Poliovirus receptor recognition: Visualization, kinetics, and thermodynamics. Brian M. McDermott, Jr. Submitted in partial fulfillment of the

Transcription

1 Poliovirus receptor recognition: Visualization, kinetics, and thermodynamics Brian M. McDermott, Jr. Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate School of Arts and Sciences COLUMBIA UNIVERSITY 2001

2 2001 Brian M. McDermott, Jr. All Rights Reserved

3 Abstract We have initiated studies to understand the interaction of poliovirus with its cellsurface receptor on a biochemical and biophysical level. This thesis contains the results of three studies. First, the kinetics and equilibrium of the poliovirus receptor interaction have been analyzed using surface plasmon resonance. Second, the interaction was visualized using cryoelectron microscopy and image reconstruction. Third, the activation energy was determined for the receptor-catalyzed transition from the native virion to the altered particle. To study the kinetics and equilibrium of poliovirus binding to the poliovirus receptor, we used surface plasmon resonance to examine the interaction of a soluble form of the receptor with poliovirus (spvr). Soluble receptor purified from mammalian cells is able to bind poliovirus, neutralize viral infectivity, and induce structural changes in the virus particle. Binding studies revealed that there are two binding sites for the receptor on the poliovirus type 1 capsid, with affinity constants at 20 C of K D1 = 0.67 µm and K D2 = 0.11 µm. The relative abundance of the two binding sites varies with temperature. At 20 C, the K D2 site constitutes approximately 46% of the total binding sites on the sensor chip, and its relative abundance decreased with decreasing temperature such that, at 5 C, the relative abundance of the K D2 site is only 12% of the total binding sites. Absolute levels of the K D1 site remained relatively constant at all temperatures tested. The two binding sites may correspond to docking sites for domain 1 of the receptor on the viral capsid, as predicted by a model of the poliovirus-receptor complex. Alternatively, the binding sites may be a consequence of structural breathing, or could result from receptor-induced conformational changes in the virus.

4 We have studied the poliovirus-receptor interaction by using cryoelectron microscopy to determine the structure at 21 Å resolution. This density map aided construction of a homology-based model of spvr and, in conjunction with the known crystal structure of the virus, allowed determination of the binding site. The virion does not change significantly in structure on binding spvr in short incubations at 4 C. We infer that the binding configuration visualized represents the initial interaction that is followed by structural changes in the virion as infection proceeds. spvr is segmented into three well-defined Ig-like domains. The two domains closest to the virion (domains 1 and 2) are aligned and rigidly connected, whereas domain 3 diverges at an angle of ~60. Two nodules of density on domain 2 are identified as glycosylation sites. Domain 1 penetrates the "canyon" that surrounds the 5-fold protrusion on the capsid surface, and its binding site involves all three major capsid proteins. The inferred pattern of virus-spvr interactions accounts for most mutations that affect the binding of Pvr to poliovirus. We have also studied the thermodynamics of the native (N) virion to the altered (A) particle transition. The altered particle results from the loss of viral protein 4 from the capsid and other structural changes in the virion. The receptor was shown to behave like a classic transition state theory catalyst for the N to A transition. The receptor accelerates the rate of the transition by lowering the activation barrier 50 Kcal/mol. These results confirm the prediction that the receptor acts much like an enzyme. The accumulated data was used to define the basic rules of picornavirus receptor interactions and to construct a model of genome release.

5 List of Tables List of Figures List of Abbreviations Acknowledgment Dedication Table of Contents iv v vii viii x Chapter I. Introduction 1 Picornavirus crystal structures 2 Picornavirus solution structures 8 Poliovirus pathogenesis and tissue tropism 9 Poliovirus host range 11 The cellular poliovirus receptor 13 Early events in poliovirus infection 17 Kinetic and affinity analysis of Picornavirus-receptor interactions 21 Picornavirus-receptor structures 25 Viral capsid sequences that regulate receptor binding 28 Pvr sequences that contact virus 30 Hydrophobic antiviral agents 31 Chapter II. Materials and Methods 32 Cells and Viruses 33 Plasmid construction 33 Establishment of a stable cell line expressing spvr 34 - i -

6 Protein expression, purification, and modification 35 Virus neutralization assay and determination of buffers for use with the optical 37 biosensor Alteration assays 37 Binding of spvr to poliovirus using an optical biosensor 38 Chapter III. Two Distinct Binding Affinities of Poliovirus for Its Cellular 39 Receptor Introduction 40 Expression and purification of spvr in mammalian cells 40 Virus neutralization and alteration activity of spvr 41 Conditions and specificity of surface plasmon resonance 44 Kinetic and equilibrium affinity analysis 49 Chapter IV. Three-dimensional structure of poliovirus receptor bound to 59 poliovirus Introduction 60 Visualization of poliovirus decorated with spvr 60 The structure of the receptor fragment 61 Contact area on poliovirus 65 Model of the receptor fragment 66 Deglycosylated spvr (dspvr) 68 - ii -

7 Chapter VI. Discussion 78 Kinetics of poliovirus interaction with spvr 79 Correlation of CryoEM structure with mutational data 83 An overview of the site of interaction 85 srr mutants possess varying degrees of resistance to spvr 86 Recombinant spvr and soluble Pvr purified from human serum neutralize 87 poliovirus at similar concentrations Kinetic analysis of the effect of poliovirus receptor on viral uncoating: the 88 receptor as a catalysis General rules for picornaviruses receptor interactions 93 Model of poliovirus interaction with cell surface 96 Conclusions and future directions 97 - iii -

8 List of Tables 1 Purification of spvr 43 2 Kinetic and affinity parameters for spvr binding to poliovirus type 1 at 20 C 54 3 Affinity for spvr binding to poliovirus type 1 and abundance of each 56 binding class at different temperatures 4 srr mutants: level of resistance to spvr, location of mutation, and phenotype 91 - iv -

9 List of Figures 1 Structure of poliovirus proteins 4 2 Schematic of poliovirus virion 6 3 Structures of the four Pvr proteins produced by alternative splicing 15 4 A model for poliovirus entry 22 5 A model of genomic RNA transfer across the cell membrane 23 6 Purity of recombinant spvr expressed in mammalian cells 42 7 Neutralization of poliovirus infectivity by spvr 45 8 Kinetics of spvr-induced conformational changes of poliovirus 46 9 Example of raw sensorgram data Effect of low ph treatment on poliovirus infectivity Specificity of spvr interaction with poliovirus on sensor chip Corrected sensorgram overlays for the interaction of decreasing 53 concentrations of spvr with immobilized poliovirus 13 Equilibrium binding sensorgrams and Scatchard analysis of the binding of 56 spvr to immobilized poliovirus 14 CryoEM of poliovirus labeled with spvr Analysis of the Reconstructions Trimming of an asparagine-linked precursor to a high-mannose structure Purity of processed recombinant sdpvr expressed in mammalian cells Neutralization of poliovirus infectivity by sdpvr and spvr CryoEM of poliovirus labeled with sdpvr 74 - v -

10 20 Neutralization of poliovirus infectivity by scd155 and spvr Neutralization of wt and srr mutants Locations of srr mutations in the capsid Diagram of domain 1 of Pvr, with β-strand labeled, abutting inferred 90 contact segments of the capsid proteins 24 Arrhenius plot for the N to A transition in the presence and absence of spvr Bind, remodel, release model for poliovirus entry vi -

11 List of Abbreviations bp BC Base pair Known loops and strands in capsid proteins are denoted by capitals C C Putative loops and strands in Pvr or ICAM-1 domains are bold italic capitals sdpvr HRV k a k d K d kd MPH Mph PVR Pvr spvr scd155 Deglycosylated spvr Human rhinovirus Association rate constant Dissociation rate constant Dissociation constant Kilodalton Murine poliovirus receptor homologue (gene) Murine poliovirus receptor homologue (protein) Poliovirus receptor (gene) Poliovirus receptor (protein) Soluble Pvr Soluble CD155 (spvr isolated from humans) - vii -

12 Acknowledgment The definition of the word mentor, according to the Oxford English Dictionary, is an experienced and trusted counselor. This word describes Vincent Racaniello well. He guided me in a balanced manner and has been both scientifically critical and encouraging. I wish to thank Saul Silverstein, Jeremy Luban, Stephen Goff, and Hamish Young for their encouragement and helpful advice over the years. Additionally, I would like to thank Saul Silverstein for his avuncular spirit, which added color to the everyday. I would like to thank James Hogle, Alasdair Steven and members of their laboratories at Harvard University and The National Institutes of Health who have been valuable collaborators. In particular, David Belnap and Simon Tsang have been generous with their insights, protocols and unpublished data. I am grateful to Roselyn Eisenberg, and Gary Cohen for allowing me to work in their laboratory and use their Biacore biosensor. In particular, I am appreciative of Ann Rux, from the Cohen lab, for assisting with Biacore data analysis. I am thankful to Shazad Majeed and Peter Kwong, from Colombia University, for collaborating with me to make sdpvr and for many interesting conversations. I am grateful to my colleagues in Vince s lab both past and present. Michael Bouchard, Yangzang Dong, Alan Dove, Ornella Flore, Julie Harris, Scott Hughes, Steven Kauder, Du Lam, Sa Liao, Yi Lin, Carl Pavel, Amy Rosenfeld, Juan Salas-Benito, Melissa Stewart, and Suhua Zhang have contributed to my scientific and personal development through thoughtful conversations, advice, humor, and friendship. Members of the Silverstein, Young and Efstratiadis labs have been generous intellectual - viii -

13 companions and they have my sincere thanks. From these labs, I especially thank Christopher Newhouse and Iaonnis Dragatsis for their friendship. Finally, I would like to thank my family and friends, especially my brother, Jonathan, and my wife, Yvonne, who have been patient, encouraging, and loving from the beginning of this endeavor. - ix -

14 Dedication In memory of Lucille, who experienced the beginning of this work, understood its difficulty, smiled, and named spvr The Great White Protein - x -

15 Chapter I. Introduction - 1 -

16 Picornavirus crystal structures Picornaviruses are naked icosahedral viruses that are typically composed of sixty copies of four polypeptide chains, designated: VP1 (viral protein 1), VP2, VP3, and VP4. VP4 is an internal capsid protein that does not have access to the surface of the protein shell in the crystal structure. VP1, VP2, and VP3 form the exposed surface of the protein shell; these three capsid proteins contain a common central structural element known as β-barrel jelly roll (Figure1). VP1-3 differ in the loops that contact the β-strands and in the amino and carboxy terminal extensions. The surface peaks are formed by loops connecting the β-strand and the carboxy terminal extensions that are exposed to the exterior of the virion. A byzantine network formed by the amino terminal extensions of VP1, VP3, and VP4 characterizes the interior of the capsid. The β-barrel jelly rolls of picornavirus structural proteins are similar in structure to the core domain of capsid proteins of a number of plants, insect, and vertebrate (+) strand RNA viruses, such as tomato bushy stunt virus (TBSV). The crystal structure of poliovirus has revealed that the proteinacious coat is organized in a hierarchical manner of repeating units. The simplest is the promoter, a heteromeric structural unit of the poliovirus capsid that contains one copy of each of the peptide chains (63). Five copies of VP1 surround the five-fold axis of symmetry, three copies of VP2 and VP3 alternate around the three-fold axis, and VP4 is located on the interior of the virion. The capsid surface is characterized by three prominent features: a mesa, the highest peak located at the five-fold axes of symmetry; the propeller, a smaller peak at the three-fold axes of symmetry; and the canyon, a surface depression that - 2 -

17 arrangements exist in many other picornaviruses, such as rhinovirus (122). Conservation of capsid structure among picornaviruses suggests information about one virus may be relevant to another. Furthermore, the features and mechanisms of virus entry, such as, receptor binding loci, receptor binding behavior, and the uncoating process may also be conserved throughout the picornavirus family. At the interface between five-fold related protomers are structural features that regulate conformational transitions during virion assembly, as well as, cell entry. The temperature-sensitive (ts) phenotype of the Sabin vaccine strain of poliovirus type 3 (P3/Sabin) is caused by a Phe to Ser mutation at residue 91 of VP3. In the virion, residue 91 of VP3 is on the surface in the canyon near the interface between five-fold related protomers (103). Results of temperature shift experiments demonstrated that the temperature sensitivity is expressed during assembly, specifically at the protomer-topentamer step (103); these results suggest that the mutations at non-permissive temperatures appear to affect the correct folding and stability of the capsid proteins and early assembly intermediates. Revertants of this temperature sensitive mutation have allowed for identification of capsid structures involved in assembly and cell entry. These structures are the hydrocarbon-binding pocket of VP1, the β-tube, and the seven-stranded β-sheet. An enigmatic feature of picornavirus capsid structure is a hydrocarbon-binding pocket in VP1. It is located beneath the canyon floor, near the interface between protomers and is believed to modulate receptor-mediated conformational transitions during entry. The pocket is normally occupied by a sausage-like hydrocarbon molecule - 3 -

18 Figure 1: Structures of poliovirus proteins A representation of the physical features of the polypeptide chain in a β-barrel jelly roll is shown at the top left. The β-strands, indicated by arrows, form two antiparallel sheets juxtaposed in a wedgelike structure. The two α-helices (purple cylinders) that surround the open end of the wedge are also conserved in location and orientation in these proteins. The VP1, VP2, and VP3 proteins each contain a central β- barrel jelly roll domain. However, the loops that connect the β-strands in this domain of the three proteins vary considerably in length and conformation, particularly at the top of the β-barrel domain, which, as represented here, corresponds to the outer surface of the capsid. The N- and C-terminal segments of the proteins, which extend from their β-barrel cores, also vary in length and structure. The very long N-terminal extension of VP3 has been truncated in this representation (Adapted from (126))

19 - 5 -

20 5x 3x 2x Figure 2: Schematic of poliovirus virion, showing the names and locations of capsid proteins and genomic RNA (From (126))

21 that resembles sphingosine, but it remains to be verified chemically. Evidence that the hydrocarbon-binding pocket is involved in conformational transitions during virion assembly and cell entry comes from studies of poliovirus mutants. Non-temperature sensitive (ts) suppressor revertants of P3/Sabin have been isolated and characterized (92). The suppressor mutation at a second site introduced a leucine in place of the phenylalanine at VP The side chain of residue VP1-132 is in a pocket that binds the endogenous lipid ligand in the hydrophobic center of VP1. Suppressing mutations are located in, or near, structures believed to regulate virion stability and conformational transitions, the interface between five-fold related protomers and the hydrocarbonbinding pocket (44). Suppressing mutations in the pocket may stabilize the interface during pentamer assembly at the non-permissive temperature by enhancing hydrocarbon binding. Two additional structures believed to be important for virion assembly are a ß- tube and a seven-stranded ß-sheet. The ß-tube is a five-stranded tube of parallel ß structure that is formed by the interaction of the amino termini of five copies of VP3 on the inner surface of the capsid at the five-fold axis of symmetry (63). The ß-tube is flanked by five copies of a three-stranded ß-sheet formed from the amino termini of VP1 and VP4 (63). The amino terminus of VP4 is myristoylated and is thought to direct the interaction between VP3 and VP4 (32). This extensive network between five-fold related protomers can only form upon pentamer assembly and is believed to stabilize pentamers. The seven-stranded ß-sheet is formed near the three-fold axis by four ß-strands of VP3 and the amino terminal extensions of VP1 from one pentamer and the amino terminal - 7 -

22 relies on the association of pentamers into larger structures and presumably stabilizes the pentamer-pentamer interaction. A third class of the aforementioned P3/Sabin ts suppressing mutations is located in the seven-stranded ß-sheet (44) and may counteract the effect of the original ts mutation by stabilizing assembly intermediates. Picornavirus solution structures Analyses of solution structures of poliovirus and rhinovirus have demonstrated dynamic capsid configurations. This is not surprising because a dynamic capsid is necessary for many aspects of the viral life cycle. Early entry related events such as cell attachment, cell entry, and nucleic acid release require many structural transitions of the viral surface. Rhinovirus has been shown to have a dynamic capsid structure; time course mass mapping has revealed that capsid proteins, which are internal in the crystal structure, are transiently exposed in solution at 25 C (86). In this comparative study, treatment with WIN 52084, an antiviral agent that inhibits rhinovirus entry, resulted not in local capsid conformational changes in the drug-binding pocket, but a global stabilization of the entire viral capsid (86). Similarly, the poliovirus capsid is a dynamic assemblage that is capable of undergoing conformational transitions at 37 C. Immunoprecipitation analysis demonstrated that it is possible to neutralize poliovirus with antibody directed against internal capsid residues. Neutralization is a result of reversible exposure of these normally internal sequences at 37 C and antibody binding (87). Since the sequences reversibly exposed at 37 C in the 160S particle are the same as - 8 -

23 conformational dynamics may play a role in cell entry (87). Poliovirus pathogenesis and tissue tropism The mode of poliovirus transmission is by the fecal-oral route (21). Virus initially multiplies in the lymphoid tissue of the pharynx and gut, which results in a transient viremia that facilitates spread of the virus to other susceptible tissues, such as brown fat and muscle (21). In most natural infections viral replication in the gut leads only to a transient viremia. Replication in secondary sites is thought to be essential for the establishment of a persistent viremia, which is necessary for spread of the virus to the central nervous system (CNS). In approximately 1% of infections the virus enters and replicates in the central nervous system, causing lesions in motor neurons of the spinal cord, the brainstem, and the motor cortex (127). The characteristic flaccid paralysis of acute paralytic poliomyelitis results from lysis of motor neurons in the spinal cord. Within an infected host, viral replication is limited to certain tissues and cells. The conventional view was that poliovirus tissue tropism was determined solely by receptor expression patterning or accessibility of cells to the virus (64). However, there is evidence that tissue tropism is determined by cellular factors other than the poliovirus receptor (Pvr), either at the cell surface or in the cytoplasm (45, 102, 119, 137). The development of a transgenic mouse model has advanced the understanding of poliovirus pathogenesis and tissue tropism (76, 120). However, unlike the case for poliovirus infection in humans, poliovirus replication is not detected in the alimentary tract of transgenic mice after oral inoculation. To determine whether Pvr is the sole determinant - 9 -

24 levels of Pvr in M cells and enterocytes were generated (152). Poliovirus was unable to replicate in the small intestine of this transgenic mouse line. These results indicate that Pvr expression is not the sole determinant of the resistance of the mouse intestine to poliovirus replication. Other cellular factors may influence the ability of poliovirus to enter or replicate in cells of the mouse alimentary tract (152). The internal ribosome entry segment (IRES), a cis-acting genomic element at the 5 end of the RNA genome, may play a role in determining tissue tropism of the virus. Cellular proteins bind to IRES sequences, some of which have been shown to be essential for IRES activity in vitro (14, 20, 59, 70, 101). Several lines of evidence support the hypothesis that the IRES participates in the tropism of the virus. A main determinant involved in neurovirulance attenuation of the Sabin vaccine strain of poliovirus type 3 was mapped to the IRES (1). Translation from the IRES of attenuated strains in vitro were shown to be specifically inhibited in cell lines of neuronal origin (55). Furthermore, poliovirus neuropathology in a mouse model was eliminated when the IRES of this virus was replaced by the IRES of rhinovirus (53). Lastly, translation from the IRES of hepatitis A virus in vitro was found to be stimulated 12-fold when fresh liver extracts were added to the assay mixture (50). In contrast to these findings, the IRES of the Theiler s virus does not determine its tropism; IRES activity, in the context of a bicistronic construct, was detected at similar levels in vivo in all examined tissues of the mouse (135)

25 Humans are the only known natural hosts of poliovirus. Monkeys are also highly susceptible to poliovirus when they are inoculated with virus directly into the central nervous system. Other animal species are generally not susceptible to poliovirus. This characteristic species specificity of poliovirus has meant that monkeys have been used as the only animal model for the study of poliovirus neurovirulance and safety testing of oral poliovirus vaccines. The majority of poliovirus strains are host-restricted; for example, P1/Mahoney causes paralysis in primates, but not in mice. In contrast, some poliovirus strains naturally cause paralysis in mice and others have been adapted; for example, the Lansing strain of poliovirus type 2 (P2/Lansing) was adapted to cause paralysis in mice after serial passage in the CNS of cotton rats (5). Intracerebral or intraspinal inoculation of mice with P2/Lansing causes a disease that resembles human poliomyelitis clinically and histopathologically (69, 83), but orally inoculated virus does not replicate in mice (83), or in mouse L cells (106). Transgenic mice expressing the human poliovirus receptor are susceptible to infection by host-restricted poliovirus strains (120), demonstrating that host restriction is, in part, at the level of the virus-receptor interaction. Molecular analysis of viral host range determinants has provided information on the virus-receptor interaction. Intertypic recombinants between mouse-adapted P2/Lansing and host-restricted P1/Mahoney demonstrated that the ability of Lansing virus to cause paralysis in mice is due to the viral capsid (83). A chimeric virus was generated to map the region of the capsid that confers mouse virulence. The region was narrowed to an eight amino stretch of VP1, the B-C

26 conferred mouse-virulence to P1/Mahoney (96, 109). Therefore, the VP1 B-C loop is an important determinant of the mouse virulence of P2/Lansing and the host-restriction of P1/Mahoney. The mechanism by which the B-C loop modulates host range is unclear. One explanation for the importance of this highly exposed portion of the viral capsid in regulating mouse-adaptation is that mouse-virulence is determined by the ability of the virus to attach to a specific receptor in the mouse nervous system; more specifically, host range may be determined by receptor recognition, and the B-C loop might be the binding site for a mouse receptor. The prominent exposure of the P2/Lansing B-C loop at the five-fold axis on the virion surface (63, 85), and the finding that few changes in the B-C loop are compatible with mouse neurovirulence (83, 108) are consistent with this hypothesis. However, host range determinants that suppress B-C loop mutations and confer mouse neurovirulence to P1/Mahoney have been identified on the interior of the capsid in the amino terminal network (36, 108), and it is unclear how they might directly regulate receptor recognition. Another possibility is that mouse neurovirulence is not determined by receptor binding. It is possible that there is a relationship between mouse adaptation of poliovirus and the involvement of VP1 B-C loop conformational changes required for infectivity (150). The mouse receptor may be able to bind to all three serotypes of poliovirus, but it may not be capable of inducing the structural changes in P1/M and P3/S required for infectivity. It is also conceivable that thevp1 B-C loop of P2/L allows the virus to enter by a different route of infection. The VP1 B-C loop conformation may have a dramatic

27 loops and possibly the flexibility of the capsid (44, 150). Structural cross talk may exist between VP1 B-C loop and conformations of other nearby sites. This situation would be similar to one observed for the foot-and-mouth disease virus. It has been proposed that the flexibility of the antigenic/receptor-binding loop (G-H loop of VP1) of foot-andmouth disease virus not only directly affects antigenicity, but also influences host cell interactions-possibly by structural cross talk. The cellular poliovirus receptor Recognition of a cell surface receptor is the first step in infection of cells by animal viruses. The poliovirus receptor is a member of the immunoglobulin superfamily that is used by all three viral serotypes to initiate infection of cells (102). The receptor is a type I integral membrane protein with a primary amino acid sequence that encodes an N- terminal secretion signal, three extracellular immunoglobulin (Ig)-like domains, a transmembrane region, and a cytoplasmic domain. There are four putative alternately spliced mrna variants of the receptor hnrna: two variants lack the transmembrane region and appear to be secreted; and, two variants that serve as poliovirus receptors encoding polypeptides of 392 and 417 A.A., differing in the lengths of their cytoplasmic domain. (Figure 3) (75, 102). The ectodomain of each of the forms contains 8 putative N- linked glycosylation sites (102) that shift the molecular mass from the predicted 43 or 45 kda, for the membrane bound forms of Pvr, to a predominant species of about 80 kda on a SDS-PAGE gel. Post-translational modification of the receptor is not necessary for entry of the virus, although destruction of the first glycosylation site by mutagenesis

28 glycosylation may partially occlude virus docking (154). The prediction that the extracellular portion of Pvr contains three Ig-like domains is based on consensus homology. An Ig-like domain is defined by sets of conserved residues that permit the polypeptide to from a globular tertiary structure called the antibody-fold. Antibody folds contain amino acids that form two sets of β-pleated sheets, each consisting of three or four anti-parallel β strands of five to ten amino acids in length (147). The interior of the fold is lined with hydrophobic residues, which alternate with outer facing hydrophilic residues. Intra-chain disulfide bonds are usually present on every domain and contribute to the stability of the structure (147). Antibody folds are broadly classified as V- or C-like, based on their relative homology to either the variable or the constant domains of immunoglobulin molecules. Domain 1 of Pvr is V-like while the other two more membrane proximal domains are C-like (102). V-like domains are generally larger than C-like domains, and usually contain an extra pair of β strands. A number of Pvr-related proteins have been identified, but not all of them function as receptors for poliovirus. As a group, most of these proteins serve as receptors for a number of viruses. All Pvr-related proteins are members of the immunoglobulin superfamily. Agm1 and Agm 2 receptors, isolated from African green monkeys, are functional receptors for poliovirus (78). Pvr mediates the entry of pseudorabies virus (PRV) and bovine herpesvirus-1 (BHV-1), but has no activity for herpes simplex virus (HSV) strains. Two other Pvr-related proteins, which do not serve as receptors for poliovirus, were originally designated poliovirus receptor-related protein 1(Prr1) (91) and poliovirus receptor-related protein 2 (Prr2) (43), serve as HSV entry mediators. Prr2 was

29 H 20 A H 20 B H 20 A 1 H 20 A 2 45 kda 43 kda 40 kda 39 kda Figure 3: Structures of the four Pvr proteins produced by alternative splicing. The membrane bound forms of Pvr are labeled H 20 A and H 20 B. The secreted forms of Pvr are labeled H 20 A 1 and H 20A 2. The putative Ig-like domains are represented by circles; leftmost and largest is the V-like domain. The green boxes represent transmembrane regions. The pink line denotes the C terminal region of the H 20 B variant, which is unique

30 mediator (HVEM) for entry. It has been renamed herpes virus entry protein B (HveB) (144). An alternate name for HveB is Mph, mouse poliovirus receptor homologue. Prr1 was found to serve as an entry receptor for HSV-1, HSV-2, PRV, and BHV-1 and was recently designated herpes virus entry protein C (HveC) (48). Pvr s cellular function and its role in development are unknown. The cytoplasmic domain of one of the Pvr isoforms is phosphorylated at a serine residue, possible by calcium/calmodulin kinase II (18). Insights into the function of Pvr can be gained by studying the function of a Pvr homologue in mice and in in vitro culture systems. Mph is a Ca 2+ -independent homophilic cell adhesion molecule. The cytoplasmic tail of Mph interacts with 1-afadin, an actin filament (F-actin)-binding protein with one PDZ domain. This interaction creates a link between the cytoplasmic tail of Mph to the actin cytoskeleton. The interaction of Mph with 1-afadin is not essential for its cis-dimerization or trans interaction (104), but is essential for the colocalization of MPH and 1-afadin with E-cadherin at cell-cell adherence junctions (141). Recently, to provide information on the function of Pvr family members, Mph was targeted for disruption in the mouse. Disruption of both alleles of the Mph gene resulted in morphologically aberrant spermatozoa. The spermatozoa contained defects in nuclear and cytoskeletal morphology, mitochondrial localization, and F-actin distribution (22). These results suggest signaling through Mph may be crucial for the cytoskeletal organization and reorganization that occur during spermatogenesis (22)

31 A model of poliovirus entry is displayed in figure 4. Shortly after poliovirus binds to cell surface Pvr, it releases its genomic contents into the host cell cytoplasm. Pvr plays two roles in virus entry: first, the receptor acts as a tether to which the viruses attach and concentrates particles on the cell surface; second, it facilitates the viral uncoating step by inducing dramatic structural changes in the virus particle (51). When poliovirus is bound to cells at 37 o C, the major population of virus is eluted as a conformationally altered form known as the A particle. A minor population also elutes off the cell surface and is termed either the 80S particle or the empty capsid. The altered form of the virion is also referred to as the 135S particle, because it sediments at 135S on a sucrose gradient, compared to 160S for the native virion. This conformational alteration results in the release of the internal capsid protein VP4 and the externalization of the N terminal extension of VP1 (46). The N-terminus of VP1 allows the altered particles to associate with lipid bilayers (46), it is also thought to be involved in formation of a pore. Multiple N termini, which have amphipathic helical character, insert into the cell membrane producing a pore through which the viral genome may pass to enter the host cell cytoplasm (Figure 5). The A particle has been proposed to be an essential intermediate in the entry of poliovirus into cells (46). Isolation of cold adapted mutants (Ca) provide evidence against the A particle as an intermediate in the entry process. The mutants infect efficiently at 25 C without formation of 135S particles. These mutants can form 135S particles at 37 C (42). These data suggest formation of 135S particles is not required for poliovirus replication

32 high concentrations of the 135S particle are capable of infecting cells in a receptor independent fashion (38). Moreover, binding of 135S-antibody complexes to the Fc receptor attached to the cell surface increases the infectivity of these particles by 2 to 3 orders of magnitude (66). Thus, the low efficiency of infection by 135S particles is due, in part, to the low binding affinity of these particles. Furthermore, it has been shown that there is an additional stage in the entry process that is associated with RNA release. This Pvr independent stage occurs after formation of the 135S particle and is rate limiting during infection at 37 C, but not at 26 C. This study also demonstrated that during infection at 26 C, the rate-limiting step is the Pvr-mediated conversion of wild-type 160S particles to 135S particles. From these data, it has been suggested that infection at 26 C by the cold-adapted viruses allows the 135S particles to be formed, but they fail to accumulate to detectable levels because the subsequent post-135s particle events occur at a much faster rate than the initial conversion of 160S to 135S. This indicates a model in which the 135S particle as an intermediate during poliovirus entry is possible (66). In the model presented in figure 4, after the virus forms the A particle some particles may release their RNA. The coincident release of the RNA genome and VP4 from the particle results in the formation of the 80S particle (Figure 4). This process is known as uncoating. Tomato bushy stunt virus (TBSV) has a capsid structure similar to poliovirus and undergoes a well-characterized particle expansion that may be similar to poliovirus alteration. Exposure of TBSV to an alkaline environment devoid of cations causes reversible expansion of the virus (81). Viral enlargement is controlled by an interface

33 intermediates similar to the intermediates formed by poliovirus. Resolution of the TBSV crystal structure, in conjunction with biochemical studies of the expanded particles, has revealed that the buried amino termini of capsid subunits are extruded through holes in the disrupted interfaces (Harrison et al., 1986). Based on these observations, analogies between TBSV expansion and poliovirus alteration have been made. It has been suggested that poliovirus is similar to the TBSV and the analogous interface between five-fold-related protomers controls poliovirus alteration. Interestingly, capsid protein VP4 and the amino terminus of VP1 are located immediately below the interface (63) and appear to be hydrogen-bonded to each other, providing a mechanism for the coordinated release of these internal components during alteration (44). A poliovirus VP4 mutant has been isolated that binds Pvr and undergoes alteration, but is blocked at a subsequent entry step (107). Other mutants with deletions in the amino terminus of VP1 have delayed RNA release (74). These mutants lend support to the idea that VP4 and the amino terminus of VP1 are involved in early entry events. Interestingly, there are significant differences between the 135S structure and the expanded structure of the TBSV. The expanded forms of the plant viruses have holes large enough to be visualized by low resolution; however, the gaps between VP1 subunits are barely large enough to allow passage of an extended polypeptide chain. This suggests that during the poliovirus alteration process the holes are larger (12). The subcellular location where the virus releases its RNA into the cytosol is unclear. Although significant progress has been made, the confounding problem in determining the subcellular location of entry is partly due to the inefficiency of virus

34 determine the subcellular location of viral genome release by standard viral metabolic labeling and cell fractionation techniques. Electron micrographs of poliovirus particles have been observed in coated pits and vesicles after adsorption, suggesting that virus might enter via receptor-mediated endocytosis (151). Further studies have been preformed to determine if poliovirus entry is dependent on clathrin, the transmembrame or cytoplasmic regions of Pvr, or a low-ph step to enter the cytoplasm. Virus entry studies were performed using HeLa cell lines that express a mutant form of dynamin; these cell lines specifically block the formation of clathrin-coated pits and vesicles without other pleotropic effects (6). It was demonstrated that human rhinovirus 14 depends on the clathrin pathway to enter Hela cells, whereas poliovirus does not (40). Moreover, poliovirus is able to infect cells using recombinant Pvr receptors that have different primary amino acid sequences in the transmembrane and cytoplasmic regions substituted from other cell surface molecules. These results indicate that Pvr amino acid sequences in the cytoplasmic and transmembrane regions are not necessary for infection in cultured cells (106, 131, 132). The requirement of a low-ph step during poliovirus entry has been investigated by using the macrolide antibiotic bafilomycin A1, which is a powerful and selective inhibitor of the vacuolar proton-atpases; it was demonstrated that poliovirus infection is not affected by the antibiotic (112). The presence of lysosomotropic agents such as chloroquine, amantadine, danaylcadaverine, and monensin during poliovirus entry did not inhibit infection (112). These data support the hypothesis that poliovirus does not depend on a low-ph step to enter the cytoplasm (112)

35 135S and 80S particles have been determined to 22 Å resolution (12). Based on these structures, the tectonic plate model has been formulated to describe the structural changes the virus undergoes during conversion catalyzed by the receptor. Domain movements of up to 9Å were detected that create gaps between adjacent subunits. The gaps formed where VP1, VP2, and VP3 subunits meet and are predicted to be the site of emergence of VP4 and the N terminus of VP1. In the transition, VP4 and the N termini of VP1 are extruded from the bottom of the canyon and arranged around the outside of the mesa, where five copies of the predicted amphipathic helices of the N terminus of VP1 would be ideally positioned for membrane insertion (12). It is possible that the N-terminal myristate of VP4 and perhaps other regions of VP4 also may be imbedded into the membrane. This could facilitate insertion of the VP1 N termini into the membrane (12). To create a channel at the five-fold axis, a transmembrane pore is formed and the VP3 plug is moved out of the way (Figure 5). The genome then exits the particle and enters the cytoplasm. During this process, the 80S particle is formed by shifts of the VP1, VP2, and VP3 subunits and the fivefold plug is restored to its original position (12). Kinetic and affinity analysis of picornavirus-receptor interactions Kinetic and affinity parameters of the various viruses with soluble forms of their receptors have been determined. Such parameters are important because they describe the interaction of virus with receptor, which enables a better understanding of the reaction and its comparison to other systems. The results of such studies, together with structural

36 Poliovirus Entry native virion 160S extracellular A particle >33 o C Pvr A particle uncoating empty capsid 80 S Figure 4. A model for poliovirus entry. 160S native virions bind to Pvr on the cell surface and, at temperatures greater than 32 C, undergo the receptor-mediated conformational transition to 135S altered particles (A particles). While a large percentage of A particles elute from the cell, some remain cell-associated and release their RNA into the host cell cytoplasm. The cellular location of uncoating, which yields the 80S particle, is unclear

37 Figure 5: A model of genomic RNA transfer across the cell membrane. VP1, VP2, VP3, and VP4 are colored cyan, yellow, red, and green, respectively. In this model, the beta-tube of VP3 (red) forms a plug at the fivefold axis that separates the virus interior from the outer surface. Attachment of the 160S particle (left) to the poliovirus receptor (three gray circles) triggers conversion to the 135S form (center). As conversion commences, cell attachment is mediated by externalized VP4 (green tubes) and the N termini of VP1 (blue tubes). The N termini emerge from the bottom of the canyon and extend along the sides of the fivefold mesa towards the apex. After the N-terminal helices of VP1 have inserted into the membrane, they rearrange to form a pore (right). To permit the RNA (purple tube) to pass through the pore into the cytoplasm, the VP3 beta-tube (red rectangle) may shift on its 40-residue tether (red tube) and the VP1 barrels could splay farther apart (From (12)

38 events in infection (61, 80, 125, 153). Surface plasmon resonance (SPR) has been used to study affinity and kinetics of the interaction of echovirus 11 with its cellular receptor decay-accelerating factor (CD55). This virus-receptor interaction is monophasic: there is a single affinity (K d ) for the interaction. The interaction was reported to be similar to cell-cell recognition molecules with a relatively low affinity of K d =3.0 µm as a result of a very fast dissociation rate (k a ~10 5 M -1.s -1 k d ~0.3 s -1 ) (84). In contrast to the Echovirus 11 virussoluble receptor interaction, rhinovirus-sicam interaction in solution and using surface plasmon resonance were determined to be biphasic: i.e., there are 2 affinities of the virus for the receptor. These SPR studies have shown that there are two different association rates and a single dissociation rate for the interaction. Each class of binding site comprises 50 % of the total at 20 C, with association rate constants of 2450 and 134 M -1 s - 1. The off rate for human rhinovirus 3 was 1.7 x 10-3 s -1 to yield a calculated dissociation constant of 0.7 and 12.5 µm (28). It has been proposed that the 2 binding classes may be a consequence of virus breathing and the receptor binding site may exist in at least 2 different conformational states. The single dissociation and biphasic association rates indicate a unique interaction site with variable conformation or accessibility to the receptor (26). Interestingly, poliovirus demonstrates two modes of binding to the surface of cells that express cell surface Pvr, a lower and a higher affinity class (8). Each of the classes could represent two different receptors, different states of the same receptor, or two states of the virus

39 The "canyon hypothesis" has been proposed to explain binding of rhinovirus to its cell receptor, intercellular adhesion molecule-1 (ICAM-1) (121). The structural similarity between poliovirus and rhinovirus raises the possibility that the poliovirus canyon may be the binding site for Pvr. The first draft of the canyon hypothesis predicted that conserved residues on the floor of the HRV-14 canyon bind to the cell receptor, but the small dimensions of the canyon should limit accessibility to antibodies. Furthermore, HRV-14 morphology should allow residues at the top of the canyon to mutate freely without affecting virus receptor interactions (121). However, the latter aspect of the hypothesis was nullified by the determination of the three-dimensional structure of intact human rhinovirus 14 complexed with Fab fragments, Fab17-IA. Fab17-IA penetrates deep within the canyon and overlaps the region the receptor binds. Hence it is unlikely that viral canyon quaternary structure evolved merely to evade immune recognition. Instead, the shape and position of the receptor binding region on rhinovirus probably dictates receptor binding and subsequent uncoating events. (139). Availability of the three-dimensional crystal structure of both rhinovirus and the first two domains of ICAM-1 has made it possible to investigate their interaction (10, 11, 79, 122). Similar to poliovirus genetic data (33, 35, 88), rhinovirus studies support the hypothesis that the canyon contains the binding site for ICAM-1. Unfortunately, no atomic resolution crystal structure of the complex between rhinovirus and its receptor is available. However, low-resolution cryoem image reconstructions of complexes of soluble ICAM-1 fragments with HRV16 (111), HRV14 (111) and HRV3 (149) have been determined. Complexes of the first two domains of ICAM-1 with HRV16 and HRV

40 ICAM-1 fragment has a dumbbell-like shape in these reconstructions (111) and is oriented roughly perpendicular to the viral surface. Atomic coordinates from the ICAM-1 receptor fragment (9) and HRV16 (110) crystal structures have been used to fit a molecular model into the CryoEM image reconstruction (10). The modeled interaction that shows the variable loops of the receptors BC, DE, and FG penetrate deep into the canyon, and the short CD loop of ICAM-1 lies against VP2 of HRV16, on the south rim of the canyon. The footprint of ICAM-1 on the HRV16 surface is essentially the same as HRV 14, and the analysis of the charge distribution on the two interacting surfaces shows remarkable complementarity (79), indicating that surface charge may play an important role in rhinovirus ICAM-1 recognition. Coxsackievirus A21 (CAV21), like HRVs, is a causative agent of the common cold. It uses the same cellular receptor, ICAM-1, as does the major group of HRV (133). It can also bind to decay-accelerating factor, although without causing infection (134). Interestingly, CAV21 differ from HRVs in that it is stable at acidic ph. Recently, the cryoem image reconstruction has been reported, at 26 Å, of CAV21 complexed with ICAM-1(148). The rendering demonstrates that domain 1 of ICAM-1 binds to a similar site as major group HRVs in the CAV21 canyon vicinity. Foot-and-mouth disease virus (FMDV) causes a highly infectious disease in cloven-hooved animals, which can be economically devastating. FMDV enters cells through receptor-mediated endocytosis followed by binding acid ph-dependent release and translocation of RNA across the endosomal membrane(24, 37, 98). The predominant cell surface ligand is heparin sulfate (HS) for certain strains of O 1 FMDV; its attachment

41 for O 1 FMDV bound to HS has revealed the binding site is a shallow depression, the pit, on the virion surface, located at the junction of the three major capsid proteins, VP1, VP2, and VP3 (47). Human rhinovirus 2 is a minor group rhinovirus that binds to the very low-density lipoprotein (VLDL)-receptor (VLDL-R), a member of the LDL-receptor family. A cryo EM and three-dimensional image reconstruction of the HRV2 and soluble fragments of the VLDL-R complex have been determined to 15Å. Surprisingly, the receptor fragments bind sites on the virus very different from major group rhinoviruses. The minor group receptor binds to the star-shaped mesa on the 5-fold axis rather than in the canyon (60). The difference in binding loci may begin to explain differences in uncoating mechanism between major and minor group rhinoviruses. More specifically, binding of ICAM-1 to HRV14 initiates rapid uncoating at physiological temperature and at neutral ph without the need for any cellular components (65, 97). In contrast, binding-ldl receptors to HRV2 does not directly promote uncoating (54, 93-95). Internalization of HRV2 into acidic an endosomal compartment is required for uncoating and transfer of viral RNA into the cytosol (115, 116, 130). It has been suggested that ICAM-1 binds to the major group HRV in a two-step mechanism (26, 79), in which the Cryo EM reconstructions of HRV complexed with soluble ICAM-1 fragments represent the initial recognition event. In a subsequent step, the receptor could move slightly to allow the north wall of the canyon, consisting of VP1 residues, to bind to domain D1. The resulting conformational change in the virion may move VP1 away from the 5-fold axis, thereby opening a channel at the pentamer vertex through which the N termini of VP1,

42 occur for HRV2 because its receptor binds to the peak at the five-fold axis. Viral capsid sequences that regulate receptor binding A detergent-solubilized form of the poliovirus receptor has been produced from insect cells, expressing a membrane bound form of the receptor. These solubilized forms of the receptor bind the virus and are capable of blocking infectivity and inducing the conversion of the 160S to the 135S and 80S products (71, 154). This system was used to identify poliovirus capsid residues involved in the virus-receptor interaction by isolating spontaneously occurring viral mutants resistant to neutralization by solubilized receptor, but remain infectious in cultured cells (33, 71). Neutralization by detergent-solubilized receptor, or membrane bound receptor, is dependent on two processes that are interrelated: receptor binding and conversion. 21 srr mutants were isolated each containing a single mutation located on the surface or the interior of the capsid (Table 4, Figure 22). One of the mutants contained a double mutation that displayed a very different phenotype. The single mutants loosely fit into one of two categories (31, 33). One class of mutants have significantly reduced affinity for the receptor on the surface of cells with or without reduced ability to undergo alteration after binding to cells. A second class of mutants has moderately reduced affinity for the receptor on the cell surface in addition to, a reduced ability to undergo alteration after binding to cells. The single double mutant has a phenotype that is not easily explained; it binds receptor with a similar affinity as wild-type virus and has an increased ability to undergo alteration upon interaction with the surface of susceptible

43 or the interface between fivefold-related promoters (Table 4). The locations of the residues altered in these mutants strongly implicate the canyon in receptor contact (Figure 22). Changes in residues 1226, 1228, 1231 and 1234, which are located at the surface of the virus in the canyon, are mutations that reduce binding affinity to the cell surface. These mutations may reduce the binding affinity by interfering with receptor contact. Mutations at internal capsid residues also reduce binding affinity. These residues are not likely to contact the receptor directly, but may affect the binding in two ways: (1) they may distort the primary site of interaction by steric interference; or, (2) they may impede the poliovirus or the receptor from undergoing the necessary structural rearrangements necessary to reach a final bound state. Further information on capsid residues that regulate interaction comes from the analysis of adapted viral mutants that grow on cells expressing mutant forms of Pvr that do not bind wild-type P1/Mahoney. Pvr mutants were constructed by substituting residues of Pvr with corresponding residues from Mph. Since Mph is not a poliovirus receptor, the described receptor mutagenesis locates regions of Pvr important for poliovirus binding. Three cell lines were derived that cannot bind wild-type virus; however, viral variants were readily isolated that can utilize the mutant Pvrs to infect cells (105). These viral mutants have expanded receptor recognition because they are still able to utilize wild-type Pvr to infect cells. Sequence analysis of the mutant viruses revealed three capsid residues that enabled poliovirus to utilize the defective receptors. Amino acid changes in adapted mutants are located in the VP1 B-C loop at the five-fold axis, at the interface between protomers near the hydrocarbon binding pocket, and in the

44 mutations and restore the virus receptor interaction or by affecting receptor contact by regulating structural transitions the capsid undergoes during receptor interaction (105). Pvr sequences that contact virus The binding site for poliovirus appears to be contained within domain 1 (77, 106, 132), which can bind poliovirus when expressed on the cell surface either alone or in a chimeric molecule fused to CD4 (89), ICAM-1 or Mph. Interaction strength is diminished when only domain 1 is present on the membrane indicating that domains 2 and 3 participate either directly or indirectly by modulating the structures of domain1. Mutagenesis of PVR domain 1 has identified the three putative contact regions for poliovirus: the C-C loop through the C strand, the border of the D strand and the DE loop, and the G strand (3, 16, 105) (Figure 23). Mutagenesis of other loops and strands has not revealed other regions that are important for binding. The use of molecular modeling in conjunction with viral genetics revealed that the C -C ridge is likely to be the main part of Pvr that contacts poliovirus. The homologous part of CD4 plays a major role in the interaction with human immunodeficiency virus type 1 (82). The D-E loop of domain 1 may also contact poliovirus, but the G strand is more distant and not likely to be directly involved with the binding site (105). A chimeric receptor containing the predicted C C D loop-strand region of Pvr was substituted into the corresponding region of Mph. L cells transformed with the chimeric receptor became permissive for P1/M but not P2/L and P3/S serotypes of poliovirus. This indicated that the three serotypes of poliovirus contact Pvr slightly differently (89). These

45 acid segment, although the contribution of conserved and similar Mph residues cannot be excluded. Hydrophobic antiviral agents Members of the picornavirus family are sensitive to a family of generally hydrophobic antiviral agents. The drugs include the WIN compounds and a class of compounds produced by Janssen Pharmaceutica. Crystallographic and cryoem image reconstruction studies of virus-drug complexes have shown that these compounds displace the pocket factor and bind in the hydrophobic core of VP1 (52, 56, 140). Drug binding has been shown to inhibit infectivity by two different mechanisms. One, drug binding to virus stabilizes virus, which in turn, inhibits structural transitions necessary for genome release. The inhibition of transition is caused by an entropic effect (142) mediated by drug not through increased rigidity of the capsid as has been previously suggested (100, 140). Second, for some rhinoviruses, drug binding induces limited local conformational changes in the loops of viral proteins in the canyon; this inhibits receptor attachment (113, 136). Recently, it has been shown that drug binding interferes with poliovirus receptor attachment at 4 C but not at physiological temperatures (42). Drug binding by poliovirus does not result in significant local structural changes at the base of the canyon (52). These results suggest inhibition of receptor attachment also may be attributed to the ability of drugs to inhibit small energy-dependent conformational alterations required for tight receptor binding

46 Chapter II. Materials and Methods

47 Cells and viruses 293-T human epithelial kidney cell line cells were propagated in Dulbecco's minimal essential medium (Life Technologies, Inc.) containing 10% fetal bovine serum (HyClone), 100 units of penicillin/ml, and 100 µg of streptomycin/ml (Life Technologies, Inc.). HeLa cells were propagated in Dulbecco's minimal essential medium containing 10% bovine calf serum, 100 units of penicillin/ml, and 100 µg of streptomycin/ml. Hybridoma cell line 711C (106) was propagated in HB basal medium plus HB101 supplement (Irvine Scientific). Poliovirus type 1 Mahoney strain, derived from an infectious cdna clone (118), was grown in HeLa cells and purified by differential centrifugation and CsCl density gradient fractionation as described in (124). The ratio of particles to plaque forming units was determined to be 250:1. Plasmid construction Polymerase chain reaction was used to amplify a portion of PVR cdna that corresponds to the ectodomain, residues DNA encoding 5 histidine residues and a termination codon were added to the 3'-end during amplification. The 5 histidine residues were added after a naturally occurring histidine (His-337) in Pvr. The following oligonucleotide primers were used: 5'-ttgagagacaattgGGAAGCGAGGAGACGCCCG-3' and 5'-gggagtgacaattgctaatggtggtgatggtgGTGCTCACTGGGAGGTCCCT-3'. Codons for the additional 5 histidine residues are shown in bold. The Pvr sequence is in capital letters. The amplified DNA product was inserted into the first cistron position of the bicistronic vector pcmv/ires/gfp, resulting in p3dpvr/ires/gfp/mp8. Expression of this DNA in mammalian cells should produce a bicistronic mrna in which the first

48 (IRES) and the second cistron, which encodes green fluorescent protein (GFP). Establishment of a Stable Cell Line Expressing spvr 293-T cells were seeded in 10-cm diameter plastic cell culture plates 1 day before use. After cells achieved 20% confluence, and 6 h before DNA transfection, the medium was changed. Ten µg of plasmid p3dpvr/ires/gfp/mp8 plus 0.1 µg of prsv-puro, a plasmid that contains the puromycin resistance gene, was introduced into 293-T cells by DNA-calcium phosphate coprecipitation (146). After 18 h of incubation at 37 C, the medium was replaced, and incubation was continued for an additional 24 h. For selection and subculturing of drug-resistant cells, 5 µg of puromycin (Sigma) was added per ml of medium. The subpopulation of puromycin-resistant 293-T cells was detached from the tissue culture plate using cell dissociation buffer (Life Technologies, Inc.), passed through a 20-µm nylon filter, chilled, and sorted on a Becton Dickinson FACStar with the excitation wavelength set at 488 nm. A small percentage of the population was sampled to determine the range of fluorescence intensity. A subpopulation of 1 X 10 7 GFP-expressing cells with relative fluorescence intensity greater than 95% of the whole population was collected on ice. GFP-positive cells were cultured for a week, and the isolation process was repeated with the following modifications. Individual cells with relative fluorescence intensity greater than 99.75% were clonally isolated and cultured. Of several cell lines obtained by this procedure, one with the highest level of secretion of

49 (data not shown). Protein expression, purification, and modification At each step of spvr purification, total protein (Pierce, Inc.), specific activity, and fold purification were determined. A unit of specific activity was defined as a 20-µl aliquot capable of neutralizing 50 of 100 plaque-forming units (pfu) in a 100-µl reaction volume. Fractions containing spvr were determined by 10% SDS-polyacrylamide gel electrophoresis and by Western blot analysis (data not shown). spvr enriched supernatant was produced by seeding 15-cm diameter plastic cell culture plates to 50% confluency with GFP-positive cells in 25 ml growth media and subsequent culturing for days without media replacement. Enriched supernatant was mixed with loading buffer A (LBA, final concentration 50 mm NaPO4, 50 mm NaOH, ph 8, 3 mm imidazole) and nickel-agarose slurry (final 4% v/v) (Qiagen, Inc.), and incubated at 4 C overnight with stirring at 1000 rpm. The slurry was packed into a column (1-cm Bio-Rad Econo-column with flow adapter) and washed with 10 volumes of wash buffer A (WBA, 10 mm imidazole in phosphate-buffered saline (PBS), 20 mm NaPO4, 150 mm NaCl, ph 7) at a flow rate of 1.5 ml/min. Bound protein was eluted with elution buffer A (EBA, 50 mm imidazole in PBS) at 4 C at a flow rate of 0.5 ml/min. The sample was dialyzed overnight against loading buffer B (LBB, 20 mm HEPES-NaOH, ph 8.0, 20 mm NaCl) prior to application to a Q-Sepharose column (1-cm Bio-Rad Econo column packed with Q-Sepharose resin charged with counter ions and preequilibrated with loading buffer) at a flow rate of 0.5 ml/min. After washing with 10 volumes of LBB, at a flow rate of

50 in 20 mm HEPES-NaOH, ph 8.0, at a flow rate 0.5 ml/min. One-ml fractions were collected, and spvr-containing samples were dialyzed against PBS, and fold purification was determined. For the production of deglycosylated spvr, the GFP-positive clonal population was expanded and grown for days on 150-cm tissue culture plates that contained 25 ml of growth medium with 100mM deoxymannojirimycin, with repeated pulses of deoxymannojirimycin every two days for ten days. A Ni agarose enrichment step was carried out as above. The eluent was dialyzed in the presence of endoglycosidase H in dialysis/ treatment buffer (D/TB: 10 mm EDTA, 300 NaCl, 20 mm Hepes, 50 mm Acetate, ph-5.9) overnight at 4 C to allow the reaction to go to completion. The ph of the reaction was adjusted to 7.5 and the concentration of CaCl and MnCl were adjusted to 5mM and 1mM, respectively. Partially glycosylated spvr was removed by application to a double media column: upper layer contained 10 ml G25 media and the bottom contained 5 ml concanamycin A (Sigma, Inc.), preequilibrated with gel filtration buffer (GFB: 300 mm NaCl, 5 mm Tris-Cl, ph 7.0, 0.02 % NaN 3 ). The endoglycosidase H was separated from sdpvr using an S-200 size exclusion column (Pharmacia, Inc.) preequilibrated with GFB. The anti-pvr monoclonal antibody 711C (105) was purified from hybridoma supernatant using Affi-Gel protein A gel according to the manufacturer's instructions (Bio-Rad)

51 Virus neutralization assay and determination of buffers for use with the optical biosensor Approximately 200 pfu of poliovirus were incubated with different concentrations of spvr in virus dilution buffer (PBS containing 0.02% bovine calf serum) for 30 min at 25 C followed by 1 h at 37 C. The virus titer was then determined by plaque assay on HeLa cell monolayers. To determine if incubation at low ph, the condition used to remove bound spvr from virus on the sensor chip, affects virus infectivity, approximately 70 pfu of poliovirus were incubated in 10 mm glycine, ph 3, or PBS, for 5 min, followed by plaque assay. Plaque assays were carried out essentially as described in (83). Alteration assays Preparation of isotopically labeled virus, purification, and alteration assays were done essentially as reported in (4, 71, 90). For alteration assays, spvr was incubated with purified poliovirus in PBS containing 1% bovine serum albumin (Sigma) in a total volume of 100 µl at 4 C overnight. The virus-receptor complexes were then shifted from 4 to 37 C for 0, 5, or 15 min, overlaid onto a 15-30% sucrose density gradient containing 0.1% bovine serum albumin, and centrifuged at 39,000 rpm for 2 h at 4 C in a Beckman SW41 rotor. Gradients were fractionated (0.6 ml) from the top to bottom, and radioactivity was measured in a liquid scintillation counter. In such assays, not all of the sample can be accounted for, probably due to the hydrophobic nature of the 135 S particle (46)

52 Binding of spvr to poliovirus using an optical biosensor Surface plasmon resonance experiments were performed on a BIAcore X and BIAcore 3000 optical biosensor (BIAcore AB) at specified temperatures. Approximately 1,200 response units of purified poliovirus were coupled to flow cell 2 (Fc2) of a CM5 sensor chip via primary amines according to the manufacturer's specifications with the following modifications. After the activation step, purified poliovirus in PBS was diluted 1:3 with 10 mm sodium acetate, ph 4.5, and injected at 2 µl/min until desired response units were coupled to the flow cell. The running buffer for the experiments was PBS containing 0.005% Tween 20 (PBS-T, ph 7.0). For kinetic analysis of the spvr-poliovirus interaction, the flow path was set to include both flow cells; the flow rate was 50 µl/min, and the data collection rate was set to high. Poliovirus was allowed to bind for a 2-min interval with a wash delay set for an additional 3 min to allow for a smooth dissociation curve. Settings for equilibrium analyses were the same as for kinetics, except that the flow rate was set to 2 µl/min. Regeneration of the virus (removal of bound spvr) was done by brief pulses of 10 mm glycine at ph 3.0 with or without 300 mm NaCl until the response was returned to base line. BIAevaluation software, version 3.0, was used to analyze the surface plasmon resonance data, using global fitting

53 Chapter III. Two Distinct Binding Affinities of Poliovirus for Its Cellular Receptor With Ann H. Rux, Roselyn J. Eisenberg, and Gary H. Cohen Department of Microbiology and Center for Oral Health Research, School of Dental Medicine, and Department of Pathobiology, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, Pennsylvania

54 We have examined equilibrium and kinetics of poliovirus binding to spvr. In the study of virus entry, determination of biochemical and biophysical parameters of virus receptor interactions are important because these parameters describe the dynamics of the interaction. In addition these parameters enable a better understanding of the reaction in comparison to other systems. The results of such studies, together with structural and genetic analyses of the virus-receptor interaction, provide a complete picture of early events in infection (61, 80, 125, 153). To study the kinetics and equilibrium of poliovirus binding to Pvr, we used surface plasmon resonance (23, 73) to examine the interaction of a soluble form of Pvr (spvr) with poliovirus. spvr expressed in and purified from mammalian cells is able to bind poliovirus, neutralize viral infectivity, and induce the formation of altered particles. Expression and purification of spvr in mammalian cells A novel approach was used to express a soluble form of the poliovirus receptor at high levels in mammalian cells for biochemical and biophysical studies. A plasmid was used that leads to the production of a bicistronic mrna upon expression in mammalian cells. The coding region of the Pvr ectodomain (with a 6-histidine tag at the C terminus, Figure 6) was placed in the first cistron position, followed by an IRES, and then the coding region for GFP in the second cistron. In a cell line stably expressing the bicistronic mrna, the intensity of GFP fluorescence is an approximate indicator of the expression of the protein in the first cistron. Fluorescence-activated cell sorter analysis was then used to isolate a clonal cell line that contains a fluorescence intensity greater than 99.75% of the GFP-positive population

55 mg/liter spvr. spvr was purified from cultured supernatant using a two-step procedure. The level of purification was determined by assaying the capacity of spvr to neutralize infectious poliovirus (71). In the first step, nickel affinity chromatography achieved 160- fold purification over the cultured supernatant (Table 1). In the second step, Q-Sepharose purification ion exchange chromatography achieved 2.3-fold purification over the previous step. At this stage of purification, spvr was the only visible band on a Coomassie Blue-stained, SDS-polyacrylamide gel (Figure 6). Edman degradation revealed that the N terminus of purified spvr begins at Asp-28 of the unprocessed precursor, as previously reported for the membrane-bound form (17). Although the predicted molecular mass of spvr is 34 kda, the purified protein migrates as a diffuse band between the 61- and 85-kDa molecular mass markers (Figure 6), suggesting that the protein is heavily glycosylated. After treatment of spvr with N-glycosidase F, which cleaves asparagine-linked glycan chains on glycoproteins, the polypeptide migrates as a smear at 34 kda, the predicted size of the non-glycosylated protein (see deglycosylated spvr section). A similar protein produced in insect cells migrated at 51 kda, probably due to less extensive glycosylation in that cell type (4). Virus neutralization and alteration activity of spvr We carried out several assays to determine whether purified spvr is biologically active. Plaque reduction assays were used to determine the efficiency of neutralization of poliovirus by spvr. Viral infectivity was reduced by 50% at 30 nm spvr (Figure 7). In

56 Figure 6: Purity of recombinant spvr expressed in mammalian cells. Left, schematic diagram of spvr, indicating the first and last amino acids of the recombinant protein. An additional 5 histidine residues were added to the C terminus of the protein. Potential N- linked glycosylation sites are designated by a ball and stick. Disulfide bonds are indicated by SS. Right, SDS-polyacrylamide gel electrophoretic analysis of purified spvr. Lane 1, 1.1 µg; lane 2, 0.6 µg; lane 3, molecular mass markers

57 Table 1: Purification of spvr Procedure Enriched cell culture supernatant Total protein Total activity Recovery Specific activity Purification mg units a % units/mg -fold 1,330 50, Nickel-agarose , , Q-Sepharose , , a 1 unit is a 10 µl aliquot capable of neutralizing 50 of 100 pfu

58 contrast, the 50% inhibitory dose for infectivity (IC 50 ) of sicam-1 for rhinovirus type 3 was 10-fold higher than spvr, 300 nm (97).One possible explanation for this difference is that the affinity of poliovirus type 1 for its soluble receptor is greater than that of rhinovirus 3 for sicam-1 (see below). We also determined if spvr is capable of inducing structural changes in poliovirus. This question was addressed by incubating spvr with poliovirus at 37 C and assaying the products by sucrose gradient centrifugation. At the concentration of spvr used, 1.8 X 10-8 M, conversion of native virus (160S) to 135S altered particles and 80 S empty capsids was nearly complete within 15 min (Figure 8). These results indicate that spvr produced in mammalian cells can efficiently bind to poliovirus and induce the structural changes associated with cell entry. Conditions and specificity of surface plasmon resonance Surface plasmon resonance allows determination of quantitative affinities (K D ), association (k a ), and dissociation (k d ) rates for the formation and dissociation of the virusreceptor complex (28, 30, 84, 128). To examine the kinetics of binding of spvr to poliovirus by surface plasmon resonance, purified poliovirus was coupled to the sensor chip surface, and spvr was injected over the chip surface. An example of raw sensorgram data is shown in Figure 9. In this experiment, flow cell 2 contained immobilized poliovirus, and flow cell 1 was activated and blocked without virus. spvr was injected, and its association with virus was followed for 2 min. At 120 s, spvr was replaced with buffer, and the dissociation of complex was followed for 3 min. The response on

59 Figure 7: Neutralization of poliovirus infectivity by spvr. Approximately 200 pfu of poliovirus were incubated with different concentrations of spvr at 37 C for 1 h. Remaining infectivity was determined by plaque assay on HeLa cell monolayers. The percent reduction of pfu was calculated relative to virus incubated with buffer only. The means ± S.D. of three experiments are shown. The IC 50 of spvr was extrapolated from a line graph of the same data

60 Figure 8: Kinetics of spvr-induced conformational changes of poliovirus. [ 35 S] Methionine-labeled poliovirus (5 X 10 9 virions) was incubated with spvr (1.6 X 10-8 M) at 4 C overnight, then shifted to 37 C for 0 (squares), 5 (triangles), or 15 (circles) min. Samples were centrifuged in a 15-30% sucrose gradient. Untreated 160 S particles and 80 S particles obtained by heating 160 S particles for 20 min at 56 C were centrifuged in a parallel gradient as markers

61 the y-axis is measured in response units. The sensorgram reveals a change in the bulk refractive index, but there was no significant background response when 1.3 µm spvr was injected over the mock-coupled control surface. In the surface plasmon resonance experiments that followed, data from flow cell 2 were subtracted from the data from flow cell 1 to correct for changes in bulk refractive index. These results demonstrate binding of spvr to poliovirus immobilized on the chip surface. The sensor chip surface was regenerated by treatment with low ph, to disrupt the virus-receptor interaction. Poliovirus remaining on the chip surface should survive these conditions, because its natural route of infection is through the acidic environment of the stomach. Two experiments were done to ensure that the sensor chips could be reused. First, unbound poliovirus was incubated in regeneration buffer (glycine buffer, ph 3) for 5 min at room temperature, and then infectivity was determined by plaque assay. As expected, this treatment did not reduce poliovirus infectivity, suggesting that conditions used for regeneration of the sensor chip do not disrupt virus structure (Figure 10). Second, repeated use and regeneration of sensor chips containing bound poliovirus did not affect sensorgrams and response levels (data not shown). To determine the specificity of the poliovirus-spvr interaction, a blocking experiment was performed using a monoclonal antibody, 711C, directed against the first domain of Pvr and which prevents poliovirus attachment to cells (105). Two concentrations of monoclonal antibody 711C were preincubated with spvr for 1 h at 4 C prior to injection onto the sensor chip containing bound poliovirus. Preincubation with

62 Figure 9. Example of raw sensorgram data. spvr (1.3 µm) was injected over a mockcoupled surface (blue line, flow cell 1) in series with a surface containing 1360 response units (RU) of poliovirus (green line, flow cell 2). At 120 s, the sample was replaced with buffer, and dissociation was followed for 2 min. The flow cell 2-flow cell 1 sensorgram (red line) represents the data corrected for changes in bulk refractive index and nonspecific binding of spvr to the chip surface

63 11). A control monoclonal antibody DL11, directed against herpes simplex virus glycoprotein D, did not inhibit the formation of the poliovirus-spvr complex (data not shown). These results indicate that the spvr-poliovirus interaction under study resembles the interaction during infection of cells, since it is mediated by domain 1 of spvr. Kinetic and equilibrium affinity analysis Determination of kinetic binding parameters for the spvr-poliovirus interaction was done at 20 C using separate injections of 2.5-fold serial dilutions of spvr onto the sensor chip containing bound poliovirus. The sensorgrams of the spvr-poliovirus interaction were imposed upon different model curves generated by global fitting analysis (Figure 12). The data fit best with the parallel reactions (2 sites) model, A + Β1 AB1, A + B2 AB2. The X 2 values generated using this model for interaction at 20 C were below 1.5, indicating an excellent fit. On the other hand, the X 2 value for a one-site binding model was 29, demonstrating a poor fit to that model. The two affinity constants calculated from the surface plasmon resonance data are 0.67 µm (K D1 ) and 0.11 µm (K D2 ) (Table 2). The calculated association rate constants are 3.6 X 10 3 M -1 s -1 (k a1 ) and 3.2 X 10 4 M -1 s -1 (k a2 ); the dissociation rate constants are 2.4 X 10 3 s -1 (k d1 ) and 3.3 X 10 3 s -1 (k d2 ) (Table 2). Binding rates were unaffected by changes in flow rate, demonstrating that the poliovirusspvr interaction is not limited by mass transport (data not shown) (72). The kinetics and affinity analysis of the rhinovirus-sicam interaction using the biosensor, as well as affinity analysis in solution, was also shown to be biphasic (28). In that study, the linear transformation method was used to analyze biosensor data on the rhinovirus-sicam

64 Figure 10: Effect of low ph treatment on poliovirus infectivity. Approximately 70 pfu of poliovirus were incubated in 10 mm glycine, ph 3, or PBS, for 5 min, followed by plaque assay. Shown is the average of two experiments

65 Figure 11: Specificity of spvr interaction with poliovirus on sensor chip surface. Anti-Pvr monoclonal antibody 711C was preincubated with spvr (4 µm) for 1 h at 4 C prior to injection onto the sensor chip containing bound poliovirus. At 180 s, the sample was replaced with buffer, and dissociation followed for 3 min. red line, no 711C; blue line, 2 µm 711C; green line, 4 µm 711C

66 also yields biphasic plots indicative of two binding sites (data not shown). Binding of spvr to the sensor chip was repeated under equilibrium conditions to confirm the existence of two classes of binding sites, and the affinity constants were determined by Scatchard analysis (129). The contact time was varied from 50 min for the lowest concentration to 10 min for the highest concentration of spvr (Fig. 13A). The Scatchard plot of the equilibrium data is curved, indicating that there are two classes of spvr-binding sites on poliovirus at 20 C, with binding affinities of 1.1 µm (K D1 ) and 0.16 µm (K D2 ) (Fig. 13B). These values are similar to those obtained by kinetic analysis (Table 2). To determine the effect of temperature on the poliovirus-receptor interaction, the kinetics experiments were repeated at 5, 10, 15, and 20 C. Higher temperatures, at which receptor-induced virus disruption occurs, were not studied because it would be difficult to interpret the biosensor data (25). With increasing temperature, the value for K D1 decreased, indicating a rise in affinity (Table 3). Binding of spvr at these sites on poliovirus is therefore endothermic. The value for K D2 did not exhibit a general increase or decrease with temperature, and therefore the thermodynamic nature of this site could not be determined. The relative abundance of the K D1 and K D2 sites at different temperatures was calculated from the kinetics data using global analysis software, assuming a parallel reactions model. At 20 C, the K D2 site constituted approximately 46% of the total binding sites on the sensor chip (Table 3, %R max2 ). The relative abundance of the K D2 site decreased with decreasing temperature. At 5 C, the relative abundance of the K D2 site is

67 Figure 12: Corrected sensorgram overlays for the interaction of decreasing concentrations of spvr with immobilized poliovirus. Data were collected at 5 Hz. Concentrations of spvr: red line, 8 µm; blue line, 3.2 µm; green line, 1.3 µm; magenta line, 0.51 µm; turquoise line, 0.21 µm. The black lines are the best global fit to the parallel reactions model (BIAevaluation 3.0 software)

68 Table 2 Kinetic and affinity parameters for spvr binding to poliovirus type 1 at 20 C Kinetic constants were measured as described in Figure 12 for binding of spvr to poliovirus type 1. k a1 k a2 k d1 x10-3 k d2 x 10-3 a K D1 K D2 M -1 s -1 s -1 µm 3,600 ± 660 b 32,000 ± 2, ± ± ± ± 0.02 a K D = k d /k a. b Mean of three experiments ± S.D

69 D1 constant at all temperatures tested

70 Figure 13: Equilibrium binding sensorgrams and Scatchard analysis of the binding of spvr to immobilized poliovirus. A, binding of spvr to immobilized poliovirus was monitored for 10 min for the injections of 13.5 (red line), 9 (blue line), 6 (green line), 4 (magenta line), and 2.67 (turquoise line) µm concentrations; 20 min for injections of 1.78 (gold line), 1.18 (black line), 0.79 (yellow line), 0.53 (pale blue line), and 0.35 (pink line) µm concentrations; and 50 min for injection of the 0.23 µm (salmon line) concentration. Arrows indicate the time points used for the Scatchard analysis. B, Scatchard analysis. C is the concentration of spvr flowed across the sensor chip surface at 20 C. The negative slope of each line is equal to each association constant; the reciprocals are the K D values. The R 2 values for the linear fit of the data were 0.95 and 0.91 for K D1 and K D2, respectively

71 - 57 -

72 Table 3 Affinity constants for spvr binding to poliovirus type 1 and abundance of each binding class at different temperatures. Temperature K D1 K D2 R max1 R max2 %R max2 a C µm RU b c (4) 10 (6) (3) 28 (8) (0.3) 45 (2) (0.6) 61 (3) 46 a R max2 /(R max1 + R max2 ) = 100. R max is the maximum receptor binding capacity, and therefore %R max2 is the relative abundance of the binding site that corresponds to K D2. b Resonance units, calculated with BIA evaluation 3.0, assuming a parallel reactions model. Mean of two experiments and range in parentheses. c Mean of two experiments

73 Chapter IV. Three-dimensional structure of poliovirus receptor bound to poliovirus With David M. Belnap*, David J. Filman, Naiqian Cheng*, Benes L. Trus*, Harmon J. Zuccola, James M. Hogle, and Alasdair C. Steven* * Laboratory of Structural Biology, National Institute of Arthritis, Musculoskeletal and Skin Diseases, Bethesda, MD 20892; Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, MA 02115; and Computational Bioscience and Engineering Laboratory, Center for Information Technology, National Institutes of Health, Bethesda, MD

74 Multiple genetic approaches have been used to study the poliovirus-pvr interaction. Poliovirus mutants have been selected for resistance to neutralization (33) with a solubilized form of Pvr or for the ability to utilize mutant Pvr (35, 88, 89, 105). Other virus mutants were generated by site-directed mutagenesis in the virus capsid (57). Analysis of these mutants suggests that the principal contact site of Pvr on the capsid is the floor of the canyon, above the hydrocarbon-binding pocket, and on the outer ("south") rim of the canyon. Additional sites that modulate receptor utilization are located at or near the peak of the mesa. Mutagenesis of Pvr DNA has revealed that the binding site for poliovirus is contained in domain 1 of Pvr (d1), the membrane-distal domain (77, 106, 132). Mutations in the predicted C'C", CC', DE, and EF loops, and the C"- and D- strands of this Ig-like domain disrupt virus binding (3, 16, 105). To complement these genetic analyses, we have analyzed the complex of poliovirus type 1 bound to spvr by cryoelectron microscopy. The resulting reconstruction revealed the manner in which domain 1 of Pvr inserts into the canyon. It also helped guide construction of a homology model of Pvr. This model, together with the known crystal structure of poliovirus (63), was used to identify specific interactions between the virus and receptor. Visualization of poliovirus decorated with spvr Cryo EM image reconstructions were generated in the following manner. Briefly, the conditions used for binding spvr to poliovirus were short incubations at 4 C to avoid converting virions to 135S particles, which occurs at higher temperatures (4, 51). The

75 micrographs were then processed into image reconstructions. With or without the addition of spvr, the most prominent feature in cryomicrographs of poliovirus is its roundish shape (Figure 14a). Faint protrusions are discernible on labeled virions and represent attached spvr molecules. The bound receptors are much more obvious after image reconstruction (see Figure 14b-d), which shows prominent protrusions. The protrusions are 115 Å long, Å wide, and extend outwards at an oblique angle relative to the capsid surface. The comparison of corresponding cross sections through the virion-spvr and virion density maps shows that the capsid is largely unchanged on binding receptor under these conditions (13). The spvr-bound virion does not show the redistribution of RNA density characteristic of the 135S particle (12). However, a subtle change in the poliovirus capsid was observed in the reconstructions: a small tunnel beneath the receptor-binding site on the floor of the canyon that extends into the "pocket-factor" binding site in VP1 was present (Figure 15b) (58). This result indicates that the pocket factor is possibly expelled during Pvr binding. The structure of the receptor fragment A representation of spvr from the difference map could be constructed because there is structural invariance of the virion on receptor binding (Figure 14e). The three Iglike domains are obvious in the reconstructions. We equate them with domains d1, d2, and d3 (102), in order of increasing distance from the capsid surface. d1 is contacted by the poliovirus capsid. d2 has two prominent nodules and is connected to d1 by a narrow region. The long axes of these two domains are aligned, and they appear to be firmly

76 Figure 14: CryoEM of poliovirus labeled with spvr. (a) Cryomicrographs of poliovirus particles complexed with (Top) and without (Bottom) spvr. Bar = 300 Å. Image reconstructions are shown of virion + spvr [in stereo (c)] and, for comparison, of the virion (b) (from (12)). The two reconstructions were overlaid in d with the respective contour levels adjusted to clarify the interaction of spvr with the virion. Bar = 100 Å. (e) Two views of a single spvr molecule extracted from the difference map (13). Domain boundaries are marked. Bar = 25 Å

77 Figure 15: Analysis of the Reconstructions (a) "Road map" representations of poliovirus (Left) and rhinovirus-14 (Right). The corresponding triangular area of the capsid surface, bounded by a 5-fold and two 3-fold icosahedral symmetry axes, is marked (Inset). The radial distances of surface residues from the virion center are color coded and contoured [see key (Top Right)]. The receptor footprints are shown in white. (b) A ribbon diagram of the spvr model is flanked by two views (68) of a single spvr molecule as portrayed in the cryoelectron microscopy density map (white cage), enclosing the model of the three spvr domains, d1 (cyan), d2 (orange), and d3 (violet). Carbohydrates attached to d2 [to N188 (Left) and N237 (Right)] and possibly to d1 are shown in brown. Also shown are the capsid proteins VP1 (blue), VP2 (yellow), VP3 (red), and VP4 (green). The tunnel beneath the spvr-binding site is evident (white arrows). "Pocket factor" is magenta. (c) The spvr sequence is mapped onto secondary structural elements of the homology model. Asn residues thought to be glycosylated are marked with asterisks. (d) Ribbon diagram showing the docking of the spvr model onto the capsid surface. Same color conventions as in b. The axes allow this view to be related to Figure 23. (e) Schematic diagram showing a possible binding configuration of poliovirus with intact membrane-bound Pvr

78 - 64 -

79 between its long axis and that of d2. Its density is lower, suggesting that d2 and d3 may be more loosely joined (13). Contact area on poliovirus The receptor fragment binds at a glancing angle, such that, its d1 domain extends into the canyon and makes contact with the capsid surface near the center of the icosahedral asymmetric unit bounded by a 5-fold axis and two 3-fold axes (Figure 15a). The receptor appears to bridge the canyon with major contact points are localized in a cleft on the "south rim of the canyon and on the side of the mesa on the "north rim," (Figures 14c, d, and 15a). The area of contact of spvr to poliovirus differs substantially from that of the rhinovirus receptor, intercellular adhesion molecule-1, to rhinovirus (111). Footprints of the interaction were created by overlaying the virion-spvr reconstruction and the difference map on the atomic-resolution coordinates of the virion, we found that the spvr-binding site includes many residues in VP1 ( , , , , , ), a few in VP2 ( , ), and several in VP3 (58-62,93, ). The footprint of spvr on the capsid surface consists of three distinct patches. Two of the patches are similar to the footprint of intercellular adhesion molecule-1 on rhinovirus (111) (Figure 15a). The third patch in the southeast corner is unique

80 The homology models for d1, d2, and d3 were fitted into the reconstruction (Figure 15b). The density map exhibits constrictions between the domains. Consequently, determining the placement of the domains was mainly a matter of fixing their orientations about their long axes. The d1 model could be fitted into the density map in either of two orientations, 180 o apart. One orientation seemed more consistent with mutational data implicating the C'C" and DE loops of Pvr, and the EF ( ) and GH ( ) loops of VP1, the EF loop of VP2 ( ), and the GH loop of VP3 ( ) as important interaction sites (Figure 23); the other orientation was less consistent with these data. The orientations of d2 and d3 were unambiguous. β-strand and loop assignments in the final model (Figure 15b Inset) are given in Figure 15c. The relative orientation of d1 and d2 in this pseudoatomic model is controversial because Pvr would be the only example of the V and C-like domains in this orientation; all other known examples of the Ig-superfamily structures are in the reverse orientation by 180 (J. Hogle, personal communication). However, the presented model of Pvr agrees with most of the mutational data and fits well into the density cage generated by the reconstructions. This controversy remains to be resolved. The pseudo atomic model of domain 1 (d1) model (residues ) fits the reconstructed density well and exhibits notable complementarity with the virus surface. Adaptation of the initial homology model to fit the density map required major changes in only one large loop (CC'). This loop, which projects laterally from the side and top of the β-sandwich in all close-sequence homologues, would protrude significantly through the envelope in any plausible orientation of d1 unless the loop were folded closer to the

81 C-strands of VP1 and the FG loop of the receptor as presently modeled, suggesting that the FG loop may adopt a different conformation than in the homologues. There is some extra density between the receptor and the virus surface in the vicinity of N105, which is a potential glycosylation site (Fig. 15 b-d); but, this feature appears too small to accommodate a full-length glycosyl modification. Interestingly, the mutant N105A, which is not glycosylatable at this site, exhibits increased virus binding and infectivity (16). There is no evidence for glycosylation at N120. In the pseudo atomic model of domain 2 (d2-residues ) the orientation was constrained by three factors. Firstly, two of the glycosylation sites should coincide with the prominent nodules that extend laterally on either side of the domain. Secondly, the convex shape of one face and concave shape of the opposite face of the domain should match between the density map and the model (see Fig. 15b). Lastly, the C terminus of the G-strand should be located near d3. In the preferred orientation, the glycosylation sites at N188 and N237 account for the lateral protrusions. The close proximity of N218 to N237 makes it difficult to determine whether N218 is glycosylated as well. The d3 model (residues ) fits its portion of the envelope well (Fig. 15b), despite the weakness of this density compared with d2 and d1. The orientation of d3 is constrained by the requirements to position its A-strand on the lower surface of d3, and to fit the large CD loop into a bulge on the top surface of the envelope. Thus positioned, the d3 model is clearly shaped similarly to its envelope (Fig. 15b). There is no compelling evidence for glycosylation at any of d3's three potential sites

82 Deglycosylated spvr (sdpvr) To resolve the controversy between our results and the results of others (58, 149) that demonstrate a reverse orientation of domain 1 of spvr in the canyon of poliovirus, we produced a deglycosylated soluble form of Pvr (sdpvr). This may resolve the controversy because glycosylation of N105 on the receptor at the spvr interface between virus and receptor may add extra density to the vicinity; these sugar moieties make modeling the receptor and its interaction difficult. Removal of all N linked sugar residues may improve the model because only amino acids would be present at the site of interaction. sdpvr was produced by expressing spvr with mainly high mannose glycosylation linkages, in mammalian cells, using 1-deoxymannojirimycin and ensuing removal of high mannose glycosylation linkages with endoglycosidase H treatment (Figures 16 & 17). sdpvr possessed biological function in agreement with previously published results that suggested glycosylation of domain 1 of membrane bound Pvr decreased virus binding (16). Virus binding was significantly enhanced when membrane bound Pvr lacked carbohydrate chains in domain 1. sdpvr neutralized poliovirus at lower concentrations than spvr in the neutralization assay, with IC 50 values of and 0.021µM, respectively (Figure 18). The initial cryoem image reconstruction, to 25 Å, of sdpvr complexed with poliovirus (Figure 19, left), demonstrated that the interacting receptor has the same projection geometry from the virus as the spvr interaction (Figure 19, right). As predicted, the "arms" of domain 2 are gone in the sdpvr structure. Therefore, the arms are likely to be formed by glycosylation. Domain 1 of sdpvr may be smaller than spvr. A reasonable explanation for this observation is that the majority of the

83 The reconstructions using sdpvr are in agreement with the enhanced neutralization capacity of sdpvr. The reason the results seem to be in agreement is because the glycosylation that causes steric hindrance at the site of interaction is not present. Recently, the naturally occurring form of soluble poliovirus receptor, scd155, was purified from human serum to ~10% purity by immunoaffinity purification and provided to us by Denis group. It is not clear the relative proportion of H 20 A 1 and H 20 A 2 soluble receptor variants present in this preparation (Figure3). A comparison of the activities of spvr and scd155 was made to determine if recombinant spvr is functioning similarly to naturally occurring receptor. A side-by-side comparison of spvr and scd155 by neutralization assay demonstrated that both proteins yield similar neutralization profiles, indicating that both proteins are functional at similar concentrations (Figure 18). This result lends strong evidence to the notion that recombinant spvr is functioning in a manner similar to Pvr purified from human serum. Since scd155 is only ~10% pure, an accurate IC 50 could not be determined. The availability of pure and quantifiable spvr allowed us to reexamine srr mutants; we determined if, and to what degree, each variant was resistant to spvr. This study is important because, in conjunction with thermodynamic and kinetic studies of the mutants, it will allow for determination of the mechanism of resistance for the mutants. srr mutants differ in several parameters: location and type of residue substitution on capsid; affinity for cell surface Pvr; and capacity to undergo alteration when contacting membrane bound Pvr (Table 4, Figure 22). Many of the srr mutants predicted the site of receptor contact to be in the canyon of the virus. To initiate studies that will allow us to

84 Figure 16: Trimming of an asparagine-linked precursor to a highmannose structure. High-mannose structures are formed first during the processing reactions, by exoglycosidase trimming of the precursor as outlined above. In the endoplasmic reticulum, Glucosidase I initiates the trimming reactions, removing the first glucose molecule from the precursor (structure I) to form structure II. Structure II is converted to structure III (cleavage of bonds 2 and 3) by glucosidase II. An endoplasmic reticulum α1,2-specific mannosidase is responsible for the next step, removal of a single mannose in the middle branch (bond 6, structure II) of the precursor. M 8 GN 2 -Asn (structure IV) is transported from the endoplasmic reticulum to the Golgi apparatus. There, the Golgi α- mannosidase I (also α1,2-specific) trims three more mannose residues (cleavage of bonds 4, 5, 7) from the precursor. Further oligosaccharide processing may occur in the Golgi apparatus to form complex and hybrid oligosaccharides. 1-deoxymannojirimycin blocks Golgi α-mannosidase I; this stops the processing of N-linked high-mannose oligosaccharides to complex oligosaccharides. Treatment of soluble receptors containing high mannose linkages with endoglycosidase H usually results in the removal of all N-linked sugar residues but the peptide proximal N- acetylglucosamine

85 - 71 -

86 kd 30 kd Figure 17: Purity of processed recombinant sdpvr expressed in mammalian cells. SDS-polyacrylamide gel electrophoretic analysis of purified sdpvr. Lane 1 and 2 are 4 µl aliquots of from fractions eluted from the S-200 gel filtration column. Right of the Coomassie blue stained gel are the positions of molecular mass standards

87 Neutralization of poliovirus by sdpvr and spvr 100 spvr sdpvr 50 0 Concentration Pvr (um) Figure 18: Neutralization of poliovirus infectivity by sdpvr and spvr. Approximately 200 pfu of poliovirus were incubated with different concentrations of sdpvr or spvr at 37 C for 1 h. Remaining infectivity was determined by plaque assay on HeLa cell monolayers. The percent remaining pfu was calculated relative to virus incubated with buffer only. The IC 50 of spvr and sdpvr were extrapolated from a line graph of the same data

88 Figure 19: CryoEM of poliovirus labeled with sdpvr. Image reconstructions are shown of virion + sdpvr (left) and virion + spvr (right) for comparison. The yellow arrow denotes a glycosylation arm of spvr. Orange arrow denotes decreased density of sdpvr at site of interaction; Pink arrow shows similar site involving spvr + virion

89 Neutralization of poliovirus with scd155 and spvr 100 spvr scd Concentration of receptor (um) Figure 20: Neutralization of poliovirus infectivity by scd155 and spvr. Approximately 200 pfu of poliovirus were incubated with different concentrations of s CD155 or spvr at 37 C for 1 h. Remaining infectivity was determined by plaque assay on HeLa cell monolayers. The percent remaining pfu was calculated relative to virus incubated with buffer only

90 Neutralization of srr mutants by spvr Concentration of spvr, um c B A R A P1/M B. Mutant 111c b a R a P1/M IC 50 ND µm Figure 21: A. Neutralization of poliovirus wt. and srr mutants by spvr. Approximately 200 pfu of poliovirus were incubated with different concentrations of spvr at 37 C for 1 h. Remaining infectivity was determined by plaque assay on HeLa cell monolayers. The percent pfu remaining was calculated relative to virus incubated with buffer only. The IC 50 of spvr was extrapolated from a line graph of the same data. B. Table of 50%inhibitory concentrations for wt and srr viral variants

91 50 for each variant. The results indicate that there is a significant difference in the level of resistance to neutralization by spvr between the variants. The level of resistance ranges from µm for wt poliovirus to greater than 0.27 µm for the mutant 111c (Figure17, Table 4)

92 Chapter VI. Discussion

93 Kinetics of poliovirus interaction with spvr To measure kinetic constants of the poliovirus-receptor interaction, we expressed and purified from mammalian cells a soluble form of the poliovirus receptor. Surface plasmon resonance was used to study binding of poliovirus with spvr. The affinities determined by biosensor are within 1 order of magnitude of the IC 50 of spvr determined by plaque assay, suggesting that the values determined by BIAcore could be the functional affinities for spvr. The results indicate that the interaction between poliovirus and spvr is biphasic. Two classes of binding site for spvr on poliovirus were detected called the K D1 site and the K D2 site. At 5 C, approximately 90% of the binding sites were K D1 sites, with a binding affinity of 1.56 µm. The fraction of K D2 sites, with a binding affinity of 0.11 µm, increases with temperature and constitutes 50% of the sites at 20 C. A biphasic binding model for poliovirus and Pvr has not been described previously. The binding affinity of poliovirus for the surface of HeLa cells was previously determined to be approximately M at 4 C (19, 34). We find that the binding affinity of the K D1 site, the predominant binding site at this temperature, is 4 orders of magnitude lower. The difference may be explained by the fact that the binding affinities calculated in the present study represent the intrinsic affinity of poliovirus for a single receptor molecule. In contrast, receptor molecules may cluster on the cell surface, increasing the apparent affinity, or avidity, of the virus-receptor interaction. Such clustering does not occur in solution (97). In another study, a single binding affinity of poliovirus for a soluble form of Pvr produced in insect cells was determined to be 4.5 X 10-8 M at 4 C (4). In those studies, binding assays were

94 the K D2 site, it is not clear why the lower affinity site was not detected. One possibility is that concentrations of spvr were not sufficiently high to detect the lower affinity site. In addition, proteins produced in insect cells and in mammalian cells have different patterns of glycosylation, which might contribute to the different results. An N-linked glycosylation site within Pvr domain 1 is known to influence its interactions with poliovirus (16) and may contact the receptor binding site on the viral capsid (13). A sideby-side comparison must be done to resolve this issue. The finding of two classes of receptor-binding sites on a virus has also been reported for rhinovirus type 3 and a soluble form of its cellular receptor, ICAM-1 (27, 28). Although the rhinovirus-sicam and poliovirus-spvr interactions are biphasic, there are significant differences in the affinity and kinetic constants. The association rates k a1 and k a2 are 25- and 13-fold higher for the poliovirus-spvr interaction than for the rhinovirus-sicam interaction at 20 C. The greater association rate of poliovirus-spvr might be due, in part, to differences in the extent of contact between virus and receptor. Three-dimensional models of virus-receptor complexes produced from cryo-electron microscopy and image reconstruction reveal that the footprint of Pvr on poliovirus is significantly larger than that of ICAM-1 on rhinovirus (13, 58, 79, 149). The extra surface area on poliovirus includes the knob of VP3 and the C terminus of VP1 from the 5-fold related promoter in the southeast corner of the road map describing the contact of Pvr on poliovirus (13). In contrast, although there are two dissociation rate constants for poliovirus-spvr, only one has been reported for the rhinovirus 3-sICAM interaction (27, 28). The dissociation rates for the poliovirus-spvr interaction are 1.5- and 2.0-fold faster

95 complex. The affinity constants for the poliovirus-spvr interaction are 19- and 6-fold greater than those reported for the rhinovirus sicam-1 complex (28). Consistent with these differences is the fact that the IC 50 of sicam-1 for rhinovirus 3 is 10-fold higher than that of poliovirus (97). However, other factors might play a role, including the number of receptors per virus particle that are required to neutralize infectivity. The effect of temperature on the interaction of poliovirus with spvr was studied. Binding at the lower affinity site, K D1, is endothermic (e.g. heat is absorbed by the complex), similar to both sites on rhinovirus (27, 28). As suggested previously, heat absorbed during the interaction of virus with receptor might help to lower the energy barrier required for uncoating of the virus particle (28). In contrast to the observations with poliovirus and rhinovirus, a single class of binding site (K D = 3.0 X 10-6 M at 20 C) was found on echovirus 11 for a soluble form of its receptor, CD55 (84). The affinity of this interaction is at least 4 times lower than either of the binding sites on poliovirus for spvr. Like most protein-protein interactions, the affinity of echovirus 11 for CD55 increases with decreased temperature, indicating that binding is exothermic. The association rate for the interaction between echovirus 11 and CD55 is faster than that of poliovirus-spvr (39- and 4.4-fold) and rhinovirus-sicam- 1 (28). One explanation for these findings is that the contact between echovirus 11 and CD55 is more extensive than that of the other two virus-receptor complexes. In addition, the binding site for CD55 on echovirus 11 might be more accessible than those of Pvr and ICAM-1, which are located in a depression on the capsid (13, 58, 149). The dissociation

96 poliovirus-spvr or rhinovirus-sicam-1 (28). These findings are consistent with a more accessible binding site for CD55 on echovirus 11, compared with the receptor-binding sites on poliovirus and rhinovirus (28, 84). In addition, it is possible that the atomic interactions between CD55 and echovirus 11 are weaker than between the other two viruses and their receptors. The faster dissociation rate of the echovirus 11 CD55 complex may be related to the finding that the interaction does not lead to structural changes of the virus particle (114), as occurs with poliovirus and rhinovirus. The lower dissociation rates for the poliovirus- and rhinovirus-receptor complexes may in part reflect the time required for structural changes to occur. Elucidation of the high resolution crystal structures of all three virus-receptor complexes should provide explanations for the differences in kinetic parameters. Why do poliovirus and rhinovirus have two classes of receptor-binding sites? One possibility is suggested by a three-dimensional model of the poliovirus spvr complex (Figure 15). In this model, domain 1 of spvr contacts two major sites on the virus surface, one in a cleft on the "south rim" of the canyon and a second on the side of the mesa on the "north rim." Whether these two contact sites correspond to the two classes of binding sites can be tested by carrying out kinetic and equilibrium binding studies on viruses with amino acid changes in these areas (33). Since all contacts of Pvr with the virus involve domain 1 (Figure 14), the finding of two classes of binding sites cannot be explained by the involvement of Pvr domains 2 and 3. Two classes of binding sites might also be a consequence of the structural flexibility exhibited by both viruses. Normally internal parts of the poliovirus and rhinovirus capsid proteins have been shown to be transiently

97 poliovirus and rhinovirus with their cellular receptors leads to irreversible and more extensive structural changes (12, 29, 65, 71). Antiviral drugs, such as WIN compounds, which replace the lipid-like molecule in the hydrophobic pocket, are believed to block uncoating of the capsid by rendering it structurally rigid. Binding of poliovirus to its cellular receptor may cause release of the lipid-like molecule from the hydrophobic pocket, allowing the capsid to undergo structural transitions necessary for binding and entry. Such structural plasticity might explain the presence of two different classes of binding sites on the virion. At lower temperatures, the higher affinity binding site is less abundant compared with the lower affinity site. At higher temperatures, the relative abundance of the higher affinity site is increased compared with the lower affinity site. One explanation for these observations is that increased breathing of the virus at higher temperatures results in the exposure of the higher affinity site. In addition, the interaction between receptor and virus may induce a conformational change in the capsid that results in exposure of the higher affinity binding site. In contrast to the findings with poliovirus and rhinovirus, binding of echovirus 11 with CD55 can be described by a simple 1:1 binding model. Such behavior, which would be expected for the interaction of two preformed binding sites, is consistent with the fact that the echovirus-cd55 interaction does not result in detectable structural changes in the capsid (114). Correlation of Cryo EM structure with mutational data The interactions described by Cryo EM and image reconstruction are consistent with mutational analyses (2, 16, 105) in the majority of mutations that affect receptor

98 loop, the C terminus of the D-strand, and the DE and FG loops of the receptor have been associated with alterations in virus binding and the ability of the receptor to support infections (2, 16, 105). However, five single-site mutations in the BC loop do not affect binding, perhaps because none of these residues is essential (16). Similarly, several mutations within the receptor footprint on the virus have been shown to alter the ability to be neutralized by spvr, bind wild-type receptor, or initiate infection with mutated receptors (33, 35, 57, 88). These include mutations in the C-strand, EF loop, and C terminus of VP1; in the N-terminal end of the smaller EF loop of VP2; and in the β-bend (residues 58-60) and GH loop of VP3. Not all of the capsid residues identified in the genetic studies cited above are involved in contacts with Pvr. Those that are not-with one exception-are buried in interfaces between subunits or on the inner surface of the capsid. The exception involves mutations in the BC loop of VP1 that have been associated with mouse adaptation, ability to use mutated receptor, and ability to establish persistent infection (reviewed in (117)). The BC loop is well outside the footprint of the receptor in the road map (Fig. 14a), and viruses in which it is replaced by a variety of heterologous sequences are viable. Alterations in the BC loop are associated with significant changes in thermal stability and the ability of the virus to undergo receptor-mediated conversions. It is therefore possible that these mutations affect cell-entry steps after receptor binding (145). A similar explanation has been proposed for mutations in residues that are buried in interfaces, in the drug-binding pocket in VP1, or on the inner surface of the protein shell (145). Indeed, most of these mutations have been shown to alter thermal stability or

99 several nonexposed residues (including residues 178 of VP3 and 177, 231, and 241 of VP1) result in significant reductions in affinity for receptor, in competition assays with wild-type virus (145). These data suggest that tight receptor binding may require minor conformational changes in the virion. This suggestion is further supported by the recent demonstration that the ability to bind receptor is ablated by capsid-binding drugs, such as WIN51711, at low temperature but not at 37 C (41). These conformational changes may be related to the "breathing" of the virus under physiological conditions (87). Since the complexes studied here were formed by brief incubations at 4 C, they may represent the initial state of receptor binding. One possible consequence of the 60 turn between the long axes of d2 and d3 is demonstrated in Figure 15e. If d3 were to emerge from the cell membrane with its long axis normal to the plane of membrane, then d1 and d2 would be inclined at an angle of 30 relative to this plane. Given the binding aspect of Pvr to the virus (Figure 14c), this configuration would orient a 5-fold axis perpendicular to the membrane, thus facilitating multiple symmetry-related attachments of receptors around this axis. Moreover, it would bring the mesa at the 5-fold axis close to the membrane. This juxtaposition is relevant to the proposal that a continuous channel is formed along the 5-fold axis of the capsid and through the membrane, as a mechanism for membrane attachment and RNA release (12). An overview of the site of interaction As forecast by chimeric receptor and point mutation analysis, domain 1 of spvr contains all contacts with the virus and approaches the capsid surface from the right as

100 pointing downward, toward the virus surface. Most of the interactions with the virus involve the BC, C'C", DE, and FG loops (Figure 23). The BC loop inserts in a groove defined by the smaller EF loop of VP2 ( ), residues at the distal end of the GH loop of VP1, and residues in the bulge at the C-terminal end of the GH loop of VP1 ( ). The DE loop contacts the residues at the N-terminal end of the GH loop of VP1 (214), the distal end of the GH loop of VP1 ( ), and the GH loop of the 5-fold related copy of VP3 ( ). The FG loop contacts the C-strand of VP1, and the C'C" loop of the receptor contacts the EF loop of VP1 ( ). The latter contacts are the only ones made with residues on the north wall of the canyon. In addition to contacts involving loops at the N-terminal end of d1, the N-terminal end of the D-strand and the C-terminal end of the E-strand (which are just below the d1- d2 junction) contact the hairpin "knob" proximal to the B-strand of VP3 (58-60) from the 5-fold related protomer and residues near the C terminus of VP1 ( ) from the adjacent protomer (Figure 23). These resides are located in the "southeast" portion of the receptor footprint (Figure 15a). srr mutants possess varying degrees of resistance to spvr Many of the srr mutants predicted the site of interaction in the canyon of the virus. To understand how the srr mutations confer resistance to neutralization by receptor we determined the IC 50 for each variant. The results indicate that there is a difference in the level of resistance to neutralization by spvr between the variants. The level of resistance ranged from µm for wt. poliovirus to greater than 0.27 µm for the

101 observed phenotypes does not exist. The location of mutation on virus, accessibility of mutation, reduction of 160S-to-135S upon binding to cell surface receptor, and a decreased affinity of the mutant for cell surface do not reveal an obvious pattern in relation to resistance to spvr (Table 4, Figure 22). In order to understand how resistance is conferred, the affinity of spvr for the variants must be determined. Furthermore, it may be necessary to determine the activation energies for the N to A conversion for the variants in the presence and absence of receptor if the mutations modulate this thermodynamic parameter instead of or in addition to decreasing the affinity for the receptor. Recombinant spvr and soluble Pvr purified from human serum neutralize poliovirus at similar concentrations In the majority of studies presented in this work recombinant spvr purified from 293 T cells was used. Limited quantities of naturally occurring soluble Pvr has been purified from human serum, scd155, and provided to us by Mark Denis laboratory. Provision of scd155 allowed us to compare the activities of recombinant and naturally occurring receptor proteins. This study was done to insure that recombinant spvr is not acting in an artifactual manner in the biochemical and biophysical experiments. A sideby-side comparison of spvr and natural scd155 purified from human serum in a neutralization assay demonstrated a remarkably similar profile and ID 50 values. This study indicated that recombinant spvr is acting in a manner similar to the receptor purified from human serum and it is suitable for biochemical and biophysical studies

102 Kinetic analysis of the effect of poliovirus receptor on viral uncoating: the receptor as a catalysis In a 1996 review (145), James Hogle and colleagues hypothesized that the poliovirus receptor may act as an enzyme using some of the energy of binding to lower this free energy barrier of conversion and allowing conversion to occur at physiological temperature. This theory is based on the observation that it is possible to reproduce the transition from intact virions to the 135S particle by heating; this suggests that native virion may represent a metastable intermediate in which the particle is trapped, by energy barriers, from the path to lower energy. We have provided Simon Tsang and James Hogle with spvr. It is a necessary reagent to explore this hypothesis. Some of the results from their experiments are presented in this section (143). The Arrhenius equation was used as a mathematical tool to determine the energy of activation (E a ) for the N to A transition. To employ the Arrhenius equation, the effect of spvr on the rate constant of the N to A transition of virus was examined by incubations at various temperatures in the presence of 0.25 um spvr. Alteration from N to A particle can be induced in vitro by warming the virus in the presence of millimolar concentrations of calcium (38) without the formation of the empty capsid. The N to A transition was monitored with an immunoprecipitation assay using a monoclonal antibody (P1 McAb) raised against residues of the capsid protein VP1. This antibody does not recognize the native virion and has little or no affinity for the 80 S particle (46, 142). In the experiments, the ratio of receptor to binding site at the

103 Figure 22. Locations of srr mutations in the capsid structure. (a) An α-carbon trace of a single poliovirus protomer viewed from the side. Sixty protomers form an entire capsid. Note the peaks at the 5-fold and 3-fold axes of symmetry, and the canyon that separates them. Surface srr mutations in VP1 are shown as yellow dots, green dots indicate internal srr mutations, and residue numbers are listed. The light blue molecule represents sphingosine bound in the hydrocarbon-binding pocket of VP1. (b) Two protomers viewed from outside the virion, showing the interface between protomers. Surface srr mutations in VP1 and VP3 are shown as yellow dots, white dots represent the positions of two Sabin 3 ts-suppressing mutations, and residue numbers are shown

104 Figure 23: Diagram of domain 1 of Pvr (cyan), with β-strands labeled, abutting inferred contact segments of the capsid proteins. Viral portions are shown as tubes, with VP1, blue; VP2, yellow; and VP3, red. Residue numbers are provided as landmarks. Black balls and colored numbers denote amino acids implicated by genetic analysis in receptor binding. Similarly, Pvr residues shown by mutation to be important for virus binding are listed (Left). The axes allow this view to be related to Figure 15d

105 Relative Resistance Mutant Name ID50 (µm) Capsid Protein Residue Number Wt Mutant Structural Location Phenotype VP1 132 Met Ile Inaccessible in lipid binding pocket 5 211b 0.18 VP1 226 Asp Gly Accessible in G-H loop VP1 234 Leu Pro Accessible in G-H loop, near opening of lipid binding pocket VP1 236 Asp Gly Partially accessible in G-H loops, near the opening of lipid binding pocket VP1 241 Ala Thr Inaccessible in H strand, in fivefold interface VP3 183 Ser Gly Partially accessible in fivefold interface 1 111c >> 0.27 VP1 228 Leu Phe Inaccessible, near surface in G-H loop 8 R VP2 215 Ser Cys Inaccessible, in fivefold interface 7 151a VP1 225 Gly Asp Inaccessible, near surface in G-H loop VP1 226 Asp Asn Accessible, in G- H loop VP1 231 Ala Val Partially accessible, in G-H loop a VP1 241 Ala Val Inaccessible, in H strand, in fivefold interface VP1 265 His Arg Internal surface, near fivefold interface, contacting VP VP3 178 Gln Leu Inaccessible, in fivefold interface 15 P1/M Greatly reduced binding to PVR; normal 160S-to- 135S alteration Greatly reduced binding to PVR; 160S-to-135S transition is reduced upon binding to receptor on cell surface Partial reduction in binding to PVR; 160S-to 135S transition appears normal Partial reduction in binding to PVR; 160Sto135S transition is reduced upon binding to receptor Surface accessibility determined by elevating the maximum radius of a sphere that contacts any atom within the residue without contacting another atom in the structure. Residues were an accessible if they could be contacted by a sphere greater than or equal to 3 Å in radius (Modified from review (31)

106 thermodynamics of poliovirus spvr interactions are complex. The available data suggest that the levels of receptor used in this study should be sufficient to guarantee high occupancy of the available sites, but are probably insufficient to guarantee full occupancy (60 sites/virion). Titration of varying concentrations of spvr demonstrated that further increases in the rate of conversion of virus could be achieved at higher concentrations (data not shown). Limited amounts of spvr were available at the time of experimentation; therefore, the concentrations are less than sufficient to guarantee full occupancy and maximal rates, and with minor caveats, as noted, the implications of the results presented below are expected to be independent of changes in occupancy. The Arrhenius equation states that for a first-order reaction obeying simple transition states kinetics, the rate constant for a reaction is exponentially dependent on the temperature: k=a exp (-E a /RT), where k is the rate constant, E a is the activation energy, R is the gas constant (1.98 kcal/mol deg), and T is the temperature in Kelvin. The preexponential factor A is described by the relation: A = (k b T/h) exp ( S /R) where K b is Boltmann s constant, h is Planck s constant, and S is the entropy difference between the ground state and the activated complex. Therefore, a plot of the natural logarithm of the first-order rate constant versus 1/RT should yield a line with a slope that is equivalent of E a and a y-intercept that is proportional to S. The virus produces a linear Arrhenius plot in the presence of receptor; this indicates that the first-order rate constants are dependent on a single exponential function as required by simple transition state theory (Figure 21). Since the Arrhenius plots are

107 temperatures. Similar behavior has been previously reported for the N to A transition of virus and virus drug complexes in the absence of receptor (142). The addition of receptor fragment decreases the E a of the N to A transition by 50 kcal/mol, from 145 kcal/mol for virus alone versus 95 kcal/mol for virus +spvr (Figure 21). Increasing the number of receptor occupancy sites occupied on the virus might be expected to induce a further decrease in E a. As predicted, the receptor is behaving like a classic transition state theory catalyst, accelerating the rate of the transition by lowering the activation barrier. A kinetic model of the N to A transition can be seen in (143). General rules for picornaviruses-receptor interactions What basic principles can be deciphered from the information known about virusreceptor interactions? Picornaviruses must undergo two processes to infect a host cell: first, they must bind to the cell; second, they must uncoat the genomic RNA and release in into the host cell cytoplasm. Picornaviruses have evolved to use two different types of receptors with different consequences on the virus uncoating and entry processes. For this discussion, receptors that only bind virions will be called binding receptors. Receptors that can both bind virus and cause the release of the virus genome will be called uncoating receptors. Viruses that use uncoating receptors such as HRV3, poliovirus, and CAV21 differ in four aspects when compared to picornaviruses that use binding receptors. One, HRV3, poliovirus, and CAV21 have receptor occupancy sites that reside within the canyon (13, 58, 111, 149). Two, the canyons of HRV3 and poliovirus are flexible and each has two

108 The two affinities are likely to reflect the plastic nature of the canyon region. Three, the soluble receptor fragments are sufficient to promote conversion of the virus (4, 29, 51, 97, 99). Four, membrane bound forms of the uncoating receptors for HRV3, poliovirus and CAV21, not only concentrate virus at the cell surface, but also promote conversion and allow entry. Binding receptors for picornaviruses contrast significantly with primary receptors. One, binding receptors do not bind the canyon. It has been suggested that the echovirus 11 receptor, DAF, binds at the 2-fold axes of symmetry of the capsid (S.M. Lea, personal communication). The human rhinovirus serotype 2 (HRV2) binds its receptor, VLDL-R, on the small star shaped dome on the icosahedral 5-fold axis (60). Two, the receptor occupancy site on the virus is preformed and stable with a single affinity for its receptor. For example, echovirus 11 has only one affinity for its receptor as determine by surface plasmon resonance using a soluble CD55 fragment (84). Third, soluble binding receptors do not promote conversion of the virus particle; for example, echovirus 7 is not converted to an altered particle by binding to its binding receptor, DAF (114). Fourth, membrane bound binding receptors act only to sequester virus on the cell surface but are not sufficient to mediate entry (15). These observations, taken together, show a clear pattern emerging. Receptor type, binding or uncoating, is governed by a distinct set of parameters: location of receptor binding on the virus surface; receptor uncoating capacity; and receptor entry mediating ability. Most importantly, uncoating receptors bind the canyon and, perhaps because of the plastic properties of this region on the protein shell, this interaction initiates uncoating. This plasticity may permit the canyon to open-up, Pinch, or

109 Figure 24: Arrhenius plots for the N to A transition in the presence and absence of spvr. Averaged values of natural log of the rates of conversion ln k from (143) were plotted as a function of 1/RT, the slope of each line is the energy of activation for conversion. Bold lines and dashed lines correspond to data collected in the presence of spvr and absence of spvr, respectively (From (143))

110 These are likely to be the changes in the protein shell necessary to initialize the release of VP4 and RNA. A possible exception to these rules is O 1 FMDV that binds hepran sulfate (HS) in the pit in the capsid surface. The pit has a similar location to the canyon on the virus surface but is shallower. HS binding to the pit does not cause conversion as would be expected by these rules. However, this finding is likely to be an exception because the depression is too shallow or because the ligand is a sugar moiety. Model of poliovirus interaction with cell surface How does poliovirus use the energy of interaction for entry? On the virus-side of the receptor, where the receptor contacts the canyon, the surface area of contact is large and specific. This seems to allow a decrease in the activation energy and catalysis of conversion of the metastable virus. On the cell surface-side of the receptor, poliovirus has been observed by electron microscopy in invaginations on the membrane (151). However, entry of poliovirus is not dependent on clathrin-mediated endocytosis (40). Therefore, we propose a simple model (termed: bind, remodel, release) for virus entry that allows for more efficient genome release into the host cells cytoplasm (Figure 25). The model is predicated on three observations: one, multiple receptors can bind virus simultaneously; two, poliovirus is found in invaginations on the cell surface; three, the free energy of interaction for a single receptor interacting with a virus is 10 Kcal/mol. Furthermore, the total energy of virus interaction with the cell surface is likely to be much higher because of multivalency. The energy generated by the multivalent

111 membrane. This would allow for membrane dimple formation similar to what has been postulated for influenza interaction with the cell surface (138). Since the virus can release its genome in the absence of membrane, its direction of release is likely to be orientation non-specific. Therefore, the gathering of the membrane around the virus particle should allow for a more efficient genome release and infection. Conclusions and future directions The result presented in this thesis, poliovirus receptor recognition: visualization; kinetics; and thermodynamics, are initial answers to questions formulated 16 years ago upon determination of the crystal structures of poliovirus and rhinovirus. In his review (7) of these crystal structures, David Baltimore discussed seven major questions that arose by the structural renderings. Two of these questions are the birthplace of this thesis: First, what imparts receptor specificity to the virion? The similarity of polio- and rhinovirus argues that the receptor specificity is encoded in the details of surface architecture, not in gross features. Second, how are particles disassembled? The receptor must somehow find a key site in the virion to affect this conversion reaction. Answers to these questions began after the poliovirus receptor was identified. Use of the receptor in conjunction with the infectious poliovirus cdna clone gave researchers the opportunity to correlate mutational analysis with detailed structural knowledge; in one key study, solubilized receptors were used to isolate viral variants that were resistant to neutralization. This provided the initial insights into how the capsid interacts with receptors

112 biochemical, and structural studies. Verifying genetic data using cryoelectron microscopy, we have used soluble Pvr to demonstrate that the canyon is the viral surface feature that encodes the receptor docking site for poliovirus. Interestingly, major group HRVs and CAV21 also use the canyon to contact their cellular receptors. The reason there are conserved canyon docking sites for some picornaviruses and not for others, such as minor group HRVs is unclear, but it is likely to be related to the mechanism of uncoating. We have shown that soluble receptor purified from mammalian cells is sufficient to disassemble poliovirus. The details of the uncoating mechanism are unknown. Determination of the crystal structure of spvr and poliovirus-spvr complex will be important steps in delineation of the mechanism. These studies will allow the interaction to be understood at the atomic level, which is important for understanding the mechanism of any biochemical reaction. Clearly the canyon is the key site that the receptor must contact to disassemble virus and the pocket factor, which is located beneath the canyon floor, plays a role in this process. How the pocket is dislodged from the virus to allow uncoating is unclear. Pocket factor may be squeezed out of the canyon by receptor induced structural capsid changes. Or, the receptor may contact the pocket factor directly and this may initiate its release. Once the atomic details of the interaction are understood it may be possible to design mutant spvr forms that are able to trap the virus in intermediate stages of uncoating. For example, if in the crystal structure three distinct regions of the receptor are contacted by the virus then mutations of specific amino acids at each of these regions separately

113 uncoating. An experimental procedure may be designed which can be used to trap the virions in interesting states such as partial RNA release, partial VP4 release, or viral expansion. These states may be visualized by Cryo EM and transitions the virus goes through may be better understood. Cryo EM image reconstruction of the conversion product catalyzed by spvr will be useful in determining the mechanism of uncoating. Although the heat conversion products, 80S and 135S, have been visualized. They may possess structural differences when compared to the products catalyzed by spvr. SPR has demonstrated that the relative abundance of the two binding sites varies with temperature and is, therefore, dynamic. It is likely the two binding sites for the receptor on the poliovirus type 1 capsid are the first clues to a mechanism of uncoating. The two binding sites may correspond to docking sites for domain 1 of the receptor on the viral capsid, as predicted by a model of the poliovirus-receptor complex. Alternatively, the binding sites may be a consequence of structural breathing, or could result from receptor-induced conformational changes in the virus. Evidence that the conversion process is the mechanism of virus genome release into the cell is compelling. Further evidence that this process is the mechanism of entry would be to demonstrate that spvr, not tethered to the cell surface, is able to mediate entry. Recent studies have shown that a soluble form of the subgroup A avian sarcoma and leucosis virus (ASLV-A) receptor (stva) is able to mediate entry of ASLV-A without being tethered to the cell surface (39). This result suggests the previously described receptor induced conformational changes on ASLV-A are likely to be involved

114 Figure 25: Model for poliovirus entry: Bind, remodel, and release