limits gene transfer to polarized airway epithelia by adeno associated v irus. J. Clin. Invest. 105 :1573 1587. 33. Edwards J., E. Mann, and D. T. Brown 1983. Conformational changes in Sindbis virus envelope proteins accompanying exposure to low pH. J. Virol. 45 :1090 1097.
122 34. Emsley P., and K. Cowtan 2004. Coot: model build ing tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60 :2126 2132. 35. Farr G. A., S. F. Cotmore, and P. Tattersall 2006. VP2 cleavage and the leucine ring at the base of the fivefold cylinder control pH dependent externalization of both the VP1 N terminus and the genome of minute virus of mice. J. Virol. 80 :161 171. 36. Farr G. A., and P. Tattersall 2004. A conserved leucine that constricts the pore through the capsid fivefold cylinder plays a central role in parvoviral infection. Viro logy 323 :243 256. 37. Farr G. A., L. Zhang, and P. Tattersall 2005. Parvoviral virions deploy a capsid tethered lipolytic enzyme to breach the endosomal membrane during cell entry. Proc. Natl. Acad. Sci. U.S.A. 102 :17148 17153. 38. Ferrari F. K., T. Sam ulski, T. Shenk, and R. J. Samulski 1996. Second strand synthesis is a rate limiting step for efficient transduction by recombinant adeno associated virus vectors. J. Virol. 70 :3227 3234. 39. Fields B. N., D. M. Knipe, and P. M. Howley ology. Lippincott Williams & Wilkins. 40. Fisher K. J., G. P. Gao, M. D. Weitzman, R. DeMatteo, J. F. Burda, and J. M. Wilson 1996. Transduction with recombinant adeno associated virus for gene therapy is limited by leading strand synthesis. J. Virol. 70 :520 532. 41. Fricks C. E., and J. M. Hogle 1990. Cell induced conformational change in poliovirus: externalization of the amino terminus of VP1 is responsible for liposome binding. J. Virol. 64 :1934 1945. 42. Friedmann T. 1999. The development of human gene therapy. CSHL Press. 43. Gao G. P., M. R. Alvira, L. Wang, R. Calcedo, J. Johnston, and J. M. Wilson 2002. Novel adeno associated viruses from rhesus monkeys as vectors for human gene therapy. Proceedings of the National Academy of Sciences 99 :1185 4 11859. 44. Gao G., L. H. Vandenberghe, M. R. Alvira, Y. Lu, R. Calcedo, X. Zhou, and J. M. Wilson 2004. Clades of Adeno Associated Viruses Are Widely Disseminated in Human Tissues. Journal of Virology 78 :6381 6388. 45. Gasteiger E., C. Hoogland, A. Gattiker, S. Duvaud, M. R. Wilkins, R. D. Appel, and A. Bairoch 2005. Protein Identification and Analysis Tools on the ExPASy Server, p. 571 607. In J.M. Walker (ed.), The Proteomics Protocols Handbook. Humana Press, Totowa, NJ.
123 46. Gerasimenko J. V., A. V. Tepikin, O. H. Petersen, and O. V. Gerasimenko 1998. Calcium uptake via endocytosis with rapid release from acidifying endosomes. Curr. Biol. 8 :1335 1338. 47. Govindasamy L., E. Padron, R. McKenna, N. Muzyczka, N. Kaludov, J. A. Chiorini, and M. Agbandje McKenna 2006. Structurally Mapping the Diverse Phenotype of Adeno Associated Virus Serotype 4. Journal of Virology 80 :11556 11570. 48. Greenfield N. J., and G. D. Fasman 1969. Computed circular dichroism spectra for the evaluation of protein c onformation. Biochemistry 8 :4108 4116. 49. Grieger J. C., S. Snowdy, and R. J. Samulski 2006. Separate basic region motifs within the adeno associated virus capsid proteins are essential for infectivity and assembly. J. Virol. 80 :5199 5210. 50. Hanecak R., B. L. Semler, C. W. Anderson, and E. Wimmer 1982. Proteolytic processing of poliovirus polypeptides: antibodies to polypeptide P3 7c inhibit cleavage at glutamine glycine pairs. Proceedings of the National Academy of Sciences of the United States of A merica 79 :3973 3977. 51. Hansen J., K. Qing, H. J. Kwon, C. Mah, and A. Srivastava 2000. Impaired intracellular trafficking of adeno associated virus type 2 vectors limits efficient transduction of murine fibroblasts. J. Virol. 74 :992 996. 52. Hansen J ., K. Qing, and A. Srivastava 2001. Adeno associated virus type 2 mediated gene transfer: altered endocytic processing enhances transduction efficiency in murine fibroblasts. J. Virol. 75 :4080 4090. 53. Harbison C. E., J. A. Chiorini, and C. R. Parrish 2008. The parvovirus capsid odyssey: from the cell surface to the nucleus. Trends Microbiol. 16 :208 214. 54. Hauck B., and W. Xiao 2003. Characterization of tissue tropism determinants of adeno associated virus type 1. J. Virol. 77 :2768 2774. 55. Hauc k B., W. Zhao, K. High, and W. Xiao 2004. Intracellular viral processing, not single stranded DNA accumulation, is crucial for recombinant adeno associated virus transduction. J. Virol. 78 :13678 13686. 56. Hermonat P. L., and N. Muzyczka 1984. Use of ad eno associated virus as a mammalian DNA cloning vector: transduction of neomycin resistance into mammalian tissue culture cells. Proceedings of the National Academy of Sciences 81 :6466 6470.
124 57. Hernando E., A. L. Llamas Saiz, C. Foces Foces, R. McKenna, I. Portman, M. Agbandje McKenna, and J. M. Almendral 2000. Biochemical and physical characterization of parvovirus minute virus of mice virus like particles. Virology 267 :299 309. 58. Inagaki K., S. Fuess, T. A. Storm, G. A. Gibson, C. F. Mctiernan, M. A. Kay, and H. Nakai 2006. Robust Systemic Transduction with AAV9 Vectors in Mice: Efficient Global Cardiac Gene Transfer Superior to That of AAV8. Mol Ther 14 :45 53. 59. Johnson J. E., and A. Schneemann 1997. Nodavirus endopeptidaseHandbook of proteoly tic enzymes. Academic Press Ltd., London. 60. Johnson J. S., M. Gentzsch, L. Zhang, C. M. P. Ribeiro, B. Kantor, T. Kafri, R. J. Pickles, and R. J. Samulski 2011. AAV exploits subcellular stress associated with inflammation, endoplasmic reticulum expansi on, and misfolded proteins in models of cystic fibrosis. PLoS Pathog. 7 :e1002053. 61. Jordan D. O. 1955. The physical properties of nucleic acidsThe Nucleic Acids. Academic Press Ltd., New York. 62. Kaludov N., K. E. Brown, R. W. Walters, J. Zabner, and J. A. Chiorini 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. Journal of Virology 75 :6884 6893. 63. Kern A., K. Schm idt, C. Leder, O. J. Mller, C. E. Wobus, K. Bettinger, C. W. Von der Lieth, J. A. King, and J. A. Kleinschmidt 2003. Identification of a heparin binding motif on adeno associated virus type 2 capsids. J. Virol. 77 :11072 11081. 64. Kerr J. R. 2006. Parvo viruses. Hodder Arnold. 65. Kim D. E., D. Chivian, and D. Baker 2004. Protein structure prediction and analysis using the Robetta server. Nucleic Acids Res 32 :W526 531. 66. Kohl N. E., E. A. Emini, W. A. Schleif, L. J. Davis, J. C. Heimbach, R. A. Dixon E. M. Scolnick, and I. S. Sigal 1988. Active human immunodeficiency virus protease is required for viral infectivity. Proc. Natl. Acad. Sci. U.S.A. 85 :4686 4690. 67. Kramer R. M., and J. D. Sharp 1997. Structure, function and regulation of Ca2+ sensit ive cytosolic phospholipase A2 (cPLA2). FEBS Lett. 410 :49 53.
125 68. Krissinel E., and K. Henrick 2007. Inference of macromolecular assemblies from crystalline state. J. Mol. Biol. 372 :774 797. 69. Kronenberg S., B. Bttcher, C. W. von der Lieth, S. Bleker and J. A. Kleinschmidt 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. 70. Lakadamyali M., M. J. Rust, and X. Zhuang 2004. Endocytosis of influenza viruse s. Microbes Infect. 6 :929 936. 71. Laskowski R. A., M. W. MacArthur, D. S. Moss, and J. M. Thornton 1993. PROCHECK: a program to check the stereochemical quality of protein structures. Journal of Applied Crystallography 26 :283 291. 72. Lerch T. F., Q. X ie, and M. S. Chapman 2010. The structure of adeno associated virus serotype 3B (AAV 3B): insights into receptor binding and immune evasion. Virology 403 :26 36. 73. Levy H. C., V. D. Bowman, L. Govindasamy, R. McKenna, K. Nash, K. Warrington, W. Chen, N. Muzyczka, X. Yan, T. S. Baker, and M. Agbandje McKenna 2009. Heparin binding induces conformational changes in Adeno associated virus serotype 2. J. Struct. Biol. 165 :146 156. 74. Lombardo E., J. C. Ramrez, J. Garcia, and J. M. Almendral 2002. Complem entary roles of multiple nuclear targeting signals in the capsid proteins of the parvovirus minute virus of mice during assembly and onset of infection. J. Virol. 76 :7049 7059. 75. Maguire A. M., F. Simonelli, E. A. Pierce, E. N. Pugh, F. Mingozzi, J. Ben nicelli, S. Banfi, K. A. Marshall, F. Testa, E. M. Surace, S. Rossi, A. Lyubarsky, V. R. Arruda, B. Konkle, E. Stone, J. Sun, J. Jacobs, L. Zelenaia, K. S. Shindler, M. G. Maguire, J. F. Wrig ht, N. J. Volpe, J. W. McDonnell, A. Auricchio, K. A. High, and J. Bennett 2008. Safety and Journal of Medicine 358 :2240 2248. 76. Mani B., C. Baltzer, N. Valle, J. M. Almendral, C. Kempf, and C. Ros 2006. Low pH dependent endosomal processing of the incoming parvovirus minute virus of mice virion leads to externalization of the VP1 N terminal sequence (N VP1), N VP2 cleavage, and uncoating of the full length genome. J. Virol. 80 :101 5 1024. 77. Maroto B., N. Valle, R. Saffrich, and J. M. Almendral 2004. Nuclear export of the nonenveloped parvovirus virion is directed by an unordered protein signal exposed on the capsid surface. J. Virol. 78 :10685 10694.
126 78. Matsui T., H. Tsuruta, and J. E. Johnson 2010. Balanced electrostatic and structural forces guide the large conformational change associated with maturation of T = 4 virus. Biophys. J. 98 :1337 1343. 79. McLaughlin S. K., P. Collis, P. L. Hermonat, and N. Muzyczka 1988. Adeno associated virus general transduction vectors: analysis of proviral structures. Journal of Virology 62 :1963 1973. 80. McNeil H. J., P. E. Shewen, R. Y. C. Lo, J. A. Conlon, and M. W. Miller 2003. Novel protease produced by a Pasteurella trehalosi seroty pe 10 isolate from a pneumonic bighorn sheep: characteristics and potential relevance to protection. Veterinary Microbiology 93 :145 152. 81. McPherson A. 1989. Preparation and analysis of protein crystals. R.E. Krieger Pub. Co. 82. Miller E. B., B. Gurda Whitaker, L. Govindasamy, R. McKenna, S. Zolotukhin, N. Muzyczka, and M. Agbandje McKenna 2006. Production, purification and preliminary X ray crystallographic studies of adeno associated virus serotype 1. Acta Crystallographica Section F Structural Biol ogy and Crystallization Communications 62 :1271 1274. 83. Mitchell M., H. J. Nam, A. Carter, A. McCall, C. Rence, A. Bennett, B. Gurda, R. McKenna, M. Porter, Y. Sakai, B. J. Byrne, N. Muzyczka, G. Aslanidi, S. Zolotukhin, and M. Agbandje McKenna 2009. Pr oduction, purification and preliminary X ray crystallographic studies of adeno associated virus serotype 9. Acta Crystallographica Section F Structural Biology and Crystallization Communications 65 :715 718. 84. Mori S., L. Wang, T. Takeuchi, and T. Kanda 2004. Two novel adeno associated viruses from cynomolgus monkey: pseudotyping characterization of capsid protein. Virology 330 :375 383. 85. Mount J. D., R. W. Herzog, D. M. Tillson, S. A. Goodman, N. Robinson, M. L. McCleland, D. Bellinger, T. C. Nichols V. R. Arruda, C. D. Lothrop, and K. A. High 2002. Sustained phenotypic correction of hemophilia B dogs with a factor IX null mutation by liver directed gene therapy. Blood 99 :2670 2676. 86. Nam H. J., B. L. Gurda, R. McKenna, M. Potter, B. Byrne, M. S alganik, N. Muzyczka, and M. Agbandje McKenna 2011. Structural studies of adeno associated virus serotype 8 capsid transitions associated with endosomal trafficking. J. Virol. 85 :11791 11799. 87. Nam H. J., M. D. Lane, E. Padron, B. Gurda, R. McKenna, E. Kohlbrenner, G. Aslanidi, B. Byrne, N. Muzyczka, S. Zolotukhin, and M. Agbandje McKenna 2007. Structure of Adeno Associated Virus Serotype 8, a Gene Therapy Vector. Journal of Virology 81 :12260 12271.
127 88. Ng R., L. Govindasamy, B. L. Gurda, R. McKenna, O. G. Kozyreva, R. J. Samulski, K. N. Parent, T. S. Baker, and M. Agbandje McKenna 2010. Structural characterization of the dual glycan binding adeno associated virus serotype 6. J. Virol. 84 :12945 12957. 89. Opie S. R., K. H. Warrington Jr, M. Agbandje McKenna, S. Zolotukhin, and N. Muzyczka 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. 90. Otwinowski Z., and W. Minor 1997. Processing of X ray diffraction data collected in oscillation mode, p. 307 326. In Macromolecular Crystallography Part A. Academic Press. 91. Pacak C. A., C. S. Mah, B. D. Thattaliyath, T. J. Conlon, M. A. Lewis, D. E. Cloutier, I. Zolotukhin, A. F Tarantal, and B. J. Byrne 2006. Recombinant Adeno Associated Virus Serotype 9 Leads to Preferential Cardiac Transduction In Vivo. Circulation Research 99 :e3 e9. 92. Di Pasquale G., and J. A. Chiorini 2006. AAV transcytosis through barrier epithelia a nd endothelium. Mol. Ther. 13 :506 516. 93. Di Pasquale G., B. L. Davidson, C. S. Stein, I. Martins, D. Scudiero, A. Monks, and J. A. Chiorini 2003. Identification of PDGFR as a receptor for AAV 5 transduction. Nat. Med. 9 :1306 1312. 94. Patterson S., an d J. S. Oxford 1986. Early interactions between animal viruses and the host cell: relevance to viral vaccines. Vaccine 4 :79 90. 95. Petit S., N. Lejal, J. C. Huet, and B. Delmas 2000. Active Residues and Viral Substrate Cleavage Sites of the Protease of the Birnavirus Infectious Pancreatic Necrosis Virus. J. Virol. 74 :2057 2066. 96. Pettersen E. F., T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, E. C. Meng, and T. E. Ferrin 2004. UCSF Chimera -a visualization system for exploratory research and analysis. J Comput Chem 25 :1605 1612. 97. Pettersen E. F., T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, E. C. Meng, and T. E. Ferrin 2004. UCSF Chimera -a visualization system for exploratory research and analysis. J Comput Chem 25 :160 5 1612.
128 98. Quesada O., B. Gurda, L. Govindasamy, R. McKenna, E. Kohlbrenner, G. Aslanidi, S. Zolotukhin, N. Muzyczka, and M. Agbandje McKenna 2007. Production, purification and preliminary X ray crystallographic studies of adeno associated virus serotyp e 7. Acta Crystallographica Section F Structural Biology and Crystallization Communications 63 :1073 1076. 99. Rabinowitz J. E., F. Rolling, C. Li, H. Conrath, W. Xiao, X. Xiao, and R. J. Samulski 2002. Cross Packaging of a Single Adeno Associated Virus ( AAV) Type 2 Vector Genome into Multiple AAV Serotypes Enables Transduction with Broad Specificity. Journal of Virology 76 :791 801. 100. Reguera J., E. Grueso, A. Carreira, C. Snchez Martnez, J. M. Almendral, and M. G. Mateu 2005. Functional relevance of amino acid residues involved in interactions with ordered nucleic acid in a spherical virus. J. Biol. Chem. 280 :17969 17977. 101. Rivera V. M., G. Gao, R. L. Grant, M. A. Schnell, P. W. Zoltick, L. W. Rozamus, T. Clackson, and J. M. Wilson 2005. Long term pharmacologically regulated expression of erythropoietin in primates following AAV mediated gene transfer. Blood 105 :1424 1430. 102. Ros C., C. Baltzer, B. Mani, and C. Kempf 2006. Parvovirus uncoating in vitro reveals a mechanism of D NA release without capsid disassembly and striking differences in encapsidated DNA stability. Virology 345 :137 147. 103. Ros C., and C. Kempf 2004. The ubiquitin proteasome machinery is essential for nuclear translocation of incoming minute virus of mice Virology 324 :350 360. 104. Rudnick J., and R. Bruinsma 2005. Icosahedral packing of RNA viral genomes. Phys. Rev. Lett. 94 :038101. 105. Sanlioglu S., P. K. Benson, J. Yang, E. M. Atkinson, T. Reynolds, and J. F. Engelhardt 2000. Endocytosis and Nuclear Trafficking of Adeno Associated Virus Type 2 Are Controlled by Rac1 and Phosphatidylinositol 3 Kinase Activation. Journal of Virology 74 :9184 9196. 106. Schmidt M., A. Voutetakis, S. Afione, C. Zheng, D. Mandikian, and J. A. Chiorini 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. 107. Scott D. L., Z. Otwinowski, M. H. Gelb, and P. B. Sigler 19 90. Crystal structure of bee venom phospholipase A2 in a complex with a transition state analogue. Science 250 :1563 1566.
129 108. Seisenberger G., M. U. Ried, T. Endress, H. Bning, M. Hallek, and C. Bruchle 2001. Real time single molecule imaging of the i nfection pathway of an adeno associated virus. Science 294 :1929 1932. 109. Shen X., T. Storm, and M. A. Kay 2007. Characterization of the relationship of AAV capsid domain swapping to liver transduction efficiency. Mol. Ther. 15 :1955 1962. 110. Singer A G., F. Ghomashchi, C. Le Calvez, J. Bollinger, S. Bezzine, M. Rouault, M. Sadilek, E. Nguyen, M. Lazdunski, G. Lambeau, and M. H. Gelb 2002. Interfacial kinetic and binding properties of the complete set of human and mouse groups I, II, V, X, and XII se creted phospholipases A2. J. Biol. Chem. 277 :48535 48549. 111. Sonntag F., S. Bleker, B. Leuchs, R. Fischer, and J. A. Kleinschmidt 2006. Adeno associated virus type 2 capsids with externalized VP1/VP2 trafficking domains are generated prior to passage t hrough the cytoplasm and are maintained until uncoating occurs in the nucleus. J. Virol. 80 :11040 11054. 112. Stahelin R. V. 2009. Lipid binding domains: more than simple lipid effectors. J. Lipid Res. 50 Suppl :S299 304. 113. Stahnke S., K. Lux, S. Uhrig F. Kreppel, M. Hsel, O. Coutelle, M. Ogris, M. Hallek, and H. Bning 2011. Intrinsic phospholipase A2 activity of adeno associated virus is involved in endosomal escape of incoming particles. Virology 409 :77 83. 114. Suikkanen S., T. Aaltonen, M. Neva lainen, O. Vlilehto, L. Lindholm, M. Vuento, and M. Vihinen Ranta 2003. Exploitation of microtubule cytoskeleton and dynein during parvoviral traffic toward the nucleus. J. Virol. 77 :10270 10279. 115. Suikkanen S., M. Antila, A. Jaatinen, M. Vihinen Ran ta, and M. Vuento 2003. Release of canine parvovirus from endocytic vesicles. Virology 316 :267 280. 116. Summerford C., and R. J. Samulski 1998. Membrane associated heparan sulfate proteoglycan is a receptor for adeno associated virus type 2 virions. J. Virol. 72 :1438 1445. 117. Thomas C. E., T. A. Storm, Z. Huang, and M. A. Kay 2004. Rapid uncoating of vector genomes is the key to efficient liver transduction with pseudotyped adeno associated virus vectors. J. Virol. 78 :3110 3122.
130 118. Thompson J. D. D. G. Higgins, and T. J. Gibson 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. 119. Vihinen Ra nta M., D. Wang, W. S. Weichert, and C. R. Parrish 2002. The VP1 N terminal sequence of canine parvovirus affects nuclear transport of capsids and efficient cell infection. J. Virol. 76 :1884 1891. 120. Walters R. W., M. Agbandje McKenna, V. D. Bowman, T. O. Moninger, N. H. Olson, M. Seiler, J. A. Chiorini, T. S. Baker, and J. Zabner 2004. Structure of Adeno Associated Virus Serotype 5. Journal of Virology 78 :3361 3371. 121. Walters R. W., S. M. P. Yi, S. Keshavjee, K. E. Brown, M. J. Welsh, J. A. Chior ini, and J. Zabner 2001. Binding of Adeno associated Virus Type 5 to 2,3 Linked Sialic Acid Is Required for Gene Transfer. Journal of Biological Chemistry 276 :20610 20616. 122. Warrington K. H. Jr, O. S. Gorbatyuk, J. K. Harrison, S. R. Opie, S. Zolotuk hin, and N. Muzyczka 2004. Adeno associated virus type 2 VP2 capsid protein is nonessential and can tolerate large peptide insertions at its N terminus. J. Virol. 78 :6595 6609. 123. Welbourn S., and A. Pause 2007. The hepatitis C virus NS2/3 protease. C urr Issues Mol Biol 9 :63 69. 124. White S., D. Scott, Z. Otwinowski, M. Gelb, and P. Sigler 1990. Crystal structure of cobra venom phospholipase A2 in a complex with a transition state analogue. Science 250 :1560 1563. 125. Wistuba A., S. Weger, A. Kern, and J. A. Kleinschmidt 1995. Intermediates of adeno associated virus type 2 assembly: identification of soluble complexes containing Rep and Cap proteins. J. Virol. 69 :5311 5319. 126. Witz J., and F. Brown 2001. Structural dynamics, an intrinsic p roperty of viral capsids. Arch. Virol. 146 :2263 2274. 127. Wobbe C. R., S. Mitra, and V. Ramakrishnan 1984. Structure of the capsid of Kilham rat virus from small angle neutron scattering. Biochemistry 23 :6565 6569. 128. Wobus C. E., B. Hgle Drr, A. G irod, G. Petersen, M. Hallek, and J. A. Kleinschmidt 2000. Monoclonal Antibodies against the Adeno Associated Virus Type 2 (AAV 2) Capsid: Epitope Mapping and Identification of Capsid Domains Involved in AAV 2 Cell Interaction and Neutralization of AAV 2 Infection. J Virol 74 :9281 9293.
131 129. Wu Z., A. Asokan, J. C. Grieger, L. Govindasamy, M. Agbandje McKenna, and R. J. Samulski 2006. Single amino acid changes can influence titer, heparin binding, and tissue tropism in different adeno associated virus se rotypes. J. Virol. 80 :11393 11397. 130. Xiao W., K. H. Warrington Jr, P. Hearing, J. Hughes, and N. Muzyczka 2002. Adenovirus facilitated nuclear translocation of adeno associated virus type 2. J. Virol. 76 :11505 11517. 131. Xie Q., W. Bu, S. Bhatia, J. Hare, T. Somasundaram, A. Azzi, and M. S. Chapman 2002. The atomic structure of adeno associated virus (AAV 2), a vector for human gene therapy. Proceedings of the National Academy of Sciences 99 :10405 10410. 132. Xue B., R. L. Dunbrack, R. W. Williams A. K. Dunker, and V. N. Uversky 2010. PONDR FIT: a meta predictor of intrinsically disordered amino acids. Biochim. Biophys. Acta 1804 :996 1010. 133. Yan Z., R. Zak, G. W. G. Luxton, T. C. Ritchie, U. Bantel Schaal, and J. F. Engelhardt 2002. Ubiquiti nation of both adeno associated virus type 2 and 5 capsid proteins affects the transduction efficiency of recombinant vectors. J. Virol. 76 :2043 2053. 134. Zdori Z., J. Szelei, M. C. Lacoste, Y. Li, S. Garipy, P. Raymond, M. Allaire, I. R. Nabi, and P. Tijssen 2001. A viral phospholipase A2 is required for parvovirus infectivity. Dev. Cell 1 :291 302. 135. Zolotukhin S., M. Potter, I. Zolotukhin, Y. Sakai, S. Loiler, T. J. Fraites Jr, V. A. Chiodo, T. Phillipsberg, N. Muzyczka, W. W. Hauswirth, T. R. Fl otte, B. J. Byrne, and R. O. Snyder 2002. Production and purification of serotype 1, 2, and 5 recombinant adeno associated viral vectors. Methods 28 :158 167. 136. (IUCr) ATSAS 2.1 towards automated and web supported small angle scattering data analysis 137. 1994. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol. Crystallogr. 50 :760 763.
132 BIOGRAPHICAL SKETCH Bala completed his Bachelor in Technology (B. Tech.) in i ndustrial b iotechnology at Anna University, Chenna i, India in 2007. His undergrad research ( under the guidance of Dr. P ennathur Gautham ) was on the systemic bioleaching of lignite by acidophilus bacteria. He joined the IDP at the University of Florida in August 2007. After a set of rotations, he joined the McKenna L ab and the Department of Biochemistry and Molecular Biology in April 2008. His doctoral thesis was on pH mediated structural transitions in Aden o Associated Viruses. His areas of interest include Structural Virology, Macromolecular Crystallography, Molecular Biophysics and Enzymology. His interests outside academics include sports, music, computers, animal care and community service.