Parvovirus Capsid Structures, Ligand Binding Interactions, And

PARVOVIRUS CAPSID STRUCTURES, LIGAND BINDING INTERACTIONS, AND

ENDOGENOUS VIRAL ELEMENTS

A Dissertation

Presented to the Faculty of the Graduate School

of Cornell University

in Partial Fulfillment of the Requirements of the Degree of

Doctor of Philosophy

Heather Merrick Callaway

August 2018

PARVOVIRUS CAPSID STRUCTURES, LIGAND BINDING INTERACTIONS, AND

ENDOGENOUS VIRAL ELEMENTS

Heather Merrick Callaway, Ph.D.

Cornell University 2018

Parvoviruses are among the simplest of viruses, with capsids composed of variants of a single structural protein. These capsids mediate many of the processes required for infection and are remarkably stable. However, antibody binding or mutations to the capsid can disrupt these processes and block infection. This dissertation discusses mutations to the canine parvovirus

(CPV) capsid that result in loss of infection, the interaction between CPV and the transferrin receptor (TfR), and parvovirus endogenous viral elements (EVEs), which contain ancient parvovirus gene sequences.

I found that point mutations to CPV VP2 residues 270, 272, 273, and 299/300 result in loss of viral infectivity. Mutation of residue 270 results in loss of a sub-molar proteolytic cleavage event in VP2 and increases capsid stability, residue 272 mutation causes loss of capsid assembly, and residue 273 mutation results in assembled capsids being trapped in the nucleus.

Mutation of VP2 residues 299 and 300, which are associated with TfR binding, to lysine disrupts the interaction between CPV and the TfR, inhibiting infection but still allowing receptor binding and uptake into cells.

I also found that CPV has different interactions with TfRs from different host species, binding strongly to some TfRs and very weakly to others, even though each TfR can mediate a successful infection. TfRs from different species also had varying levels of occupancy on CPV capsids, with up to 12 black-backed jackal TfRs, but only 1-2 feline TfRs binding to each capsid, and it is possible that the feline TfR induces a conformational change in CPV that inhibits binding of additional TfRs. Antibody binding could also disrupt the CPV/TfR binding interaction, suggesting a possible mechanism of virus neutralization.

To examine the structure and function of ancient parvovirus capsids, I expressed the VP2 gene from three different parvovirus EVEs. VP2 expressed from an EVE in the M. spretus genome assembled into capsids, and I determined that these capsids were highly stable, could bind to N-Acetylneuraminic acid, and were endocytosed into murine cells.

The results in this dissertation provide new information about parvovirus infection, receptor binding interactions, and evolution, and further the understanding of how infection occurs and may be disrupted.

BIOGRAPHICAL SKETCH

Heather Callaway attended White Station High School and graduated in 2009. She went to college at Emory University and graduated summa cum laude in 2013 with a double major in biology and physics. Throughout college, Heather worked in several research laboratories, with projects including doing behavioral experiments on rats, examining how hypoxia affects the growth of cancer cells, and sequencing rhodopsin and other photoreceptor genes in fish. Her undergraduate thesis work focused on understanding how cellular proliferation in Drosophila melanogaster was affected by mutations to the gene domino.

Heather became interested in working with viruses after learning about bacteriophages in high school biology. Once enrolled in the Ph.D. program at Cornell University, Heather rotated in the labs of Dr. Joel Baines, Dr. John March, and Dr. Colin Parrish. Since joining Colin

Parrish’s lab, Heather has been working on understanding how parvoviruses interact with receptors and antibodies, and what makes them infectious. After graduating from Cornell,

Heather will do a postdoc in the lab of Dr. Erica Ollmann Saphire, working on rabies virus glycoprotein structure.

iii

For my parents, Dawn and Jay,

and my sister, Erin

ACKNOWLEDGMENTS

Thank you to everyone who helped me throughout my time in graduate school. I am very grateful to have had Colin Parrish as a mentor. Colin has provided me with excellent feedback on my research, experimental design, and writing, and I have become a better scientist because of his guidance. To Wendy Weichert, I owe a debt of gratitude for training me on how to do all of the biochemical assays I needed for my research, for help in troubleshooting my experiments, and for excellent technical support. Thank you for keeping us sane and the lab running, and for everything else you do.

I am grateful to my committee members, Dr. Linda Nicholson, Dr. Gary Whittaker, and

Dr. Bettina Wagner, for their help and guidance over the past five years, particularly on experimental design. I am also very grateful to Shelagh Johnston for all the help with sorting out scheduling and travel, as well as for keeping me on my toes. I am not sure I will ever be able to look at the California highway system the same way again.

To Karen, Matt, Melissa, Becky, Brynn, Ian, Brian, Esteban, Furkat, Ed, Chip, and

Wayne – thank you for your friendship, for making the lab a fun place to work, and for testing out all of my culinary experiments, good and bad. Thank you to all the folks at Baker who have made this such a wonderful place to work.

I owe thanks, lastly, to my parents, for everything they do, and to my sister, for keeping me sane and from falling too far down the rabbit hole into scientific jargon. I agree – tea should be steeped, not incubated.

TABLE OF CONTENTS

Biographical sketch ...... iii

Acknowledgements ...... v

List of figures ...... x

List of tables ...... xii

CHAPTER ONE: Introduction ...... 1

1.1 Emergence and importance of canine parvovirus ...... 2

1.2 Viral proteins and capsid structure ...... 3

1.3 Parvovirus receptor binding ...... 7

1.3a Canine parvovirus and the transferrin receptor ...... 7

1.3b Sialic acid binding...... 10

1.4 Parvovirus infection ...... 11

1.4a Uptake, endosomal escape, and intracellular trafficking ...... 11

1.4b DNA uncoating, RNA splicing, and protein expression ...... 13

1.4c Capsid assembly, genome replication, and packaging ...... 14

1.4d Nuclear egress and release from the cell ...... 17

1.5 Evolution and host adaptation of canine parvovirus ...... 18

1.6 Parvovirus/antibody interactions ...... 20

1.6a Antibody recognition of canine parvovirus...... 20

1.6b Mechanisms of antibody neutralization ...... 23

1.7 Biophysical and biochemical properties of parvovirus capsids ...... 24

1.8 Critical structures in parvovirus capsids ...... 25

1.9 Endogenous viral elements ...... 27

1.9a Origin of endogenous viral elements ...... 27

1.9b Identification and dating of endogenous viral elements ...... 28

1.9c Reconstruction of endogenous viral elements...... 29

1.9d Host suppression of endogenous viral elements ...... 30

1.9e Host benefits of endogenous viral elements...... 31

1.10 Dissertation overview ...... 34

1.11 References ...... 35

CHAPTER TWO: Parvovirus capsid structures required for infection: mutations controlling receptor recognition and protease cleavages ...... 52

2.1 Abstract ...... 53

2.2 Importance ...... 53

2.3 Introduction ...... 54

2.4 Materials and Methods ...... 58

2.5 Results ...... 64

2.6 Discussion ...... 80

2.7 Acknowledgements ...... 85

2.8 References ...... 86

CHAPTER THREE: Complex and dynamic interactions between parvovirus capsids, transferrin receptors, and antibodies control cell infection and host range ...... 95

vii

3.1 Abstract ...... 96

3.2 Importance ...... 96

3.3 Introduction ...... 97

3.4 Results ...... 102

3.5 Discussion ...... 121

3.6 Methods...... 125

3.7 Acknowledgements ...... 133

3.8 References ...... 134

CHAPTER FOUR: Examination and reconstruction of ancient, endogenous parvovirus capsid proteins in rodent genomes ...... 144

4.1 Abstract ...... 145

4.2 Importance ...... 145

4.3 Introduction ...... 146

4.4 Results ...... 149

4.5 Discussion ...... 162

4.6 Methods...... 166

4.7 Acknowledgements ...... 172

4.8 References ...... 173

CHAPTER FIVE: Summary and conclusions ...... 177

5.1 Mutations to critical capsid structures ...... 178

5.2 Receptor and antibody binding interactions ...... 179

viii

5.3 Parvovirus endogenous viral elements...... 182

5.4 Conclusions and future directions ...... 184

LIST OF FIGURES

Figure 1.1. The CPV genome and mRNA transcripts ...... 4

Figure 1.2. CPV capsid structure ...... 6

Figure 1.3. The structure of the TfR ...... 8

Figure 1.4. Parvovirus genome replication ...... 16

Figure 1.5. Antigenic sites on the CPV capsid ...... 21

Figure 2.1. Protein structure, surface morphology, and viability of the external capsid surface

mutations ...... 66

Figure 2.2. Binding and uptake of 4μg/mL CPV or VLP into NLFK cells determined by

immunofluorescence assay ...... 69

Figure 2.3. Flow cytometry showing binding and uptake assay of 4μg/mL CPV or VLP into

NLFK cells ...... 70

Figure 2.4. Binding of CPV or VLP to soluble feline TfR using bio-layer interferometry ..... 71

Figure 2.5. Structure and sequence conservation of the capsid region on the interior surface of

CPV that was mutated ...... 75

Figure 2.6. Capsid assembly, viability, and proteolytic cleavage of the mutations changing

residues on the internal capsid surface...... 76

Figure 2.7. Stability of wild-type or mutant VLPs determined using tryptophan fluorescence

spectroscopy or differential scanning fluorometry ...... 79

Figure 2.8. Intracellular localizations of the Cys273Ser mutant compared to wild-type capsids

...... 81

Figure 3.1. Composition of ligands and VLPs used in this study ...... 103

Figure 3.2. Binding kinetics of parvovirus capsids to TfRs from different species...... 107

Figure 3.3. Relative infectivity of CPV and FPV on cells transfected to express

different TfRs ...... 109

Figure 3.4. Binding kinetics of Fabs or TfRs...... 112

Figure 3.5. Competition between scFv-Fcs and Fabs or TfRs and Fabs for capsid binding .. 115

Figure 3.6. Effects of the initial CPV/ligand interaction on subsequent bbjTfR or

felTfR binding ...... 118

Figure 3.7. Binding kinetics of mutant VLPs with TfRs, scFv-Fcs, and Fabs ...... 119

Figure 3.8. Summary of major findings ...... 123

Figure 4.1. Endogenous viral element insertions ...... 150

Figure 4.2. Sequence similarity between the endogenous viral elements and the most closely

related extant virus ...... 152

Figure 4.3. Expression and assembly of chimeras constructed by combining BglII and

PhoI restriction enzyme digestions of the Rattus Norvegicus EVE with

those of CPV-2 ...... 155

Figure 4.4. Construction and expression of chimeric VLPs consisting of Rattus Norvegicus EVE

surface loops incorporated into CPV VP-2 ...... 156

Figure 4.5. Cellular uptake and negative stain EM of Loop 4 chimeric R. Norvegicus EVE/CPV-

2 VLPs ...... 159

Figure 4.6. Expression and assembly of A. sylvaticus and M. spretus EVE VLPs...... 160

Figure 4.7. Characterization of the assembled Mus Spretus EVE VLPs ...... 161

LIST OF TABLES

Table 2.1. Fitting constants for CPV and VLP external capsid mutants, derived from fitting bio-

layer interferometry sensorgrams to Eq.1 ...... 73

Table 3.1. Differences in VP2 protein sequences for viruses and VLPs, relative to CPV-2 .. 105

Table 4.1. Percentage of O-linked sialic acids on murine erythrocytes...... 163

Table 4.2. Specimens used to determine the presence of the rat EVE ...... 168

xii

CHAPTER ONE

Introduction

Parvoviruses are extremely small, single-stranded DNA viruses that infect a wide variety of hosts, ranging from invertebrates to mammals, and are divided into two types: autonomous parvoviruses and dependoviruses. Autonomous parvoviruses, which include canine parvovirus

(CPV), feline panleukopenia virus (FPV), mink enteritis virus (MEV), the minute virus of mice

(MVM), human parvovirus B19, and many others, are capable of independent replication in a host. Dependoviruses, which include the numerous types of adeno-associated viruses (AAV) and their relatives, require a helper virus in order to replicate. Many parvoviruses cause subclinical infection or result in mild disease, but others can be fatal. CPV and its close relatives, FPV and

MEV, fall into the latter category. CPV and MEV can both cause severe enteritis in their hosts, while FPV infection often causes leukocyte depletion (1). CPV has also been reported to cause myocarditis in puppies and FPV to cause neurological damage in kittens (2–4).

In addition to causing disease, parvovirus infection can also result in the formation of endogenous viral elements (EVEs), which are segments of viral DNA that are integrated into the host genome and vertically transmitted from parent to offspring. What, if any, biological purpose parvovirus EVEs have remains unknown, but these integrated sequences can provide information about the structure and function of ancient parvovirus proteins and the long-term evolution of the parvoviruses.

1.1 Emergence and importance of canine parvovirus

Canine parvovirus was first reported in 1978, although some estimates place the time of its emergence at as early as 1968 (1,5). Soon after it was first reported, it was discovered that

CPV was a parvovirus and was similar enough to FPV, a long-established cat parvovirus, that

FPV antisera cross reacted with it (6). CPV continued to evolve after its emergence, and by the

early 1980s, a new antigenic variant of CPV had completely replaced the original CPV strain from 1978, CPV-2 (7). This new strain, CPV-2a, and its minor variants (sometimes designated

CPV-2b and CPV-2c) continue to circulate today in populations of unvaccinated canines and other susceptible carnivores (8–10).

CPV’s emergence and subsequent evolution provide an important example of host jumping and adaptation, where viral mutations can be observed in real time. Because of the simplicity of the CPV capsid, which mediates most of the early steps of infection, it is possible to determine which residues are responsible for changes in tropism and infectivity and predict how the virus might change in the future. The capsid can also be studied to learn more about structures required for infection, many of which have yet to be discovered, and which may have analogs in more complex viruses, as well as in other parvoviruses.

1.2 Viral proteins and capsid structure

CPV virions are composed of a ~5100bp single-stranded DNA genome and a nonenveloped capsid that is 26 nm in diameter. The genome encodes two non-structural (NS1 and

NS2) and two structural (VP1 and VP2) proteins, with overlapping open reading frames for

NS1/NS2 and VP1/VP2 (Figure 1.1). The structural proteins make up the capsid, while the nonstructural proteins are involved in genome replication and packaging (NS1) or have an unknown, yet dispensable, function in CPV infection (NS2) (11). There is an additional open reading frame overlapping the VP1/VP2 genes in all autonomous parvoviruses, which encodes a protein called

SAT and has been shown to facilitate viral spread in PPV (12). However, what, if any, roles it plays during CPV infection have not yet been reported.

The CPV capsid is a T=1 icosahedron composed of 60 copies of the viral structural

Figure 1.1. The CPV genome and mRNA transcripts. A) The CPV genome (13), showing the P4 and P38 promoters and the open reading frames encoding the non-structural genes (NS1/NS2) and the structural genes (VP1/VP2). B) Transcripts of the CPV genome, with the encoded gene labeled and indicated by a colored rectangle. RNA splice sites are also shown.

proteins (10% VP1 and 90% VP2). VP1, the minor capsid protein, and VP2, the major capsid protein, share 584 amino acids that form the basic structure of the icosahedron. VP1 contains an additional 143 N-terminal amino acids that have phospholipase A2 (PLA2) enzymatic activity and a nuclear localization signal (NLS), which are required for infection but not capsid assembly.

The shared residues of VP1 and VP2 assemble into an 8-stranded β-barrel, with long loops between some of the strands (Figure 1.2) (14). These connecting loops form the exterior surface of the capsid, while the β-barrel faces inwards and gives the capsid structural stability. Repeating motifs in the ssDNA genome interact with residues on the β-barrel on the interior surface of the capsid, creating regions of electron density that can be visualized in crystal structures of CPV

(15).

On the exterior surface of the capsid, pores located at the 5-fold axes of symmetry serve as portals through which homo-hexamers of NS1 package the viral genome (16,17). NS1 has been shown to remain associated with the 5' terminus of the genome after packaging, and it, along with approximately 24 base pairs of DNA, remain outside of the capsid and exposed to nucleases after genome encapsulation (18,19). Other features on the exterior surface of the capsid include raised "spikes" at the 3-fold axis of symmetry and "dimples" at the 2-fold axis of symmetry, which serve as antigenic sites and also contain the receptor binding site (14) (Figure

1.2).

In full capsids, where the genome has been packaged into the capsid, the N-terminus of

VP2 can protrude from the 5-fold pore (except the pore through which the genome was packaged), resulting in the cleavage of 20 amino acids and producing VP3 (20). VP1 can also be exposed through the 5-fold pore, but does so only after capsids have entered the endosome and the endosomes have acidified or the capsids have been exposed to harsh conditions (heat, urea,

Figure 1.2. CPV capsid structure. A) One VP2 monomer (PDB ID: 2CAS) viewed from the side.

Surface exposed portions of loops are color coded. B) A VP2 monomer viewed from the front, with surface exposed portions of loops labeled and color coded. C) A VP2 monomer shown in context of an assembled capsid, with surface exposed portions of loops color coded. One 5-fold axis of symmetry is indicated by a pentagon, two 3-fold axes of symmetry are indicated by triangles, and one 2-fold axis of symmetry is indicated by an oval.

etc.) in vitro (20,21). While empty capsids do not undergo proteolytic cleavage to produce VP3

(20) and lack the genome, they are still able to bind to receptors and enter cells.

1.3 Parvovirus receptor binding

1.3a Canine parvovirus and the transferrin receptor

The transferrin receptor type-1 (TfR) is the cellular receptor that CPV uses to enter and infect cells (22). The TfR is a single-pass transmembrane protein, consisting of two identical subunits that dimerize to form a butterfly-shaped protein (Figure 1.3A). Each half of the TfR dimer contains an apical domain, a protease-like domain, a helical domain, and a transmembrane/stalk domain. The two subunits interface at the helical domain and are held together by hydrogen bonds and van der Waals interactions there, as well as by two disulfide bridges in the transmembrane/stalk domain, outside of the cell membrane (23,24). The interactions in the helical domain alone are sufficient to hold the subunits together, as TfR cleaved from the membrane crystallizes as a dimer (25) and soluble forms of the TfR lacking the transmembrane domain are secreted as dimers (24).

Transferrin, a protein responsible for transporting iron from the blood stream into the cell, binds to 1-2 molecules of iron and then attaches to the TfR across the protease-like domain at neutral pH (26,27). Two transferrin proteins typically attach to the TfR dimer at a time, one on each protease-like domain (27). The TfR/transferrin complex is then endocytosed into the cell via clathrin-mediated endocytosis (28). Iron molecules are released from transferrin upon endosomal acidification, likely with the help of a chelating agent present in the endosome, and the TfR/transferrin complex is then recycled back to the cell surface (26). Upon exposure to neutral pH outside of the cell, apo-transferrin (transferrin with no iron bound) is released from

Figure 1.3. The structure of the TfR. A) The TfR ectodomain (PDB ID: 1SUV), with the apical domain colored green, the protease-like domain colored red, and the helical domain colored yellow. The transmembrane/stalk domain consists of residues 1-121 and extends from the protease-like domain, but is not depicted. B) The TfR ectodomain bound to two iron-loaded transferrin proteins (blue) (PDB ID: 1SUV).

the TfR, allowing a new, iron-bound transferrin proteins to attach to the TfR (26,29). In the intestine, the iron-binding protein HFE (hemochromatosis protein) binds across the TfR protease-like domain similarly to transferrin to facilitate iron uptake (30,31).

CPV binds to the transferrin receptor apical domain, which is not sterically occluded after

Tf or HFE binds to the TfR (27,32) (Figure 1.3B). On the surface of the capsid, the putative TfR binding site overlaps portions of both the 3-fold "spikes" and the 2-fold "dimple" (33) and mutations to these regions of the capsid have been shown to affect host cell tropism and viral uptake into cells. At the 2-fold "dimple", Gly299Glu mutation inhibits uptake of CPV into canine cells, preventing infection (34), while Ala300Asp mutation is a host adaptation mutation that has been observed in raccoons that have been infected in the wild and, rarely, after culture of CPV-2 in feline cells (34–36). Other mutations around the 2-fold “dimple” at VP2 residues 297-301 have been observed in nature, with residue variation depending on the carnivore species CPV was isolated from (35,37,38). Around the 3-fold "spike", mutations to VP2 residues 87, 93, 101,

103, 224, and 426 have been observed in CPV isolates from different carnivore hosts (35,38), and experimental mutations at residues 93 and 103 have been shown to result in the loss of canine host cell tropism (39).

The interaction between CPV and the TfR is complex, and much about it is still not understood. CPV infection requires more than simple adhesion for uptake into cells, as modifications to the TfR that block infection still allow viral uptake into cells. These modifications include both point mutations to the TfR apical domain and chimeric receptors, where the entire apical domain has been replaced with an antibody fragment. Leu221Lys mutation of the feline TfR, in particular, has been shown to allow viral uptake into cells, but not infection (32). Similarly, when the TfR ectodomain was replaced with single-chain variable

fragments (scFvs) derived from either of two different anti-CPV monoclonal antibodies, cellular uptake occurred without infection (40).

Although an estimated 12 to 24 TfR binding sites exist on the capsid surface, depending on the angle at which the TfR binds, both in vitro and in vivo studies have shown that five or fewer copies of the feline TfR bind to CPV capsids (33,41). While TfR membrane attachment and cellular membrane curvature could explain low TfR occupancy on CPV capsids in cell binding experiments, soluble TfRs incubated with CPV capsids in solution show similar results, indicating that either a limited number of TfR binding sites are available or that the feline TfR induces conformational changes in CPV that inhibit the attachment of additional TfRs. If the former is true, the limited number of TfR binding sites must be clustered on one side of the capsid in order for the cell binding experiments, where binding is constrained by membrane curvature, to yield the same results as the experiments with soluble TfR.

1.3b Sialic acid binding

MVM and PPV both use sialic acid as their primary cellular receptor. MVM binds to α2-

3 and α2-8 N-linked sialic acids (42,43), while PPV binds to α2-3 and α2-6 N- and O-linked sialic acids (44). The MVM sialic acid binding site has been characterized, and capsids that have mutations at residues involved in that binding interaction have been shown to have lower infectivity on neuraminidase treated cells than wild-type MVM particles, suggesting a lower binding affinity for sialic acid (45). While the sialic acid binding site for PPV has not yet been identified, one study revealed that PPV particles could also be taken up into cells via macropinocytosis, possibly with the help of a yet unknown cellular receptor in addition to sialic acid (44).

CPV binds to N-glycolylneuraminic acid (Neu5Gc) (46), a modified form of sialic acid that has an N-linked glycolyl group at the carbon 5 position. However, western dog breeds lack the CMAH enzyme that produces Neu5Gc (46), and mutation of VP2 residues 323, 375, and

377, which control sialic acid binding, do not result in loss of infectivity in dog cells (47,48).

One possible explanation for CPV sialic acid binding is that this functionality was important for the CPV/FPV common ancestor, and CPV retained it after its emergence.

1.4 Parvovirus infection

Parvoviruses infect cells by first binding to a cellular receptor (TfR, sialic acid, etc.) and getting taken up into endosomes via receptor mediated endocytosis. In the endosome, the capsid’s VP1 N-termini become exposed and mediate escape from the nucleus and trafficking to the nucleus, where DNA uncoating occurs. Viral transcription, RNA splicing, and transcription of RNA into protein follow, along with capsid assembly and genome replication, resulting in the production of infectious virions. The progeny viruses then escape the nucleus and the cell, and are free to bind to a new cellular receptor, repeating the cycle. Failure to complete any of these steps results in a loss of infection, and can be caused antibody binding, improper interaction with the viral receptor, or mutations to viral proteins. Much is still unknown about how infection is mediated, and studying how loss of infection occurs can provide new information about how viruses work.

1.4a Uptake, endosomal escape, and intracellular trafficking

Although CPV uses the TfR to enter both canine and feline cells, it has different patterns of uptake depending on cell type. Virions adhere primarily to the body of feline cells (Crandell-

Rees Feline Kidney (CRFK) cells), but bind to the filopodia of canine cells (Cf2Th, an immortalized canine thymus cell line) (49). Approximately 15 seconds after binding to the TfR,

CPV is taken up into the cell via clathrin-mediated endocytosis (41), and rapidly becomes associated with Rab 5, 7, and 11 positive endosomal compartments (49). These endosomes then travel to the cytoplasm near the nucleus in a dynamin and microtubule dependent manner

(50,51). CPV and the TfR can remain associated inside the endosome at least 4 hours, as antibodies targeting the TfR cytoplasmic tail injected into the cell cytoplasm can inhibit infection up to this point (22), but some parvovirus capsids (such as AAV) have also been reported to reach the nucleus as early as 15 minutes after being added to cells (52). While inside the endosome, CPV releases the N-terminus of VP1 through the pore at the 5-fold axis of symmetry.

This process requires both the prior cleavage of the VP2 N-termini on adjacent capsid subunits, to clear the 5-fold pore, and endosomal acidification (53,54). Inhibition of endosomal acidification with Bafilomycin A1, chloroquine, or ammonium chloride prevents the exposure of

VP1 and results in loss of infectivity, but treatment of purified capsids with acidic pH does not result in VP1 exposure or restore infectivity, suggesting that cellular factors may also play a role in VP1 uncoating (50,55). The exposed VP1 N-terminus, which contains PLA2 enzymatic activity, then disrupts the endosomal membrane, allowing capsids to escape into the cytoplasm.

It was shown that the holes that CPV opens in the endosomal membrane are likely small and/or localized as they do not allow escape of endocytosed alpha-sarcin (~18kDa) (50).

Upon release into the cytoplasm near the nucleus, a nuclear localization signal (NLS) also located on the VP1 N-terminus allows CPV to traffic to the nucleus. Microinjection of monoclonal anti-capsid antibodies into the cell cytoplasm blocks infection, indicating that intact capsids, as opposed to uncoated DNA, traffic through the cytoplasm (51). A small fraction of

capsids enters the nucleus (49), where uncoating occurs and viral DNA is released for transcription and replication.

1.4b DNA uncoating, RNA splicing, and protein expression

The process by which DNA is uncoated from parvovirus capsids in the nucleus is still incompletely understood, but some experiments have shown that removing Ca2+ and Mg2+ from

MVM capsids with ethylenediaminetetraacetic acid (EDTA) results in the 3’ to 5’ release of the genome at physiological temperature (56). Cellular conditions that might trigger the removal of divalent cations from parvovirus capsids include low cytoplasmic Ca2+ concentration, an influx of dNTPs during cellular genome replication, or an as yet unknown cellular chelator (56).

Human parvovirus B19 has been shown to release its DNA in vitro when exposed to endosomal pH (57), which suggests that capsid conformational changes triggered by acidification may also play a role in uncoating. Studies with both MVM and B19 have shown that capsids remain associated with the 5’ end of viral DNA after uncoating, which the authors hypothesize may help with trafficking into specific nuclear compartments (56,57).

Once the parvovirus genome has reached the nucleus and is uncoated, transcription occurs, and several splicing events produce viral mRNA transcripts. As with MVM and other autonomous parvoviruses, CPV mRNA contains separate promoters for the nonstructural and structural proteins (P4 and P38 promoters, respectively) as well as two introns, resulting in a total of three distinct mRNAs after splicing (11,58,59) (Figure 1.1B). R1 mRNA, which encodes

NS1, is formed when RNA is transcribed from the nonstructural gene promoter (P4) and only the small intron is removed. R2, which encodes NS2, is also transcribed from the nonstructural gene promoter (P4) and results from the splicing of both the large and small introns. R3 mRNA, which

encodes VP1 and VP2, is transcribed from the structural gene promoter (P38) and has the small intron removed (Figure 1.1B). Approximately 5 times more VP2 protein is produced than VP1 both in vitro and in vivo (58,60), which must result from translational differences, as both genes share the same mRNA transcript. Activation of the P38 promoter also requires prior NS1 expression, suggesting a mechanism that regulates the timing of capsid production (61,62).

A non-structural protein, designated SAT, has been observed in PPV, with analogs in all autonomous, vertebrate parvoviruses (protoparvoviruses), including CPV (12). SAT knockout results in slower growth in tissue culture (12) and has been implicated in the endoplasmic reticulum stress response during infection and virion release (63). The adeno-associated viruses, which are more distantly related to autonomous parvoviruses like CPV, PPV, and MVM, produce transcripts for assembly-activating protein (AAP) from the mRNA encoding the structural proteins (64). AAP, which is expressed from an alternative open reading frame in the structural protein mRNA, has been shown to target capsid proteins to the nucleus and facilitate capsid assembly (64). No similar open reading frame exists for CPV or MVM, however, indicating that this protein is restricted to the adeno-associated viruses or dependoviruses.

1.4c Capsid assembly, genome replication, and packaging

VP1 and VP2 capsid subunits are produced in the cytoplasm, where they assemble into trimers before trafficking to the nucleus for assembly upon the induction of S phase of the cell cycle (65–67). Trimers consist of either three VP2 subunits or two VP2 subunits and one VP1 subunit, and the two trimer forms exist in roughly equimolar amounts (65,67). VP1 and VP2 N- termini are concealed from antibodies in the cytoplasm, and recent studies have shown that trimer subunits are phosphorylated in the cytoplasm, possibly causing the obscuring of the N-

termini and/or preventing trimers from assembling into capsids prematurely (65). There is evidence that the cellular protein Raf1 causes this phosphorylation, as parvovirus capsid proteins expressed in insect cells, which do not express Raf1, assemble in the cytoplasm instead of the nucleus (68,69). If the insect cells are transfected to express Raf1, however, assembly occurs in the nucleus (68).

After capsids have assembled in the nucleus during early to mid-S phase of the cell cycle, viral genome replication occurs (66). Parvoviruses use ‘rolling-hairpin’ replication to make new copies of the genome. First, complementary bases at each end of the genome pair with the DNA beside them, forming short hairpins on each end of the genome (Figure 1.4A) (70). Host DNA polymerases and enzymes (including DNA polymerase δ and proliferating cell nuclear antigen

(PCNA)) initiate replication in a 5’ to 3’ direction, unfolding the short tandem repeat at the opposite end of the viral genome as it progresses, and resulting in a double-stranded DNA hairpin (Figure 1.4B) (66,70,71). The complementary bases at the end of the new hairpin fold back again (Figure 1.4C) and host DNA polymerases initiate replication, forming another hairpin containing two concatemeric, double-stranded parvovirus genomes (Figure 1.4D), and this process repeats, forming a double-stranded tetrameric hairpin (Figure 1.4E and F). Individual double-stranded DNA genomes are excised from these intermediates via the formation and separation of complex, hetero-cruciform DNA structures, which involves NS1 and host enzymes

(70) (Figure 1.4G). The positive sense strand of the double-stranded DNA genome can then form a hairpin, beginning another round of replication and displacing the negative sense strand of the genome (72) (Figure 1.4H and I). The displaced negative sense strand can also serve as a template for DNA replication, or it can be packaged into an assembled capsid (72). Positive sense, single-stranded DNA genomes are generated when NS1 nicks the replication origin of the

Figure 1.4. Parvovirus genome replication. A-F) Replication of parvoviral DNA by host polymerases, starting with a negative sense DNA genome. G-I show production of negative sense, single-stranded DNA genomes for encapsidation starting from a double-stranded monomeric genome, which results from the steps depicted in A-F. Colored arrows indicate the direction of DNA synthesis, 5’ and 3’ ends of DNA are labeled, positive (+) and negative (-) sense genomes are indicated, and left (L) and right (R) hairpins are indicated. Information compiled from (70,72).

double-stranded dimer intermediate (Figure 1.4D), allowing host polymerases to use the negative sense strand as a template and displacing the positive sense strand (72).

Once viral capsids have assembled in the nucleus and single-stranded genomes have been generated, NS1 hexamers package the single-stranded DNA genome through a 5-fold pore in a

3’ to 5’ direction (56). Some autonomous parvoviruses, including MVM, preferentially package negative sense ssDNA genomes (70)(69)(Cotmore and Tattersall, 1996), while the adeno- associated viruses show no preference for positive or negative sense DNA (70). After the DNA is packaged into the capsid, NS1 remains attached to the 5’ end via a covalent bond at residue

Tyr210 (MVM NS1) (18,73).

1.4d Nuclear egress and release from the cell

Many aspects about how parvoviruses exit cells are still unknown, but growing evidence supports a model of active transport of full capsids from the nucleus prior to release from the cell or cell lysis. In order to exit the nucleus, MVM requires externalized VP2 tails, which occurs only in DNA containing capsids, and phosphorylated surface residues, as well as the nuclear exportin CRM1, which interacts with NS2 in a yet unknown way to facilitate nuclear egress (74–

76).

Furthermore, while the majority of parvovirus capsids produced during an infection appear to be released from the cell upon cell lysis, some MVM capsids have been observed to be secreted from cells prior to lysis (77,78). After leaving the nucleus, MVM full capsids are incorporated into vesicles by the perinuclear endoplasmic reticulum, trafficked through the endoplasmic reticulum to the Golgi, where they undergo phosphorylation, and then transported to the plasma membrane for extracellular release (77). This process has been shown to be

dependent on NS1-mediated phosphorylation of gelsolin, which triggers the depolymerization of actin filaments, and which in turn has been hypothesized to promote microtubule-dependent vesicle transport to the plasma membrane (78). Transport of MVM capsids through the endoplasmic reticulum and the Golgi has also been reported to accelerate cell lysis, suggesting that this process regulates when parvoviruses trigger cell death (77).

1.5 Evolution and host adaptation of canine parvovirus

CPV-2 emerged from feline panleukopenia virus, or a virus closely related to it, as early as 1968 (5). CPV and FPV share over 99% of their sequence identity, and while FPV can infect numerous carnivore species, dogs are resistant to infection as FPV cannot bind to the canine TfR.

The primary differences between CPV and FPV are changes to the major capsid protein VP2, which include Asn80Arg, Lys93Asn, Thr101Ile, Val103Ala, Val232Ile, Tyr303Phe, Asp323Asn,

Asp375Asn, Asn564Ser, and Ala568Gly. Only the mutations to residues 93 and 323 are required to give FPV canine cell tropism, however, and replacement of residues 93 or 323 in CPV-2 with the corresponding FPV residue, either singly or in combination, results in the loss of the canine host range (39). These mutations are believed to allow CPV to accommodate a bulky glycan on the canine TfR apical domain (residue 384) that prevents FPV binding (79). FPV can bind to and infect cells expressing the TfR from the black-backed jackal, the most closely related species to dogs that does not have this glycosylation site, but when the black-backed jackal TfR is modified to contain the glycosylation site (Lys384Asn mutation), FPV loses most of its ability to bind to and infect those cells (79). While the reverse experiment, mutating the canine TfR to remove the glycosylation site, was not performed, the canine and black-backed jackal TfRs differ in only 5 amino acids on the apical domain (12 amino acids total), suggesting that the results would be

similar.

Only a few years after CPV-2 emerged, it acquired an additional 4 amino acid mutations on VP2 that allowed it to better infect dogs (Met87Leu, Ile101Thr, Ala300Gly, and Asp305Tyr

(80). This new variant, designated CPV-2a, replaced CPV-2 worldwide by 1980 and could also infect cats, whereas CPV-2 could only infect cat cells (81). Evolutionary analysis of CPV isolates from raccoons suggest that CPV-2 passed through raccoons as an intermediate host, gaining mutations to VP2 residues 87, 101, and 300, before passing back into dogs and acquiring additional mutations to produce CPV-2a (35). Experiments where CPV-2 was mutated to contain all 16 combinations of the 4 residues distinguishing CPV-2 from CPV-2a showed that CPV-2 likely had to acquire the four mutations in a specific order, as many combinations were much less fit than wild-type CPV-2 in feline cells and some were even non-viable (82).

CPV-2a continued to evolve after replacing CPV-2, and additional mutations to VP2 have been observed at residue 426, with Asn426Asp resulting in a variant designated CPV-2b and 426Glu resulting in CPV-2c (80,83). Numerous other mutations to VP2 have also been observed in CPV isolated from different host species or passaged on different species’ cells, particularly around residues 103, 300, and 375, where mutations have been shown to impact host range or virus binding (38). Mutation to residue 300 has also been shown to select for additional mutations to nearby residues when CPV is grown on different host cells (37), marking it as a major determinant of host cell tropism as well as a residue that overlaps the TfR binding site

(33). Residue 103 is located adjacent to residue 300 and also overlaps the TfR binding site, while residue 375 affects binding to Neu5Gc sialic acid, which is absent in Western dog breeds, but is present in many carnivore species.

1.6 Parvovirus/antibody interactions

1.6a Antibody recognition of canine parvovirus

Monoclonal antibodies (Mabs) generated against CPV and FPV recognize one of three potential antigenic sites on the capsid surface. The "A-site" is located around the 3-fold axis of symmetry and the "B-site" is located around the 2-fold axis of symmetry, and both overlap regions on the capsid that have been associated with TfR binding (Figure 1.5). Antibodies that bind around the 5-fold axis of symmetry have not yet been characterized for CPV or FPV, but have been reported for other parvoviruses (84,85). While a number of labs have generated monoclonal and polyclonal anti-CPV antibodies, the most structurally characterized anti-CPV antibodies are monoclonals derived either from mice immunized with CPV (Mabs 1-25) or rats immunized with FPV (Mabs A-F) (86–90).

Antibodies recognizing the "A-site" include Mabs 6, 14, and B, as determined by cryo- electron microscopy (cryo-EM) structures (91) and Mabs 7, 13, and I, which lose binding activity to viruses that have mutations to A-site residues (90,92). The shared footprint for Mabs

6, 14, and B includes VP2 residues 93, 222-224, and 423-428, with residues 93, 222, and 224 predicted to directly interact with the three antibodies (91). Viral escape mutations to antibodies that target the "A-site" include residues 93 (Mabs 7, 13, 14, and 2A9), 222 (Mabs 6,14, B, and I), and 224 (all A-site Mabs tested), indicating that these residues are directly involved in the virus/antibody interaction (92,93).

Antibodies recognizing the "B-site" include Mabs 8, 15, 16, E, and F, as determined by cryo-EM structures (91,94), and antibodies 1, D, and J, which lose capsid binding activity when

B-site residues are mutated (92). The shared footprint for Mabs 8, 15, 16, E, and F includes VP2

Figure 1.5. Antigenic sites on the CPV capsid. A) Antigenic sites shown in the context of an assembled CPV capsid. Shared “A-site” residues are colored in red, with residues predicted to

directly interact with antibodies shaded dark red. Shared “B-site” residues are colored blue, with residues predicted to directly interact with antibodies shaded dark blue. Residues 93 and 300, which are predicted to interact with the TfR, are colored green. B) Magnification of A). C)

Stereographic projection of the capsid surface generated with the program RIVEM (95) and colored as in A).

residues 298, 300, 301, 302, and 387, with residues 300 and 302 predicted to directly interact with all five antibodies (91). Viral escape mutations at the "B-site" include residues 299, 300 (all

B-site antibodies except Mab F) and 302 (Mabs 1, 8, 16, and D) (34,92).

Because of the high sequence similarity between FPV, CPV, and similar parvoviruses including mink enteritis virus (MEV), many of the monoclonal antibodies generated against CPV or FPV will recognize one or more of the other viruses. Antibodies 15, 16, E, and F, in particular, have been shown to recognize all three of these viruses well (90). On the other end of the spectrum, antibodies 7, 11, 13, 14, and 25 have been shown to only bind strongly to CPV (90).

This differential binding ability has been used to distinguish viral serotypes including CPV-2,

CPV-2a, and CPV-2b (7,80,96).

1.6b Mechanisms of antibody neutralization

Antibodies have been shown to neutralize unenveloped viruses in numerous different ways, including by blocking receptor binding sites, crosslinking capsids, either inhibiting or prematurely triggering uncoating of the genome, and targeting the viruses for destruction by immune cells (97). All of the monoclonal antibodies derived from the mouse anti-CPV and rat anti-FPV immunizations (Mabs 1-25 and A-F) that have been tested are capable of neutralizing virus in tissue culture if they are at high concentrations (89,90,93). Some of this neutralization is likely by crosslinking virions in solution, as all of these antibodies can inhibit hemagglutination of red blood cells, which requires either viral crosslinking or near-complete steric blocking of capsid surface structures (89,90).

In some cases, antigen binding fragments (Fabs) derived from these antibodies are also capable of neutralizing infection, even though they are monovalent and incapable of

crosslinking. Most of these Fabs neutralize virus only at very high concentrations (93), indicating that they inhibit infection by binding over all of the available capsid surfaces and obstructing the receptor binding site or preventing uncoating of capsids. Fab E (derived from Mab E), however, has been reported to neutralize capsids at a ratio as low as 6.5 Fabs per capsid, whereas other

Fabs neutralize at close to 30-60 copies per capsid, the maximum number of Fabs that can bind depending on angle of attachment and steric hindrance (93). Cryo-EM reconstruction of Fab

E/CPV complexes has shown that Fab E binds across three different capsid protein subunits at once, crosslinking them, and also induces a conformation change in the capsid structure (94).

This could result in inhibition of capsid uncoating during infection or increase the capsid stability and prevent necessary conformational changes, thus neutralizing infection. Some anti-CPV monoclonal antibodies were reported to have low hemagglutination inhibition activity, but normal viral neutralization in tissue culture (90), suggesting that, like with Mab E, they may also have additional means of neutralizing viral capsids aside from crosslinking capsids.

1.7 Biophysical and biochemical properties of parvovirus capsids

Canine parvovirus is remarkably stable and can withstand temperatures of up to 70°C

(98), urea concentrations to 4 M (20) and changes in pH ranging from 3.1 to 11.4 (69) before the capsid will begin to disassemble. Full capsids (which have packaged genomes) are more resistant to urea than empty capsids, with some intact particles detectible at 7 M concentration, but full and empty capsids are equally susceptible to changes in pH and thermal denaturation (69,98).

Other parvoviruses, including AAVs (99), MVM (100), and parvovirus B19 (101) also have high thermal stability, with temperatures triggering capsid disassociation ranging from 65-90°C depending on the virus.

Assembled CPV capsids are largely resistant to external proteases, with Proteinase K and trypsin incubation resulting in some cleavage products, but leaving most capsid subunits intact

(20,98). In spite of CPV's resistance to external proteases, capsids that are denatured and run on an SDS-PAGE gel show numerous sub-molar proteolytic cleavage fragments (98), suggesting capsids are proteolytically cleaved prior to assembly or undergo auto-proteolytic cleavage after assembly, as has been reported for AAV serotypes 1-9 after exposure to pH 5.5 (102).

Although CPV capsids are intact and stable at endosomal pH, incubation at pH 5.5 has been shown to result in a conformational change to the capsid focused around surface loop 4, which contains VP2 residues 359-397 and is responsible for sialic acid binding and hemagglutination of red blood cells (47,48,103). AAV capsid structure also undergoes a conformational change upon exposure to endosomal pH, with minor shifts to capsid surface loops II, IV, and VI and the capsid/DNA interactions becoming disordered, possibly in preparation for genome uncoating (104). Shifts were also seen in AAV interior surface residues

R392, Y707, E566, and H529, which are associated with the auto-proteolytic cleavage seen in

AAV capsids (102,104).

1.8 Critical structures in parvovirus capsids

Although parvovirus capsids are robust and can survive harsh environmental conditions, single mutations to the major capsid protein can result in loss of infectivity if they occur in critical structures on the capsid. These mutations can either prevent capsid subunits from assembling or disrupt one of the capsid functions during the infectious cycle. Aside from the receptor binding sites, where mutations affect host cell tropism or eliminate infection altogether

(34,37,105,106), other critical structures include the 5-fold pore (69,107), the N-terminus (108),

a VP2 nuclear localization determinant near the C-terminus (109), and, for the adeno-associated viruses, residues that trigger pH-dependent, autocatalytic cleavage of the major capsid protein

(102) and residues that are required for genome transcription (110).

Mutations that affect capsid assembly occur around the 5-fold pore and the VP2 N- and

C-termini. In CPV, point mutations around the 5-fold pore (Val522Gly or

Lys536Thr/Lys538Thr) result in loss of capsid assembly (69), and in MVM, the point mutation

Leu172Trp has been shown to block the 5-fold pore, inhibiting genome packaging and preventing infection even though capsids assemble (107). Parvovirus B19 capsids that have 24 or more amino acids truncated from the VP2 N-terminus are unable to assemble, as are capsids that had a deletion of VP2 residue 25 (108). Furthermore, if B19 VP2 has more than 30 N-terminal amino acids truncated, assembly of co-expressed truncated and wild-type capsids is also inhibited (108). Multiple different point mutations to the MVM C-terminal VP2 protein (residues

528-538) have been shown to prevent capsid assembly by preventing VP2 from trafficking to the nucleus (109).

In the adeno-associated viruses, capsid surface residues termed the ‘pH quartet’ (AAV2 and 8 VP1 residues 389, 526, 563, 704) undergo a conformational change upon exposure to pH

5.5 (104). Mutations to these conserved residues result in up to a 6-log-fold reduction to infectivity, and recent studies have shown that they are required for autocatalytic cleavage of

AAV capsid proteins at acidic pH, but that they are not catalytic residues of a protease (102).

Another mutation to the AAV capsid protein (Y704A) causes loss of infectivity by decreasing viral transcription (110,111), suggesting that the AAV capsid protein is involved in transcription.

No studies on the autonomous parvoviruses have yet reported either pH-dependent viral protease activity or capsid-dependent genome transcription, but it is possible that their capsid proteins

perform similar functions during infection.

1.9 Endogenous viral elements

1.9a Origin of endogenous viral elements

Endogenous viral elements (EVEs) are segments of viral nucleic acid that have been converted into dsDNA, integrated into the genome of a germline cell, and vertically transferred from parent to offspring. Once integrated into the host genome, EVEs acquire mutations at the same rate as the host, rather than of circulating viruses under selection, creating what is effectively a molecular fossil of an ancient virus. These ‘fossils’ can provide information about how old particular genera of viruses are, what range of species they infected, how different viruses evolved over long-term time scales (millions of years instead of decades), and how hosts containing EVE sequences have responded to either disrupt or maintain viral genes. Some characterized EVEs, for example, appear to have undergone selection to maintain open reading frames, resulting in a lower mutation rate than otherwise expected, and RNA-sequencing for multiple species has revealed that genes from some EVEs are transcribed (112).

EVEs derived from retroviruses, which encode their own reverse transcriptase and integrase, are most common, with some studies estimating that they comprise up to 8% of the human genome (113), but EVEs have been reported from many classes of viruses (ssDNA, dsDNA, ssRNA, dsRNA, etc.) (114). Double-stranded DNA viruses and single-stranded DNA viruses, which have a double-stranded DNA intermediate during replication, might become integrated into the host genome during double-stranded break repair. Alternatively, if a cell is co- infected with a retrovirus or expresses reverse transcriptase or integrase from an endogenous retrovirus (retrotransposon), RNA might be converted into dsDNA and integrated into the host

genome. Sometimes, non-retroviral EVEs are flanked by retroviral insertion sequences (long terminal repeats (LTRs)) (115), supporting the hypothesis of retrovirus involvement.

Parvoviruses and circoviruses are disproportionately represented among endogenized

DNA viruses, possibly because they encode an endonuclease that could help facilitate integration into the genome (116). These nucleases (NS1 or Rep for autonomous parvoviruses and the adeno-associated viruses, respectively) recognize specific sequences in viral DNA and cleave it during viral genome replication, but host cell DNA sometimes has matching sequences that the nuclease can target. AAV2, for example, is predisposed to integrate into human chromosome 19 during infection and targets a site that closely resembles one of the terminal repeats at the end of its genome (117,118). The parvoviral EVE sequences that have been examined so far, however, frequently lacked their 5’ and 3’ terminal repeats and were flanked by LTRs or other TE sequences, suggesting an alternative method of integration (115). Regions containing TEs are unstable and are prone to dsDNA breaks as TEs move/replicate (115,119–121), and host machinery could potentially use dsDNA parvoviral replication intermediates to repair those breaks, resulting in parvoviral EVEs flanked by TEs.

1.9b Identification and dating of endogenous viral elements

Retroviral endogenous viral elements, including LINEs, SINEs, and retrotransposons are common in mammalian genomes and are readily identified by their multiple integrations in the host genome (122–124). Non-retroviral EVEs, however, frequently do not contain open reading frames or LTRs and have few or no duplicates within the genome, making this method of identification difficult. Instead, these EVEs are discovered by compiling a library of all the existing genomes for a particular class of virus, identifying conserved motifs, and searching

through deep sequencing databases for sequences that contain those conserved motifs (114).

Because many species are infected by exogenous viruses at the time of sequencing, the results are then checked for sequence segments that overlap the end of the putative EVE and the host genome, to make sure that the EVE was integrated (115). Frameshift mutations or stop codons located in the EVE sequence provide further support of integration, as exogenous viruses containing these mutations would not be viable (115). Expressed sequence tag (EST) deep sequencing databases can be searched similarly, looking for EVE gene expression, but must be matched by sequence to an EVE in the host genome in order to differentiate EVE transcripts from exogenous virus transcripts.

The age of endogenous viral elements can be approximated in several ways. If the EVE insertion is in the same location in multiple host genomes, it can be inferred that it is at least as old as the most recent common ancestor that each of those hosts shared, but no older than the most recent common ancestor shared between those species and species that do not contain the

EVE (114). For EVEs that contain LTRs, including retroviruses and retrotransposons, age of insertion can be approximated by comparing the 5’ and 3’ LTRs, which are identical to each other in exogenous viruses, to one another to determine the number of mutations per base pair that have accumulated in the EVE. If the host DNA mutation rate is known, the ratio of mutations per base pair can be converted into estimated age (125). For EVEs that do not contain

LTRs, a similar technique can be used if stop codons or other identifiable mutations (insertions, deletions, etc.) are present in the genome, with an estimated maximum sensitivity of 90 million years if the EVE is under neutral selection (117).

1.9c Reconstruction of endogenous viral elements

In the case of recent EVE insertions, where relatively few mutations have occurred, it is possible to repair the EVE sequence and express proteins in order to study how ancient viruses functioned. Human endogenous retrovirus K (HERV-K), an extinct retrovirus that integrated into the human genome less than five million years ago (126), was repaired by deriving a consensus sequence from multiple human genome sequences and synthesizing it (127). Although the complete genome was never tested because of concerns about viral release, experiments with individual proteins showed that they were functional. Expression of the gag and protease proteins resulted in the assembly of virus-like particles, pseudotyping of the envelope protein onto HIV-1 particles allowed entry into cells, and expression of reverse transcriptase and integrase showed that the viral replication machinery was still functional (126,127).

For the parvoviruses, one study describes an in silico reconstruction of a capsid derived from an adeno-associated virus EVE that was integrated into marsupial (macropodid) genomes approximately 30 million years ago (116). While the insertions in some macropodid species had large deletions of 1.2-1.6kbp of the genome, others contained the complete genome, allowing for comparison to extant AAVs, with which the EVE shared around 60% amino acid sequence identity of its capsid protein. Most of the residues that were conserved between the marsupial

EVE and modern AAVs were located within the β-barrel on the interior surface of the capsid, while variable residues were located on exposed surface loops (116), a trend that also emerges when comparing extant AAVs to one another. The Rep protein was reported to be most similar to that of AAV5, with the phospholipase domain, rolling circle replication motifs, and nuclease domain being readily apparent (116).

1.9d Host suppression of endogenous viral elements

Some endogenous viral elements, transposable elements in particular, are detrimental to the host if expressed at the wrong time or inserted into the wrong place in the genome, disrupting critical gene function and causing diseases, including cancer. To avoid this, host cells suppress transposable elements using several mechanisms, including DNA methylation, cytidine deamidation, and small RNAs (128). Selective methylation of transposons in host DNA occurs during gametogenesis and prevents transposons from being transcribed. In mice, knocking out the cytosine methyltransferase has been shown to result in the reactivation of transcriptional elements and the death of male germ cells (129). Cytidine deaminases, including the APOBEC3 family of deaminases, can deamidate transposable element cDNA, resulting in its degradation or in the mutation of the transposon (128). Small RNAs, though primarily described in plants and invertebrates, have also been reported to inhibit expression of L1 retrotransposons in mammals

(130). L1 retrotransposons contain sense and antisense promoters in their 5’ untranslated regions

(UTR), giving rise to L1 mRNA transcripts and chimeric transcripts of adjacent genes, which also contain portions of the L1 5’ UTR (130). These mRNA transcripts can base pair to produce dsRNA, which can then be recognized and processed by Dicer and used to degrade L1 mRNA transcripts (130).

1.9e Host benefits of endogenous viral elements

In some cases, the host has been reported to co-opt endogenized viral sequences for its own use, particularly for regulation of gene expression (131), showing that some EVEs are beneficial to the host and may be under selection to be maintained. Transmissible elements derived from retroviral DNA including long interspersed nuclear elements (LINEs), short interspersed nuclear elements (SINEs), and endogenous retroviruses with LTRs have been shown

to contain transcription factors as wells as promoters and enhancers, which can directly affect gene expression (131). Depending on how these transmissible elements are inserted into the genome, they can also isolate a gene from its regulatory elements, introduce or remove splicing sites, add nonsense mutations, or convert DNA to heterochromatin (131).

Some species have been reported to use whole genes or series of genes of viral origin.

Female parasitic wasps have specialized ovary cells that contain endogenized viral DNA and produce polydnavirus particles. These viral particles are injected into lepidopteran larvae along with wasp eggs, and package genes that manipulate the larvae’s immune system so that it tolerates the wasp eggs. The polydnavirus particles, believed to originate from nudiviruses that integrated into the wasp genome, do not contain encapsulated structural genes, nor are they capable of replicating in caterpillars. Instead, the genes encoding polydnavirus structural genes are found in the wasp genome, in a separate location from the genes which are incorporated into viral particles (132). This exaptation of the nudivirus genome has occurred at least three separate times in different species of parasitic wasps, as evidenced by the unique position of nudivirus

DNA in the genomes of three distinct species, and supports a hypothesis of convergent evolution in the acquisition of polydnavirus EVEs (132).

In humans, the placentally expressed genes Synsytin-1 and Synsytin-2 are responsible for cell-to-cell fusion in the trophoblast and syncytiotrophoblast, respectively (133). Syncytin genes, which are derived from retroviral envelope/fusion proteins, have been exapted at least 5 separate times for use by different mammalian species including humans, mice, ruminants, carnivores, and rabbits/hares, with insertions into the host genome ranging from 12 to 80 million years ago

(133). Knockout of syncytin-A in mice, which corresponds to Synsytin-1 in humans, prevents proper placental development and results in embryonic death (133). In both mice and humans,

syncytin expression is regulated by the placental transcription factor GCM1, showing that these hosts have extensively repurposed the retroviral gene.

Recent studies of endogenous retroviruses have suggested that EVEs can also function as restriction factors targeting modern-day viruses in a process known as ‘receptor interference’

(134). Some species, including cats, mice, mink, sheep, and chickens (134), have acquired integrated copies of retrovirus envelope genes which are similar to those of circulating retroviruses but lack the transmembrane domain that anchors the envelope protein to a membrane. When these genes are expressed, cells secrete a soluble form of the envelope protein

(and the signal peptide domain) that can bind to the host receptors on adjacent cells. When the secreted envelope proteins have attached to host receptors, it blocks circulating viruses from binding to them for uptake or infection (135). This mechanism has also been used for regulation of syncytins in mammalian placental tissue, where the protein Suppressyn competes with syncytin for binding to the ASCT2 receptor (134).

Other studies have shown that endogenous viral element genes from bornaviruses and filoviruses are expressed and are under selective pressure to maintain open reading frames (112).

Although the mechanism that mediates it is unknown, it appears that expression of these ORFs protects the host from bornavirus or filovirus infection, as there is a positive correlation between species that have bornavirus or filovirus EVE insertions and resistance to infection by modern- day strains of these viruses (112). Mice that are genetically modified to express the bornavirus nucleocapsid protein gene in neurons, furthermore, become resistant to bornavirus infection

(136), indicating that these expressed proteins may actively inhibit viral replication.

For the parvoviruses, almost all EVEs reported thus far contain stop codons and frameshift mutations that disrupt open reading frames, and while genes from some parvovirus

EVEs have been transcribed into mRNA, the presence of these mutations makes it unlikely that the translated proteins maintain their functionality (114–116,137). Some of these truncated proteins might interfere with infection of related, extant parvoviruses, providing a form of immunity to the host, but no reports confirming or disproving this hypothesis have been published yet.

1.10 Dissertation overview

This dissertation examines the structure and function of canine parvovirus capsids and how the interactions of those capsids with monoclonal antibodies or TfRs from different species either facilitate or inhibit infection of cells. It also investigates parvovirus endogenous viral elements from three different species of rodents and explores the structures and functions of capsids derived from those endogenous sequences.

Chapter 2 discusses mutations to the CPV major capsid protein that result in loss of viral infectivity and examines how the mutations cause that loss, with impacted functions and properties including receptor binding, capsid stability, proteolytic cleavage, and nuclear trafficking.

Chapter 3 discusses the interactions between CPV, the TfR, and different monoclonal antibodies, and how those interactions impact infection.

Chapter 4 discusses the repair and expression of the major capsid protein of parvovirus endogenous viral elements from three different rodent species, as well as the properties of assembled EVE-derived parvovirus capsids.

Chapter 5 summarizes the results of chapters 2-4 and discusses the implications to the field of virology.

REFERENCES

1. Pollock RVH, Parrish CR. Canine Parvovirus. In: Comparative Pathobiology of Viral

Diseases. Boca Raton, Florida: CRC Press; 1985. p. 145–77.

2. Csiza CK, Scott FW, De Lahunta A, Gillespie JH. Pathogenesis of feline panleukopenia

virus in susceptible newborn kittens I. Clinical signs, hematology, serology, and virology.

Infect Immun. 1971 Jun;3(6):833–7.

3. Csiza CK, De Lahunta A, Scott FW, Gillespie JH. Pathogenesis of Feline Panleukopenia

Virus in Susceptible Newborn Kittens II. Pathology and Immunofluorescence. Infect

Immun. 1971 Jun;3(6):838–46.

4. Stuetzer B, Hartmann K. Feline parvovirus infection and associated diseases. Vet J Lond

Engl 1997. 2014 Aug;201(2):150–5.

5. Shackelton LA, Parrish CR, Truyen U, Holmes EC. High rate of viral evolution associated

with the emergence of carnivore parvovirus. Proc Natl Acad Sci U S A. 2005 Jan

11;102(2):379–84.

6. Carmichael LE. An annotated historical account of canine parvovirus. J Vet Med B Infect

Dis Vet Public Health. 2005 Oct;52(7–8):303–11.

7. Parrish CR, O’Connell PH, Evermann JF, Carmichael LE. Natural variation of canine

parvovirus. Science. 1985 Nov 29;230(4729):1046–8.

8. Canuti M, Rodrigues B, Whitney HG, Lang AS. Introduction of canine parvovirus 2 into

wildlife on the Island of Newfoundland, Canada. Infect Genet Evol J Mol Epidemiol Evol

Genet Infect Dis. 2017 Nov;55:205–8.

9. Folitse RD, Kodie DO, Amemor E, Dei D, Tasiame W, Burimuah V, et al. DETECTION

OF CANINE PARVOVIRUS ANTIGEN IN DOGS IN KUMASI, GHANA. Afr J Infect

Dis. 2018;12(1):28–32.

10. Quino Quispe R, Luna Espinoza L, Rímac Beltrán R, Rosadio Alcántara R, Maturrano

Hernández L. Canine parvovirus types 2a and 2c detection from dogs with suspected

parvoviral enteritis in Peru. Virusdisease. 2018 Mar;29(1):109–12.

11. Wang D, Yuan W, Davis I, Parrish CR. Nonstructural protein-2 and the replication of

canine parvovirus. Virology. 1998 Jan 20;240(2):273–81.

12. Zádori Z, Szelei J, Tijssen P. SAT: a late NS protein of porcine parvovirus. J Virol. 2005

Oct;79(20):13129–38.

13. Parrish CR. Mapping specific functions in the capsid structure of canine parvovirus and

feline panleukopenia virus using infectious plasmid clones. Virology. 1991

Jul;183(1):195–205.

14. Tsao J, Chapman MS, Agbandje M, Keller W, Smith K, Wu H, et al. The three-

dimensional structure of canine parvovirus and its functional implications. Science. 1991

Mar 22;251(5000):1456–64.

15. Xie Q, Chapman MS. Canine parvovirus capsid structure, analyzed at 2.9 A resolution. J

Mol Biol. 1996 Dec 6;264(3):497–520.

16. James JA, Escalante CR, Yoon-Robarts M, Edwards TA, Linden RM, Aggarwal AK.

Crystal structure of the SF3 helicase from adeno-associated virus type 2. Struct Lond Engl

1993. 2003 Aug;11(8):1025–35.

17. Smith RH, Spano AJ, Kotin RM. The Rep78 gene product of adeno-associated virus

(AAV) self-associates to form a hexameric complex in the presence of AAV ori

sequences. J Virol. 1997 Jun;71(6):4461–71.

18. Cotmore SF, Tattersall P. A genome-linked copy of the NS-1 polypeptide is located on the

outside of infectious parvovirus particles. J Virol. 1989 Sep;63(9):3902–11.

19. Prasad KM, Trempe JP. The adeno-associated virus Rep78 protein is covalently linked to

viral DNA in a preformed virion. Virology. 1995 Dec 20;214(2):360–70.

20. Weichert WS, Parker JS, Wahid AT, Chang SF, Meier E, Parrish CR. Assaying for

structural variation in the parvovirus capsid and its role in infection. Virology. 1998 Oct

10;250(1):106–17.

21. Sonntag F, Bleker S, Leuchs B, Fischer R, Kleinschmidt JA. Adeno-associated virus type

2 capsids with externalized VP1/VP2 trafficking domains are generated prior to passage

through the cytoplasm and are maintained until uncoating occurs in the nucleus. J Virol.

2006 Nov;80(22):11040–54.

22. Parker JS, Murphy WJ, Wang D, O’Brien SJ, Parrish CR. Canine and feline parvoviruses

can use human or feline transferrin receptors to bind, enter, and infect cells. J Virol. 2001

Apr;75(8):3896–902.

23. Jing SQ, Trowbridge IS. Identification of the intermolecular disulfide bonds of the human

transferrin receptor and its lipid-attachment site. EMBO J. 1987 Feb;6(2):327–31.

24. Lawrence CM, Ray S, Babyonyshev M, Galluser R, Borhani DW, Harrison SC. Crystal

structure of the ectodomain of human transferrin receptor. Science. 1999 Oct

22;286(5440):779–82.

25. Turkewitz AP, Schwartz AL, Harrison SC. A pH-dependent reversible conformational

transition of the human transferrin receptor leads to self-association. J Biol Chem. 1988

Nov 5;263(31):16309–15.

26. Dautry-Varsat A, Ciechanover A, Lodish HF. pH and the recycling of transferrin during

receptor-mediated endocytosis. Proc Natl Acad Sci U S A. 1983 Apr;80(8):2258–62.

27. Eckenroth BE, Steere AN, Chasteen ND, Everse SJ, Mason AB. How the binding of

human transferrin primes the transferrin receptor potentiating iron release at endosomal

pH. Proc Natl Acad Sci U S A. 2011 Aug 9;108(32):13089–94.

28. Dautry-Varsat A. Receptor-mediated endocytosis: the intracellular journey of transferrin

and its receptor. Biochimie. 1986 Mar;68(3):375–81.

29. Klausner RD, Ashwell G, van Renswoude J, Harford JB, Bridges KR. Binding of

apotransferrin to K562 cells: explanation of the transferrin cycle. Proc Natl Acad Sci U S

A. 1983 Apr;80(8):2263–6.

30. Bennett MJ, Lebrón JA, Bjorkman PJ. Crystal structure of the hereditary

haemochromatosis protein HFE complexed with transferrin receptor. Nature. 2000 Jan

6;403(6765):46–53.

31. Parkkila S, Waheed A, Britton RS, Feder JN, Tsuchihashi Z, Schatzman RC, et al.

Immunohistochemistry of HLA-H, the protein defective in patients with hereditary

hemochromatosis, reveals unique pattern of expression in gastrointestinal tract. Proc Natl

Acad Sci U S A. 1997 Mar 18;94(6):2534–9.

32. Palermo LM, Hueffer K, Parrish CR. Residues in the apical domain of the feline and

canine transferrin receptors control host-specific binding and cell infection of canine and

feline parvoviruses. J Virol. 2003 Aug;77(16):8915–23.

33. Hafenstein S, Palermo LM, Kostyuchenko VA, Xiao C, Morais MC, Nelson CDS, et al.

Asymmetric binding of transferrin receptor to parvovirus capsids. Proc Natl Acad Sci U S

A. 2007 Apr 17;104(16):6585–9.

34. Parker JS, Parrish CR. Canine parvovirus host range is determined by the specific

conformation of an additional region of the capsid. J Virol. 1997 Dec;71(12):9214–22.

35. Allison AB, Harbison CE, Pagan I, Stucker KM, Kaelber JT, Brown JD, et al. Role of

multiple hosts in the cross-species transmission and emergence of a pandemic parvovirus.

J Virol. 2012 Jan;86(2):865–72.

36. Llamas-Saiz AL, Agbandje-McKenna M, Parker JS, Wahid AT, Parrish CR, Rossmann

MG. Structural analysis of a mutation in canine parvovirus which controls antigenicity and

host range. Virology. 1996 Nov 1;225(1):65–71.

37. Allison AB, Organtini LJ, Zhang S, Hafenstein SL, Holmes EC, Parrish CR. Single

Mutations in the VP2 300 Loop Region of the Three-Fold Spike of the Carnivore

Parvovirus Capsid Can Determine Host Range. J Virol. 2016;90(2):753–67.

38. Allison AB, Kohler DJ, Ortega A, Hoover EA, Grove DM, Holmes EC, et al. Host-

specific parvovirus evolution in nature is recapitulated by in vitro adaptation to different

carnivore species. PLoS Pathog. 2014 Nov;10(11):e1004475.

39. Chang SF, Sgro JY, Parrish CR. Multiple amino acids in the capsid structure of canine

parvovirus coordinately determine the canine host range and specific antigenic and

hemagglutination properties. J Virol. 1992 Dec;66(12):6858–67.

40. Hueffer K, Palermo LM, Parrish CR. Parvovirus infection of cells by using variants of the

feline transferrin receptor altering clathrin-mediated endocytosis, membrane domain

localization, and capsid-binding domains. J Virol. 2004 Jun;78(11):5601–11.

41. Cureton DK, Harbison CE, Cocucci E, Parrish CR, Kirchhausen T. Limited transferrin

receptor clustering allows rapid diffusion of canine parvovirus into clathrin endocytic

structures. J Virol. 2012 May;86(9):5330–40.

42. Halder S, Cotmore S, Heimburg-Molinaro J, Smith DF, Cummings RD, Chen X, et al.

Profiling of glycan receptors for minute virus of mice in permissive cell lines towards

understanding the mechanism of cell recognition. PloS One. 2014;9(1):e86909.

43. Nam H-J, Gurda-Whitaker B, Gan WY, Ilaria S, McKenna R, Mehta P, et al.

Identification of the sialic acid structures recognized by minute virus of mice and the role

of binding affinity in virulence adaptation. J Biol Chem. 2006 Sep 1;281(35):25670–7.

44. Boisvert M, Fernandes S, Tijssen P. Multiple pathways involved in porcine parvovirus

cellular entry and trafficking toward the nucleus. J Virol. 2010 Aug;84(15):7782–92.

45. López-Bueno A, Rubio M-P, Bryant N, McKenna R, Agbandje-McKenna M, Almendral

JM. Host-selected amino acid changes at the sialic acid binding pocket of the parvovirus

capsid modulate cell binding affinity and determine virulence. J Virol. 2006

Feb;80(3):1563–73.

46. Löfling J, Lyi SM, Parrish CR, Varki A. Canine and feline parvoviruses preferentially

recognize the non-human cell surface sialic acid N-glycolylneuraminic acid. Virology.

2013 May 25;440(1):89–96.

47. Barbis DP, Chang SF, Parrish CR. Mutations adjacent to the dimple of the canine

parvovirus capsid structure affect sialic acid binding. Virology. 1992 Nov;191(1):301–8.

48. Tresnan DB, Southard L, Weichert W, Sgro JY, Parrish CR. Analysis of the cell and

erythrocyte binding activities of the dimple and canyon regions of the canine parvovirus

capsid. Virology. 1995 Aug 1;211(1):123–32.

49. Harbison CE, Lyi SM, Weichert WS, Parrish CR. Early steps in cell infection by

parvoviruses: host-specific differences in cell receptor binding but similar endosomal

trafficking. J Virol. 2009 Oct;83(20):10504–14.

50. Parker JS, Parrish CR. Cellular uptake and infection by canine parvovirus involves rapid

dynamin-regulated clathrin-mediated endocytosis, followed by slower intracellular

trafficking. J Virol. 2000 Feb;74(4):1919–30.

51. Vihinen-Ranta M, Yuan W, Parrish CR. Cytoplasmic trafficking of the canine parvovirus

capsid and its role in infection and nuclear transport. J Virol. 2000 May;74(10):4853–9.

52. Seisenberger G, Ried MU, Endress T, Büning H, Hallek M, Bräuchle C. Real-time single-

molecule imaging of the infection pathway of an adeno-associated virus. Science. 2001

Nov 30;294(5548):1929–32.

53. Farr GA, Cotmore SF, Tattersall P. 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. 2006 Jan;80(1):161–71.

54. Mani B, Baltzer C, Valle N, Almendral JM, Kempf C, Ros C. 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. 2006 Jan;80(2):1015–24.

55. Basak S, Turner H. Infectious entry pathway for canine parvovirus. Virology. 1992

Feb;186(2):368–76.

56. Cotmore SF, Hafenstein S, Tattersall P. Depletion of virion-associated divalent cations

induces parvovirus minute virus of mice to eject its genome in a 3’-to-5’ direction from an

otherwise intact viral particle. J Virol. 2010 Feb;84(4):1945–56.

57. Ros C, Baltzer C, Mani B, Kempf C. Parvovirus uncoating in vitro reveals a mechanism of

DNA release without capsid disassembly and striking differences in encapsidated DNA

stability. Virology. 2006 Feb 5;345(1):137–47.

58. Cotmore SF, Tattersall P. The autonomously replicating parvoviruses of vertebrates. Adv

Virus Res. 1987;33:91–174.

59. Jongeneel CV, Sahli R, McMaster GK, Hirt B. A precise map of splice junctions in the

mRNAs of minute virus of mice, an autonomous parvovirus. J Virol. 1986 Sep;59(3):564–

73.

60. Cotmore SF, Sturzenbecker LJ, Tattersall P. The autonomous parvovirus MVM encodes

two nonstructural proteins in addition to its capsid polypeptides. Virology. 1983

Sep;129(2):333–43.

61. Christensen J, Cotmore SF, Tattersall P. Minute virus of mice transcriptional activator

protein NS1 binds directly to the transactivation region of the viral P38 promoter in a

strictly ATP-dependent manner. J Virol. 1995 Sep;69(9):5422–30.

62. Niskanen EA, Kalliolinna O, Ihalainen TO, Häkkinen M, Vihinen-Ranta M. Mutations in

DNA binding and transactivation domains affect the dynamics of parvovirus NS1 protein.

J Virol. 2013 Nov;87(21):11762–74.

63. Mészáros I, Tóth R, Olasz F, Tijssen P, Zádori Z. The SAT Protein of Porcine Parvovirus

Accelerates Viral Spreading through Induction of Irreversible Endoplasmic Reticulum

Stress. J Virol. 2017 15;91(16).

64. Sonntag F, Schmidt K, Kleinschmidt JA. A viral assembly factor promotes AAV2 capsid

formation in the nucleolus. Proc Natl Acad Sci U S A. 2010 Jun 1;107(22):10220–5.

65. Gil-Ranedo J, Hernando E, Valle N, Riolobos L, Maroto B, Almendral JM. Differential

phosphorylation and n-terminal configuration of capsid subunits in parvovirus assembly

and viral trafficking. Virology. 2018 Mar 7;518:184–94.

66. Gil-Ranedo J, Hernando E, Riolobos L, Domínguez C, Kann M, Almendral JM. The

Mammalian Cell Cycle Regulates Parvovirus Nuclear Capsid Assembly. PLoS Pathog.

2015 Jun;11(6):e1004920.

67. Riolobos L, Reguera J, Mateu MG, Almendral JM. Nuclear transport of trimeric assembly

intermediates exerts a morphogenetic control on the icosahedral parvovirus capsid. J Mol

Biol. 2006 Mar 31;357(3):1026–38.

68. Riolobos L, Valle N, Hernando E, Maroto B, Kann M, Almendral JM. Viral oncolysis that

targets Raf-1 signaling control of nuclear transport. J Virol. 2010 Feb;84(4):2090–9.

69. Yuan W, Parrish CR. Canine parvovirus capsid assembly and differences in mammalian

and insect cells. Virology. 2001 Jan 20;279(2):546–57.

70. Cotmore SF, Tattersall P. Parvovirus DNA Replication. In: DNA Replication in

Eukaryotic Cells. Cold Spring Harbor Laboratory Press; 1996. p. 799–813.

71. Bashir T, Horlein R, Rommelaere J, Willwand K. Cyclin A activates the DNA polymerase

delta -dependent elongation machinery in vitro: A parvovirus DNA replication model.

Proc Natl Acad Sci U S A. 2000 May 9;97(10):5522–7.

72. Cotmore SF, Tattersall P. Parvovirus. In: DNA Replication and Human Disease. New

York: Cold Spring Harbor Laboratory Press; 2006. p. 593–608.

73. Nüesch JP, Cotmore SF, Tattersall P. Sequence motifs in the replicator protein of

parvovirus MVM essential for nicking and covalent attachment to the viral origin:

identification of the linking tyrosine. Virology. 1995 May 10;209(1):122–35.

74. Maroto B, Valle N, Saffrich R, Almendral JM. Nuclear export of the nonenveloped

parvovirus virion is directed by an unordered protein signal exposed on the capsid surface.

J Virol. 2004 Oct;78(19):10685–94.

75. Miller CL, Pintel DJ. Interaction between parvovirus NS2 protein and nuclear export

factor Crm1 is important for viral egress from the nucleus of murine cells. J Virol. 2002

Apr;76(7):3257–66.

76. Wolfisberg R, Kempf C, Ros C. Late Maturation Steps Preceding Selective Nuclear

Export and Egress of Progeny Parvovirus. J Virol. 2016 01;90(11):5462–74.

77. Bär S, Rommelaere J, Nüesch JPF. Vesicular transport of progeny parvovirus particles

through ER and Golgi regulates maturation and cytolysis. PLoS Pathog. 2013

Sep;9(9):e1003605.

78. Bär S, Daeffler L, Rommelaere J, Nüesch JPF. Vesicular egress of non-enveloped lytic

parvoviruses depends on gelsolin functioning. PLoS Pathog. 2008 Aug 15;4(8):e1000126.

79. Kaelber JT, Demogines A, Harbison CE, Allison AB, Goodman LB, Ortega AN, et al.

Evolutionary reconstructions of the transferrin receptor of Caniforms supports canine

parvovirus being a re-emerged and not a novel pathogen in dogs. PLoS Pathog.

2012;8(5):e1002666.

80. Parrish CR, Aquadro CF, Strassheim ML, Evermann JF, Sgro JY, Mohammed HO. Rapid

antigenic-type replacement and DNA sequence evolution of canine parvovirus. J Virol.

1991 Dec;65(12):6544–52.

81. Truyen U, Evermann JF, Vieler E, Parrish CR. Evolution of canine parvovirus involved

loss and gain of feline host range. Virology. 1996 Jan 15;215(2):186–9.

82. Stucker KM, Pagan I, Cifuente JO, Kaelber JT, Lillie TD, Hafenstein S, et al. The role of

evolutionary intermediates in the host adaptation of canine parvovirus. J Virol. 2012

Feb;86(3):1514–21.

83. Nakamura K, Sakamoto M, Ikeda Y, Sato E, Kawakami K, Miyazawa T, et al. Pathogenic

potential of canine parvovirus types 2a and 2c in domestic cats. Clin Diagn Lab Immunol.

2001 May;8(3):663–8.

84. Kailasan S, Garrison J, Ilyas M, Chipman P, McKenna R, Kantola K, et al. Mapping

Antigenic Epitopes on the Human Bocavirus Capsid. J Virol. 2016 May;90(9):4670–80.

85. McCraw DM, O’Donnell JK, Taylor KA, Stagg SM, Chapman MS. Structure of adeno-

associated virus-2 in complex with neutralizing monoclonal antibody A20. Virology. 2012

Sep 15;431(1–2):40–9.

86. Burtonboy G, Bazin H, Delferriere N. Rat hybridoma antibodies against canine

parvovirus. Arch Virol. 1982;71(4):291–302.

87. Horiuchi M, Mochizuki M, Ishiguro N, Nagasawa H, Shinagawa M. Epitope mapping of a

monoclonal antibody specific to feline panleukopenia virus and mink enteritis virus. J Vet

Med Sci. 1997 Feb;59(2):133–6.

88. Nakamura M, Nakamura K, Miyazawa T, Tohya Y, Mochizuki M, Akashi H. Monoclonal

antibodies that distinguish antigenic variants of canine parvovirus. Clin Diagn Lab

Immunol. 2003 Nov;10(6):1085–9.

89. Parrish CR, Carmichael LE, Antczak DF. Antigenic relationships between canine

parvovirus type 2, feline panleukopenia virus and mink enteritis virus using conventional

antisera and monoclonal antibodies. Arch Virol. 1982;72(4):267–78.

90. Parrish CR, Carmichael LE. Antigenic structure and variation of canine parvovirus type-2,

feline panleukopenia virus, and mink enteritis virus. Virology. 1983 Sep;129(2):401–14.

91. Hafenstein S, Bowman VD, Sun T, Nelson CDS, Palermo LM, Chipman PR, et al.

Structural comparison of different antibodies interacting with parvovirus capsids. J Virol.

2009 Jun;83(11):5556–66.

92. Strassheim ML, Gruenberg A, Veijalainen P, Sgro JY, Parrish CR. Two dominant

neutralizing antigenic determinants of canine parvovirus are found on the threefold spike

of the virus capsid. Virology. 1994 Jan;198(1):175–84.

93. Nelson CDS, Palermo LM, Hafenstein SL, Parrish CR. Different mechanisms of antibody-

mediated neutralization of parvoviruses revealed using the Fab fragments of monoclonal

antibodies. Virology. 2007 May 10;361(2):283–93.

94. Organtini LJ, Lee H, Iketani S, Huang K, Ashley RE, Makhov AM, et al. Near-Atomic

Resolution Structure of a Highly Neutralizing Fab Bound to Canine Parvovirus. J Virol.

2016 Nov 1;90(21):9733–42.

95. Xiao C, Rossmann MG. Interpretation of electron density with stereographic roadmap

projections. J Struct Biol. 2007 May;158(2):182–7.

96. Mochizuki M, Ohshima T, Une Y, Yachi A. Recombination between vaccine and field

strains of canine parvovirus is revealed by isolation of virus in canine and feline cell

cultures. J Vet Med Sci. 2008 Dec;70(12):1305–14.

97. Reading SA, Dimmock NJ. Neutralization of animal virus infectivity by antibody. Arch

Virol. 2007;152(6):1047–59.

98. Nelson CDS, Minkkinen E, Bergkvist M, Hoelzer K, Fisher M, Bothner B, et al. Detecting

small changes and additional peptides in the canine parvovirus capsid structure. J Virol.

2008 Nov;82(21):10397–407.

99. Bennett A, Patel S, Mietzsch M, Jose A, Lins-Austin B, Yu JC, et al. Thermal Stability as

a Determinant of AAV Serotype Identity. Mol Ther Methods Clin Dev. 2017 Sep

15;6:171–82.

100. Boschetti N, Wyss K, Mischler A, Hostettler T, Kempf C. Stability of minute virus of

mice against temperature and sodium hydroxide. Biol J Int Assoc Biol Stand. 2003

Sep;31(3):181–5.

101. Sánchez-Rodríguez SP, Münch-Anguiano L, Echeverría O, Vázquez-Nin G, Mora-Pale M,

Dordick JS, et al. Human parvovirus B19 virus-like particles: In vitro assembly and

stability. Biochimie. 2012 Mar;94(3):870–8.

102. Salganik M, Venkatakrishnan B, Bennett A, Lins B, Yarbrough J, Muzyczka N, et al.

Evidence for pH-dependent protease activity in the adeno-associated virus capsid. J Virol.

2012 Nov;86(21):11877–85.

103. Simpson AA, Chandrasekar V, Hébert B, Sullivan GM, Rossmann MG, Parrish CR. Host

range and variability of calcium binding by surface loops in the capsids of canine and

feline parvoviruses. J Mol Biol. 2000 Jul 14;300(3):597–610.

104. Nam H-J, Gurda BL, McKenna R, Potter M, Byrne B, Salganik M, et al. Structural studies

of adeno-associated virus serotype 8 capsid transitions associated with endosomal

trafficking. J Virol. 2011 Nov;85(22):11791–9.

105. López-Bueno A, Segovia JC, Bueren JA, O’Sullivan MG, Wang F, Tattersall P, et al.

Evolution to pathogenicity of the parvovirus minute virus of mice in immunodeficient

mice involves genetic heterogeneity at the capsid domain that determines tropism. J Virol.

2008 Feb;82(3):1195–203.

106. Parrish CR, Carmichael LE. Characterization and recombination mapping of an antigenic

and host range mutation of canine parvovirus. Virology. 1986 Jan 15;148(1):121–32.

107. Plevka P, Hafenstein S, Li L, D’Abrgamo A, Cotmore SF, Rossmann MG, et al. Structure

of a packaging-defective mutant of minute virus of mice indicates that the genome is

packaged via a pore at a 5-fold axis. J Virol. 2011 May;85(10):4822–7.

108. Kawase M, Momoeda M, Young NS, Kajigaya S. Modest truncation of the major capsid

protein abrogates B19 parvovirus capsid formation. J Virol. 1995 Oct;69(10):6567–71.

109. Lombardo E, Ramírez JC, Agbandje-McKenna M, Almendral JM. A beta-stranded motif

drives capsid protein oligomers of the parvovirus minute virus of mice into the nucleus for

viral assembly. J Virol. 2000 Apr;74(8):3804–14.

110. Aydemir F, Salganik M, Resztak J, Singh J, Bennett A, Agbandje-McKenna M, et al.

Mutants at the 2-Fold Interface of Adeno-associated Virus Type 2 (AAV2) Structural

Proteins Suggest a Role in Viral Transcription for AAV Capsids. J Virol. 2016

15;90(16):7196–204.

111. Salganik M, Aydemir F, Nam H-J, McKenna R, Agbandje-McKenna M, Muzyczka N.

Adeno-associated virus capsid proteins may play a role in transcription and second-strand

synthesis of recombinant genomes. J Virol. 2014 Jan;88(2):1071–9.

112. Belyi VA, Levine AJ, Skalka AM. Unexpected inheritance: multiple integrations of

ancient bornavirus and ebolavirus/marburgvirus sequences in vertebrate genomes. PLoS

Pathog. 2010 Jul 29;6(7):e1001030.

113. Kapoor A, Simmonds P, Lipkin WI. Discovery and characterization of mammalian

endogenous parvoviruses. J Virol. 2010 Dec;84(24):12628–35.

114. Katzourakis A, Gifford RJ. Endogenous viral elements in animal genomes. PLoS Genet.

2010 Nov 18;6(11):e1001191.

115. Liu H, Fu Y, Xie J, Cheng J, Ghabrial SA, Li G, et al. Widespread endogenization of

densoviruses and parvoviruses in animal and human genomes. J Virol. 2011

Oct;85(19):9863–76.

116. Smith RH, Hallwirth CV, Westerman M, Hetherington NA, Tseng Y-S, Cecchini S, et al.

Germline viral “fossils” guide in silico reconstruction of a mid-Cenozoic era marsupial

adeno-associated virus. Sci Rep. 2016 Jul 5;6:28965.

117. Belyi VA, Levine AJ, Skalka AM. Sequences from ancestral single-stranded DNA viruses

in vertebrate genomes: the parvoviridae and circoviridae are more than 40 to 50 million

years old. J Virol. 2010 Dec;84(23):12458–62.

118. Urcelay E, Ward P, Wiener SM, Safer B, Kotin RM. Asymmetric replication in vitro from

a human sequence element is dependent on adeno-associated virus Rep protein. J Virol.

1995 Apr;69(4):2038–46.

119. Bill CA, Summers J. Genomic DNA double-strand breaks are targets for hepadnaviral

DNA integration. Proc Natl Acad Sci U S A. 2004 Jul 27;101(30):11135–40.

120. Miller DG, Trobridge GD, Petek LM, Jacobs MA, Kaul R, Russell DW. Large-scale

analysis of adeno-associated virus vector integration sites in normal human cells. J Virol.

2005 Sep;79(17):11434–42.

121. Miller DG, Petek LM, Russell DW. Adeno-associated virus vectors integrate at

chromosome breakage sites. Nat Genet. 2004 Jul;36(7):767–73.

122. Adams JW, Kaufman RE, Kretschmer PJ, Harrison M, Nienhuis AW. A family of long

reiterated DNA sequences, one copy of which is next to the human beta globin gene.

Nucleic Acids Res. 1980 Dec 20;8(24):6113–28.

123. Britten RJ, Kohne DE. Repeated sequences in DNA. Hundreds of thousands of copies of

DNA sequences have been incorporated into the genomes of higher organisms. Science.

1968 Aug 9;161(3841):529–40.

124. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, et al. Initial

sequencing and analysis of the human genome. Nature. 2001 Feb 15;409(6822):860–921.

125. Dangel AW, Baker BJ, Mendoza AR, Yu CY. Complement component C4 gene intron 9

as a phylogenetic marker for primates: long terminal repeats of the endogenous retrovirus

ERV-K(C4) are a molecular clock of evolution. Immunogenetics. 1995;42(1):41–52.

126. Dewannieux M, Harper F, Richaud A, Letzelter C, Ribet D, Pierron G, et al. Identification

of an infectious progenitor for the multiple-copy HERV-K human endogenous

retroelements. Genome Res. 2006 Dec;16(12):1548–56.

127. Lee YN, Bieniasz PD. Reconstitution of an infectious human endogenous retrovirus. PLoS

Pathog. 2007 Jan;3(1):e10.

128. Levin HL, Moran JV. Dynamic interactions between transposable elements and their

hosts. Nat Rev Genet. 2011 Aug 18;12(9):615–27.

129. Bourc’his D, Bestor TH. Meiotic catastrophe and retrotransposon reactivation in male

germ cells lacking Dnmt3L. Nature. 2004 Sep 2;431(7004):96–9.

130. Yang N, Kazazian HH. L1 retrotransposition is suppressed by endogenously encoded

small interfering RNAs in human cultured cells. Nat Struct Mol Biol. 2006

Sep;13(9):763–71.

131. Rebollo R, Romanish MT, Mager DL. Transposable elements: an abundant and natural

source of regulatory sequences for host genes. Annu Rev Genet. 2012;46:21–42.

132. Bézier A, Herbinière J, Lanzrein B, Drezen J-M. Polydnavirus hidden face: the genes

producing virus particles of parasitic wasps. J Invertebr Pathol. 2009 Jul;101(3):194–203.

133. Lavialle C, Cornelis G, Dupressoir A, Esnault C, Heidmann O, Vernochet C, et al.

Paleovirology of “syncytins”, retroviral env genes exapted for a role in placentation.

Philos Trans R Soc Lond B Biol Sci. 2013 Sep 19;368(1626):20120507.

134. Malfavon-Borja R, Feschotte C. Fighting fire with fire: endogenous retrovirus envelopes

as restriction factors. J Virol. 2015 Apr;89(8):4047–50.

135. Ito J, Watanabe S, Hiratsuka T, Kuse K, Odahara Y, Ochi H, et al. Refrex-1, a soluble

restriction factor against feline endogenous and exogenous retroviruses. J Virol. 2013

Nov;87(22):12029–40.

136. Rauer M, Götz J, Schuppli D, Staeheli P, Hausmann J. Transgenic mice expressing the

nucleoprotein of Borna disease virus in either neurons or astrocytes: decreased

susceptibility to homotypic infection and disease. J Virol. 2004 Apr;78(7):3621–32.

137. Arriagada G, Gifford RJ. Parvovirus-derived endogenous viral elements in two South

American rodent genomes. J Virol. 2014 Oct;88(20):12158–62.

CHAPTER TWO

Parvovirus capsid structures required for infection: mutations controlling receptor

recognition and protease cleavages

Heather M. Callaway, Kurtis H. Feng, Donald W. Lee, Andrew B. Allison, Melissa Pinard,

Robert McKenna, Mavis Agbandje-McKenna, Susan Hafenstein, Colin R. Parrish. 2017.

Parvovirus capsid structures required for infection: mutations controlling receptor recognition

and protease cleavages. Journal of Virology. Jan 3;91(2).

2.1 Abstract

Parvovirus capsids are small but complex molecular machines responsible for undertaking many of the steps of cell infection, genome packing, and cell-to-cell as well as host- to-host transfer. The details of parvovirus infection of cells are still not fully understood, but the processes must involve small changes in the capsid structure that allow the endocytosed virus to escape from the endosome, pass through the cell cytoplasm, and deliver the ssDNA genome to the nucleus where viral replication occurs. Here, we have examined capsid substitutions that eliminate canine parvovirus (CPV) infectivity, and have identified how those mutations changed the capsid structure or altered interactions with the infectious pathway. Amino acid substitutions on the exterior surface of the capsid (Gly299Lys/Ala300Lys) altered the binding of the capsid to the transferrin receptor type 1 (TfR), particularly during virus disassociation from the receptor, but still allowed efficient entry into both feline and canine cells without successful infection.

These substitutions likely control specific capsid structural changes resulting from TfR binding required for infection. A second set of changes on the interior surface of the capsid reduced viral infectivity by >100-fold, and included two cysteine residues and neighboring residues. One of these substitutions, Cys270Ser, modulates a VP2 cleavage event found in ~10% of the capsid proteins which also was shown to alter capsid stability. A neighboring substitution, Pro272Lys, significantly reduced capsid assembly, while a Cys273Ser change appeared to alter capsid transport from the nucleus. These mutants reveal additional structural details that explain cell infection processes of parvovirus capsids.

2.2 Importance

Parvoviruses are commonly found in both vertebrate and invertebrate animals and cause

widespread disease. They are also being developed as oncolytic therapeutics and as gene therapy vectors. Most functions involved in infection or transduction are mediated by the viral capsid, but the structure-function correlates of the capsids and their constituent proteins are still incompletely understood, especially in relation to identifying capsid processes responsible for infection and release from the cell. Here we characterize the functional effects of capsid protein mutations that result in the loss of virus infectivity, giving a better understanding of the portions of the capsid that mediate essential steps in successful infection pathways, and how those contribute to viral infectivity.

2.3 Introduction

The parvovirus infectious cycle is a complex process that includes cell receptor binding, endosomal and cytoplasmic trafficking, nuclear transport, DNA replication within the nucleus, capsid assembly, nuclear export, and escape or release of the newly formed virions (1, 2).

Parvoviral capsid proteins are therefore required to perform a variety of complex functions in order to complete these steps, and those capsids are assembled from only a few forms of a single viral protein. The capsid encloses the viral genome and protects it during transfer between cells and between animal hosts, but also performs a variety of functions required to deliver the genome to the nucleus for infection (2-10). Some of these capsid-dependent processes are host- specific, and result in viral host range variation that determines an animal's susceptibility to different viruses. Many parvovirus capsid structures have been determined to near atomic resolution (11-14), but the details of how the various components of the capsid and their structures mediate cell infection and the other steps in the replication cycle that involve dynamic changes in the capsids are still poorly understood.

The canine parvovirus (CPV) capsid is a T=1 icosahedron that is assembled from 60 copies of combinations of the viral proteins VP1 and VP2. A VP3 form is generated in full

(DNA-containing capsids) through the proteolytic cleavage of ~19 residues from the N-terminus of a proportion of the VP2 molecules (15). VP1 is an extended form of VP2 that has 143 additional N-terminal amino acids which include a nuclear-targeting sequence and a PLA2 enzyme domain (6, 16, 17). VP1 comprises ~10% of the capsid subunits (15, 18). The surface of the CPV capsid is structurally complex, and includes features such as pores passing to the capsid exterior at the fivefold axes, raised surface regions (spikes) around the threefold axes, and surface valley regions located about the twofold axes and fivefold axes of symmetry (the dimple and canyon, respectively) (19-21). The capsid binds the transferrin receptor type 1 (TfR), and

TfR is the essential and apparently only receptor required by CPV and its relatives for binding to and infection of carnivore host cells (22, 23). The capsid may also bind sialic acids, specifically the modified form N-glycolylneuraminic acid (Neu5Gc), which is present in some hosts but not others, and while that binding is not an essential receptor interaction, it may reduce infectivity of cells (24). The capsid stimulates the formation of and is the target for host antibodies that efficiently block infection and aid in recovery from disease (25, 26).

The TfR binds the capsid through a structure on the threefold spikes that appears to center around VP2 residue 300, but involves other residues that are 20 to 30Å apart, suggesting a broad interaction with the capsid (27, 28). The threefold spikes are also the major antibody binding motif, and the antibody binding sites have been divided into two somewhat distinct groupings which are termed the A- and B-sites (29, 30). The modified sialic acid N-glycolyl neuraminic acid (Neu5Gc) binds to a site within the twofold dimple (31, 32).

CPV infects cells by binding to the TfR on the cell membrane, quickly followed by

clathrin-mediated endocytosis (8, 23, 33). This results in exposure of the capsid to lower pH as the endosome acidifies, which likely causes the release of some of the capsid-bound calcium ions and movement of capsid loops that also facilitate the release of the VP1 unique region (34).

Many changes in the capsid structure resulting from low pH treatment (to ~pH 5.0 or possibly lower) appear to be reversible (8, 34), and do not appear to alter the structure in a way that changes the overall capsid susceptibility to external protease treatment (15, 35). However, small or asymmetric changes may be occurring that result from TfR binding, perhaps in combination with low pH or proteolytic cleavage of a proportion of capsid proteins after assembly, and which are necessary for cell infection by the virus.

While viruses related to feline panleukopenia virus (FPV) have been infecting a wide variety of carnivore hosts for many years (commonly including domestic cats), the variant that infects dogs (CPV) arose during the 1970s initially through the acquisition of mutations that changed surface residues in two regions of the capsid – VP2 residues 93Lys to Asn (Lys93Asn), and 323Asp to Asn (Asp323Asn) (36). An additional region that controls canine and other host ranges was subsequently defined by examining mutations of residues 299 (Gly to Glu or Asp), or

300 (Ala or Gly to Asp) (37-39). The mutation Ala 300Asp has been observed repeatedly after natural transfer of CPV-2 from dogs to raccoons or during passage of CPV strains in raccoon cells (40), and that change also arose (along with a change of residue 301) during CPV-2 passage in feline cells in vitro (41, 42). The presence of 300Asp also results in reduced canine in vivo and in vitro infection, and in decreased binding to the feline TfR (43-45). The 300Asp substitution results in more extensive structural changes in the capsid than are seen for the 300Ala or Gly, due to the formation of a salt bridge between VP2 Asp300 and Arg81 on adjacent subunits, which stabilizes the extended loop that contains residue 300 in a new position (44). The change

of the adjacent VP2 residue 299 from Gly to Glu results in a CPV capsid that no longer binds the canine TfR or infects canine cells, but that does not significantly reduce the efficiency of infection of feline cells (39, 41). Residues 299 and 300 fall in the middle of one of the two principle antibody binding sites (the B-site) on the capsid, and may also influence the binding of monoclonal antibodies (MAb) that recognize that site (11, 30, 38, 46).

The CPV capsid is very stable and retains infectivity after exposure to temperatures above 60°C (35), yet it likely changes side-chain and loop conformations during cell infection.

Small differences in the structures and dynamics of various parvovirus and adeno-associated virus (AAV) capsids have been reported (35, 47, 48). Although the crystal structures of empty particles or VLPs appear structurally highly similar to infectious DNA-containing capsids when different crystal structures are compared, some differences in the structures and dynamics have been revealed (49-51). Residues on the interior of the capsid recognize and interact with the ssDNA genome by recognizing the splayed out DNA bases (11, 14, 52), and changes in the capsid-DNA interactions may control the packaging of the DNA, as well as its release of the genome during infection. The cleavage of VP2 to VP3 in full capsids may also play a role in facilitating DNA release, and has been shown to alter the capsid infectivity of minute virus of mice (MVM) (53-55). Submolar proteolytic cleavages were consistently observed within a small proportion of the capsid proteins (35), although the specific sites cleaved, the origins of the cleavages, or any functions were not defined.

The CPV capsid may also undergo structural changes after assembly, possibly upon binding to the host TfR or to host antibodies, either outside the cell or after uptake into an endosome and exposure to low pH. However, it does not appear to contain a pH-sensitive domain that changes structure in the acidic environment of the endosome that are reported for

adeno-associated virus (AAV) capsids, and which triggers an autocatalytic cleavage event when those viruses encounter low pHs (48).

Here we define key positions and functions within the parvovirus capsid involved in the cell infectious process by revealing the functional effects of single or double changes in the CPV capsid protein which make the capsid non-infectious. Some changes altered receptor binding and blocked infection after cell entry, while others were associated with a submolar cleavage of the capsid protein, which altered the interior of the capsid and changed its stability.

2.4 Materials and Methods

Mutant viruses and VLPs. Mutations were introduced into the full length infectious plasmid clone of the CPV-2 genome (56) using the Phusion mutagenesis method (New England

Biolabs) according to the manufacturer's instructions, and then sequenced to confirm the mutations. Plasmids were transfected into NLFK cells using Lipofectamine (Invitrogen) according to the manufacturer's protocol, after which viral infectivity was determined or viruses recovered. Mutations introduced included the VP2 substitutions of H137A, C270S, K271A,

K271R, C273S, G299E, G299K, A300D, A300K, and G299K/A300K.

The CPV VP2 sequence was used for production of virus like particles (VLPs) by expression in High Five insect cells with the FastBac plasmid system (Invitrogen), and mutations were also introduced using the Phusion mutagenesis method. Plasmids were sequenced to verify the presence of the mutations, then transformed into DH10Bac E. coli to generate bacmids.

Bacmids were transfected into Sf9 cells with Cellfectin II (Invitrogen) according to the manufacturer's protocol to generate baculovirus. Mutations included the VP2 substitutions of

H137A, C270S, K271A, K271R, P272K, C273S, and G299K/A300K.

Cells, viruses and VLP preparations. Norden Laboratory Feline Kidney (NLFK) cells were grown in 1:1 McCoy’s/Lebovitz L15 media (Corning) with 5% fetal calf serum (FCS). Sf9 and High5 insect cells were grown in Grace's Insect Medium (Invitrogen) with 10% FCS, and

Express Five Serum Free Media (Invitrogen), respectively.

The CPV-2 strain of virus was used to prepare full (DNA containing) and empty particles by growth of the virus in NLFK cells, and capsids were purified as described previously (11).

VLPs were prepared by expressing CPV-2 VP2 in High5 cells using baculoviruses as described above, followed by purification. Briefly, after 4 days of incubation with baculovirus, High5 cells were re-suspended in 25 ml of lysis buffer (10mM NaCl; 10mM Tris-HCl (pH 7.5); 15mM

MgCl2; 5mM Triton X-100; 0.4 × M221 Protease Inhibitor (Amresco)), then freeze-thawed 4 times. Lysates were treated with a Branson sonifier 250 (output 3, duty cycle 50% for 30s) then debris was pelleted at 12,400×g for 30 mins, and the supernatant collected. Capsids were then either purified by 1) polyethylene glycol (PEG) precipitation and sucrose gradient centrifugation as described previously (11), followed by size exclusion chromatography in a BioRad A5M column, or 2) by centrifuging on a CsCl step gradient (3.5 ml of 1.25 g/cm3, 3 ml of 1.35 g/cm3, and 0.5 ml of 1.5 g/cm3 CsCl), followed by banding in a isopycnic gradient formed from 1.35 g/cm3 CsCl, as described in Jager et. al. (57).

Production of soluble feline TfR ectodomain. BHK cells expressing the soluble 6-His- tagged feline TfR ectodomain were grown in roller bottles with Dulbecco's Modified Eagle

Medium (DMEM)/F12 (Lonza) with 10% FCS. DMEM/F12 media was then replaced with production media (Pro293A-CDM (Lonza) with added L-glutamine (2mM) and butyric acid

(1mM)), and media was collected at one day intervals (37).

Molecular images and homology modeling. Images of the major capsid protein were

generated in UCSF Chimera (version 1.10.1) (RBVI). The homology model of the

Gly299Lys/Ala300Lys mutant was generated via the mutagenesis function in PyMOL (version

1.7.4) (Schrödinger, LLC).

Hemagglutination Inhibition (HAI) assays. HAI assays were used to determine the relative activity of monoclonal antibodies (MAbs) for VLPs. Four HA units of VLP were added to MAbs 14, E and 8 (25, 26) that were diluted two-fold in Bis-Tris Buffered Saline (BTBS) (25

μl total volume) in a 96-well V-bottom plate and incubated for 1 h at room temperature, then 50

μl of 0.5% feline red blood cells were then added to each well and the plates were incubated at

4°C for 2 h.

TCID50 assays. 2.5μg of mutant virus infectious plasmid was transfected into NLFK cells seeded at 2 × 104 cells/cm2 in 9cm2 plates, as described above. 4 days after transfection, supernatant was collected and serially passaged twice at 1:4, with supernatant collected at 4 days after each passage. For virus titrations, NLFK cells seeded at 2 × 104 cells/cm2 in 96-well plates were incubated with 50 μl of 10-fold diluted viruses for 1 h at 37°C, then 200 μl of tissue culture media was added and cells were incubated at 37°C. After 2 days, cells were fixed, incubated for

1h with rabbit polyclonal anti-FPV antibody, washed, incubated with goat anti-rabbit horse radish peroxidase (HRP) antibody, and stained with amino ethyl carbazol (AEC) substrate (20mg

AEC in 0.035M sodium acetate pH 5.0, with 0.015% H2O2). Infected wells were determined and used to calculate TCID50 by the automatic drop through method using the NIH ID50 Server available at http://www.ncbi.nlm.nih.gov/CBBresearch/Spouge/html_ncbi/html/id50/id50.html

(58). Three separate transfections and serial passages were performed for each virus.

Bio-layer interferometry of capsid-TfR binding kinetics. Binding kinetics between the

CPV capsids or VLPs and the purified feline TfR ectodomain were determined in a ForteBio

BLITz bio-layer interferometer, using 240 μg/ml of capsid (as CPV empty capsids or insect cell expressed VLPs). The assay procedure is as follows: 1) hydration of Nickel Nitrilotriacetic Acid

(NiNTA) tips for 10 min in kinetics buffer (PBS with 0.02% ovalbumin and 0.02% Tween-20),

2) 30s baseline wash in kinetics buffer, 3) 300s TfR loading, 4) 60s baseline wash in kinetics buffer, 5) 300s capsid association, and 6) 300s analyte disassociation in PBS with 0.02% Tween-

20. The feline TfR was added as a culture supernatant prepared from BHK cell expression, which contained the expressed dimeric ectodomain with a 6-His tag, and that was loaded onto tips to give 0.8nm of binding. For each capsid or VLP sample, experiments were repeated 6 times to obtain an averaged kinetic curve with standard error of the mean.

Binding and uptake analysis using immunofluorescence assays (IFA). NLFK cells seeded on cover slips at 2 × 104 cells/cm2 were incubated in DMEM with 0.1% bovine serum albumin (BSA) for 30 min at 37°C. Cells were incubated with 40 μl of 4 μg/ml CPV capsids or

VLPs diluted in DMEM with 0.1% BSA for 1 h at 37°C, then fixed in 4% paraformaldehyde

(PFA) for 10 min at room temperature. Cells were washed with PBS, and incubated with polyclonal rabbit anti-FPV and mouse anti-TfR cytoplasmic tail (Affymetrix) antibodies for 1 h at room temperature. Cells were then washed with PBS, incubated with goat anti-rabbit

Alexa594 and goat anti-mouse Alexa 488 for 1 h at room temperature and mounted on slides using Prolong Gold Antifade with 4',6-diamidino-2-phenylindole (DAPI) (Life Technologies).

Cells were imaged using a Nikon TE300 fluorescence microscope.

Flow cytometry. NLFK cells seeded at 2 × 104 cells/cm2 in 9 cm2 plates were incubated in DMEM with 0.1% BSA for 30 min at 37°C. Cells were then incubated with 500μl of 4μg/ml of CPV or VLP diluted in DMEM with 0.1% BSA for 1 h at 37°C. Cells were trypsinized and centrifuged at 500 × g for 5 min, washed with PBS and fixed in 4% PFA, washed with

permeabilization buffer (PBS with 0.1% BSA, 0.1% Triton-X 100, and 0.1% NaN3) and incubated with 1.5 μg/ml MAb 14 for 30 min at room temperature. Cells were washed with permeabilization buffer, incubated with goat anti-mouse Alexa647 for 30 min at room temperature, then washed with permeabilization buffer and resuspended in PBS. Measurements were taken using an A94291 Gallios flow cytometer (Beckman Coulter) running Kaluza

Analysis Software (Beckman Coulter) and graphs were prepared using FlowJo V10 (FlowJo,

LLC).

Immunofluorescence assays to monitor capsid assembly. NLFK cells seeded at 2 ×

104 cells/cm2 on cover slips in 9 cm2 plates were transfected with 2.5μg of infectious plasmid with Lipofectamine (Invitrogen) according to the manufacturer's instructions. Four days later, the cells were fixed in 4% PFA for 10 min at room temperature, washed with PBS, then incubated with polyclonal rabbit anti-FPV antibody (recognizing VP2 subunits and assembled capsids) or mouse MAb 8 (recognizing only the assembled capsid) for 1 h at room temperature. Cells were washed with PBS, incubated with secondary antibody (goat anti-rabbit Alexa488 or goat anti- mouse Alexa 488 as appropriate) for 1 h at room temperature, then mounted on slides using

Prolong Gold Antifade with DAPI (Life Technologies). Cells were imaged with a Nikon TE300 fluorescence microscope.

Quantitative western blotting. Samples of 40 μg CPV or VLP were run on 10% SDS-

PAGE gels and transferred to nitrocellulose membranes. Membranes were blocked in Odyssey blocking buffer (LI-COR) for 1 h at room temperature, then incubated with primary antibody: polyclonal rabbit anti-FPV antibody, polyclonal mouse antibody prepared against VP2 peptide

258-270 (15), or monoclonal mouse antibody P1F6 against the CPV-2 C-terminus (15).

Membranes were incubated either for 1 h at room temperature or overnight at 4°C, washed in

PBS, and incubated with secondary antibody (goat anti-rabbit IRDye 800CW or goat anti-mouse

IRDye 680RD (LI-COR)) for 1 h at room temperature. Membranes were washed again in PBS and imaged on an Odyssey infrared imager (LI-COR).

Peptide mapping. Wild-type CPV capsid proteins were run on a 10% SDS-PAGE gel and stained with Coomassie blue. Gels were stored in distilled water and sent to the Cornell

Biotechnology Resource Center Proteomics Core, where the N-terminal ~30kDa proteolytic cleavage fragment was excised from the gel, digested in separate preparations with trypsin and

Glu-C enzymes, and peptide mapped using an Orbitrap fusion or Orbitrap elite mass spectrometer.

Quantification of capsid nuclear or cytoplasmic localization. NLFK cells seeded at a density of 4 × 104 cells/cm2 in 9 cm2 plates were transfected with 2.5μg of wild-type or

Cys273Ser infectious plasmid with Lipofectamine (Invitrogen) according to the manufacturer's instructions. Cells were fixed with 4% paraformaldehyde 24 h later, then stained with MAb8

(recognizing assembled capsids) and goat anti-mouse Alexa488 secondary antibody in permeabilization buffer (see above), and mounted with ProLong Anti-Fade Gold containing

DAPI (Life Technologies). Cells imaged using a Nikon TE300 fluorescence microscope were scored according to whether assembled capsids localized only to the nucleus or were also visible in the cytoplasm. Data were from 6 replicates.

Tryptophan fluorescence spectroscopy. CPV empty capsids or VLPs were diluted to

0.5 μg/ml in 500 μl PBS (pH 7.4). Tryptophan fluorescence was measured using 1cm path length cuvettes in a Cary Eclipse Spectrophotometer (Agilent Technologies), with 5 nm excitation and emission bandwidths. Samples were excited at 295 nm and read at 331 nm, as described previously (35), and measurements were taken every 0.5°C as samples were heated from 25°C to

85°C at 0.5°C/min. The melting temperature (TM) for each sample was determined by finding the minimum of the first derivative of the resulting curve using GraphPad Prism 6 (GraphPad

Software, Inc.).

Differential Scanning Fluorometry (DSF). CPV empty capsids or VLPs were diluted to

0.1-0.2 mg/ml with PBS (pH 7.4), 100 mM citric acid (pH 5.5), or 10mM EGTA (pH 8.0) to a total volume of 22.5 μl and incubated for one h at room temperature. 2.5 μl of 1% SYPRO

Orange dye (Life Technologies) was added to samples in a 96-well qPCR plate, and samples were then loaded into a StepOne Plus qPCR machine (Applied Biosystems). Measurements were taken every 0.2°C as samples were heated from 25°C to 85°C (1°C/min). TM for each sample was determined by finding the maximum of the first derivative of the resulting curve using

GraphPad Prism 6 (GraphPad Software, Inc.).

2.5 Results

To better understand the capsid controls of cell infection, we examined a panel of point mutations in the CPV capsid, of which VP2 mutants Gly299Lys/Ala300Lys, Pro272Lys, and

Cys273Ser did not result in the production of viable viruses and Cys270Ser resulted in 100-fold less viable virus. These mutations were prepared in infectious plasmid clones of the viruses and tested for sustained growth of the viruses after transfection into NLFK cells, which are susceptible to all of the natural isolates of CPV. Mutations blocking viral infectivity included those changing residues on the capsid surface (VP2 299 and 300), and others on the interior of the capsid (VP2 270, 272, and 273).

To allow us to produce sufficient particles to study the properties of the capsids containing substitutions that produced non-viable viruses, VLPs were prepared from the VP2

gene expressed in insect cells, and the capsids were purified and compared to both wild-type

VLPs and wild-type empty capsids recovered from cell infections.

VP2 residue 299/300 surface mutations allow feline TfR binding, but block infection.

VP2 residues 299 and 300 are within a surface-exposed loop (Figure 2.1A and B) that previously has been shown to play a crucial role in the infectious interactions with both the canine and feline TfRs (22, 44, 56). Some mutations to VP2 residue 300 have been structurally characterized (Gly300 and Asp300) (Figure 2.1C) and have been shown to minimally affect the capsid surface structure, with the greatest change being the formation of a salt bridge between

Asp300 and Arg81. We predict that other mutations to residues 299 and 300 will similarly affect only the local capsid surface structure, and we have constructed a homology model of what a

Gly299Lys/Ala300Lys mutation may look like (Figure 2.1C).

To compare the properties of infectious and non-infectious CPV mutants, Gly299Glu and

Ala300Asp empty capsids were compared as controls, as both are viable in NLFK cells (Figure

2.1D), but show loss of binding to the canine TfR (Gly299Glu), or reduced infectivity in canine cells (Gly299Glu and Ala300Asp) (22, 39, 44). Both of these mutants also showed decreased binding by some MAbs recognizing the B antigenic site (Gly299Glu and Ala300Asp), indicating that they create sufficient alteration of the capsid surface to block antibody binding (30, 39, 44,

46) (Figure 2.1E).

The Gly299Lys/Ala300Lys double mutant exhibited a contrasting profile, as the capsid assembled (identified by VLP hemagglutination of feline red blood cells), but it was not viable in

NLFK feline cells (Figure 2.1D) and VLPs showed a complete loss of binding of the B-site specific MAb 8 (Figure 2.1E). By fluorescence microscopy, the Gly299Lys/Ala300Lys mutant

VLPs showed significant binding and uptake into feline cells, and although levels appeared

Figure 2.1. Protein structure, surface morphology, and viability of the external capsid surface mutations. (A) CPV VP2 full capsid (PDB ID: 4DPV) (exterior surface facing out) with VP2 residues 299 and 300 displayed in yellow. (B) Magnification of region of (A) showing surrounding subunits (gray), and residues within 5 Å of Gly299 and Ala300 displayed and labeled. (C) From left to right: WT CPV full capsid (Ala300) (PDB ID: 4DPV); Gly300 CPV

(PDB ID: 4QYK); Asp300 CPV, with salt bridge to Arg81 (PDB ID: 1IJS); and a homology 66

model of Gly299Lys/Ala300Lys. (D) Reactivity of wild-type or mutant capsids with mAbs 14

(A-site), 8 (B-site), and E (B-site), using a hemagglutination inhibition assay (HAI). Data were collected from 3 independent experiments and analyzed via ANOVA. (E) Relative infectivity and growth of mutant CPVs compared to wild-type, measured by TCID50, over three passages in

NLFK feline cells. Data was collected from 3 independent experiments and analyzed via

ANOVA. Error bars display standard error of the mean, and asterisks denote p<0.05.

lower than seen for wild-type CPV or VLPs, they were higher than the levels seen for either

Gly299Glu or Ala300Asp, which produced viable and naturally transmissible capsids (Figure

2.2). When levels of virus binding and uptake into NLFK cells were determined by flow cytometry (Figure 2.3), we observed similar results to the immunofluorescence assay, with wild- type CPV capsids and VLPs showing the highest levels of cell association after 60 mins of incubation at 37°C, followed by the Gly299Lys/Ala300Lys mutant VLPs, while the Gly299Glu and Ala300Asp mutants showed low levels of binding and uptake that were similar to the isotype control (Figure 2.3).

In vitro binding analysis. In order to determine whether changing residues 299 or 300 affected affinity of virus for the feline transferrin receptor (fTfR), we used bio-layer interferometry to measure the binding and unbinding rates of CPV and VLPs to fTfR-coated probe surfaces. Both CPV and VLPs showed a fast association phase and a biphasic disassociation phase, with an initial fast disassociation followed by a slower disassociation rate

(Figure 2.4). VLPs with Gly299Lys/Ala300Lys substitutions had association curves comparable to wild-type CPV and VLPs, but the dissociation curves for Gly299Lys/Ala300Lys VLPs were much steeper (Figure 2.4). Capsids with Gly299Glu and Ala300Asp substitutions had diminished association rates and increased dissociation rates. Variations in binding rates could not be explained by variations in capsid concentrations, because these were kept constant for all samples.

To provide a more quantitative comparison of dissociation behaviors, we sought kinetic models that could fit the dissociation curves. Characterizing the dissociation curves can also be useful for determining if adhesion-strengthening mechanisms are present, as demonstrated by related work on tracking contact times between individual parvoviruses and TfRs (43). Because

Figure 2.2. Binding and uptake of 4μg/mL CPV or VLP into NLFK cells determined by immunofluorescence assay. (Top) Images show a merge of TfR expression (detecting the cytoplasmic tail) (green), CPV capsids (red) and DAPI/nuclei (blue) for wild-type or mutant

CPV or VLP or (bottom) CPV capsids only.

Figure 2.3. Flow cytometry showing binding and uptake assay of 4μg/mL CPV or VLP into

NLFK cells. (A) Representative flow cytometry histogram and (B) quantification of mean fluorescence intensity (MFI) from 3 experiments with 2 replicates, analyzed via ANOVA. Error bars display standard error of the mean, and asterisks denote p<0.05.

Figure 2.4. Binding of CPV or VLP to soluble feline TfR using bio-layer interferometry.

Sensorgram data shows association of CPV or VLPs to feline TfR for 5 min of incubation, followed by disassociation for 5 min. Curves and error bars represent the averaged data from 6 replicates. Fitted dissociation curves using the adhesion-strengthening model are shown as black lines. Error bars display standard error of the mean.

the dissociation curves here did not fit the basic 1:1 binding model, we used the adhesion- strengthening model (43, 59). This model assumes that the binding force can change with contact time, perhaps due binding-triggered conformational changes of the capsid (43). However, the initial adhesion-strengthening model only applies to studies where contact times for each capsid particle can be measured, whereas here we measured the collective loss of many capsids from the surface over time. We therefore used the ensemble-averaged form of the model (Eq. 1) as explained elsewhere (59).

′ 푡 =푡푑푖푠푠−∆푡 t−t′ −A N (t) = ∑ B [ln ] Eq. 1 diss ∆t 푡′=0

In Eq. 1, parameter A reflects how fast a virus-receptor contact can transition into a stronger contact, with lower A values representing faster transition times and stronger contacts.

Parameter B reflects how many binding attempts are made by virus particles, which include diffusion-limited transport of the virus to the surface and also rebinding attempts made by weakly-bound viruses. Results of the fits are provided in Table 2.1. We are mainly interested in comparing A values amongst the mutated capsids to determine if certain residues affect the ability of parvovirus to form stronger bonds with the TfR. Wild-type VLPs and CPV both have the lowest A values of ~1 and 2 respectively, which suggests that these capsids quickly form strong contacts with the TfR. Capsids with the Gly299Lys/Ala300Lys double mutation have a moderately high A value, suggesting a high delay time for forming strong contacts with the TfR.

Capsids with Gly299Glu and Ala300Asp mutations have the highest A value, suggesting that these capsids have the most difficulty undergoing changes that stabilizes contact with TfR.

Overall, these results agree with those derived from the virus binding and uptake studies shown above.

Table 2.1. Fitting constants for CPV and VLP external capsid mutants, derived from fitting bio-

layer interferometry sensorgrams to Eq.1.

Fit of Averaged Curve Aa Ba R 2 WT CPV 2.12±0.21 0.026±0.004 0.996 G299E CPV 4.68±0.86 1.100±1.054 0.952 A300D CPV 4.51±0.46 1.000±0.215 0.962 WT VLP 1.02±0.33 0.003±0.002 0.936 G299K/A300K VLP 3.56±0.50 0.333±0.224 0.985

a Values are ± SD.

Mutations on the interior capsid surface that affect infection. Other mutations that affected viral infectivity were found on the interior of the capsid (Figure 2.5A-C), and did not affect receptor binding by the VLPs. Mutations examined were clustered in a region that contains the only four Cys residues in the capsid (VP2 residues 270, 273, 490, 494). Cys490 and Cys494 form an intra-chain disulfide bond, while Cys270 and Cys273 are free and point away from each other in crystal structures of the capsid (Figure 2.5A-C) (14). The 270-273 amino acid region of

CPV VP2 is exposed on the internal surface of the capsid, and while it is not known to be in direct contact with the DNA or to differ greatly in position between full and empty capsids, the position of residue 269 is influenced by the loop containing VP2 residues 490-494, which directly interacts with viral DNA (Figure 2.5C) (11, 19, 34, 52). Cys490 is conserved in other parvoviruses, but the sequences surrounding this region of the capsid are not conserved in other parvoviruses such as the minute virus of mice (MVM) or porcine parvovirus (PPV) (Figure

2.5D), where residues that align with 270 are Thr and with 273 are either Val or Leu.

The Cys270Ser and Cys273Ser substitutions in CPV allowed capsid expression and assembly (Figure 2.6A), but significantly impaired viral infectivity. The Cys270Ser change resulted in a 100-fold decrease in viral titer after 3 passages compared to wild-type CPV, while the Cys273Ser mutation completely eliminated infection in NLFK cells (Figure 2.6B).

Pro272Lys mutation also eliminated viral infectivity in NLFK cells (Figure 2.6B), but the low infectivity appears to be due to that substitution preventing capsid assembly, leading to a degradation of capsid subunits in the cytoplasm after transfection of the infectious plasmid

(Figure 2.6A) and greatly decreased VLP production after baculovirus expression of the mutant

VP2 (data not shown).The Lys271Ala and Lys271Arg substitutions were also located between

Cys270 and Cys273, but they showed capsid assembly and viral infectivity similar to the wild-

Figure 2.5. Structure and sequence conservation of the capsid region on the interior surface of

CPV that was mutated. (A) CPV VP2 empty capsid (PDB ID: 2CAS) (interior surface facing out), with VP2 residues 270-273 and 137 colored orange, residues within 5 Å of 270-273 and

137 colored green, and associated DNA colored red. (B) Magnification of (A), with residues

270-273 and 137 labeled. (C) Diagram of (B), showing the overall structural arrangement, the relative positions of residues 270-273, 137 and residues within 5 Å of those. (D) Amino acid sequence alignment generated using Clustal Omega for CPV (UniProtKB Q11213) and the parvoviruses porcine parvovirus (PPV) (UniProtKB P18546) and minute virus of mice (MVM)

(UniProtKB P07320). Residues labeled in (C) are underlined.

Figure 2.6. Capsid assembly, viability, and proteolytic cleavage of the mutations changing residues on the internal capsid surface. (A) Immunofluorescence assay of NLFK cells four days after transfection with WT or mutant CPV infectious plasmid. Cells were stained with MAb 8

(which recognizes only assembled capsids) or a polyclonal anti-VP1/2 antibody (recognizing both monomeric VPs and capsids). (B) Relative infectivity of mutant CPVs compared to wild- type, measured by TCID50, over three passages in NLFK feline cells. Data was collected from 3 independent experiments and analyzed via ANOVA. (C) Western blot of wild-type CPV empty capsids or wild-type or mutant VLPs. Peptides were stained with polyclonal anti-VP1/2 antibody

(top), monoclonal antibody recognizing a peptide containing residues 362-373 (middle, red

outline), or a polyclonal antibody recognizing a peptide containing residues 258-270 (bottom, green outline). (D) Diagram showing results of peptide mapping experiments, with the identified proteolytic cleavage site located between residues Asp269 and Cys270. (E) Diagram showing the proteolytic cleavage site in the context of the surrounding protein structure. Error bars display standard error of the mean.

type (Figure 2.6A and B).

The Cys270Ser mutant assembled normal levels of capsids in NLFK cells and VLPs in insect cells, but lacked a submolar proteolytic cleavage fragment when capsid proteins in the

VLPs were examined by Western blotting (Figure 2.6C). This cleavage product occurs in ~10% of the wild-type VP2 proteins in either capsids or VLPs, and mass spectrometry peptide mapping after trypsin and Glu-C protein digestion of wild-type capsids showed that the cleavage generating that peptide occurred between Asp269 and Cys270 (Figure 2.6D and E).

The Ser270 mutation directly altered that cleavage recognition site, but the source of the proteolytic activity was not clear. We asked whether Cys270 might itself be a component of a cysteine protease, and therefore examined the surrounding region for His or Asp/Glu residues which could complete a catalytic triad of residues typically observed in cysteine proteases (60,

61). There were no Asp or Glu residues within 5Å of Cys270, but His137 was close enough to

Cys270 to interact with it, and in some cases, Cys/Ser-His dyads have formed functional proteases (60). However, changing His137 to Ala did not change the formation of the VP2 peptide compared to wild-type VLPs (Figure 2.6C) or result in a loss of infectivity (data not shown).

The Cys270Ser substitution significantly increased the melting temperature of the viral capsid compared to wild-type VLPs, as measured by both tryptophan fluorescence spectroscopy and differential scanning fluorometry (Figure 2.7). This increase was also present when capsids were incubated at pH 5.5, which has previously been shown to induce a conformational change in the capsid (34), and when capsids were incubated in 10mM EGTA, which removes two calcium ions per subunit (34).

The Cys273Ser substitution allowed capsids to assemble efficiently as VLPs (data not

Figure 2.7. Stability of wild-type or mutant VLPs determined using tryptophan fluorescence spectroscopy or differential scanning fluorometry. (A) Tryptophan fluorescence spectroscopy of

WT, C270S, and C273S VLPs heated from 25°C to 85°C at 0.5°C/min at pH 7.4. (B) Melting temperatures of the different capsids determined from tryptophan fluorescence spectroscopy experiments in (A). (C) Differential scanning fluorometry (DSF) of WT, C270S, and C273S

VLPs heated from 25°C to 85°C at 0.5°C/min at pH 7.4. (D) Melting temperatures of the different capsids determined from DSF experiments at pH 7.4 (as in C), pH 5.5, and with 10mM

EGTA. Data was taken from 3 independent experiments and analyzed via ANOVA. Error bars display standard error if the mean, and asterisks denote p<0.05.

shown), but resulted in complete loss of viability. At 24 h after transfecting NLFK cells with infectious plasmid, unassembled capsid subunits were localized throughout the cell cytoplasm, but assembled capsids were located primairly in the nucleus, unlike wild-type virus, where assembled capsids were found in both the nucleus and cytoplasm (Figure 2.8A and B).

2.6 Discussion

Here we studied two capsid structures, one surface exposed and one internally located, to reveal new details about how the parvovirus capsid controls the successful infection of cells. In both cases the effects appear to be due to changes in the dynamic properties of the capsid that are required for successful cell infection. The Gly299Lys/Ala300Lys mutations on the outer surface of the capsid resulted in a non-infectious particle, despite the mutant being able to bind the TfR and to enter cells at apparently normal rates. Those changes therefore appear to interfere with a capsid structural change that is required for infection after the initial receptor binding. The internal capsid substitution of Cys270 to Ser also inhibited cell infection, but it was associated with loss of a proteolytic cleavage product and increased capsid stability, again indicating a structure that functions after receptor binding and cell entry. The Pro272Lys change likely decreased infectivity because it interfered with efficient capsid assembly, while the specific role of the Cys273Ser mutation is not clear, but it appears to prevent egress of assembled capsids from the nucleus.

Key structural interactions between the TfR and the capsid are required for infection. The external surface changes show that there are essential interactions between the

TfR and the CPV capsid in addition to simple binding that are required for cell infection. Most of our studies were of the Gly299Lys/Ala300Lys double mutant, but the Ala300Lys mutation alone

Figure 2.8. Intracellular localizations of the Cys273Ser mutant compared to wild-type capsids.

(A) IFA showing nuclear localization of capsid subunits detected with a polyclonal anti-VP1/2 antibody (left column) and assembled capsids detected with MAb 8 (right column) 24 h post- transfection. (B) Percentage of cells that had assembled capsids only in the nucleus out of all transfected cells, from a total of 6 transfections in 3 independent experiments, analyzed via

ANOVA. Error bars display standard error of the mean, and asterisks denote p<0.05.

also eliminated infectivity. Since the nonviable Gly299Lys/Ala300Lys mutant bound to TfR and entered feline cells to higher levels than either of the viable Gly299Glu or Ala300Asp mutants, and residues 299 and 300 are located directly on the receptor binding site, there must be an additional interaction between CPV and the TfR that is required for infection. We suspect that the 300 residue governs a conformation change that is triggered by specific receptor binding events that precede viral genome delivery. In support of this hypothesis, previous work has suggested that parvoviruses require structural cues from the receptor to infect cells, as replacement of the TfR ectodomain with a scFv (single-chain variable fragment) allows CPV to enter cells, but prevents infection (62). A substitution in the apical domain of the feline TfR

(Leu221Lys) also results in reduced infection despite allowing efficient cell binding and uptake of CPV capsids (63). The TfR-capsid interaction is likely highly asymmetric, with only a small number of TfRs binding per capsid – indeed, capsids were shown to only engage one or two

TfRs during initial cell binding and entry (33), and biochemical and imaging analysis also show fewer than 5 TfR bound per capsid in vitro (27). In summary, one or a few TfRs bind the capsid to allow uptake and infection, but the structural interaction with the capsid is highly specific and involves contacts between the TfR apical domain, which likely include residue 221 of the feline

TfR, and the three-fold spike surface, including residue 300.

These results add to our previous understanding of how the specific structures of the CPV capsid controls infection and host cell tropism, where variation in the capsid controls the host range (23, 36, 37, 40, 44). Residue 300, which we showed here controls the virus after cell entry, has also been shown in other cases to control receptor binding, tropism and host range. One of those processes was the gaining of the canine host range by CPV in the 1970s (45). The structural challenge encountered appears to have been particularly difficult as the canine TfR has

an additional glycan within the virus-binding region of its apical domain, and changes in three different areas of the capsid have been shown to influence binding to that TfR and infection of canine cells (36, 39, 64). The Lys93Asn and Asp323Asn substitutions in the original CPV allowed canine TfR binding and canine cell infection, while the single change of Gly299 to Glu blocked those activities (39). Variation of VP2 residue 300 and nearby residues are associated with the specific adaptation to other hosts or their cultured cells, most likely due to small accommodations to allow efficient binding and infection using the variant TfRs of the different host animals (37). Altogether, it is clear that infection requires specific interactions between the capsid and the TfR that involve not just binding but also the ability to induce small asymmetric allosteric changes in the capsid that are essential for successful infection.

Internal capsid structures controlling infection. A second structural region that appeared to undergo key changes that were required for infection was identified on the interior of the capsid. We confirmed that about 10% the VP2 (and presumably VP1) molecules in assembled capsids were cleaved between Asp269 and Cys270. The substitution of Ser for Cys at residue 270 resulted in loss of infectivity, and there was little or no proteolysis of that site and a small but significant increase in capsid melting temperature, indicating increased capsid stability.

This suggests that proteolytic cleavage of a small number of capsid subunits may give the capsid a specific structure or conformational flexibility required for infection. Many different viruses use proteolytic cleavage of one or more of the viral capsid proteins for maturation into an infectious form, including poliovirus (65), rhinovirus (66), flockhouse virus (67), and adenovirus

(68). It is unclear what introduces the submolar proteolytic cleavage of the CPV VP2 at Cys270, but this must be either an autocatalytic cleavage or due to a cellular protease.

The Cys270 cleavage occurred in capsids from virus-infected cells as well as in VLPs to

similar levels, so any cellular protease must be present in feline (NLFK) cells and insect (High

Five) cells. The identity of a catalytic structure in the capsid that would cleave at Asp269/Cys270 is not obvious. The His137Ala substitution had no effect on formation of the submolar cleavage product, and the presence of a Cys/Lys dyad protease was not supported by the finding that the

Lys271Ala mutation had no effect on the formation of the submolar cleavage product, and there are no other lysine residues are nearby. Although there are reports of cysteine/serine/threonine residues using the N-terminal amine group as a base during proteolysis (60, 61), it is much more common to find the conserved catalytic triad (serine/cysteine, histidine and glutamate/aspartate)

(60). The cleavage in CPV has parallels in the capsids of various AAV serotypes, which have a variety of low pH-activated protease cleavages that may also activate the capsid for infection, where the sources of the apparent auto-proteolytic activities are still unclear (48).

Of the other mutations examined, the Pro272Lys mutation prevented assembly of capsids both in the virus and in VLPs, likely due to disruption of the local structure of the capsid. The

Cys273Ser mutation resulted in a non-viable virus that could not be passaged successfully, and those capsids appeared to be primarily nuclear localized, in contrast to the wild-type CPV or other mutations where capsids were distributed in both the cytoplasm and nucleus. Similar results were seen in the minute virus of mice (MVM) when serine residues on the N-terminus of

VP2 were mutated to prevent phosphorylation, and the capsids could no longer be recognized by nuclear export machinery (69). Residue 273 is on the interior of the capsid and distant from the

5-fold pores, so the Cys273Ser change seems unlikely to act through such a mechanism, and additional experiments will be needed to determine its role in the CPV infection and replication process.

Summary. These studies confirm that the parvovirus capsid is a finely tuned molecular

machine, and its highly specialized functioning can be disrupted by as little as a single amino acid change in critical positions. Here we show that small and likely submolar asymmetric changes within the capsid control key steps of infection – some being triggered by receptor binding, while others appear to involve proteolytic cleavages. Because all the steps in infection, replication, and release are mediated by a single protein structure (shared by VP1 and VP2), understanding those functions will require analysis of fine-scale and often submolar processes, and defining their different roles in the replication cycle and in controlling cell-cell and animal- animal transmission.

2.7 Acknowledgements

We thank Wendy Weichert for excellent technical support.

REFERENCES

1. Berns KI. 1990. Parvovirus replication. Microbiol Rev 54:316-329.

2. Harbison CE, Chiorini JA, Parrish CR. 2008. The parvovirus capsid odyssey: from the

cell surface to the nucleus. Trends Microbiol 16:208-214.

3. Basak S, Turner H. 1992. Infectious entry pathway for canine parvovirus. Virology

186:368-376.

4. Cohen S, Behzad AR, Carroll JB, Panté N. 2006. Parvoviral nuclear import: bypassing

the host nuclear-transport machinery. The Journal of General Virology 87:3209-3213.

5. Duan D, Li Q, Kao AW, Yue Y, Pessin JE, Engelhardt JF. 1999. Dynamin is required for

recombinant adeno-associated virus type 2 infection. J Virol 73:10371-10376.

6. Farr GA, Zhang L-g, Tattersall P. 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.

7. Girod A, Wobus CE, Zádori Z, Ried M, Leike K, Tijssen P, Kleinschmidt JA, Hallek M.

2002. The VP1 capsid protein of adeno-associated virus type 2 is carrying a

phospholipase A2 domain required for virus infectivity. The Journal of General Virology

83:973-978.

8. Parker JS, Parrish CR. 2000. Cellular uptake and infection by canine parvovirus involves

rapid dynamin-regulated clathrin-mediated endocytosis, followed by slower intracellular

trafficking. J Virol 74:1919-1930.

9. Vihinen-Ranta M, Yuan W, Parrish CR. 2000. Cytoplasmic trafficking of the canine

parvovirus capsid and its role in infection and nuclear transport. J Virol 74:4853-4859.

10. Whittaker GR, Helenius A. 1998. Nuclear import and export of viruses and virus

genomes. Virology 246:1-23.

11. Agbandje M, McKenna R, Rossmann MG, Strassheim ML, Parrish CR. 1993. Structure

determination of feline panleukopenia virus empty particles. Proteins 16:155-171.

12. Llamas-Saiz AL, Agbandje-McKenna M, Wikoff WR, Bratton J, Tattersall P, Rossmann

MG. 1997. Structure determination of minute virus of mice. Acta Crystallographica

Section D, Biological Crystallography 53:93-102.

13. Simpson AA, Hébert B, Sullivan GM, Parrish CR, Zádori Z, Tijssen P, Rossmann MG.

2002. The structure of porcine parvovirus: comparison with related viruses. J Mol Biol

315:1189-1198.

14. Xie Q, Chapman MS. 1996. Canine parvovirus capsid structure, analyzed at 2.9 A

resolution. J Mol Biol 264:497-520.

15. Weichert WS, Parker JS, Wahid AT, Chang SF, Meier E, Parrish CR. 1998. Assaying for

structural variation in the parvovirus capsid and its role in infection. Virology 250:106-

117.

16. Vihinen-Ranta M, Wang D, Weichert WS, Parrish CR. 2002. The VP1 N-terminal

sequence of canine parvovirus affects nuclear transport of capsids and efficient cell

infection. J Virol 76:1884-1891.

17. Zádori Z, Szelei J, Lacoste MC, Li Y, Gariépy S, Raymond P, Allaire M, Nabi IR,

Tijssen P. 2001. A viral phospholipase A2 is required for parvovirus infectivity. Dev Cell

1:291-302.

18. Rhode SL. 1985. Nucleotide sequence of the coat protein gene of canine parvovirus. J

Virol 54:630-633.

19. Tsao J, Chapman MS, Agbandje M, Keller W, Smith K, Wu H, Luo M, Smith TJ,

Rossmann MG, Compans RW. 1991. The three-dimensional structure of canine

parvovirus and its functional implications. Science (New York, NY) 251:1456-1464.

20. Wikoff WR, Wang G, Parrish CR, Cheng RH, Strassheim ML, Baker TS, Rossmann MG.

1994. The structure of a neutralized virus: canine parvovirus complexed with neutralizing

antibody fragment. Structure (London, England: 1993) 2:595-607.

21. Wu H, Rossmann MG. 1993. The canine parvovirus empty capsid structure. J Mol Biol

233:231-244.

22. Hueffer K, Parker JSL, Weichert WS, Geisel RE, Sgro J-Y, Parrish CR. 2003. The

natural host range shift and subsequent evolution of canine parvovirus resulted from

virus-specific binding to the canine transferrin receptor. J Virol 77:1718-1726.

23. Parker JS, Murphy WJ, Wang D, O'Brien SJ, Parrish CR. 2001. Canine and feline

parvoviruses can use human or feline transferrin receptors to bind, enter, and infect cells.

J Virol 75:3896-3902.

24. Löfling J, Lyi SM, Parrish CR, Varki A. 2013. Canine and feline parvoviruses

preferentially recognize the non-human cell surface sialic acid N-glycolylneuraminic

acid. Virology 440:89-96.

25. Parrish CR, Carmichael LE. 1983. Antigenic structure and variation of canine parvovirus

type-2, feline panleukopenia virus, and mink enteritis virus. Virology 129:401-414.

26. Parrish CR, Carmichael LE, Antczak DF. 1982. Antigenic relationships between canine

parvovirus type 2, feline panleukopenia virus and mink enteritis virus using conventional

antisera and monoclonal antibodies. Arch Virol 72:267-278.

27. Hafenstein S, Palermo LM, Kostyuchenko VA, Xiao C, Morais MC, Nelson CDS,

Bowman VD, Battisti AJ, Chipman PR, Parrish CR, Rossmann MG. 2007. Asymmetric

binding of transferrin receptor to parvovirus capsids. Proc Natl Acad Sci U S A

104:6585-6589.

28. Hueffer K, Govindasamy L, Agbandje-McKenna M, Parrish CR. 2003. Combinations of

two capsid regions controlling canine host range determine canine transferrin receptor

binding by canine and feline parvoviruses. J Virol 77:10099-10105.

29. Harbison CE, Lyi SM, Weichert WS, Parrish CR. 2009. Early steps in cell infection by

parvoviruses: host-specific differences in cell receptor binding but similar endosomal

trafficking. J Virol 83:10504-10514.

30. Strassheim ML, Gruenberg A, Veijalainen P, Sgro JY, Parrish CR. 1994. Two dominant

neutralizing antigenic determinants of canine parvovirus are found on the threefold spike

of the virus capsid. Virology 198:175-184.

31. Parrish CR, O'Connell PH, Evermann JF, Carmichael LE. 1985. Natural variation of

canine parvovirus. Science (New York, NY) 230:1046-1048.

32. Tresnan DB, Southard L, Weichert W, Sgro JY, Parrish CR. 1995. Analysis of the cell

and erythrocyte binding activities of the dimple and canyon regions of the canine

parvovirus capsid. Virology 211:123-132.

33. Cureton DK, Harbison CE, Cocucci E, Parrish CR, Kirchhausen T. 2012. Limited

transferrin receptor clustering allows rapid diffusion of canine parvovirus into clathrin

endocytic structures. J Virol 86:5330-5340.

34. Simpson AA, Chandrasekar V, Hébert B, Sullivan GM, Rossmann MG, Parrish CR.

2000. Host range and variability of calcium binding by surface loops in the capsids of

canine and feline parvoviruses. J Mol Biol 300:597-610.

35. Nelson CDS, Minkkinen E, Bergkvist M, Hoelzer K, Fisher M, Bothner B, Parrish CR.

2008. Detecting small changes and additional peptides in the canine parvovirus capsid

structure. J Virol 82:10397-10407.

36. Chang SF, Sgro JY, Parrish CR. 1992. Multiple amino acids in the capsid structure of

canine parvovirus coordinately determine the canine host range and specific antigenic

and hemagglutination properties. J Virol 66:6858-6867.

37. Allison AB, Organtini LJ, Zhang S, Hafenstein SL, Holmes EC, Parrish CR. 2016. Single

Mutations in the VP2 300 Loop Region of the Three-Fold Spike of the Carnivore

Parvovirus Capsid Can Determine Host Range. J Virol 90:753-767.

38. Organtini LJ, Lee H, Iketani S, Huang K, Ashley RE, Makhov AM, Conway JF, Parrish

CR, Hafenstein S. 2016. The near-atomic resolution structure of a highly neutralizing Fab

bound to canine parvovirus. J Virol.

39. Parker JS, Parrish CR. 1997. Canine parvovirus host range is determined by the specific

conformation of an additional region of the capsid. J Virol 71:9214-9222.

40. Allison AB, Harbison CE, Pagan I, Stucker KM, Kaelber JT, Brown JD, Ruder MG, Keel

MK, Dubovi EJ, Holmes EC, Parrish CR. 2012. Role of multiple hosts in the cross-

species transmission and emergence of a pandemic parvovirus. J Virol 86:865-872.

41. Hueffer K, Parrish CR. 2003. Parvovirus host range, cell tropism and evolution. Curr

Opin Microbiol 6:392-398.

42. Parrish CR, Have P, Foreyt WJ, Evermann JF, Senda M, Carmichael LE. 1988. The

global spread and replacement of canine parvovirus strains. The Journal of General

Virology 69 ( Pt 5):1111-1116.

43. Lee DW, Allison AB, Bacon KB, Parrish CR, Daniel S. 2016. Single-Particle Tracking

Shows that a Point Mutation in the Carnivore Parvovirus Capsid Switches Binding

between Host-Specific Transferrin Receptors. J Virol 90:4849-4853.

44. Llamas-Saiz AL, Agbandje-McKenna M, Parker JS, Wahid AT, Parrish CR, Rossmann

MG. 1996. Structural analysis of a mutation in canine parvovirus which controls

antigenicity and host range. Virology 225:65-71.

45. Parrish CR, Carmichael LE. 1986. Characterization and recombination mapping of an

antigenic and host range mutation of canine parvovirus. Virology 148:121-132.

46. Hafenstein S, Bowman VD, Sun T, Nelson CDS, Palermo LM, Chipman PR, Battisti AJ,

Parrish CR, Rossmann MG. 2009. Structural comparison of different antibodies

interacting with parvovirus capsids. J Virol 83:5556-5566.

47. Rayaprolu V, Kruse S, Kant R, Venkatakrishnan B, Movahed N, Brooke D, Lins B,

Bennett A, Potter T, McKenna R, Agbandje-McKenna M, Bothner B. 2013. Comparative

analysis of adeno-associated virus capsid stability and dynamics. J Virol 87:13150-

13160.

48. Salganik M, Venkatakrishnan B, Bennett A, Lins B, Yarbrough J, Muzyczka N,

Agbandje-McKenna M, McKenna R. 2012. Evidence for pH-dependent protease activity

in the adeno-associated virus capsid. J Virol 86:11877-11885.

49. Carreira A, Menéndez M, Reguera J, Almendral JM, Mateu MG. 2004. In vitro

disassembly of a parvovirus capsid and effect on capsid stability of heterologous peptide

insertions in surface loops. The Journal of Biological Chemistry 279:6517-6525.

50. Castellanos M, Pérez R, Carrasco C, Hernando-Pérez M, Gómez-Herrero J, de Pablo PJ,

Mateu MG. 2012. Mechanical elasticity as a physical signature of conformational

dynamics in a virus particle. Proc Natl Acad Sci U S A 109:12028-12033.

51. Hernando E, Llamas-Saiz AL, Foces-Foces C, McKenna R, Portman I, Agbandje-

McKenna M, Almendral JM. 2000. Biochemical and physical characterization of

parvovirus minute virus of mice virus-like particles. Virology 267:299-309.

52. Chapman MS, Rossmann MG. 1995. Single-stranded DNA-protein interactions in canine

parvovirus. Structure (London, England: 1993) 3:151-162.

53. Carrasco C, Carreira A, Schaap IaT, Serena PA, Gómez-Herrero J, Mateu MG, de Pablo

PJ. 2006. DNA-mediated anisotropic mechanical reinforcement of a virus. Proc Natl

Acad Sci U S A 103:13706-13711.

54. Cotmore SF, Tattersall P. 2005. Encapsidation of minute virus of mice DNA: aspects of

the translocation mechanism revealed by the structure of partially packaged genomes.

Virology 336:100-112.

55. Farr GA, Cotmore SF, Tattersall P. 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.

56. Parrish CR, Aquadro CF, Strassheim ML, Evermann JF, Sgro JY, Mohammed HO. 1991.

Rapid antigenic-type replacement and DNA sequence evolution of canine parvovirus. J

Virol 65:6544-6552.

57. Jager L, Hausl MA, Rauschhuber C, Wolf NM, Kay MA, Ehrhardt A. 2009. A rapid

protocol for construction and production of high-capacity adenoviral vectors. Nat Protoc

4:547-564.

58. Spouge JL. 1992. Statistical analysis of sparse infection data and its implications for

retroviral treatment trials in primates. Proc Natl Acad Sci U S A 89:7581-7585.

59. Lee DW, Hsu H-L, Bacon KB, Daniel S. 2016. Image Restoration and Analysis of

Influenza Virions Binding to Membrane Receptors Reveal Adhesion-Strengthening

Kinetics. PLoS One 11:e0163437.

60. Ekici OD, Paetzel M, Dalbey RE. 2008. Unconventional serine proteases: variations on

the catalytic Ser/His/Asp triad configuration. Protein Science: A Publication of the

Protein Society 17:2023-2037.

61. Polgár L. 2005. The catalytic triad of serine peptidases. Cellular and molecular life

sciences: CMLS 62:2161-2172.

62. Hueffer K, Palermo LM, Parrish CR. 2004. Parvovirus infection of cells by using variants

of the feline transferrin receptor altering clathrin-mediated endocytosis, membrane

domain localization, and capsid-binding domains. J Virol 78:5601-5611.

63. Goodman LB, Lyi SM, Johnson NC, Cifuente JO, Hafenstein SL, Parrish CR. 2010.

Binding site on the transferrin receptor for the parvovirus capsid and effects of altered

affinity on cell uptake and infection. J Virol 84:4969-4978.

64. Palermo LM, Hafenstein SL, Parrish CR. 2006. Purified feline and canine transferrin

receptors reveal complex interactions with the capsids of canine and feline parvoviruses

that correspond to their host ranges. J Virol 80:8482-8492.

65. Holland JJ, Kiehn ED. 1968. Specific cleavage of viral proteins as steps in the synthesis

and maturation of enteroviruses. Proc Natl Acad Sci U S A 60:1015-1022.

66. Arnold E, Luo M, Vriend G, Rossmann MG, Palmenberg AC, Parks GD, Nicklin MJ,

Wimmer E. 1987. Implications of the picornavirus capsid structure for polyprotein

processing. Proc Natl Acad Sci U S A 84:21-25.

67. Oliveira AC, Gomes AM, Almeida FC, Mohana-Borges R, Valente AP, Reddy VS,

Johnson JE, Silva JL. 2000. Virus maturation targets the protein capsid to concerted

disassembly and unfolding. The Journal of Biological Chemistry 275:16037-16043.

68. Weber J. 1976. Genetic analysis of adenovirus type 2 III. Temperature sensitivity of

processing viral proteins. J Virol 17:462-471.

69. Maroto B, Valle N, Saffrich R, Almendral JM. 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.

CHAPTER THREE

Complex and dynamic interactions between parvovirus capsids, transferrin receptors and

antibodies control cell infection and host range.

Heather M. Callaway, Kathrin Welsch, Wendy Weichert, Andrew B. Allison, Susan L.

Hafenstein, Kai Huang, Sho Iketani, and Colin R. Parrish. 2018. Complex and dynamic interactions between parvovirus capsids, transferrin receptors and antibodies control cell infection and host range. Journal of Virology. In press.

3.1 Abstract

Antibody and receptor binding are key virus/host interactions that control host range and determine the success of infection. Canine and feline parvovirus capsids bind the transferrin receptor type-1 (TfR) to enter host cells, and specific structural interactions appear necessary to prepare the stable capsids for infection. Here we define the details of binding, competition, and occupancy of wild-type and mutant parvovirus capsids with purified receptors and antibodies.

TfR/capsid binding interactions depended on the TfR species and varied widely, with no direct relationship between binding affinity and infection. Capsids bound feline, raccoon, and black- backed jackal TfRs at high affinity, but barely bound canine TfRs, which mediated infection efficiently. TfRs from different species also occupied capsids to different levels, with an estimated 1-2 feline TfRs but 12 black-backed jackal TfRs binding each capsid. Multiple alanine substitutions within loop 1 on the capsid surface reduced TfR binding, but substitutions within loop 3 did not, suggesting that loop 1 directly engaged the TfR and loop 3 sterically affected that interaction. Binding and competition between different TfRs and/or antibodies showed complex relationships. Both antibodies 14 and E competed capsids off TfRs, but antibody E could also compete capsids off itself and antibody 14, likely by inducing capsid structural changes. In some cases, the initial TfR or antibody binding event affected subsequent TfR binding, suggesting that capsid structure changes occur after TfR or antibody binding and may impact infection. This shows that precise, host-specific TfR/capsid interactions, beyond simple attachment, are important for successful infection.

3.2 Importance

Host receptor binding is a key step during viral infection, and may control both infection

and host range. In addition to binding, some viruses require specific interactions with host receptors in order to infect, and anti-capsid antibodies can potentially disrupt these interactions, leading to neutralization. Here, we examine the interactions between parvovirus capsids, the receptors from different hosts, and anti-capsid antibodies. We show that interactions between parvovirus capsids and host-specific TfRs vary in both affinity and in the numbers of receptors bound, with complex effects on infection. In addition, antibodies binding to two sites on the capsids had different effects on TfR/capsid binding. These experiments confirm that receptor and antibody binding to parvovirus capsids are complex processes and the infection outcome is not determined simply by the affinity of attachment.

3.3 Introduction

Interactions of animal viruses with host cell receptors are key events that control cell infection, tissue tropism, and host range. Specific virus-receptor interactions lead to cellular attachment and initiate cell entry and infection, which generally occur through receptor-mediated endocytosis. However, in addition to the tethering process, receptor binding may also result in structural changes to the capsid that are important for allowing binding to secondary receptors or for controlling later intracellular steps in infection (1–5). Variation in different host receptors also creates host range barriers, and virus adaption to those receptors may be associated with host range expansion (6–8). Other important virus-host interactions occur through antibody binding, where host antibodies bind viral proteins and can directly neutralize viral infectivity or protect against infection through a variety of different mechanisms (9–12) . Despite the importance of these different interactions, we still lack a detailed understanding of how receptor or antibody binding dynamics control cell binding and infection, cause disease, or influence transmission to

new hosts.

Canine parvovirus (CPV) and feline panleukopenia virus (FPV) have single-stranded

DNA genomes of about 5,100 bases that encode at least 4 genes, including the major capsid protein VP2, and the minor capsid protein, VP1. The genomes are packaged into small, nonenveloped and stable capsids consisting of 60 subunits, ~90% of which are VP2 and the remaining ~10% of which are VP1. CPV and FPV show natural variations in host range that are primarily controlled by differences in viral capsids and their interactions with the host receptor, the transferrin receptor type-1 (TfR) (13–16). The TfR is the natural cellular receptor for iron- loaded transferrin (Tf) and the hemochromatosis protein (HFE), which regulates transferrin binding and iron uptake in the intestine (17, 18). Once bound to the TfR, diferric Tf or viral capsids are efficiently transported into cells via clathrin-mediated endocytosis (19, 20). While Tf recycles to the cell surface after releasing iron, the virus/TfR complex appears to be retained in the endosomal system for a prolonged period before entering the cytoplasm and releasing viral

DNA into the nucleus (21). Both Tf and HFE bind on the side or underneath the TfR protease- like and helical domains (22, 23), while CPV binds to the apical domain (15, 24), making it unlikely that either Tf or HFE binding to the TfR would affect CPV binding.

Canine parvovirus type-2 (CPV-2) arose in the mid-1970s as a variant of a virus similar to FPV (25, 26). In 1978, CPV spread world-wide, causing a pandemic of disease among dogs, wolves, and coyotes, and during 1979 and 1980 a virus variant, designated CPV-2a and containing 5 mutations on or near the capsid surface, arose and replaced the original CPV-2 strain worldwide (27–29). In previous studies, we have shown that CPV acquired the canine host range as the result of mutations to a small number of residues on the capsid surface (30, 31). Two of those mutations (VP2 residues 93Lys to Asn and 323Asp to Asn) acted together to give CPV

the ability to bind to the canine TfR and infect canine cells and dogs (13, 31), although CPV has been shown to bind to the canine TfR with much lower affinity than to the feline TfR (13, 14).

The canine TfR differs from other host receptors both in amino acid sequence and in having an additional (fourth) glycosylation within the apical domain (attached to Asn384 of the canine TfR sequence), which blocked the binding of FPV-like viruses (15, 32). Other mutations that blocked the infection of canine cells were observed in VP2 surface loop 3, around residue 300 (33, 34).

The CPV-derived viruses in dogs have continued to evolve since the emergence of the CPV-2a common ancestor, and many different studies have reported VP2 sequences from infected dogs or other carnivore hosts from different regions of the world (35–38). Those studies show that a variety of mutations have become widespread in the virus population, including a number of changes in the capsid surface that fall close to or within the predicted receptor binding site or antigenic epitopes.

The CPV-like viruses have broad host ranges, and naturally infect many different animals within the Order Carnivora, with infections and sometimes disease being commonly observed in dogs, mink, raccoons, foxes, domestic cats, large cats, and skunks (26, 37, 39, 40). However, these hosts differ naturally in their susceptibility to infection by FPV, CPV and closely related viruses, and infection of different hosts, including raccoons, mink, foxes, was often associated with selection of additional mutations that alter residues on the capsid protein structure (37, 39).

Similar mutations also arose after passaging viruses in cultured cells from different hosts, showing that the selection occurs at the level of the host cell, most likely the receptor (37).

Those previous studies therefore show that TfR/capsid interactions are highly specific, and that capsid mutations affect the functional interactions between capsids and the TfR. The variant interactions resulting from virus and TfR differences may involve both the affinity of

capsid binding to the TfR, as well as additional structural interactions essential for virus infectivity of cells. This was seen most clearly in studies where we showed that some capsid variants bound efficiently to the TfR and were endocytosed into cells, but did not result in infection (41). However, the details of the TfR/capsid interactions are not yet well understood, including the stoichiometry of receptor binding, the structures involved in attachment, and whether the number of bound receptors is important for successful infection. For CPV and FPV infection, the number of receptors that mediate binding during cell infection appears to be very low. The T=1 parvovirus capsid should display 60 equivalent receptor binding sites, with an estimated maximum occupancy of between 12 and 24 TfRs per capsid after accounting for steric interference between neighboring receptors (42). However, it was seen that single CPV particles bound less than 5 feline TfR molecules on the surface of cells prior to endocytosis and infection

(43), and incubating CPV capsids with purified feline TfR ectodomains resulted in only 1 to 4

TfR being co-purified with each capsid on average (44).

Antibodies bind to a number of overlapping epitopes on parvovirus capsids, and for CPV and FPV, monoclonal antibodies have been classified into two broad groups recognizing two different structures on the capsids (45). Eight different antigen binding fragments (Fabs) derived from monoclonal antibodies were examined for their interactions with the CPV and FPV capsids using cryo-electron microscopy (cryoEM), and their combined footprints covered more than 60% of the viral surface (46). Three of those antibodies recognized the “A” site and bound close to the top of a raised region surrounding the threefold axis of the capsid (the threefold spike), while the other five antibodies tested recognized the “B” site and attached to the side of the threefold spike

(also called the “shoulder”) (46, 47). In contrast to the TfRs, most of the Fabs tested appeared to fully or nearly fully occupy the 60 possible binding sites on the capsids when added in excess

100

(46). However, those Fabs differed significantly in their neutralization efficiencies, with some being poorly neutralizing even when present in excess of the binding sites, and others neutralizing the virus with as few as 5 to 10 Fabs per capsid (48). The reasons for the large differences in neutralization efficiency are not yet fully understood, but the high resolution structure of a complex between the Fab of one highly neutralizing antibody (FabE) and the CPV capsid showed structural variation of non-bound regions of the capsid, suggesting that the Fab triggered allosteric changes that interfered with infection (47).

Other parvoviruses and adeno-associated viruses (AAV) use a variety of different receptors for binding and infection of cells. The AAVs use a protein receptor (the AAV receptor), which they bind through different domains, in addition to a variety of other glycoprotein or glycan receptors, including heparin sulfate proteoglycans and sialic acids (49–

51). Those distinct interactions determine the tissue tropism of the AAVs and the gene delivery vectors derived from them (52). Cell infection and the tissue tropisms of the minute virus of mice

(MVM) and H1 viruses are in some cases determined by variation in the specific interactions of the capsids with the host sialic acids (53, 54). Complex virus/receptor interactions also occur in many other systems. For example, some New World arenaviruses differ in their cell and host tropism due to variation in their interactions with host TfRs (55–57). Picornavirus receptors may trigger structural changes in the capsid after binding, either at high occupancy or when groups of receptors bind to one region of the capsid (58). HIV gp120 also undergoes conformational changes after binding CD4 that reveal the binding site for the second receptor, either CXCR4 or

CCR5 (2, 4, 5). Likewise, antibody binding has been shown to induce changes in the structures of some viral proteins after binding (9, 10, 12, 59).

Here, we further explore this key area in the natural history of animal viruses by

101

examining the specific interactions of the parvovirus capsids and their receptors involved in infection and in determining viral host range. We use a variety of approaches to reveal the dynamic interactions between parvovirus variants and TfRs from different hosts, as well as with antibodies against defined capsid epitopes. These interactions proved to be surprisingly variable and were influenced by mutations in both capsids and TfRs, which altered the binding affinity and occupancy of TfR on the capsids.

3.4 Results

Here, we define the specific interactions of TfRs from various animal hosts with capsids of viruses from different hosts, or with engineered virus-like particles (VLPs), in order to explain how those interactions alter binding and control infection. The receptors examined were derived from cats (Felis catus) (felTfR), dogs (Canis lupus familiaris) (k9TfR), raccoons (Procyon lotor)

(racTfR), black-backed jackals (Canis mesomelas) (bbjTfR), and humans (Homo sapiens)

(huTfR). These receptors have all been shown to have distinct interactions with capsids when expressed on cells (14, 16, 32, 39, 60), including the canine TfR, which acts as a host range barrier to FPV binding and infection in dogs (60). The protein sequences of these TfRs are largely conserved, but loops with low sequence conservation exist in the apical domain, where

CPV has been shown to bind (Figure 3.1A). Antibodies against two different sites on the capsids were also examined, both as Fabs and after cloning and expression as single-chain variable fragments fused to antibody constant fragments (scFv-Fcs). The soluble TfR, an scFv-Fc, and the engineered VLPs are diagrammed in Figure 3.1, along with an outline of how they were used in the binding studies carried out here, and Table 3.1 lists the differences between the different parvoviruses and VLPs used in this study. Figure 3.1 also contains silver stains of proteins to

102

Figure 3.1. Composition of ligands and VLPs used in this study. A) Structure of the human

TfR (PDB ID: 1suv) with residues color coded by sequence conservation among the canine, feline, raccoon, black-backed jackal, and human TfRs. B) Soluble TfR, consisting of the transferrin receptor type 1 ectodomain fused to a 6x His-tag. Helical, apical, and protease-like

103

domains are labeled. C) Silver stain showing purified TfRs. D) scFv-Fc structure, with variable heavy (VH), variable light (VH), heavy constant 2 (CH2), and heavy constant 3 (CH3) domains labeled. E) Silver stain showing purified scFv-Fcs and Fabs. Figures (F-J) show the design of the bio-layer interferometry binding experiments used throughout the study. F) A capsid binding to soluble TfR, which is attached to a Ni-NTA biosensor via a 6x-his tag. G) Soluble TfRs with

His-tags removed are shown binding to capsids that are attached to a Ni-NTA biosensor via a

TfR with an intact His-tag. H) A capsid binding to a scFv-Fc, which is attached to a Protein A biosensor via its Fc domain. I) Fabs are shown binding to a capsid that is attached to a Protein A biosensor via a scFv-Fc. J) Fabs binding to a capsid that is attached to a Ni-NTA biosensor via a

TfR with an intact His-tag. J-N) CPV capsid mutations used in the study. J) A canine parvovirus VP2 (PDB ID: 2CAS) subunit is shown with mutated capsid surface residues in loop

1 (yellow) and loop 3 (red) highlighted. L) Magnification of (H), focusing on residues in loop 3.

M) Magnification of (H), focusing on residues in loop 1. N) An intact CPV-2 capsid, showing the positioning of loop 1 and loop 3 after capsid assembly. O) Differential scanning fluorimetry experiment showing average capsid melting temperature as a measure of stability. Results are from three independent experiments except for N39K VLPs where limited sample quantity permitted only one measurement. Error bars show ± standard error. P) Silver stain showing purified VLPs. Q) Silver stain showing purified empty capsids.

104

Table 3.1. Differences in VP2 protein sequences for viruses and VLPs, relative to CPV-2.

Differences from CPV-2 CPV-2 - R80N, N93K, I101T, A103V, I232V, F303Y, N323D, FPV N375D, S564N, G568A A300D CPV-2 A300D CPV-2a M87L, I101T, A300G, D305Y, N426D N93K CPV-2 VLPs N93K Loop 1 Mutant VLPs N85A, M87A, V92A, N93A Loop 3 Mutant VLPs S293A, Q296A, S297A, T301A, N302A, D305A N85A, M87A, V92A, N93A, S293A, Q296A, S297A, Double Loop Mutant VLPs T301A, N302A, D305A

105

show purity (Figure. 3.1C, E, P, and Q) and a differential scanning fluorimetry assay on engineered VLPs to show that they are intact and at least as stable as wild-type VLPs (Figure.

3.1O).

Capsid/TfR Binding. Binding studies were carried out using bio-layer interferometry, where the ligands (TfRs or scFv-Fcs) were bound to Ni-NTA biosensors (His-tagged TfRs) or

Protein A biosensors (scFv-Fcs) (Figure 3.1F-J). When TfRs from different species were immobilized on Ni-NTA biosensors and incubated with purified capsids, the results depended on the combinations of capsid and TfR examined, though the results of each combination were quite reproducible (Figure 3.2). After attachment, a proportion of the capsids appeared to be loosely bound and were quickly released from the TfRs when washed in buffer, while the remaining capsids stayed attached to receptors for long periods (Figure 3.2). The feline TfR bound to all capsids, as expected from previous studies (13, 14, 41), and the raccoon TfR bound capsids to similar levels (Figure 3.2B and C). For the feline and raccoon TfRs, CPV-2 capsids bound to the highest levels, while FPV and a point mutant of CPV-2 with a VP2 Ala300Asp mutation, which occurs during natural infection of raccoons or (more rarely) during prolonged passage in cultured feline cells (27, 39, 61), bound to lower levels. The more recent strain of CPV (CPV-2a) bound to the lowest level. The canine TfR showed very low levels of binding to all capsids tested in this assay, close to the background (Figure 3.2D), as has been observed previously (13, 14). The black-backed jackal TfR was included in these studies because the black-backed jackal is the closest relative to dogs that lacks the canine TfR’s FPV-blocking glycosylation site (32). The black-backed jackal TfR showed the highest level of binding of any of the TfRs tested, regardless of the capsid form examined (Figure 3.2E). Compared to the feline, raccoon, and black-backed jackal TfRs, the human TfR bound all the capsids tested to very low levels, only

106

Figure 3.2. Binding kinetics of parvovirus capsids to TfRs from different species. Analysis of the binding of capsids (CPV-2, A300D CPV-2, FPV, or CPV-2a) to the TfR ectodomains from different species (cat, dog, raccoon, black-backed jackal, or human). A) A sample bio-layer interferometry experiment. Ni-NTA biosensors were washed (step 1), incubated with TfR ectodomains to 0.8nm of measured binding (step 2), washed again (step 3), incubated with capsids at 240 µg/mL to measure TfR/capsid association (step 4), and washed to measure

TfR/capsid disassociation (step 5). Results are measured as the change in wavelength (in nm) over time. Results of capsids binding to feline TfR (B), raccoon TfR (C), canine TfR (D), black- backed jackal TfR (E), and human TfR (F) are shown. Error bars show ± standard error.

107

slightly higher than the canine TfR (Figure 3.2F).

Infectivity of parvoviruses on TRVb cells expressing different TfRs. TRVb cells, modified Chinese Hamster Ovary (CHO) cells without endogenous transferrin receptor, were transfected with plasmids to stably express the feline, canine, raccoon, or black-backed jackal

TfRs (Figure 3.3A-B) and were then inoculated with parvoviruses. While TfR-free TRVb cells completely resisted infection as expected (16), all of the receptors transfected into TRVb cells added some susceptibility to infection, with the level of infection varying significantly depending on the specific virus and expressed TfR tested (Figure 3.3C-G). There was also no direct connection between the affinity of capsid binding to the TfR (Figure 3.2) and the level of infection. When CPV-2 capsids were incubated with fixed TRVb cells expressing different TfRs

(Figure 3.3H), binding approximated the results observed via bio-layer interferometry.

Previously published studies examining CPV binding to TRVb cells expressing the TfR or canine or feline cell lines showed similar results (13, 14), further validating the results in Figure

3.2.

Antibody and TfR binding kinetics and competition. To better understand the capsid binding of different ligands in these studies, we also compared the binding of two well characterized monoclonal anti-capsid antibodies and the feline and black-backed jackal TfRs.

The antibodies Mab14 and MabE bind to the “A” and “B” sites, respectively, of the capsid surface (46). Mab14 and MabE were either purified from hybridomas and cleaved to form Fabs or expressed as scFv-Fcs in insect cells. The cloned scFv-Fcs have the binding properties of the original antibodies (62), can be modified easily, and bind to Protein A bio-layer interferometry probes with the same high affinity as the human IgG1 Fc.

When scFv-Fc14 or E were attached to biosensors and incubated with CPV-2 capsids,

108

Figure 3.3. Relative infectivity of CPV and FPV on cells transfected to express different

TfRs. TRVb cells (CHO-derived cells that lack endogenous TfR) were transfected with TfR cDNA from different species. A) A sample flow cytometry plot showing expression of TfR on wild-type NLFK cells and transfected TRVb cells. B) Percent of cells positive for TfR expression, averaged over three independent experiments. (C-G) Cells were seeded at 2×104

2 cells/cm in 96-well plates and inoculated with 200 TCID50 units of CPV-2, A300D CPV-2, 109

FPV, or CPV-2a. Cultures were incubated for 48 hours, fixed, and stained with an anti-NS1 antibody to determine how many cells were infected. Infected cells in each well of wild-type

NLFK cells (C), or TRVb cells expressing feline (D), canine (E), raccoon (F), or black-backed jackal TfR (G) were counted as a measurement of infectivity. H) Binding assay showing attachment of 10µg/mL CPV-2 capsids to fixed TRVb cells expressing different TfRs (top) and staining of permeabilized TRVb cells with an anti-TfR antibody to visualize TfR expression

(bottom). Error bars show ± standard error. * p<0.05, ** p<0.01, *** p<0.001, ****p<0.0001.

110

capsids showed high on-rates and very low off-rates after binding to the scFv-Fc (Figure 3.4A).

After binding capsids to scFv-Fcs and washing, we incubated the scFv-Fc/capsid complexes with four concentrations of Fabs derived from the same antibody that had been used to generate the scFv-Fc, allowing us to calculate the affinity of Fab binding. Graphs were globally fit assuming a

-8 4 1:1 Fab:binding site ratio and showed binding affinities of 2.74  10 M (ka=8.99  10 1/M*s;

-3 -8 5 -3 kd=2.47 10 1/s) for Fab 14 and 2.16  10 M for Fab E (ka=2.28  10 1/M*s; kd=4.91  10

1/s) (Figure 3.4B-C). Fab14 fit the 1:1 model well in this assay, but the FabE fit was poorer, likely due to a conformational change it has previously been shown to induce in capsids after binding (47).

We also examined the binding of the different TfRs to CPV-2 capsids. In this study, feline or black-backed jackal TfRs were attached to Ni-NTA biosensors via a His-tag, incubated with CPV-2 capsids, washed, and incubated with four concentrations of soluble TfR of the same type (with cleaved His-tags) (Figure 3.4D). Additional black-backed jackal TfRs bound

-7 efficiently to tethered capsids (Figure 3.4E), yielding a binding affinity of 5.32  10 M (ka=2.16

4 -2  10 1/M*s; kd=1.15 10 1/s). However, the additional feline TfRs showed very low levels of binding (Figure 3.4F), even though the initial affinity of CPV for the His-tagged feline TfR on biosensors was high (Figure 3.2B).

Competition for binding between antibodies or TfRs and antibodies. We used bio- layer interferometry to measure how Fab binding affected capsid attachment to either scFv-Fcs or TfRs. Capsids were attached to biosensors with scFv-Fc14, scFv-FcE, black-backed jackal

TfR, or feline TfR and then incubated with very high concentration (4800nM) Fab14 or E. There was significant variation in the abilities of the two Fabs to compete CPV or FPV capsids off the biosensor when they were tethered through either scFv-Fc14, scFv-FcE, the black-backed jackal 111

Figure 3.4. Binding kinetics of Fabs or TfRs. Level of binding of Fabs or TfRs to CPV-2 empty capsids. A) A sample graph showing binding of multiple Fabs to capsids that had been attached to biosensors via scFv-Fcs derived from the same antibody (e.g. FabE and scFv-FcE;

Fab14 and scFv-Fc14). Protein A probes were washed (step 1), incubated with scFv-Fc (step 2), washed again (step 3), incubated with CPV-2 empty capsids (step 4), and washed (step 5), resulting in capsids attached to probes by a scFv-Fc. scFv-Fc/CPV complexes were then incubated with Fabs (step 6), followed by a wash (step 7). B) Fab14 binding to CPV/scFv-Fc14 complexes, with Fab14 concentrations ranging from 37.5 nM to 300nM. C) FabE binding to

CPV/scFv-FcE complexes, with FabE concentrations ranging from 18.75nM to 150nM. D) A sample graph where CPV-2 capsids are attached to Ni-NTA probes via soluble TfR (steps 1-5), followed by incubation of TfR/CPV complexes with purified TfRs with cleaved His-tags (step 6) and a wash (step 7). E) Black-backed jackal TfR binding to CPV/bbjTfR complexes, with concentrations of purified, His-tag cleaved TfR ranging from 200nM to 1600nM. F) Feline TfR 112

binding to felTfR/CPV complexes, with concentrations of purified, His-tag cleaved TfR ranging from 200nM to 1600nM.

113

TfR, or the feline TfR (Figure 3.5). The species of TfR holding capsids to biosensors (black- backed jackal or feline) and the type of parvovirus tethered to biosensors (CPV or FPV) also affected these results (Figure 3.5E-F and H-I). FabE was particularly efficient in competing capsids off all ligands used to attach it to the biosensor, while Fab14 showed only a mild effect on capsid/TfR binding (Figure 3.5B-I). Fab14 also showed both the lowest competitive effect and level of binding to the FPV capsid (Figure 3.5F and I), as expected since it is largely specific for CPV capsids compared to FPV (45). Overall, the data strongly supports the model where binding of the highly neutralizing FabE induces changes in the capsid that result in capsid detachment from different ligands, including TfRs and other Fabs.

The TfR occupancy on capsids. Saturating capsids with Fabs and/or TfRs also provided information about the level of occupancy of the different TfRs on capsids. TfR binding levels were calibrated by comparison with Fab binding (Figure 3.5B), as Fabs 14 and E are known to fully occupy the capsids when Fab/capsid complexes were examined via cryoEM (46). In this assay, steric blocking by the tethering scFv-Fc and the proximity to the biosensor tip may reduce the number of available binding sites, so we conservatively estimated that around 50 Fabs would be bound to the capsid at saturating levels. We assumed that the 4800nM concentration of Fab14 resulted in complete saturation of available binding sites (~50 sites) on scFv-Fc/CPV-2 complexes in Figure 3.5B, and that the 1600nM incubation of black-backed jackal TfR with black-backed jackal TfR/CPV-2 complexes in Figure 3.4E also resulted in saturation. Comparing the level of Fab14 or black-backed jackal TfR binding to CPV-2, relative to the amount of CPV-

2 bound to biosensors, allowed us to estimate that approximately 12 black-backed jackal TfRs bound to CPV-2, assuming that TfR dimers were complexed with two transferrin proteins, as has been observed by biochemical analysis and by cryo-electron microscopy of CPV complexed with

114

Figure 3.5. Competition between scFv-Fcs and Fabs or TfRs and Fabs for capsid binding.

Capsids were attached to Ni-NTA probes or Protein A probes with TfR or scFv-Fc, respectively.

Attached capsids were then incubated with Fabs at high concentration (4800nM) to determine whether the Fabs could compete capsids off of the ligand that bound them to the probe (scFv-

Fc14, scFv-FcE, felTfR, or bbjTfR). A) A sample graph where capsids are attached to Protein A probes via a scFv-Fc (steps 1-5), incubated with Fabs (step 6) and washed (step 7). B) Fab14 or

FabE incubation with scFv-Fc14/CPV-2 complexes. C) Fab14 or FabE incubation with scFv-

FcE/CPV-2 complexes. D) A sample graph where capsids are attached to Ni-NTA probes via the

115

black-backed jackal TfR (steps 1-5), incubated with Fabs (step 6) and washed (step 7). E) Fab14 or FabE incubated with bbjTfR/CPV-2 complexes. F) Fab14 or FabE incubated with bbjTfR/FPV complexes. G) A sample graph where capsids are attached to probes via the feline

TfR (steps 1-5), incubated with Fabs (step 6) and washed (step 7). H) Fab14 or FabE incubated with felTfR/CPV-2 complexes. I) Fab14 or FabE incubated with felTfR/FPV complexes. Error bars show ± standard error. Statistics were calculated for the 300s and 600s time points using the student’s t-test. * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001.

116

the black-backed jackal TfR (H. Lee and S.L. Hafenstein, unpublished data). Using the same calculations and assuming that 1600nM of feline TfR saturated felTfR/CPV-2 complexes in

Figure 3.4F, we estimated that approximately 1-2 feline TfRs bound to CPV-2.

Influence of initial capsid binding event on attachment of additional TfRs. When felTfR/CPV-2, black-backed jackal TfR/CPV-2, or scFv-Fc14/CPV-2 complexes were incubated with soluble feline or black-backed jackal TfRs, we saw that the initial ligand binding event tethering CPV-2 capsids to biosensors affected subsequent attachment of soluble TfR, even though the same amount of capsid was loaded onto biosensors (Figure 3.6). Black-backed jackal

TfR/CPV-2 complexes bound the highest levels of black-backed jackal or feline TfR, with felTfR/CPV-2 complexes showing slightly lower levels of binding (Figure 3.6B and C).

However, while scFv-Fc14/CPV-2 complexes bound black-backed jackal TfR to similar levels as felTfR/CPV-2 complexes during the association phase, there was a significantly greater drop-off of black-backed jackal TfR during the disassociation phase (Figure 3.6B). Significantly higher drop-off during the disassociation phase also occurred with the feline TfR and scFv-Fc14/CPV-2 complexes (Figure 3.6C).

Capsid residues influencing the TfR binding interaction. To examine the capsid surface for structures involved in the interactions with the TfRs and antibodies 14 or E, we mutated groups of surface-exposed residues in the VP2 structure – four residues in loop 1 (VP2 residues N85A, M87A, V92A, and N93A), six residues in loop 3 (VP2 residues S293A, Q296A,

S297A, T301A, N302A, and D305A), or all 10 mutations together (Figure 3.1J-N and Table

3.1). When CPV-2-derived VLPs containing the four surface residue changes within loop 1 were used in binding studies, we saw greatly reduced binding to the feline TfR (Figure 3.7B). In contrast, CPV-2-derived VLPs with six substitutions in surface loop 3 showed similar or even

117

Figure 3.6. Effects of the initial CPV/ligand interaction on subsequent bbjTfR or felTfR binding. bbjTfR, felTfR, or scFv-Fc14 were used to attach CPV-2 capsids to probes, so that the total amount of CPV attached to the probe, regardless of the ligand attaching it, was equal to

1nm after washing (steps 1-5). Bound capsids were then incubated with 1600nM of bbjTfR or felTfR with cleaved His-tags (step 6) and washed (step 7). A) A sample binding curve, showing capsids loaded to 1nm of binding. B) Incubation of bbjTfR (with cleaved His-tag) with bbjTfR/CPV-2, felTfR/CPV-2, or scFv-Fc14/CPV-2 complexes. C) Incubation of felTfR (with cleaved His-tag) with bbjTfR/CPV-2, felTfR/CPV-2, or scFv-Fc14/CPV-2 complexes. Error bars show ± standard error. Statistics were calculated for the 300s and 600s time points using

ANOVA. * p<0.05, ** p<0.01, *** p<0.001.

118

Figure 3.7. Binding kinetics of mutant VLPs with TfRs, scFv-Fcs, and Fabs. Mutant VLPs were incubated with feline TfR, canine TfR, black-backed jackal TfR, scFv-Fc14, scFv-FcE,

Fab14, and FabE to determine how capsid surface loops affect ligand binding interactions. A) A sample graph where Ni-NTA probes are washed (step 1), incubated with TfR (step 2), washed

(step 3), incubated with VLPs (step 4), and washed again (step 5). B-D) Results of mutant VLPs binding to the feline (B), canine (C), and black-backed jackal (D) TfRs. E) A sample graph where VLPs are attached to Protein A probes via scFv-Fc14 (steps 1-5), incubated with Fab14

(4800nM) (step 6) and washed (step 7). F) The initial binding interaction between scFv-Fc14 and 119

VLPs (steps 4-5). G) Binding of 4800nM Fab14 to scFv-Fc14/VLP complexes (steps 6-7). (H-J)

As in (E-G), but with scFv-FcE replacing scFv-Fc14 and FabE replacing Fab14. Error bars show

± standard error.

120

increased binding to the feline TfR compared to wild type CPV-2 VLPs (Figure 3.7B). Loop 3 mutant VLPs also bound the canine TfR to higher levels than any of the other VLPs (Figure

3.7C), and all VLPs bound strongly to the black-backed jackal TfR (Figure 3.7D). We did not test the loop 3 mutant for infectivity due to concerns about a possible host range expansion of that virus.

scFv-Fc14 and Fab14 bound to all of the mutant capsids except for those with substitutions within its footprint, which included loop 1 mutant VLPs, double loop mutant VLPs, and Asn93Lys VLPs (Figure 3.7F and G). In this experiment, Fab14 binding was more affected by these substitutions than scFv-Fc binding, likely due to the monomeric binding of the Fab forms (Figure 3.7G). The loop 3 mutant VLPs bound to scFv-FcE as well as wild-type VLPs, even though the antibody E binding footprint fell over that group of mutations (Figure 3.7I) (46,

47), and FabE competed all the different VLPs off of scFv-FcE (Fig3. 7J).

3.5 Discussion

These studies further explain the underlying interactions between CPV capsids and different host TfRs that resulted in the emergence and pandemic spread of CPV in dogs during

1978, as well as the subsequent spread and evolution of that virus within dogs and other hosts during the last 40 years. These specific capsid/TfR contacts and interactions are clearly key to the process of cell infection, as they control the virus’s natural host range and also overlap the antibody binding sites on capsids, so that many antibody and receptor interactions are affected by the same mutations (13, 39, 41, 47, 63).

Capsid/TfR Interactions. These studies reveal new molecular details about the interactions between closely related parvovirus capsids and the receptors from susceptible hosts,

121

and in particular show the dynamic nature of these interactions (Figure 3.8). They confirm that there is no direct connection between the affinity of capsid binding to the TfR and the success of cell infection. This was most clearly seen in a comparison of the feline and canine TfRs, which mediate CPV-2a infection to similar levels (Figure 3.3D and E), even though the canine TfR shows extremely low binding affinity (Figure 3.2B and D). This effect has also been previously observed in studies of a CPV-2 mutant with lysine substitutions of VP2 residues 299 and 300, which bound the feline TfR but did not infect cells (41).

We also revealed more details about the binding site on the capsid and the ways in which viral residues control both the specific interactions with the TfR and the infectious process.

Comparing the naturally occurring viruses showed that the capsids of CPV-2 (the first virus strain that spread world-wide in dogs in 1978 (27)) bound to TfR with higher affinity than the

CPV-2a capsids, which had substitutions of VP2 residues 87, 300, and 305 on the capsid surface

(27, 29, 64). CPV-2a arose naturally during 1979, and completely replaced CPV-2 within dogs worldwide by the end of 1980, again showing that the lower receptor binding affinity was widely selected in nature. We were surprised to find that CPV-2 capsids with 6 mutations in VP2 loop 3 bound the feline and other TfRs to similar or greater level than wild-type VLPs (Figure 3.7). This capsid surface loop was previously implicated in TfR binding because growth in different hosts or host cells may readily select for changes in loop 3 residues (including VP2 residues 297, 299,

300, 301, 302) (27–29, 33, 37, 39). However, the strong TfR binding of the loop 3 mutant VLP suggests that most of loop 3 is not directly involved in TfR binding, but it may instead sterically interfere with TfR binding or control the structural transitions required for capsid infectivity. In contrast, loop 1 appears to play a direct role in TfR binding, as mutating 4 residues in that loop greatly reduced attachment, as has been suggested by previous structural and mutational studies

122

Figure 3.8. Summary of major findings. A) Both strong and weak binding interactions between the capsid and the TfR at capsid surface loop 1 result in infection. Antibody binding can disrupt the capsid/TfR binding interaction and block infection. B) Both the black-backed jackal

TfR, which binds to the capsid at high occupancy, and the feline TfR, which binds to the capsid at low occupancy, facilitate successful infection. C) When high concentrations of FabE are incubated with scFv-Fc/capsid complexes, FabE competes with scFv-Fcs for capsid binding and

FabE/capsid complexes are released from biosensors.

123

and by its role in controlling canine host range (13, 31, 65).

The occupancy of TfR on the capsid was another feature of the capsid/receptor interaction revealed here. It was previously reported that CPV capsids engage less than five feline TfRs when they attach to the cell surface (43), and incubating capsids with soluble feline

TfR ectodomains prepared from insect cells showed binding of fewer than 5 TfRs on average

(44). The binding of fewer than 4 feline TfRs per capsid was confirmed using bio-layer interferometry. However, we also showed that receptor occupancy was controlled at least in part by the receptor form tested, as many more black-backed jackal TfRs bound per CPV capsid tethered to biosensors (Figure 3.4). Modeling the maximum occupancy of the TfR on the capsid indicates the maximum is 12 or 24 ectodomain dimers, depending on how the TfR binds to the capsid, due to the large size of the TfR and steric interference between adjacent binding sites. In our study, some binding sites on the capsid would have been blocked by the tethering TfR or scFv-Fc and biosensor, so that the black-backed jackal TfR may be occupying close to the maximum number of sites on the bound capsid. Possible explanations for the low levels of binding of the feline TfR include a limited number of preexisting binding sites on the capsid, or that binding of the feline TfR to one or two positions changes the capsid to prevent the binding of additional TfRs. Here, we show that feline TfR binding may block the attachment of at least some additional TfRs, as fewer black-backed jackal TfRs associated with the capsids that were attached to biosensors via the feline TfR, compared to capsids that were attached via the black- backed jackal TfR (Figure 3.6).

Antibody Binding and Neutralization. We also examined potential mechanisms of neutralization by antibodies, using Fabs derived from whole IgGs and cloned versions expressed as scFv-Fcs. Both antibodies we tested bound to similar high affinities, as expected, with both

124

high ka and low kd rates, and capsids remained attached to scFv-Fcs with little or no detectable release over 5 minutes of washing (Figure 3.4). However, FabE efficiently competed with scFv-

Fcs for capsid binding, resulting in release of FabE/capsid complexes from biosensors and indicating that FabE was actively changing the capsid to cause release from the scFv tethering it to the biosensor. Similarly, both Fabs were able to compete the capsids off the feline and black- backed jackal TfRs, suggesting that receptor detachment is a mechanism of virus neutralization.

Summary. These results provide a clearer view of the complex and dynamic processes of receptor and antibody binding to these superficially simple capsids. In different combinations, the variability of capsid interactions with receptors from different hosts shows that there may be variation in the affinity, occupancy, and infection-mediating interactions, or all of those at once.

This highlights the finely tuned processes that mediate efficient infection and host range variability. For these viruses, variations in these processes control the natural host range, including the changes in the capsid that allowed functional canine TfR binding, which resulted in infection of dog cells and the emergence of CPV as a pandemic virus that continues to circulate today. These types of refined virus/receptor interactions are likely the rule rather than the exception, and a better understanding of the rules that apply in general and in particular cases will allow us to both develop anti-viral strategies and better anticipate the emergence of viruses with new host ranges or pathogenic properties mediated by host receptors.

3.6 Methods

Cells. Norden Laboratory Feline Kidney (NLFK) cells (Norden Laboratories) were derived as a single cell clone of the Crandell Rees Feline Kidney (CRFK) cell line (66), and were received directly from Norden Laboratories in 1980. TRVb cells (obtained from the laboratory of

125

Dr. Timothy McGraw) are Chinese Hamster Ovary (CHO) cells without endogenous TfR (67).

BHK (Baby Hamster Kidney) cells (obtained from the laboratory of Dr. Anne Mason), Sf9

(Spodoptera frugiperda) cells (Invitrogen), and High Five (Trichoplusia ni) cells (Boyce

Thompson Institute, clone BTI-TN-551-4) were also used in this study. NLFK cells were used to prepare infectious viruses and capsids and were grown in 1:1 McCoys/Leibovitz’s L15 media

(Corning) with 5% fetal calf serum (FCS). TRVb cells were grown in Ham’s F12 media

(Corning) with 5% FCS. TRVb cells were transfected with pcDNA3.1(-) derived plasmids containing cDNAs of the feline (Felis catus), canine (Canis lupus familiaris), raccoon (Procyon lotor), or black-backed jackal (Canis mesomelas) TfR using Transit-X2 (Mirus) according to the manufacturer's protocol. Cells were selected for plasmid transfection and TfR expression with

G418 (Corning) and were used to measure infectivity of viruses on different TfRs expressed in the same cell background.

BHK cells were used for production of soluble TfR (described below) and were grown in

DMEM/F-12 (Lonza) with 10% FCS. Sf9 and High Five cells were used for growth of baculovirus stocks and for production of proteins from baculoviruses, respectively. Sf9 cells were grown in Grace's supplemented Insect Media (Gibco) with 10% FCS at 28°C, while High

Five cells were grown in Express5 Serum Free Media (Gibco) at 28°C.

Quantification of TfR expression on transfected TRVb cells. TRVb cells transfected to express TfRs from different species were fixed with 4% paraformaldehyde, permeabilized with PBS containing 0.5% bovine serum albumin and 0.5% Triton X-100, and stained with a mouse anti-TfR cytoplasmic tail antibody (Life Technologies, #QB213080) and a goat anti- mouse Alexa488 conjugated antibody (Life Technologies, #A11029). Cell staining was then quantified using a Guava EasyCyte plus flow cytometer (Millipore). Three biological replicates

126

were performed and the percentage of TfR positive cells was quantified using FlowJo (FlowJo,

LLC, v.7.6.5).

Viruses and Virus-like Particles. CPV-2, A300D CPV-2, CPV-2a (containing all the mutations described in (29, 68) as well as VP2 426Asp), and FPV were each produced from infectious plasmid clones (69, 70). Viruses were either left in tissue culture supernatant for infectivity assays or purified for binding assays using standard methods involving banding on sucrose gradients to separate the empty and full (DNA containing) capsids (30, 42). Titers of virus in supernatant were determined via TCID50 assays on NLFK cells and TRVb cells expressing the feline TfR.

For testing infectivity on TRVb cells, TRVb cells stably expressing the feline, canine, raccoon, or black-backed TfR were seeded at 2  104 cells/cm2 in 96 well plates (0.32 cm2 per well). 200 TCID50 units (measured on TRVb cells expressing the feline TfR) of CPV-2, FPV,

A300D CPV-2, or CPV-2a were added to each TfR-expressing TRVb cell line. After inoculation, cells were incubated for 2 days at 37°C, when cells were fixed and stained with an anti-NS1 monoclonal antibody that recognizes an epitope that is identical on all four parvovirus strains tested (71). Viral growth was determined by counting the number of infected cells per well.

Three independent biological replicates were performed, and the results were analyzed via

ANOVA.

Virus-like particles (VLPs) were prepared by expression of CPV-2 VP2 from baculovirus vectors. Mutant VLPs that contained groups of substitutions within capsid surface loops 1 and 3, which have been associated with the control of TfR binding (13, 31, 33, 37, 65), were also prepared (Figure 3.1J-N). One mutant included six substitutions in VP2 Loop 3 (Ser293Ala,

Gln296Ala, Ser297Ala, Thr301Ala, Asn302Ala, and Asp305Ala) (Figure 3.1L), another

127

included four substitutions in the surface-exposed region of VP2 Loop 1 (Asn85Ala, Met87Ala,

Val92Ala, and Asn93Ala) (Figure 3.1M), and a third variant contained all 10 surface substitutions in both loops 1 and 3. Wild-type VP2, Asn93Lys VP2, and the three mutant forms were prepared as VLPs by expression of VP2 in HighFive cells, and purified as described previously (41).

Expression and purification of TfR ectodomains. A plasmid clone containing the soluble human TfR ectodomain (residues 121 to 760 (the C-terminus)) fused to a 6x-His tag on its N-terminus was obtained from Dr. Ann Mason, University of Vermont (72) (Figure 3.1B).

The plasmid was stably transfected into BHK cells and the TfR ectodomain was secreted into the culture medium. Equivalent clones that expressed the feline, canine, black-backed jackal, and raccoon TfR ectodomains were produced after cloning those sequences in place of the human

TfR. Plasmids were transfected into BHK cells using Transit-X2 (Mirus) according to the manufacturer's protocol and positive cells were selected with 500μg/mL methotrexate. Cells were incubated with Pro-293a chemically defined media (Lonza) with 1mM butyric acid and

2mM L-glutamine in order to produce protein. Supernatants were collected after 2-3 days of culture and stored at 4°C. TfRs were purified by incubating supernatants with Ni-NTA beads

(ThermoFisher) for at least 1 hr and then washing three times with 25mM imidazole. Bound TfR was cleaved on Ni-NTA beads overnight with Factor Xa protease (New England Biolabs) according to the manufacturer's protocol. Protein concentrations were determined with a BCA

Protein Assay Kit (Pierce).

Preparation of single chain fragment variable (scFv)-Fcs and antigen binding fragments (Fabs). The heavy and light variable domains of two previously characterized Mabs, a mouse IgG-2a anti-CPV antibody Mab14 (73), and a rat IgG-2b anti-FPV antibody MabE (45)

128

were joined by a linker sequence of three repeats of (Gly-Gly-Gly-Ser) to prepare a scFv. The baculovirus gp68 signal sequence was added to the N-terminus, the scFv was joined through its

C-terminus to the Fc domain of human IgG1 with a flexible linker containing a thrombin cleavage site, and a 6x-His tag was added at the C-terminus (Figure 3.1D). Clones were expressed in High Five cells using the Bac-to-bac expression system (Invitrogen) according to the manufacturer’s protocol and protein was collected after 2-3 days of incubation of baculovirus with High Five cells at 28°C. The scFv-Fc was isolated from supernatant with Protein G affinity chromatography, eluted at pH 3.0, and immediately neutralized with Tris-HCl to pH 7.0. Eluted fractions were pooled and buffer exchanged into PBS with a 10kDa Amicon Ultra Centrifugal

Filter (Millipore).

Fabs were produced by digesting purified MabE or Mab14 with immobilized papain beads (ThermoFisher) according to the manufacturer’s instructions. The antibody constant region

(Fc) was separated from antibody 14 Fabs by incubating cleaved Mab14 with Protein A CL-4B

Sepharose beads (Pharmacia Biotech) in 50mM Tris, pH 7.0. The MabE Fc was removed by running cleaved MabE through a DEAE-sephadex A25 (Sigma) column in 0.01M phosphate buffer, pH 7.8 and collecting Fabs in the column flow-through. Incompletely cleaved proteins were removed via size exclusion chromatography on a Sephadex G100 (Pharmacia) column in

PBS. Fabs were then concentrated using an Amicon 10kDa centrifugal filter (Millipore).

Determination of protein purity and VLP stability. Protein purity for TfRs, scFv-Fcs,

Fabs, VLPs, and parvovirus empty capsids was determined by silver staining samples run on

10% SDS-PAGE gels with a Pierce Silver Stain Kit (Thermo Scientific) according to the manufacturer’s protocol. Stability of mutant VLPs was determined by differential scanning fluorimetry of purified VLPs, as previously described (41). Three replicates were performed for

129

each sample except for the N93K VLPs, where limited sample quantity allowed only one measurement.

Measurement of TfR and scFv binding using bio-layer interferometry. To examine the specific interactions involved in antibody and TfR binding, we bound the capsids and ligands in different combinations, as diagrammed in Figure 3.1F-J and in figures throughout the paper.

Binding kinetics and dynamics were measured using a BLItz bio-layer interferometer (ForteBio).

Ni-NTA biosensors were used to bind TfRs and Protein A biosensors were used to bind scFv-

Fcs. To determine the affinities of capsids and TfRs or capsids and scFv-Fcs, biosensors were first blocked and hydrated in kinetics buffer (PBS with 0.02% ovalbumin and 0.02% Tween-20).

Basic binding experiments to determine capsid/TfR or capsid/scFv-Fc affinity (Figure 3.1F and

H) were performed as follows: 30s baseline in kinetics buffer, 300s of loading TfR or scFv-Fc,

60s wash in kinetics buffer, then 300s association with capsids (240μg/mL) in PBS, and 300s disassociation in kinetics buffer. TfRs in supernatant (with uncleaved His-tags) were loaded onto

Ni-NTA biosensors to a consistent binding level of 0.8 nm, while purified scFv-Fcs were loaded to approximately 0.4 nm (0.83 μg/mL of protein).

Comparison of TfR affinity between different CPV strains or CPV-2 VLPs. We compared the relative binding affinities of feline, canine, raccoon, black-backed jackal, and human TfRs by loading TfR in supernatant to 0.8nm of binding, as described above, followed by incubation with CPV-2, CPV-2a, FPV, or A300D CPV-2 capsids or CPV-2 VLPs at 240μg/mL.

Three independent replicates were performed for each TfR/capsid combination.

Binding of CPV-2 to transfected TRVb cells. Wild-type TRVb cells and TRVb cells transfected to express different TfRs were seeded on glass cover slips at a density of 2 × 104 cells/cm2 and grown overnight. Cells were incubated in DMEM with 0.1% bovine serum

130

albumin (BSA) for 30 min at 37°C, then washed with PBS and fixed in 4% paraformaldehyde for

10 min. Cells were then incubated with 10µg/mL CPV-2 empty capsids diluted in DMEM with

0.1% BSA for 1 hr at 37°C, stained with anti-capsid monoclonal antibody 8, which does not detach capsids from fixed cells (14, 46, 73), and mounted on slides using ProLong Gold antifade reagent with DAPI (Invitrogen). Additional cover slips were permeabilized in PBS with 0.5%

BSA and 0.5% Triton X-100 after fixation and stained with a mouse anti-TfR cytoplasmic tail antibody (Life Technologies, #QB213080) and a goat anti-mouse Alexa594 conjugated antibody

(Life Technologies, #A11005).

Determination of Fab/CPV and TfR/CPV affinity constants and approximation of number of TfRs bound per capsid. Affinity constants for Fab/CPV and TfR/CPV (cleaved His- tag) interactions were determined through multi-step binding experiments (Figure 3.1G and I), where CPV capsids were first immobilized on biosensors with either scFv-Fcs or TfRs, exactly as in the basic binding experiments, described above and in Figs. 3.1G, 3.1I, and 3.2A. These complexes then underwent two additional steps to give 1:1 capsid/ligand binding interactions:

300s incubation with Fabs in PBS or His-tag-cleaved TfRs, followed by 300s disassociation in kinetics buffer. Four different concentrations of Fab or TfR were used to generate four graphs per combination, which were then fitted using 1:1 global fitting in the BLItz Pro Software

(ForteBio). Experiments were performed in duplicate, and one replicate was shown and analyzed.

The number of copies of TfR that bound to each capsid was approximated by comparing binding of TfR to TfR/CPV complexes with binding of Fab14 to scFv-Fc14/CPV complexes, under conditions where ligand binding was at or near saturation (1600 nM of TfR at 300s (Figure

3.4E and F) and 4800 nM of Fab14 at 300s (Figure 3.5B)). Levels of ligand binding (His-tag-

131

cleaved TfR or Fab14) were normalized relative to the loading of CPV onto biosensors and adjusted to account for the difference in molecular weight between the TfR and Fab14. We conservatively estimated that Fab14 saturated our system at 50 Fabs/capsid due to steric hindrance of scFv-Fc14 holding capsids to biosensors and multiplied 50 Fabs/capsid by the normalized and adjusted ratio of TfR vs. Fab14 binding to capsids to make our estimate of TfR copies bound per capsid.

Competition between scFv-Fcs and Fabs or TfRs and Fabs for capsid binding. To determine whether or not antibodies could affect capsid interactions with the TfR or other antibodies, we designed multi-step binding experiments where capsids were attached to biosensors with one ligand, and then incubated with either the same ligand or a different ligand. scFv-Fc14, scFv-FcE, feline TfR, or black-backed jackal TfR was used to attach CPV-2 capsids to biosensors, as described above. After a 300s wash step, complexes were incubated with either

4800 nM of FabE or Fab14 for 300s, followed by an additional 300s wash step. Three independent replicates were performed and binding at 300s and 600s was analyzed via ANOVA.

Effects of the initial CPV/ligand binding interaction on subsequent TfR binding. To examine whether the ligand holding CPV to biosensors affected subsequent binding of TfR, we bound scFc-Fc14, feline TfR, or black-backed jackal TfR to biosensors, as described for basic binding experiments, above. The concentrations of these ligands were adjusted so that all three ligands bound 1nm of CPV-2 capsids to biosensors (Figure 3.6A). Ligand/CPV complexes were then incubated with either black-backed jackal TfR (cleaved His-tag, 1600 nM) or feline TfR

(cleaved His-tag, 1600 nM) and washed in kinetics buffer for 300s. Three independent replicates were performed and binding at 300s and 600s was analyzed via ANOVA.

132

3.7 Acknowledgements

This work was supported by NIH grant R01 AI092571 to CRP and SLH. HMC was supported by NSF-GRFP DGE-1650441. The funders had no role in study design, data collection and interpretation, or decision to submit the work for publication.

133

REFERENCES

1. Bönsch C, Zuercher C, Lieby P, Kempf C, Ros C. 2010. The globoside receptor triggers

structural changes in the B19 virus capsid that facilitate virus internalization. J Virol

84:11737–11746.

2. Chen B, Vogan EM, Gong H, Skehel JJ, Wiley DC, Harrison SC. 2005. Structure of an

unliganded simian immunodeficiency virus gp120 core. Nature 433:834–841.

3. Kirschner AN, Sorem J, Longnecker R, Jardetzky TS. 2009. Structure of Epstein-Barr

virus glycoprotein 42 suggests a mechanism for triggering receptor-activated virus entry.

Struct Lond Engl 1993 17:223–233.

4. Sattentau QJ, Moore JP. 1991. Conformational changes induced in the human

immunodeficiency virus envelope glycoprotein by soluble CD4 binding. J Exp Med

174:407–415.

5. Zhang W, Canziani G, Plugariu C, Wyatt R, Sodroski J, Sweet R, Kwong P, Hendrickson

W, Chaiken I. 1999. Conformational changes of gp120 in epitopes near the CCR5 binding

site are induced by CD4 and a CD4 miniprotein mimetic. Biochemistry (Mosc) 38:9405–

9416.

6. Li F. 2013. Receptor recognition and cross-species infections of SARS coronavirus.

Antiviral Res 100:246–254.

7. Ohishi K, Suzuki R, Maeda T, Tsuda M, Abe E, Yoshida T, Endo Y, Okamura M,

Nagamine T, Yamamoto H, Ueda M, Maruyama T. 2014. Recent host range expansion of

canine distemper virus and variation in its receptor, the signaling lymphocyte activation

molecule, in carnivores. J Wildl Dis 50:596–606.

8. Xiong X, McCauley JW, Steinhauer DA. 2014. Receptor binding properties of the

134

influenza virus hemagglutinin as a determinant of host range. Curr Top Microbiol

Immunol 385:63–91.

9. Hernandez R, Paredes A, Brown DT. 2008. Sindbis virus conformational changes induced

by a neutralizing anti-E1 monoclonal antibody. J Virol 82:5750–5760.

10. Lee H, Brendle SA, Bywaters SM, Guan J, Ashley RE, Yoder JD, Makhov AM, Conway

JF, Christensen ND, Hafenstein S. 2015. A cryo-electron microscopy study identifies the

complete H16.V5 epitope and reveals global conformational changes initiated by binding

of the neutralizing antibody fragment. J Virol 89:1428–1438.

11. Reading SA, Dimmock NJ. 2007. Neutralization of animal virus infectivity by antibody.

Arch Virol 152:1047–1059.

12. Wien MW, Filman DJ, Stura EA, Guillot S, Delpeyroux F, Crainic R, Hogle JM. 1995.

Structure of the complex between the Fab fragment of a neutralizing antibody for type 1

poliovirus and its viral epitope. Nat Struct Biol 2:232–243.

13. Hueffer K, Govindasamy L, Agbandje-McKenna M, Parrish CR. 2003. Combinations of

two capsid regions controlling canine host range determine canine transferrin receptor

binding by canine and feline parvoviruses. J Virol 77:10099–10105.

14. Palermo LM, Hafenstein SL, Parrish CR. 2006. Purified feline and canine transferrin

receptors reveal complex interactions with the capsids of canine and feline parvoviruses

that correspond to their host ranges. J Virol 80:8482–8492.

15. Palermo LM, Hueffer K, Parrish CR. 2003. Residues in the apical domain of the feline and

canine transferrin receptors control host-specific binding and cell infection of canine and

feline parvoviruses. J Virol 77:8915–8923.

16. Parker JS, Murphy WJ, Wang D, O’Brien SJ, Parrish CR. 2001. Canine and feline

135

parvoviruses can use human or feline transferrin receptors to bind, enter, and infect cells. J

Virol 75:3896–3902.

17. Mason AB, Byrne SL, Everse SJ, Roberts SE, Chasteen ND, Smith VC, MacGillivray

RTA, Kandemir B, Bou-Abdallah F. 2009. A loop in the N-lobe of human serum

transferrin is critical for binding to the transferrin receptor as revealed by mutagenesis,

isothermal titration calorimetry, and epitope mapping. J Mol Recognit JMR 22:521–529.

18. Trinder D, Olynyk JK, Sly WS, Morgan EH. 2002. Iron uptake from plasma transferrin by

the duodenum is impaired in the Hfe knockout mouse. Proc Natl Acad Sci U S A

99:5622–5626.

19. Parker JS, Parrish CR. 2000. Cellular uptake and infection by canine parvovirus involves

rapid dynamin-regulated clathrin-mediated endocytosis, followed by slower intracellular

trafficking. J Virol 74:1919–1930.

20. Zhao N, Enns CA. 2012. Iron transport machinery of human cells: players and their

interactions. Curr Top Membr 69:67–93.

21. Harbison CE, Lyi SM, Weichert WS, Parrish CR. 2009. Early steps in cell infection by

parvoviruses: host-specific differences in cell receptor binding but similar endosomal

trafficking. J Virol 83:10504–10514.

22. Bennett MJ, Lebrón JA, Bjorkman PJ. 2000. Crystal structure of the hereditary

haemochromatosis protein HFE complexed with transferrin receptor. Nature 403:46–53.

23. Cheng Y, Zak O, Aisen P, Harrison SC, Walz T. 2004. Structure of the human transferrin

receptor-transferrin complex. Cell 116:565–576.

24. Goodman LB, Lyi SM, Johnson NC, Cifuente JO, Hafenstein SL, Parrish CR. 2010.

Binding site on the transferrin receptor for the parvovirus capsid and effects of altered

136

affinity on cell uptake and infection. J Virol 84:4969–4978.

25. Parrish CR, Kawaoka Y. 2005. The origins of new pandemic viruses: the acquisition of

new host ranges by canine parvovirus and influenza A viruses. Annu Rev Microbiol

59:553–586.

26. Truyen U, Gruenberg A, Chang SF, Obermaier B, Veijalainen P, Parrish CR. 1995.

Evolution of the feline-subgroup parvoviruses and the control of canine host range in vivo.

J Virol 69:4702–4710.

27. Parrish CR, Have P, Foreyt WJ, Evermann JF, Senda M, Carmichael LE. 1988. The global

spread and replacement of canine parvovirus strains. J Gen Virol 69 ( Pt 5):1111–1116.

28. Parrish CR, O’Connell PH, Evermann JF, Carmichael LE. 1985. Natural variation of

canine parvovirus. Science 230:1046–1048.

29. Stucker KM, Pagan I, Cifuente JO, Kaelber JT, Lillie TD, Hafenstein S, Holmes EC,

Parrish CR. 2012. The role of evolutionary intermediates in the host adaptation of canine

parvovirus. J Virol 86:1514–1521.

30. Agbandje M, McKenna R, Rossmann MG, Strassheim ML, Parrish CR. 1993. Structure

determination of feline panleukopenia virus empty particles. Proteins 16:155–171.

31. Chang SF, Sgro JY, Parrish CR. 1992. Multiple amino acids in the capsid structure of

canine parvovirus coordinately determine the canine host range and specific antigenic and

hemagglutination properties. J Virol 66:6858–6867.

32. Kaelber JT, Demogines A, Harbison CE, Allison AB, Goodman LB, Ortega AN, Sawyer

SL, Parrish CR. 2012. Evolutionary reconstructions of the transferrin receptor of

Caniforms supports canine parvovirus being a re-emerged and not a novel pathogen in

dogs. PLoS Pathog 8:e1002666.

137

33. Allison AB, Organtini LJ, Zhang S, Hafenstein SL, Holmes EC, Parrish CR. 2016. Single

Mutations in the VP2 300 Loop Region of the Three-Fold Spike of the Carnivore

Parvovirus Capsid Can Determine Host Range. J Virol 90:753–767.

34. Llamas-Saiz AL, Agbandje-McKenna M, Parker JS, Wahid AT, Parrish CR, Rossmann

MG. 1996. Structural analysis of a mutation in canine parvovirus which controls

antigenicity and host range. Virology 225:65–71.

35. Zhou P, Zeng W, Zhang X, Li S. 2017. The genetic evolution of canine parvovirus - A

new perspective. PloS One 12:e0175035.

36. Miranda C, Thompson G. 2016. Canine parvovirus: the worldwide occurrence of antigenic

variants. J Gen Virol 97:2043–2057.

37. Allison AB, Kohler DJ, Ortega A, Hoover EA, Grove DM, Holmes EC, Parrish CR. 2014.

Host-specific parvovirus evolution in nature is recapitulated by in vitro adaptation to

different carnivore species. PLoS Pathog 10:e1004475.

38. Maya L, Calleros L, Francia L, Hernández M, Iraola G, Panzera Y, Sosa K, Pérez R. 2013.

Phylodynamics analysis of canine parvovirus in Uruguay: evidence of two successive

invasions by different variants. Arch Virol 158:1133–1141.

39. Allison AB, Harbison CE, Pagan I, Stucker KM, Kaelber JT, Brown JD, Ruder MG, Keel

MK, Dubovi EJ, Holmes EC, Parrish CR. 2012. Role of multiple hosts in the cross-species

transmission and emergence of a pandemic parvovirus. J Virol 86:865–872.

40. Steinel A, Parrish CR, Bloom ME, Truyen U. 2001. Parvovirus infections in wild

carnivores. J Wildl Dis 37:594–607.

41. Callaway HM, Feng KH, Lee DW, Allison AB, Pinard M, McKenna R, Agbandje-

McKenna M, Hafenstein S, Parrish CR. 2017. Parvovirus Capsid Structures Required for

138

Infection: Mutations Controlling Receptor Recognition and Protease Cleavages. J Virol

91.

42. Nelson CDS, Minkkinen E, Bergkvist M, Hoelzer K, Fisher M, Bothner B, Parrish CR.

2008. Detecting small changes and additional peptides in the canine parvovirus capsid

structure. J Virol 82:10397–10407.

43. Cureton DK, Harbison CE, Cocucci E, Parrish CR, Kirchhausen T. 2012. Limited

transferrin receptor clustering allows rapid diffusion of canine parvovirus into clathrin

endocytic structures. J Virol 86:5330–5340.

44. Hafenstein S, Palermo LM, Kostyuchenko VA, Xiao C, Morais MC, Nelson CDS,

Bowman VD, Battisti AJ, Chipman PR, Parrish CR, Rossmann MG. 2007. Asymmetric

binding of transferrin receptor to parvovirus capsids. Proc Natl Acad Sci U S A 104:6585–

6589.

45. Parrish CR, Carmichael LE. 1983. Antigenic structure and variation of canine parvovirus

type-2, feline panleukopenia virus, and mink enteritis virus. Virology 129:401–414.

46. Hafenstein S, Bowman VD, Sun T, Nelson CDS, Palermo LM, Chipman PR, Battisti AJ,

Parrish CR, Rossmann MG. 2009. Structural comparison of different antibodies

interacting with parvovirus capsids. J Virol 83:5556–5566.

47. Organtini LJ, Lee H, Iketani S, Huang K, Ashley RE, Makhov AM, Conway JF, Parrish

CR, Hafenstein S. 2016. Near-Atomic Resolution Structure of a Highly Neutralizing Fab

Bound to Canine Parvovirus. J Virol 90:9733–9742.

48. Nelson CDS, Palermo LM, Hafenstein SL, Parrish CR. 2007. Different mechanisms of

antibody-mediated neutralization of parvoviruses revealed using the Fab fragments of

monoclonal antibodies. Virology 361:283–293.

139

49. Agbandje-McKenna M, Kleinschmidt J. 2011. AAV capsid structure and cell interactions.

Methods Mol Biol Clifton NJ 807:47–92.

50. Pillay S, Meyer NL, Puschnik AS, Davulcu O, Diep J, Ishikawa Y, Jae LT, Wosen JE,

Nagamine CM, Chapman MS, Carette JE. 2016. An essential receptor for adeno-

associated virus infection. Nature 530:108–112.

51. Pillay S, Zou W, Cheng F, Puschnik AS, Meyer NL, Ganaie SS, Deng X, Wosen JE,

Davulcu O, Yan Z, Engelhardt JF, Brown KE, Chapman MS, Qiu J, Carette JE. 2017.

AAV serotypes have distinctive interactions with domains of the cellular receptor AAVR.

J Virol.

52. Srivastava A. 2016. In vivo tissue-tropism of adeno-associated viral vectors. Curr Opin

Virol 21:75–80.

53. López-Bueno A, Rubio M-P, Bryant N, McKenna R, Agbandje-McKenna M, Almendral

JM. 2006. Host-selected amino acid changes at the sialic acid binding pocket of the

parvovirus capsid modulate cell binding affinity and determine virulence. J Virol

80:1563–1573.

54. Pittman N, Misseldine A, Geilen L, Halder S, Smith JK, Kurian J, Chipman P, Janssen M,

Mckenna R, Baker TS, D’Abramo A, Cotmore S, Tattersall P, Agbandje-McKenna M.

2017. Atomic Resolution Structure of the Oncolytic Parvovirus LuIII by Electron

Microscopy and 3D Image Reconstruction. Viruses 9.

55. Abraham J, Kwong JA, Albariño CG, Lu JG, Radoshitzky SR, Salazar-Bravo J, Farzan M,

Spiropoulou CF, Choe H. 2009. Host-species transferrin receptor 1 orthologs are cellular

receptors for nonpathogenic new world clade B arenaviruses. PLoS Pathog 5:e1000358.

56. Radoshitzky SR, Abraham J, Spiropoulou CF, Kuhn JH, Nguyen D, Li W, Nagel J,

140

Schmidt PJ, Nunberg JH, Andrews NC, Farzan M, Choe H. 2007. Transferrin receptor 1 is

a cellular receptor for New World haemorrhagic fever arenaviruses. Nature 446:92–96.

57. Sarute N, Ross SR. 2017. New World Arenavirus Biology. Annu Rev Virol 4:141–158.

58. Lee H, Shingler KL, Organtini LJ, Ashley RE, Makhov AM, Conway JF, Hafenstein S.

2016. The novel asymmetric entry intermediate of a picornavirus captured with nanodiscs.

Sci Adv 2:e1501929.

59. Lazear E, Whitbeck JC, Ponce-de-Leon M, Cairns TM, Willis SH, Zuo Y, Krummenacher

C, Cohen GH, Eisenberg RJ. 2012. Antibody-induced conformational changes in herpes

simplex virus glycoprotein gD reveal new targets for virus neutralization. J Virol

86:1563–1576.

60. Hueffer K, Parker JSL, Weichert WS, Geisel RE, Sgro J-Y, Parrish CR. 2003. The natural

host range shift and subsequent evolution of canine parvovirus resulted from virus-specific

binding to the canine transferrin receptor. J Virol 77:1718–1726.

61. Hueffer K, Parrish CR. 2003. Parvovirus host range, cell tropism and evolution. Curr Opin

Microbiol 6:392–398.

62. Yuan W, Parrish CR. 2000. Comparison of two single-chain antibodies that neutralize

canine parvovirus: analysis of an antibody-combining site and mechanisms of

neutralization. Virology 269:471–480.

63. Strassheim ML, Gruenberg A, Veijalainen P, Sgro JY, Parrish CR. 1994. Two dominant

neutralizing antigenic determinants of canine parvovirus are found on the threefold spike

of the virus capsid. Virology 198:175–184.

64. Parrish CR, Aquadro CF, Strassheim ML, Evermann JF, Sgro JY, Mohammed HO. 1991.

Rapid antigenic-type replacement and DNA sequence evolution of canine parvovirus. J

141

Virol 65:6544–6552.

65. Govindasamy L, Hueffer K, Parrish CR, Agbandje-McKenna M. 2003. Structures of host

range-controlling regions of the capsids of canine and feline parvoviruses and mutants. J

Virol 77:12211–12221.

66. Crandell RA, Fabricant CG, Nelson-Rees WA. 1973. Development, characterization, and

viral susceptibility of a feline (Felis catus) renal cell line (CRFK). In Vitro 9:176–185.

67. McGraw TE, Greenfield L, Maxfield FR. 1987. Functional expression of the human

transferrin receptor cDNA in Chinese hamster ovary cells deficient in endogenous

transferrin receptor. J Cell Biol 105:207–214.

68. Parrish CR, Aquadro CF, Carmichael LE. 1988. Canine host range and a specific epitope

map along with variant sequences in the capsid protein gene of canine parvovirus and

related feline, mink, and raccoon parvoviruses. Virology 166:293–307.

69. Parker JS, Parrish CR. 1997. Canine parvovirus host range is determined by the specific

conformation of an additional region of the capsid. J Virol 71:9214–9222.

70. Parrish CR. 1991. Mapping specific functions in the capsid structure of canine parvovirus

and feline panleukopenia virus using infectious plasmid clones. Virology 183:195–205.

71. Yeung DE, Brown GW, Tam P, Russnak RH, Wilson G, Clark-Lewis I, Astell CR. 1991.

Monoclonal antibodies to the major nonstructural nuclear protein of minute virus of mice.

Virology 181:35–45.

72. Byrne SL, Leverence R, Klein JS, Giannetti AM, Smith VC, MacGillivray RTA,

Kaltashov IA, Mason AB. 2006. Effect of glycosylation on the function of a soluble,

recombinant form of the transferrin receptor. Biochemistry (Mosc) 45:6663–6673.

73. Parrish CR, Carmichael LE, Antczak DF. 1982. Antigenic relationships between canine

142

parvovirus type 2, feline panleukopenia virus and mink enteritis virus using conventional antisera and monoclonal antibodies. Arch Virol 72:267–278.

143

CHAPTER FOUR

Examination and reconstruction of ancient, endogenous parvovirus capsid proteins in

rodent genomes

Heather M. Callaway, Christian Urbina, Karen Barnard, Robert Dick, Robert J. Gifford, Colin R.

Parrish. Examination and reconstruction of ancient, endogenous parvovirus capsid proteins in rodent genomes. Manuscript in preparation.

144

4.1 Abstract

Parvovirus-derived endogenous viral elements (EVEs) have been found in the genomes of many different animal species. While some viral integration events appear to have occurred more than 50 million years ago, others likely occurred much more recently. Here, we further investigate the properties of integrated viral sequences derived from autonomous parvoviruses and describe their relationships to contemporary viruses. We did not find any intact capsid protein open reading frames in the integrated viral sequences, but were able to identify some that could be repaired to form full-length sequences with relatively few changes. These sequences were found in the genomes of Rattus norvegicus (brown rat), Mus spretus (Algerian mouse), and

Apodemus sylvaticus (wood mouse). The sequence in the genome of R. norvegicus was also in the R. rattus, R. tanezumi, R. exulans, R. everetti genomes, indicating that it was over 2.3 million years old, but the M. spretus and A. sylvaticus EVEs were not found in the published genomes of other mouse species, indicating more recent insertions. Repaired VP2 sequences from the three different species were expressed in mammalian and insect cells. The M. spretus VP2 sequence assembled into capsids, which had high thermal stability, bound N-acetyl neuraminic acid

(Neu5Ac), and entered murine L cells. While the repaired VP2 sequences from R. norvegicus and A. sylvaticus did not assemble, chimeric sequences that combining capsid surface loops from

R. norvegicus with the canine parvovirus VP2 gene did, allowing some of that EVE’s structures and functions to be examined.

4.2 Importance

Parvovirus EVEs incorporated into the genomes of different animals represent ancient viruses that infected the ancestors of those animals millions of years ago, but we know little

145

about their properties. By expressing the capsid proteins of parvovirus EVEs found in the genomes of three different rodents, we can examine their structures and functions, including ligand binding, what cells and host species they infected, and whether antibodies generated against modern parvoviruses can recognize EVE-derived capsid proteins or capsids. Here, we show that one of the EVE VP2 sequences can assemble into capsids that have biophysical and sialic acid binding properties similar to extant parvoviruses. Chimeras of a contemporary virus and an EVE that failed to form capsids allow us to examine the properties of its different surface loops and explore why the repaired sequence failed to assemble into capsids.

4.3 Introduction

Endogenous viral elements (EVEs) are generated when viral nucleic acid is incorporated into the genome of a host organism. EVEs range in size from small gene fragments to whole viral genomes, and they serve as molecular ‘fossils’ if integrated into germ line cell DNA and vertically transmitted from parent to offspring. EVEs are derived from many different classes of viruses, particularly retroviruses, which encode reverse transcriptase and integrase proteins that facilitate their integration into host DNA. Of the non-retrovirus derived EVEs, parvoviruses are relatively highly represented, and likely derive from non-homologous recombination, possibly facilitated by their ssDNA nickase activity (1).

The Parvoviridae comprise a family of small, single stranded DNA (ssDNA) viruses with genomes of around 5,000 nucleotides. They infect many different animals, ranging from invertebrates to mammals, causing disease in many hosts and likely also causing subclinical infections in many others (2,3). The different subfamilies of Parvoviridae include the parvoviruses (Parvovirinae) and the densoviruses (Densovirinae). The genomes of the

146

parvoviruses contain two large genes and a variety of smaller genes expressed through alternative splicing or through the use of small transcription units. One large gene encodes the non-structural protein (called non-structural protein 1 (NS1) or replication protein (Rep)), which becomes covalently attached to the 5’ end of linear, viral DNA and performs a number of functions. Those functions include acting as a helicase and a site-specific DNA nickase, as well as controlling gene expression from the viral and other promoters. Smaller versions of NS1 or

Rep are formed through internal initiation of translation or alternative splicing of the viral messages, creating proteins that may perform a subset of the functions of the large protein, or which provide additional functions such as assisting in viral capsid transport within the cells (4–

7). The capsid proteins, designated virus protein (VP) or capsid protein (Cap) depending on the virus, are encoded by the other large open reading frame. There are a number of variants, including a smaller version of the protein (VP2 in the autonomous parvoviruses) between 60 to

70 kDa which self-assembles into a stable T=1 capsid (8–10). The full-length versions of the capsid proteins (VP1) have the same sequence as the VP2 protein, but have additional sequences on the N-terminus. VP1 is incorporated into the assembled capsids in varying but low amounts

(around 10%) in a manner that is structurally similar to VP2. The N-terminal unique region of

VP1 provides functions required for infection, including a phospholipase A2 enzyme activity and a nuclear import motif (11,12). The non-coding regions of the viruses contain promoters for expression of one or more of the viral transcripts, polyadenylation sites, transcriptional regulatory sequences, sites involved in DNA replication, and the viral genomic termini, which contain terminal repeats that form hairpin or cruciform structures and allow the replication of the viral DNA by host cell DNA polymerases (13–16).

Here, we investigate the properties of endogenous parvovirus sequences integrated into

147

the genomes of many different animals. These sequences generally appear to be ancient, with some being dated to 50 million years old or older based on their orthologous positions in the genomes of divergent species (1,17–19). In other cases, the sequences have thus far only been found in the genomes of single animal species, suggesting either that they were integrated after the divergence of the last common ancestral animal for which sequences are available, or that the integrated copies were deleted from other genomes (18,19). Parvovirus EVEs most likely derive from non-homologous recombination of ancient parvoviruses into germ cell genomes, but other mechanisms of integration, including targeted insertion into genomes facilitated by the viral nickase (NS1 or Rep), have also been suggested (1). Retrovirus-derived sequences or transposable elements flank some parvovirus insertions, suggesting that the inherent stability of regions containing transposons or retrotransposons may facilitate the insertion of parvovirus or other DNA into host genomes (19). Parvovirus EVEs generally resemble host non-coding DNA or pseudogenes, showing degeneration that is generally proportional to the length of time they have been integrated, and most do not contain long intact open reading frames due to the presence of single base mutations, or deletions or insertions of varying sizes that result in stop codons or frameshifts within the remaining sequence (17–19). A minority of parvovirus EVEs, however, contain intact, transcribed open reading frames, indicating that some insertions may be of benefit to the host and under positive selection to be maintained (17,19).

Although the presence of parvovirus-derived EVE sequences in eukaryotic genomes has been recognized for a number of years and more are being found regularly as new genome sequences are obtained, we know little about their biology, or whether they play roles in the functions or evolutionary success of their hosts. There has been relatively little analysis of their properties and how they may relate to contemporary viruses, or the deep evolutionary history of

148

the parvoviruses. Here, we examine three parvovirus-derived EVEs that are present in rodent genomes, describe their relationships within the parvovirus family, and examine their properties in capsid production and function. The three VP2-related sequences examined are related to extant viruses, and were repaired to give intact open reading frames. The proteins were tested for their ability to assemble into capsids or to be expressed as domains on virus-like particles, and the capsids were tested for cell binding, stability, sialic acid binding, and recognition by antibodies derived against an extant parvovirus.

4.4 Results

Genome mining, examination of hosts related to R. norvegicus, and reconstitution of parvovirus EVE capsid genes. Whole genome shotgun (WGS) sequence data of 113 mammal species were screened to identify endogenous parvovirus capsid genes and pseudogenes. Close- to-intact parvovirus VP2 sequences were chosen for reconstruction from the genomes of the brown rat (R. norvegicus), the Western Mediterranean mouse (M. spretus), and the wood mouse

(A. sylvaticus). Each integration occurred at a distinct genomic locus and can therefore be assumed to have arisen independently. All three EVE sequences clearly grouped within the extant Protoparvovirus genus in phylogenies (Figure 4.1A). Notably, all three rodent EVEs were more closely related to viruses isolated from non-rodent host species than they were to those that have been isolated from rodents, such as the minute virus of mice (MVM). To investigate when the rat endogenous viral element integrated during the Rattus speciation, we used PCR to screen for a ~500bp sequence of the EVE within the DNA of five species within the genus: R. rattus, R. tanezumi, R. exulans, R. everetti, as well R. norvegicus as a control. We were able to recover sequences from all those species, putting the age of the R. norvegicus EVE insertion at least, 2.3

149

Figure 4.1. Endogenous viral element insertions. A) Phylogenetic tree containing the three rodent EVE sequences and extant parvoviruses. B-D) Map of the parvovirus endogenous viral elements used in this study, showing sequence overlap with the most closely related extant virus and frameshift mutations or deletions that disrupt open reading frames. B) The Rattus

Norvegicus EVE compared to CPV-2. C) The Mus Spretus EVE compared to porcine parvovirus. D) The Apodemus Sylvaticus EVE compared to porcine parvovirus.

150

million years ago. The VP2 sequences from all three EVEs contained insertion or deletion frameshift mutations, and the A. sylvaticus sequence also contained a total of five in-frame stop codons and a stretch of 52 bases that were missing from the sequence (Figure 4.1B-D).

Canine parvovirus (CPV) is the extant parvovirus most closely related to the R. norvegicus EVE, with those sequences sharing 68.4% amino acid identity. Most of the differences between CPV and the R. norvegicus EVE were located on exterior capsid surface loops, and the β-strands that form the interior surface of the parvovirus capsid were largely conserved (Figure 4.2A). The M. spretus and A. sylvaticus EVEs were both most closely related to porcine parvovirus (PPV), which infects swine, sharing 81.7% and 71.8% amino acid sequence identity with that virus, respectively. Most differences between the M. spretus or A. sylvaticus EVEs and PPV also occurred in the exterior capsid surface loops, while interior surface β-strands were largely conserved (Figure 4.2B and C).

R. norvegicus VP2 expression and capsid assembly. After repairing each sequence, the

VP2 genes were synthesized and inserted into expression plasmids for transfection into mammalian cells or baculovirus expression in insect cells. When the repaired R. norvegicus VP2 sequence was expressed in the HEK293 mammalian cell line, protein expression was readily detected in transfected cells by both immunofluorescence assays and Western blots with a rabbit polyclonal antibody prepared against the capsids of feline panleukopenia virus (FPV), which is closely related to CPV. However, electron microscopy of purified samples showed no virus-like particles (VLPs). Baculovirus expression of the R. norvegicus VP2 gene, which allows for higher levels of protein expression than the mammalian cell system, produced similar results, with protein expression being detectible in insect cell lysates, but the capsid protein appearing degraded, indicating a lack of assembly.

151

Figure 4.2. Sequence similarity between the endogenous viral elements and the most closely related extant virus. A) (Top) The Rattus Norvegicus EVE VP2 sequence, with residues identical to CPV-2 colored green and residues that differ from CPV-2 colored black. Surface loops 1-5 and restriction enzyme sites are shown, and amino acid position is marked. (Bottom) CPV-2 VP2 protein (PDBID: 2CAS) with shared residues with the Rattus Norvegicus EVE colored green

(bottom left), shown on the capsid exterior (bottom middle), and shown on the capsid interior

(bottom right). B) (Top) The Mus Spretus EVE VP2 sequence, with residues identical to PPV colored red and residues that differ from PPV colored black. Amino acid position is marked. 152

(Bottom) PPV VP2 protein (PDBID: 1K3V) with shared residues with the Mus Spretus EVE colored red (bottom left), shown on the capsid exterior (bottom middle), and shown on the capsid interior (bottom right). C) (Top) The Apodemus Sylvaticus EVE VP2 sequence, with residues identical to PPV colored blue and residues that differ from PPV colored black. Amino acid position is marked. (Bottom) PPV VP2 protein (PDBID: 1K3V) with shared residues with the

Apodemus Sylvaticus EVE colored blue (bottom left), shown on the capsid exterior (bottom middle), and shown on the capsid interior (bottom right).

153

R. norvegicus EVE sequence chimeras. We created chimeric VP2 sequences with portions of the CPV gene replaced with the corresponding segments from the R. norvegicus

EVE. For our first attempt to get VLPs to assemble from chimeras, we used shared restriction enzyme sites between CPV and the R. norvegicus EVE to divide VP2 into thirds, and then ligated the sequences together in all 6 possible combinations and expressed them in insect cells

(Figure 4.3). While each of the chimeras expressed protein (Figure 4.3A), most was either degraded low molecular weight products, or a smear of higher molecular weight products suggestive of protein degradation. Negative stain electron microscopy of purified fractions also revealed no capsids (Figure 4.3C-H).

A second set of chimeras had individual loops from the CPV-2 exterior capsid surface replaced with the corresponding sequence of the R. norvegicus EVE (Figure 4.4A and B).

Replacing all of the surface loops in combination allowed us to effectively pseudotype the surface of the CPV capsid, resulting in a VLP that potentially had the same antigenic sites and receptor binding site as the R. norvegicus EVE (Figure 4.4B). We tested the chimeras with individual surface loop substitutions for assembly by expressing capsid protein in insect cells and then staining those cells with monoclonal anti-CPV antibodies that recognize assembled capsids

(Figure 4.4C). The binding sites of these antibodies on CPV capsids have been well characterized (20); antibody 14 recognizes the “A” antigenic site on the 3-fold spikes on the capsid, composed of loops 1, 2, and parts of loop 4, while antibody 8 recognizes the “B” antigenic site which is composed of loop 3 and parts of loop 4. When tested for antibody binding, the chimeric VLPs were able to bind to the monoclonal antibodies which recognized the remaining CPV surface, but not those where the epitope had been substituted for one or more R. norvegicus EVE loops (Figure 4.4C). The chimera containing all five surface loops failed to

154

Figure 4.3. Expression and assembly of chimeras constructed by combining BglII and PhoI restriction enzyme digestions of the Rattus Norvegicus EVE with those of CPV-2. A) Western blot showing capsid proteins chimeras or CPV-2 expressed in insect cells and stained with a polyclonal anti-CPV antibody. B) Electron micrograph of CPV-2 VLPs showing assembled capsids. C-H) Electron micrographs of chimeras between the Rattus Norvegicus EVE and CPV-

2, showing that capsid proteins do not assemble. Diagrams of the chimeras are also shown, with the VP2 gene divided at the BglII and PhoI restriction enzyme sites, and the Rattus Norvegicus

EVE represented in green.

155

Figure 4.4. Construction and expression of chimeric VLPs consisting of Rattus Norvegicus EVE surface loops incorporated into CPV VP-2. A) VP2 from the Rattus Norvegicus EVE, showing the location of 5 surface loops and their differences from CPV-2. Numbering of residues is from

CPV-2. B) Diagrams of the Rattus Norvegicus EVE surface loop/CPV-2 chimeras. Rattus

Norvegicus EVE surface loops are colored green and shown on the CPV-2 VP2 backbone

(PDBID: 2CAS) (left), and shown on the capsid surface (right). C) Immunofluorescence assay

156

showing expression of chimeric VLPs with a polyclonal anti-CPV antibody (red), and assembly of VLPs with monoclonal antibodies that only recognize assembled capsids (green). ‘B-site’ (top row) and ‘A-site’ (bottom row) monoclonal antibody staining are both shown.

157

assemble into VLPs, suggesting that the R. norvegicus surface loops may also contribute to assembly defects when in combination.

Characterization of R. norvegicus chimeric EVEs. We purified chimeras containing loop 4, the largest of the capsid surface loops and which also contains residues responsible for sialic acid binding in CPV. Electron microscopy of the purified surface loop 4 chimera confirmed capsid assembly (Figure 4.5), and incubation of chimeras with NLFK feline kidney cells and TfR deficient TRVb cells (modified Chinese Hamster Ovary cells) showed that surface loop 4 altered patterns of cellular uptake compared to CPV-2 capsids (Figure 4.5). CPV-2 VLPs depend on the TfR for entry into cells and accumulate within endosomes around the nucleus shortly after uptake, loop 4 chimeric VLPs accumulated in a single region in the capsid cytoplasm, had far fewer particles enter NLFK cells than wild-type CPV-2 capsids, and entered

TRVb cells, which do not express the TfR (Figure 4.5). These results indicate that loop 4 substitutions prevent the TfR binding for chimeric capsids, and that loop 4 may contain the R. norvegicus EVE’s receptor and/or sialic acid binding site.

Expression of M. spretus and A. sylvaticus EVE sequences and testing of M. spretus capsid structure and function. M. spretus and A. sylvaticus VP2 EVE sequences were also expressed in insect cells, and while both EVEs produced capsid proteins that were recognized by an anti-CPV polyclonal antibody (Figure 4.6A), only the M. spretus EVE assembled into capsids

(Figure 4.6C-D). The M. spretus capsids were highly stable, with an average melting temperature of 82°C determined by differential scanning fluorimetry (DSF) experiments (Figure 4.7A and B).

While changing pH from 7.4 to 5.5 affected the melting temperature of CPV-2 VLPs, which were used as controls, the overall melting temperature of the M. spretus VLPs was not affected.

The derivative curve from M. spretus VLPs did, however, show a minor peak that was affected

158

Figure 4.5. Cellular uptake and negative stain EM of Loop 4 chimeric R. Norvegicus EVE/CPV-

2 VLPs. (Top) CPV-2 VLP and Loop 4 chimeric VLP (bottom) uptake into NLFK cells or TRVb cells, and negative stain EM of purified capsids. For uptake assays, capsids are stained with a polyclonal anti-FPV antibody (red) and the transferrin receptor (TfR) is stained with monoclonal anti-TfR cytoplasmic tail antibody (green).

159

Figure 4.6. Expression and assembly of A. sylvaticus and M. spretus EVE VLPs. A) Western blot showing capsid proteins expressed in insect cells and stained with a polyclonal anti-CPV antibody. B-D) Negative stain electron microscopy of CPV-2 VLPs (B), A. sylvaticus EVE VLPs

(C), and M. spretus EVE VLPs (D).

160

Figure 4.7. Characterization of the assembled Mus Spretus EVE VLPs. A) Differential scanning fluorometry experiment showing derivative melting curves for CPV-2 VLPs and Mus Spretus

EVE VLPs at pH 7.4 and 5.5. B) Quantification of capsid melting temperature as determined by differential scanning fluorometry. Results of three independent experiments are shown. Error bars denote ± standard error. C) Western blot and D) silver stain of purified CPV-2 or Mus

Spretus EVE VLPs. The western blot was stained with a polyclonal anti-CPV antibody. E)

Uptake of fluorescently labeled Mus Spretus EVE VLPs into murine L cells. F)

Hemagglutination assay of CPV-2 VLPs and Mus Spretus EVE VLPs on mouse red blood cells

(top) and hemagglutination of Mus Spretus EVE VLPs on mouse red blood cells treated with a broad-spectrum neuraminidase (bottom).

161

by changes in pH (Figure 4.7A) which may represent a transition in the capsid structure prior to full disassembly. M. spretus VLP stability was also confirmed by western blotting and silver staining (Figure 4.7C and D), where M. spretus VLPs formed far fewer low molecular weight proteolytic cleavage products than did CPV-2 VLP controls.

Fluorescently labeled M. spretus capsids were taken up into murine L cells and were distributed throughout cells in a perinuclear location suggestive of endocytosis (Figure 4.7E). M. spretus capsids also hemagglutinated mouse erythrocytes (Figure 4.7F), which contained sialic acids composed of over 97% of forms of N-acetyl neuraminic acid (Neu5Ac) (Table 4.1), suggesting that sialic acid form is recognized by the virus. Neuraminidase treatment of mouse erythrocytes eliminated hemagglutination by M. spretus capsids (Figure 4.7F), showing that binding was to sialic acids.

4.5 Discussion

The initial discovery of integrated sequences related to autonomous parvoviruses in the genomes of various animals was unexpected, as autonomous parvoviruses do not go through an integrated stage in their life cycles. However, it is now clear that numerous animal genomes contain one or more endogenous parvovirus sequences, many of which represent remnants that appear to have been integrated into the genomes of their ancestral hosts millions of years ago

(1,17–19). Whether these integrated sequences benefit the host or play any biological role is not known. While endogenized viral sequences have frequently been used for host functions (21–

24), the degenerated state of the parvoviral insertions examined here makes it unlikely that they express viral proteins or peptides.

Our goal here was to examine the properties of three endogenized major capsid protein

162

Table 4.1. Percentage of O-linked sialic acids on murine erythrocytes.

O-linked sialic % Abundance acid type Neu5Ac 60.50% Neu5,9Ac2 16.60% Neu5,7,8/9Ac3 13.90% Neu5,7Ac2 3.40% Neu5,8Ac2 2.90% Neu5Gc 2.70%

163

gene sequences found in mouse and rat genomes in order to better understand the properties of the ancient viruses that they were derived from. The sequences of the three viruses that we examined were only observed in the genomes of individual species and closely related taxa. For the murid rodents (A. sylvaticus and M. spretus), there are a number whole genome sequences from related species available, but none contain these EVE sequences, indicating that these

EVEs were acquired within the last several million years. Furthermore, the sites corresponding to the M. spretus and A. sylvaticus EVE insertions in these related species showed no evidence of the integrated sequences, making it unlikely that the insertions were lost from those mice. There are few genomic sequences of rats closely related to R. norvegicus in databases, so we examined

DNA extracted from the tissues of four closely related rodents for the R. norvegicus EVE sequence. We found that sequence in the DNA of five of the Rattus species examined, indicating that it was likely acquired over 2.3 million years ago. These sequences therefore likely represent viruses that are thousands or millions of years older than the contemporary viruses for which we have sequences.

Both the R. norvegicus and A. sylvaticus VP2 proteins failed to assemble into capsids when first repaired. The A. sylvaticus VP2 sequence contained a relatively large, 52 base deletion that we filled in with nucleotides from the corresponding region in the consensus sequence of closely related contemporary viruses, and incompatible residues within that sequence may have prevented capsid assembly. Other mutations within the A. sylvaticus or R. norvegicus EVE sequences could have also prevented assembly. Studies with extant parvoviruses show that single point mutations within the major capsid protein can inhibit capsid assembly, especially when they occur to the well-conserved residues that form the interior of the capsid or the 5-fold pore

(25–28). Our CPV/R. norvegicus EVE chimeras support the latter hypothesis, as chimeras which

164

contained sequences from the R. norvegicus EVE capsid interior or 5-fold pore failed to assemble (Figure 4.3). However, replacing the unconserved, exterior capsid surface loops allowed chimeras to assemble (Figure 4.4). Each of the three R. norvegicus EVE segments from the first set of chimeras prevented capsids from assembling, suggesting that all regions contain mutations that prevent capsid assembly.

The R. norvegicus/CPV chimera with all 5 EVE surface loops together failed to assemble into capsids, even though the chimeras containing individual surface loops could. Surface loop 5 chimeras, however, appeared to have difficulty assembling, as not all cells that stained positive for capsid protein expression were also positive for assembled capsids, and this loop could have impeded assembly of the complete chimera. The loop 4 chimera, which contained the largest capsid surface loop, altered the uptake of VLPs into feline and hamster cells (Figure 4.5), even though loops 1, 2, and 3, whose residues have been implicated in receptor binding in wild-type

CPV capsids, were unchanged. This suggests that the loop 4 substitution changed the position of one or more of these loops and rendered the chimeric VLPs unable to bind to the feline TfR.

That this chimera could still enter TRVb cells, which do not express the transferrin receptor, indicates that it either entered cells nonspecifically (perhaps through macropinocytosis) or that loop 4 of the R. norvegicus EVE contains a receptor binding site that allows the chimera to enter cells.

The M. spretus EVE produced capsids directly after being repaired, and those capsids were highly stable at both neutral and endosomal pH. Minor peaks on the DSF curve could be the result of partial unfolding of the capsid or the disassociation of capsid subunits (dimers, trimers, or pentamers) from the whole, both of which would expose hydrophobic residues and result in increased fluorescence. M. spretus capsids had cellular uptake patterns similar to those

165

of extant parvoviruses with capsids having a perinuclear localization an hour after uptake (29)

(Figure 4.7), and since those capsids bind Neu5Ac sialic acids, they likely serve as the receptor for cellular uptake.

Summary. Here we reveal the characteristics of ancient parvoviruses that were recovered as molecular fossils from host genomes. One of the sequences assembled into VLPs and we could therefore examine its capsid stability, sialic acid binding, and cell uptake. Others could be assembled as components of CPV sequences, and those showed us some surface features of the ancient virus capsids. Studying these capsids provides an understanding of how parvoviruses have evolved over long period – likely over thousands or millions of years. These capsids also show that the M. spretus virus capsids bound to sialic acids and entered mouse cells, as expected since the original virus presumably had the host range for rodents. It is still unknown whether or not endogenized parvovirus sequences provide any benefit to their hosts, and future experiments could focus on whether or not expression of truncated proteins or RNA from these sequences has any effect on the hosts or infecting viruses.

4.6 Methods

Genome sequence analysis. The database-integrated genome screening (DIGS) tool (30)

(available at http://giffordlabcvr.github.io/DIGS-tool/) was used to identify parvovirus-related sequences in whole genome shotgun sequence data of 113 mammal species. A subset of whole genome shotgun contigs that contained both complete or near-complete parvovirus capsid gene sequences and a region of host genomic sequence were selected for further analysis. A combination of automated procedures (MUSCLE (31,32), BLAST (33)) and manual adjustments were used to align parvovirus EVEs to the most closely related extant virus sequence among a

166

set of representative parvovirus genomes, allowing us to infer the likely ancestral open reading frames (ORFs). The original alignments were also used to construct phylogenies using maximum likelihood (ML) as implemented in RAxML (34). Parameters for phylogenetic analysis were selected using ProtTest (35).

Analysis of the viral sequences and reconstruction of open reading frames. Three endogenized viral sequences with extended capsid protein genes, derived from the genomes of

Rattus norvegicus, Mus spretus and Apodemus sylvaticus, were chosen for protein expression.

These sequences had relatively complete open reading frames, which could be repaired with some confidence, and had homology to contemporary infectious viruses. To repair the parvovirus

EVEs, VP2 sequences were aligned with the sequences of a group of the most closely related contemporary infectious viruses and the open reading frames were modified to remove stop codons, insertions, and deletions, recreating the open reading frame (diagrammed in Figure 1).

Where necessary to repair likely deletions, we introduced an amino acid or sequence that was most similar to the consensus of the closely related contemporary viral sequences.

Identification of EVE sequences in DNA species related to R. norvegicus. To obtain more information about the possible presence of the sequence within the R. norvegicus genome, we obtained DNA from other R. norvegicus samples (from laboratory rats), as well as from bio- banked samples of R. rattus, R. tanezumi, R. exulans, and R. everetti from the Field Museum of

Natural History, Chicago (Table 4.2). DNA was extracted from tissues using a Tissue DNA extraction kit (Omega Bio-tek) and eluted in water. Primers that recognized the EVE sequence were used to perform PCR (5'-AGATCAGGCTACTATACTAGCTCCGATTC-3' and 5'-

AACCAACTTACTACTACCCAAGACAACA-3', as described in (19)), with a total of 100µg of

DNA, and 35 cycles of PCR with Q5 High Fidelity DNA Polymerase (NEB). The ~500 base pair

167

Table 4.2. Specimens used to determine the presence of the rat EVE.

Species Number of samples Field Museum of Natural History ID # Rattus rattus 4 (2M, 2F) 222595, 198176, 181086, 179191 Rattus exulans 4 (2M, 2F) 198761, 168962, 188484, 168964 Rattus everetti 4 (2M, 2F) 198872, 198796, 196061, 191246 Rattus tanezumi 4 (2M, 2F) 198768, 198797, 194747, 178427

168

DNA products amplified were sequenced using Sanger sequencing to confirm the presence of the

EVE.

Cells. Human embryonic kidney 293 (HEK293) cells, Sf9 (Spodoptera frugiperda) cells

(Invitrogen), and High Five (Trichoplusia ni) cells (Boyce Thompson Institute, clone BTI-TN-

551-4) were used in this study. HEK293 cells were grown in DMEM with 10% fetal calf serum

(FCS) at 37°C, Sf9 cells were grown in Grace’s Supplemented Insect Media (Gibco) with 10%

FCS at 28°C, and High Five cells were grown in Express5 Serum Free Media (Gibco) at 28°C.

Norden Laboratory Feline Kidney (NLFK) cells, TRVb cells (modified Chinese Hamster Ovary cells selected to not express the TfR), and L cells (a murine fibroblast cell line obtained from the laboratory of John Parker) were also used in this study. NLFK cells were grown in 1:1

McCoy’s/Lebovitz media with 5% FCS at 37°C, L cells were grown in Eagle’s Minimal

Essential Medium (EMEM) with 10% FCS at 37°C, and TRVb cells were grown in Ham’s/F12 media with 10% FCS at 37°C.

Expression of endogenous viral element capsid proteins. The complete, repaired VP2 sequences were codon optimized for mammalian cell expression and synthesized along with the

12 base pairs preceding canine parvovirus VP2 to favor protein expression. Sequences were then cloned into the pCDNA(-) plasmid vector under the control of the CMV immediate early promoter for expression in HEK293 cells. The genes were also cloned into the pFastBac1 plasmid (Invitrogen) under the control of the polyhedron promoter for baculovirus expression in

High Five cells. Baculoviruses were prepared by transfection of bacmids into Sf9 cells, and then proteins prepared by inoculation of High Five insect cells, as described previously (25).

For VP2 capsid protein purification, cells were collected 3 days after transfection in

HEK293 cells with plasmids or 4 days after infection with baculovirus. Supernatants from

169

transfected HEK293 cells were pelleted at 100,000×g for 2 hrs. Virus-like particles (VLPs) were purified from infected High Five cells as described previously (25). The resulting fractions were examined via negative stain electron microscopy with 2% uranyl acetate to determine whether capsids were present.

Construction of R. norvegicus EVE/CPV-2 chimeric VP2 sequences. Some of the VP2 sequences did not assemble into capsids when expressed from plasmids after transfection into

293 cells or after infection of High Five cells with baculovirus containing VP2 sequences. To examine the nature of the assembly defect(s) in the R. norvegicus EVE VP2 gene sequence and to allow definition of the structures of the protein domains, we prepared recombinants that combined the R. norvegicus EVE VP2 sequence with that of the CPV-2 VP2 gene. The first set of chimeras were constructed by digesting the R. Norvegicus and CPV-2 VP2 genes with BglII and PhoI restriction enzymes (NEB), dividing the genes roughly into thirds, and ligating the sequences together in all 6 possible combinations, as shown in Figure 4.3. The second set of chimeras was created by cloning surface loops from the R. norvegicus EVE into the CPV-2 VP2 backbone, as depicted in Figure 4. Five loops were individually cloned into the backbone

(Loop1, VP2 residues 82-104; Loop 2, residues 214-244; Loop 3, residues 290-314; Loop 4, residues 346-449; Loop 5, residues 546-584) and a complete chimera containing all 5 surface loops was also created. For individual surface loop chimeras, capsid assembly was determined by incubating capsids with anti-CPV monoclonal antibodies 8 and 14, which recognize two different antigenic sites on CPV capsids and only bind to assembled particles (20).

Examination of assembled capsid properties. The M. spretus EVE VLPs and wild-type

CPV-2 VLPs were tested for stability using differential scanning fluorometry. 2.5µg of VLPs in either PBS (pH 7.4) or 0.1M citric acid (pH 5.5) were combined with 2.5 µL of 1% SYPRO

170

orange dye (Life Technologies) and heated from 25°C to 95°C at a rate of 1°C/min using a

StepOnePlus Real-Time PCR thermocycler (Applied Biosystems). Melting temperature was determined by taking the first derivative of the resulting curve and finding its maximum. Results from three independent experiments were analyzed via ANOVA using GraphPad Prism (v7.04,

GraphPad Software, Inc.).

Purified CPV-2 and M. Spretus EVE VLPs were separated on a 10% SDS-PAGE gel and either silver stained using a Pierce Silver Stain kit (Thermo) according to the manufacturer’s instructions or transferred onto a nitrocellulose membrane and stained with a rabbit polyclonal anti-CPV primary antibody and a goat anti-rabbit HRP secondary antibody.

Capsids were also tested for hemagglutination (HA) with mouse erythrocytes. Capsids were serially diluted 1:1 in 25µL of Bis-Tris buffered saline (BTBS) pH 6.2, then 50µL of 0.5% v/v mouse erythrocytes in BTBS was added and mixtures were incubated at 4°C overnight. For neuraminidase (NA)-treated erythrocytes, sialic acids were cleaved by overnight incubation at

37°C with 40U of α-2,3,6,8,9 Neuraminidase A (NEB) in PBS pH 7.4. The sialic acid forms on mouse erythrocytes was were examined by analysis of DMB-labelled sialic acids on a Dionex

UltiMate 3000 high-performance liquid chromatography system.

Cell binding and uptake of M. spretus EVE VLPs was determined by fluorescently labeling capsids with Alexa-488 using Alexa Fluor 488 5-SDP Ester (Life Technologies) according to the manufacturer’s protocol. 10µg of capsids were added to mouse L cells seeded at

1×105 cells/cm2on glass cover slips, and incubated for 1 h at 37°C. Cells were fixed in 4% paraformaldehyde for 10 min, mounted on slides using ProLong Gold Antifade Mountant with

DAPI (Life Technologies), and imaged using a Nikon TE300 epifluorescence microscope. Cell binding and uptake of Loop 4 chimeric VLPs was determined by incubating 4µg/mL capsids

171

with NLFK or TRVb cells seeded at 2×104 cells/cm2 on cover slips. Cells were fixed and then stained with a rabbit anti-CPV polyclonal antibody and/or a mouse anti-TfR monoclonal antibody (Life Technologies, #QB213080), followed by staining with goat anti-rabbit Alexa594 and/or goat anti-mouse Alexa488 secondary antibodies. Cells were mounted on slides and imaged in the same way as the L cells.

4.7 Acknowledgements

We would like to thank Wendy Weichert for providing excellent technical support. We also thank the Field Museum of Natural History for providing tissue samples.

172

REFERENCES

1. Smith RH, Hallwirth CV, Westerman M, Hetherington NA, Tseng Y-S, Cecchini S, et al.

Germline viral “fossils” guide in silico reconstruction of a mid-Cenozoic era marsupial

adeno-associated virus. Sci Rep. 2016 Jul 5;6:28965.

2. Cotmore SF, Agbandje-McKenna M, Chiorini JA, Mukha DV, Pintel DJ, Qiu J, et al. The

family Parvoviridae. Arch Virol. 2014 May;159(5):1239–47.

3. Steinel A, Parrish CR, Bloom ME, Truyen U. Parvovirus infections in wild carnivores. J

Wildl Dis. 2001 Jul;37(3):594–607.

4. Mendelson E, Trempe JP, Carter BJ. Identification of the trans-acting Rep proteins of

adeno-associated virus by antibodies to a synthetic oligopeptide. J Virol. 1986

Dec;60(3):823–32.

5. Miller CL, Pintel DJ. Interaction between parvovirus NS2 protein and nuclear export

factor Crm1 is important for viral egress from the nucleus of murine cells. J Virol. 2002

Apr;76(7):3257–66.

6. Redemann BE, Mendelson E, Carter BJ. Adeno-associated virus rep protein synthesis

during productive infection. J Virol. 1989 Feb;63(2):873–82.

7. Wang D, Yuan W, Davis I, Parrish CR. Nonstructural protein-2 and the replication of

canine parvovirus. Virology. 1998 Jan 20;240(2):273–81.

8. Hernando E, Llamas-Saiz AL, Foces-Foces C, McKenna R, Portman I, Agbandje-

McKenna M, et al. Biochemical and physical characterization of parvovirus minute virus

of mice virus-like particles. Virology. 2000 Feb 15;267(2):299–309.

9. López de Turiso JA, Cortés E, Martínez C, Ruiz de Ybáñez R, Simarro I, Vela C, et al.

Recombinant vaccine for canine parvovirus in dogs. J Virol. 1992 May;66(5):2748–53.

173

10. Ruffing M, Zentgraf H, Kleinschmidt JA. Assembly of viruslike particles by recombinant

structural proteins of adeno-associated virus type 2 in insect cells. J Virol. 1992

Dec;66(12):6922–30.

11. Girod A, Wobus CE, Zádori Z, Ried M, Leike K, Tijssen P, et al. The VP1 capsid protein

of adeno-associated virus type 2 is carrying a phospholipase A2 domain required for virus

infectivity. J Gen Virol. 2002 May;83(Pt 5):973–8.

12. Vihinen-Ranta M, Wang D, Weichert WS, Parrish CR. The VP1 N-terminal sequence of

canine parvovirus affects nuclear transport of capsids and efficient cell infection. J Virol.

2002 Feb;76(4):1884–91.

13. Ahn JK, Pitluk ZW, Ward DC. The GC box and TATA transcription control elements in

the P38 promoter of the minute virus of mice are necessary and sufficient for

transactivation by the nonstructural protein NS1. J Virol. 1992 Jun;66(6):3776–83.

14. Cotmore SF, Tattersall P. An asymmetric nucleotide in the parvoviral 3’ hairpin directs

segregation of a single active origin of DNA replication. EMBO J. 1994 Sep

1;13(17):4145–52.

15. Krauskopf A, Resnekov O, Aloni Y. A cis downstream element participates in regulation

of in vitro transcription initiation from the P38 promoter of minute virus of mice. J Virol.

1990 Jan;64(1):354–60.

16. Tyson JJ, Chen KC, Lederman M, Bates RC. Analysis of the kinetic hairpin transfer

model for parvoviral DNA replication. J Theor Biol. 1990 May 22;144(2):155–69.

17. Arriagada G, Gifford RJ. Parvovirus-derived endogenous viral elements in two South

American rodent genomes. J Virol. 2014 Oct;88(20):12158–62.

18. Katzourakis A, Gifford RJ. Endogenous viral elements in animal genomes. PLoS Genet.

174

2010 Nov 18;6(11):e1001191.

19. Liu H, Fu Y, Xie J, Cheng J, Ghabrial SA, Li G, et al. Widespread endogenization of

densoviruses and parvoviruses in animal and human genomes. J Virol. 2011

Oct;85(19):9863–76.

20. Hafenstein S, Bowman VD, Sun T, Nelson CDS, Palermo LM, Chipman PR, et al.

Structural comparison of different antibodies interacting with parvovirus capsids. J Virol.

2009 Jun;83(11):5556–66.

21. Bézier A, Herbinière J, Lanzrein B, Drezen J-M. Polydnavirus hidden face: the genes

producing virus particles of parasitic wasps. J Invertebr Pathol. 2009 Jul;101(3):194–203.

22. Ito J, Watanabe S, Hiratsuka T, Kuse K, Odahara Y, Ochi H, et al. Refrex-1, a soluble

restriction factor against feline endogenous and exogenous retroviruses. J Virol. 2013

Nov;87(22):12029–40.

23. Lavialle C, Cornelis G, Dupressoir A, Esnault C, Heidmann O, Vernochet C, et al.

Paleovirology of “syncytins”, retroviral env genes exapted for a role in placentation.

Philos Trans R Soc Lond B Biol Sci. 2013 Sep 19;368(1626):20120507.

24. Malfavon-Borja R, Feschotte C. Fighting fire with fire: endogenous retrovirus envelopes

as restriction factors. J Virol. 2015 Apr;89(8):4047–50.

25. Callaway HM, Feng KH, Lee DW, Allison AB, Pinard M, McKenna R, et al. Parvovirus

Capsid Structures Required for Infection: Mutations Controlling Receptor Recognition

and Protease Cleavages. J Virol. 2017 Jan 15;91(2).

26. Kawase M, Momoeda M, Young NS, Kajigaya S. Modest truncation of the major capsid

protein abrogates B19 parvovirus capsid formation. J Virol. 1995 Oct;69(10):6567–71.

27. Lombardo E, Ramírez JC, Agbandje-McKenna M, Almendral JM. A beta-stranded motif

175

drives capsid protein oligomers of the parvovirus minute virus of mice into the nucleus for

viral assembly. J Virol. 2000 Apr;74(8):3804–14.

28. Yuan W, Parrish CR. Canine parvovirus capsid assembly and differences in mammalian

and insect cells. Virology. 2001 Jan 20;279(2):546–57.

29. Parker JS, Murphy WJ, Wang D, O’Brien SJ, Parrish CR. Canine and feline parvoviruses

can use human or feline transferrin receptors to bind, enter, and infect cells. J Virol. 2001

Apr;75(8):3896–902.

30. Zhu H, Dennis T, Hughes J, Gifford RJ. Database-integrated genome screening (DIGS):

exploring genomes heuristically using sequence similarity search tools and a relational

database. bioRxiv. 2018 Apr 25;

31. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high

throughput. Nucleic Acids Res. 2004;32(5):1792–7.

32. Edgar RC. MUSCLE: a multiple sequence alignment method with reduced time and space

complexity. BMC Bioinformatics. 2004 Aug 19;5:113.

33. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool.

J Mol Biol. 1990 Oct 5;215(3):403–10.

34. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of

large phylogenies. Bioinforma Oxf Engl. 2014 May 1;30(9):1312–3.

35. Darriba D, Taboada GL, Doallo R, Posada D. ProtTest 3: fast selection of best-fit models

of protein evolution. Bioinforma Oxf Engl. 2011 Apr 15;27(8):1164–5.

176

CHAPTER FIVE

Summary and conclusions

177

Despite their small size and seeming simplicity, parvoviruses must still undergo all of the steps of infection that viruses with larger genomes do. Because parvoviruses encode variants of only two proteins, their capsids, rather than a series of specialized proteins, carry out all of these steps up to delivering the viral genome to a host cell nucleus, and undergo a number of complex interactions with receptors and other host factors in order to do so. Capsids must also be stable enough to protect the viral genome from the environment, but flexible enough to release the genome when the conditions are right. Mutations or host factors, such as antibodies, that disrupt any of these processes can result in a loss of infection. The studies in this dissertation explore how the canine parvovirus (CPV) capsid facilitates infection, from receptor binding and proteolytic cleavage events to the effects of antibody binding. This dissertation also examines parvovirus endogenous viral elements (EVEs), ancient viral sequences that allow determination of what parvoviruses were like millions of years ago. These works improve the understanding of parvovirus infection and evolution, and provide targets that antiviral drugs or antibodies could potentially use to neutralize infection of parvoviruses.

5.1 Mutations to critical capsid structures

Point mutations to critical structures of the parvovirus capsid result in loss of infectivity.

VP2 Cys270Ser mutation results in the loss of a proteolytic cleavage site between residues 269 and 270 on the interior surface of the capsid, decreasing infectivity roughly 100-fold compared to wild-type virus and increasing the stability of capsids. Because the proteolytic cleavage occurs both in wild-type viruses produced in mammalian cells and virus-like particles produced in insect cells, it is likely that the capsid undergoes autoproteolytic cleavage. The initial analysis of the capsid structure suggested that Cys270 and the nearby residue His137 might form the

178

protease, but mutation of His137 to alanine had no effect on viral infectivity or proteolytic cleavage. Subsequent analysis of CPV capsids with high resolution mass spectrometry showed that Cys270 was carboxymethylated, and it is possible that the carboxyl group and residue

Asp269 acted as the catalytic dyad in an acid protease to trigger proteolytic cleavage. Additional experiments disrupting this putative protease and exploring why only a fraction of capsid proteins are cleaved are needed to fully understand the proteolytic cleavage event.

VP2 Cys273Ser mutation reduces viral infectivity over 106-fold, and results in a significant increase in capsids that remained trapped in the nucleus after assembly. The mechanism responsible for this increase is unknown, however, and seems unrelated to the proteolytic cleavage event at residue 270. Pro272Lys mutation similarly causes an over 106-fold loss of infectivity, and capsid proteins containing this mutation appear to degrade in the cell cytoplasm, suggesting an assembly defect that prevents capsids from forming.

Each of these mutations disrupts a critical function during CPV infection, which likely has analogs in other parvoviruses and perhaps even in other types of viruses. Better understanding how these mutations interfere with infection provides basic knowledge about how viruses work and highlights potential antiviral drug targets for this and other viruses.

5.2 Receptor and antibody binding interactions

CPV and the transferrin receptor (TfR) have a complex interaction that determines whether or not CPV can infect cells. This interaction involves more than strength of adhesion, as has been shown by several studies that report that CPV can enter cells with specific mutations or alterations to the TfR, but is unable to infect them. Mutation of CPV VP2 residues 299 and 300, which are associated with TfR binding, to lysine results in a 106-fold loss of infectivity even

179

though capsids retain their ability to bind to the TfR and enter cells. More detailed examination of the TfR/capsid interaction with this mutant revealed that even though it bound strongly to the

TfR and had an association curve that was similar to wild-type capsids in bio-layer interferometry experiments, its disassociation curve was significantly different. While wild-type capsids had a disassociation curve that showed they remained strongly attached to the TfR (after an initial drop on the curve as weak associations came unbound), the mutant capsids showed a much steeper disassociation curve, indicating that they interacted with the TfR differently than wild-type capsids. During an infection, capsids are endocytosed into cells immediately after binding to the TfR, which allows even viruses with low affinity for the TfR to enter cells, but the

299Lys/300Lys mutant’s loss of infectivity indicates that a TfR/capsid interaction after endocytosis is required for infection. One explanation for these results is that the TfR induces a conformational change in the CPV capsid after binding to it that activates the capsid for infection.

TfRs of different species have different interactions with CPV capsids. The feline TfR binds to CPV strongly, but the canine TfR binds to CPV extremely weakly due to a bulky glycan on its apical domain, where virus binding occurs. TfRs from the black-backed jackal, the most closely related species to dogs without the inhibitory glycan, bind to CPV even more strongly than the feline TfR. While CPV capsids are taken up into cells expressing the black-backed jackal TfR to higher levels than cells expressing the feline or canine TfRs, all of these TfRs allow infection. Furthermore, greater strength of TfR binding does not correlate with higher levels of infection, and cells expressing the feline TfR or the black-backed jackal TfR both support CPV-2 growth to similar levels.

The maximum number of TfRs that can occupy the capsid surface also differs depending

180

on species. Between 1 and 5 feline TfRs bind to CPV-2 capsids. In contrast, up to 12 black- backed jackal TfRs can occupy each capsid (one receptor per 5-fold axis of symmetry), the maximum number of TfRs predicted to fit around the capsid due to steric constraints.

Interestingly, if CPV capsids are first bound to a single copy of the feline TfR, there is a small reduction in the number of black-backed jackal TfRs that can bind to capsids, compared to CPV capsids that were first bound to a single copy of the black-backed jackal TfR. This suggests that the feline TfR induces a local conformational change in the capsid that prevents the attachment of additional TfRs, and could explain why few copies of the feline TfR bind to the capsid even though the binding interaction itself is strong.

CPV VP2 residues Asn93 (on capsid surface loop 1) and Ala300 (on capsid surface loop

3) have been implicated in the TfR binding interaction in numerous studies and vary depending on the host CPV is isolated from. Mutation of residue 93 and three other residues on loop 1 to alanine results in a large and significant decrease in binding to the TfR. Mutation of 6 amino acids surrounding residue 300 to alanine, however, not only does not decrease the binding affinity to the TfR, but actually may strengthen it. This suggests that while loop 1 is directly involved in the capsid/TfR binding interaction, loop 3 may sterically or biochemically inhibit that interaction, as substituting all of the large, charged, and/or polar residues on loop 3 to alanine improved binding strength. The loop 3 mutant capsids also bind much more strongly to the canine TfR, implying that loop 3 residues on wild-type capsids may clash with the glycan on the canine TfR apical domain.

Antibodies 14 and E bind to canine parvovirus with 10-fold higher affinity than the strongest CPV/TfR interaction. At sufficiently high concentrations, this difference in affinity allows antibodies (or antigen binding fragments (Fabs)) to compete CPV capsids off of the TfR.

181

Fab E, produced from a monoclonal antibody of the same name, is particularly efficient at competing CPV capsids off of the TfR, and can also successfully compete capsids away from scFv-Fcs (single-chain variable fragments fused to an antibody constant fragment) derived from monoclonal antibodies. Fab E, unlike many other Fabs that have been tested, can neutralize CPV at very low concentrations, which suggests that it in some way alters the capsid in order to neutralize it, rather than sterically blocking all of its receptor binding sites, as is possible at higher concentrations. A cryo-EM structure of CPV and Fab E confirms that Fab E induces a conformational change in CPV after binding, and this change to capsid structure is likely what makes Fab E so successful at competing CPV capsids off of other ligands.

Together, these results illustrate the complex nature of the CPV/TfR interaction, which involves far more than adhesion. Likely, TfRs from different species bind to CPV in approximately the same way and location, but slight differences either in the fit of the TfR on

CPV, in a conformational change induced in the CPV capsid, or both, determine if a particular species’ TfR can facilitate infection. Antibody binding may also have differential effects on the

CPV/TfR binding interaction, depending on what species the TfR is derived from and the strength of binding. However, structural characterization of the different CPV/TfR complexes is necessary to determine if either of these hypotheses are correct.

5.3 Parvovirus endogenous viral elements

Ancient parvoviral sequences, preserved in host genomes as endogenous viral elements

(EVEs), provide information about how parvoviruses have evolved over millions of years. Some

EVE sequences can be repaired to remove stop codons and frameshift mutations, and capsid proteins expressed from those sequences can be examined to determine what ligand binding and

182

biochemical properties ancient parvoviruses had. Capsid proteins expressed from the Mus spretus EVE sequence assemble into virus-like particles (VLPs) with a size and morphology similar to modern parvoviruses. These VLPs have high structural stability, even among parvovirus capsids, and their capsid proteins do not appear to undergo much proteolytic cleavage, which may account for some of the structural stability. M. spretus VLPs are also capable of binding to N-acetylneuraminic acid (Neu5Ac), which is present on the surface of murine cells, and are taken up into murine cells in a pattern reminiscent of modern parvoviruses.

Capsid proteins expressed from Apodemus sylvaticus and Rattus norvegicus EVE sequences do not assemble into capsids after removal of stop codons and frameshift mutations, but if individual R. norvegicus surface loops are substituted into the major capsid protein of CPV, chimeric VLPs can assemble and show altered patterns of uptake into cells.

The difficulty of getting capsids to assemble from some of the endogenized sequences indicates that there may be other mutations distributed throughout the sequences that prevent capsids from assembling. This, along with the multiple mutations that disrupt open reading frames, implies that expression of intact parvovirus capsid proteins does not benefit (and may actually harm) host cells, even in the absence of viral replication. Any advantages EVEs provide to host cells must therefore be mediated either by truncated proteins, if EVEs are transcribed and expressed, or at the DNA or RNA level.

Analysis of virus-like particles derived from endogenized parvoviral sequences gives a greater understanding of how ancient parvoviruses behaved, what ligands they bound to, and what range of species they may have infected. This information can help predict how parvoviruses might change in the future, and which regions of their capsids might remain the same. These capsids, which have never been exposed to the immune systems of modern species,

183

might also serve as gene therapy vectors or aid in their development.

5.4 Conclusions and future directions

Taken as a whole, these studies show that in spite of its small size, canine parvovirus is remarkably complex. Although parvovirus capsids are extremely stable in the environment, single mutations to the capsid protein can disrupt the machinery that allows CPV to infect cells.

Likewise, small changes that disrupt the interaction between CPV and the TfR can result in a loss of infectivity even though capsids are still taken up into cells. TfR binding likely activates

CPV for infection, possibly through a conformational change to the capsid protein, and antibody binding that interferes with this process, either through crosslinking capsid subunits or inducing a conformational change of its own, should be particularly effective at blocking infection. Analysis of capsid proteins and VLPs derived from parvovirus EVEs furthers the understanding of parvovirus ligand binding, assembly requirements, and sensitivity to mutations or changes to surface loops.

Additional experiments are needed to define the precise nature of the interaction between

CPV and the TfR and determine whether or not receptor binding induces a conformational change in capsid structure. Further work, examining what impact endogenous parvoviral sequences have on infection by extant parvoviruses or assembly of ancient parvovirus capsid proteins (expressed as VLPs), is also needed to advance the understanding of why hosts maintain

EVE sequences and what, if any, purpose EVEs serve in the host.

184