Development of Instantaneous Protection against SARS-CoV with Implications for. Multiple RNA Viruses

Transcription

1 Development of Instantaneous Protection against SARS-CoV with Implications for Multiple RNA Viruses By Hatem A. Elshabrawy B.S., Cairo University, Cairo, Egypt, 2003 THESIS Submitted as partial fulfillment of the requirements for the degree of Doctor of Philosophy in Microbiology and Immunology in the Graduate College of the University of Illinois at Chicago, 2012 Chicago, Illinois Defense Committee: Dr. Bellur S. Prabhakar, Chair and Advisor Dr. Alan Mclachlan Dr. Deepak Shukla, Ophthalmology and Visual Sciences/Microbiology and Immunology Dr. Bin He Dr. Michael Caffrey, Biochemistry and Molecular Genetics

2 I would like to dedicate this thesis to my parents and my wife who have always supported and encouraged me to achieve my goals. Their love and support have helped me in facing challenges and difficulties during my doctoral degree study. ii

3 ACKNOWLEDGEMENTS I would like to thank my thesis advisor Dr. Bellur S. Prabhakar for being a great mentor, giving me the opportunity to work in his lab, his support during my studies and teaching me to be an independent scientist. I would like to thank my committee members, Dr. Alan Mclachlan, Dr. Bin He, Dr. Deepak Shukla, and Dr. Michael Caffrey for their constructive suggestions. Dr. Mclachlan helped me a lot when I was transferring from the Department of Pharmacognosy and Medicinal Chemistry, College of Pharmacy at UIC to the Department of Microbiology and Immunology. I would like to thank the members of my laboratory who have made working in the lab a pleasurable experience. I would like to thank my parents who have always supported and encouraged me to be the best I can be in my life. My father has always been a very good friend to me and has been a very reliable supporter. He has given me very helpful advice and I have benefitted from his broad experience in life. My mother has always provided me with love and care that have taught me to be patient during difficult times. I would like to thank my wife who stayed with me during my graduate studies. She encouraged me to be optimistic and taught me to be patient. She took care of my kids while I was at work during late hours. This would not have been possible without the support of my parents and my wife. HAE iii

4 PREFACE The purpose of this study is to develop and identify neutralizing human monoclonal antibodies against a wide range of Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV) clinical isolates and antiviral drugs of therapeutic value against multiple RNA viruses e.g. SARS-CoV, Ebola virus (EBOV), Hendra virus (HeV), and Nipah virus (NiV). The aforementioned viruses are zoonotic viruses that emerged to cause outbreaks of lethal infections in humans. Currently, there are no effective treatments available. Only instantaneous protection using drugs or neutralizing monoclonal antibodies can protect against these acute viral infections giving the individual s immune system sufficient time to mount an effective immune response and eliminate the infection. The best way to provide passive immunotherapy is the use of homologous monoclonal antibodies with defined antiviral targets. However, the antiviral therapy (antibodies or antiviral drugs) should be effective against a wide range of viral variants that are likely to arise during the course of infection. This study shows the identification of broad spectrum human monoclonal antibodies that can neutralize a wide range of SARS-CoV variants. Additionally, this study describes the identification of small chemical compounds which exhibit antiviral activity against SARS-CoV, EBOV, HeV, and NiV. If these compounds can be optimized, they can then be used therapeutically against those fatal viral infections. iv

5 TABLE OF CONTENTS CHAPTER PAGE I. Introduction. 1 A. RNA Viruses and Acute Infections Coronaviruses. 2 a. Taxonomy and Morphology... 2 b. Coronaviruses and Infection..4 c. SARS-CoV genome: Organization, Replication Strategy, and Transcription...6 d. SARS-CoV Life Cycle e. The SARS-CoV S protein.. 9 f. Tropism.13 g. SARS-CoV Outbreak and Pathology.. 14 h. SARS-CoV Evolution..15 i. Immune responses to SARS-CoV Innate Immunity Cell Mediated Immunity SARS-CoV and Humoral Immunity.. 18 a. General Properties of Antibodies b. Monoclonal Antibody Technology c. SARS-CoV and Antibodies d. SARS-CoV and Passive Immunotherapy Ebola Virus. 26 a. Taxonomy and Morphology b. Evolution. 27 c. Genome Organization and Replication d. Ebola Viral Proteins e. Ebola Virus Transmission, and Disease f. Ebola Virus Glycoprotein Structure, and Functional Organization 32 g. Ebola Virus Entry into Host Cells h. Prevention of Ebola Virus Infection Henipaviruses a. Taxonomy and Morphology 38 b. Genome, Viral proteins, and Entry into Target Cells. 38 c. Current Targeted Therapeutics for Henipaviruses B. Current Antiviral Strategies for SARS-CoV, Ebola, and Henipaviruses Monoclonal Antibodies (mabs) HR2 Specific Peptides Antiviral Drugs.. 43 C. Current Study II. Materials and Methods 46 A. Characterization of broad spectrum neutralizing human monoclonal antibodies v

6 TABLE OF CONTENTS (continued) CHAPTER PAGE (HmAbs) against a wide range of SARS-CoV clinical isolates Purpose and Rationale Materials and Methods..47 a. Cells..47 b. Construction of expression plasmids for SARS-CoV S1-IgG and full length spike (S) protein mutants...48 c. Construction of S-ectodomain, S2, HR1 and HR2 domains expression plasmids.. 49 d. Expression and purification of SARS-CoV S1-IgG Urbani and mutant proteins as well as S-ectodomain, S1, S2, HR1 and HR2 domain proteins. 50 e. Purification of the non S1 binding neutralizing human mabs f. Enzyme Linked Immunosorbent Assay (ELISA).. 51 g. Competitive ELISA inhibition assay h. ELISA using Rabbit anti-sars CoV S protein immune serum.. 52 i. Production of SARS S- Urbani and mutant pseudotyped viruses. 52 j. Establishment of 293/ACE2 stable cell line k. In vitro pseudotyped virus neutralization assay B. Development of broad spectrum antiviral drugs against SARS-CoV, Ebola, and Henipaviruses Purpose and Rationale Materials and Methods. 57 a. Cells. 57 b. Viral and host derived peptides synthesis. 57 c. Mass spectrometry d. Optimization of the High Throughput Screening Assay (HTSA) e. Screening for inhibitors of cathepsin L mediated cleavage of viral glycoproteins derived labeled peptides.. 59 f. Preparation of pseudotyped viruses 60 g. In vitro pseudovirus inhibition assay.. 60 h. IC50 determination 61 i. Cytotoxicity assay (MTT assay) j. Enzyme kinetics.. 62 III. Results A. Characterization of broad spectrum neutralizing human monoclonal antibodies (HmAbs) against a wide range of SARS-CoV clinical isolates Expression of SARS-CoV S1 IgG1 Urbani and mutant proteins The S1 proteins containing RBD sequences of Sin845, GD01, and GZ0402 isolates show low binding to S1 specific neutralizing HmAbs, while that of GZ-C isolate shows higher binding The S proteins containing RBD sequences of Sin845, GD01, GZ0402 vi

7 TABLE OF CONTENTS (continued) CHAPTER PAGE and GZ-C isolates do not affect psudovirus entry Pseudoviruses containing S proteins with RBD sequences of Sin845, GD01 and GZ0402 isolates escape neutralization while GZ-C shows enhanced neutralization by S1 specific HmAbs Differential reactivity of non-s1 binding HmAbs with S ectodomain, S2 domain, HR1 and HR2 regions suggest multiple mechanisms of virus neutralization Combinations of SARS-CoV HmAbs targeted to different regions of the S glycoprotein more efficiently inhibit the entry of RBD surrogate clinical isolates B. Development of broad spectrum antiviral drugs against SARS-CoV, Ebola, and Henipaviruses Synthesis of viral and host proteins derived peptides that contain the natural cathepsin L cleavage sites The synthesized peptides are efficiently cleaved at the expected sites by cathepsin L Optimization of the High Throughput Screening Assay (HTSA) The EBOV, Hev, NiV, and host pro-npy derived peptides can be used to screen for inhibitory compounds in the HTSA Pro-NPY derived peptide is rapidly cleaved than viral derived peptides HTS of small molecules library identifies potential inhibitors of catl cleavage Screening of the top 50 hits identified potential broad spectrum inhibitors of catl mediated cleavage of viral peptides HTSA selected inhibitors showed differential inhibition to pseudotyped virus entry Compound and its derivative are not cytotoxic and their actions are specific IC50 determination of compound and against EBOV pseudotyped virus shows that compound is more potent than Compound is a mixed inhibitor for cathepsin L enzymatic reaction 104 IV. Discussion Cited Literature 123 Vita 142 vii

8 LIST OF TABLES TABLE I DIFFERENTIAL REACTIVITY OF 39 SARS-CoV NON-S1 NEUTRALIZING HUMAN MONOCLONAL ANTIBODIES TO SPIKE PROTEIN FRAGMENTS TABLE II HUMAN MONOCLONAL ANTIBODIES TO HR1 AND HR2 CAN EFFICIENTLY NEUTRALIZE SURROGATE CLINICAL ISOLATES...75 TABLE III SCREENING OF 5000 COMPOUNDS LIBRARY AGAINST SARS-CoV DERIVED PEPTIDE...92 TABLE IV SCREENING of 50 HITS AGAINST ALL VIRAL AND HOST DERIVED PEPTIDES.95 PAGE viii

9 LIST OF FIGURES PAGE Figure 1 Taxonomic classification of coronaviruses 3 Figure 2 SARS-CoV virion structure 5 Figure 3 Structural organization and features of the Ebola virus glycoprotein, GP.. 33 Figure 4 Comparative sequence analysis of the receptor binding domain of spike proteins in SARS-CoV clinical isolates 64 Figure 5 Expression and purification of SARS-CoV S1 proteins (aa ) Figure 6 Reactivity of the18 neutralizing HmAbs with SARS CoV S1 proteins. 67 Figure 7 Reactivity of Urbani SARS-CoV-S protein antibodies with Urbani and mutant S1proteins 69 Figure 8 Pseudoviruses expressing the spike glycoprotein of clinical isolates can enter cells with equal efficiency as HIV/S Figure 9 In vitro pseudovirus neutralization assay 72 Figure 10 Expression and purification of SARS-CoV-S protein domains. 73 Figure 11 Combinations of HmAbs more efficiently inhibit the entry of SARS-CoV RBD surrogate clinical isolates.. 77 Figure 12 Synthesis of host and viral derived peptides 79 Figure 13 SARS-CoV and EBOV derived peptides are cleaved efficiently by cathepsin L (catl) at the expected sites 81 Figure 14 Nipah and Hendra virus derived peptides are cleaved by catl 82 ix

10 LIST OF FIGURES (continued) PAGE Figure 15 Labeled SARS-CoV and EBOV derived peptides are cleaved efficiently by catl at the expected sites 83 Figure 16 Labeled Nipah and pro-npy derived peptides are cleaved efficiently by catl at the expected sites.. 84 Figure 17 Optimization of HTSA.. 86 Figure 18 Cleavage of labeled viral and host derived peptides by catl. 88 Figure 19 Dose dependent cleavage of viral and host derived peptides by catl. 90 Figure 20 High throughput screening assay statistics Figure 21 Pseudovirus entry Inhibition assay 97 Figure 22 Chemical structures of the inhibitory compounds identified by pseudovirus inhibition assay.. 98 Figure 23 Compound and its derivative , unlike and , have no cytotoxic effect on 293FT cells. 100 Figure 24 Compound and its derivative have no inhibitory effect on VSVG pseudotyped virus entry. 102 Figure 25 IC50 determinations of compound and its derivative Figure 26 Compound is a mixed enzyme inhibitor 105 x

11 LIST OF ABBREVIATIONS RNA: ribonucleic acid ssrna: single-stranded RNA dsrna: double-stranded RNA SARS-CoV: severe acute respiratory syndrome coronavirus CoV: coronavirus IBV: avian infectious bronchitis virus MHV: murine hepatitis virus TGEV: Transmissible gastroenteritis virus hcov-229e: human cornavirus 229E hcov-oc43: human coronavirus OC43 BCoV: bovine coronavirus CCoV: canine coronavirus EToV: equine torovirus TCoV: turkey coronavirus E.M.: electrom microscopy S protein: spike protein M protein: membrane protein E protein: envelope protein N protein: nucleocapsid protein SARS: severe acute respiratory syndrome ORF: open reading frame HE: hemagglutinin esterase xi

12 DMV: double membrane vesicle TRS: transcription regulatory sequence CS: conserved core sequence CatL: cathepsin L pp1a: replicase polyprotein 1a pp1ab: replicase polyprotein 1ab RBD: receptor binding domain RBM: receptor binding motif HR1: heptad repeat 1 HR2: heptad repeat 2 ARDS: acute respiratory distress syndrome MBL: mannan binding lectin IFN: interferon CTL: cytotoxic T lymphocytes CDR: complementary determining region Fab: fraction antigen binding Fc: fraction crystalline mab: monoclonal antibody EGFR: epidermal growth factor EBOV: ebola virus MARV: Marburg virus HeV: hendra virus NiV: Nipah virus xii

13 HF: hemorrhagic fever DC: dendritic cell GP: glycoprotein xiii

14 SUMMARY Severe acute respiratory syndrome coronavirus (SARS-CoV), Ebola (EBOV), Hendra (HeV) and Nipah (NiV) viruses are highly infectious zoonotic (animal-borne) RNA viruses. SARS-CoV belongs to family Coronaviridae and causes severe acute respiratory syndrome (SARS) that initially originated in the Guangdong province of China in late 2002, spread rapidly around the world along international air-travel routes, and resulted in approximately 10% mortality in infected individuals in different parts of the world. EBOV belongs to family Filoviridae and has been identified as the causative agent of severe hemorrhagic fever with human case fatality rate exceeding 90% in large outbreaks. NiV and HeV are closely related and belong to the genus Henipaviruses within the Paramyxoviridae family and were first identified as the etiologic agents responsible for an outbreak of fatal encephalitis among pig farmers in Malaysia and Singapore in 1999 with a case fatality rate of 40%. The high virulence of these viruses and the absence of effective therapeutic modalities and vaccines have led to their classification as Biosafety level 4 (BSL4) pathogens except for the SARS-CoV which is a Biosafety level 3 (BSL3) virus. Bats are the natural reservoir of SARS-CoV, EBOV, HeV and NiV and thus these viruses can persist in nature and pose an ongoing threat to the public health. The diseases caused by the above mentioned viruses are acute suggesting that treatment modalities that provide instantaneous protection (passive immunotherapy or antiviral drugs) would be ideal. However, viral variants that arise due to copying errors commonly found in RNA viruses should be considered. Accordingly, an ideal therapy should be effective against a broad range of viral variants. Numerous studies in patients xiv

15 and animal models have shown that infection with SARS-CoV leads to the induction of neutralizing antibodies. These antibodies have been shown to play a significant role in clearing the SARS-CoV infection. There is strong evidence that neutralizing antibodies against SARS-CoV could have a therapeutic potential. A DNA vaccine study in BALB/c mice demonstrated that the antibody response, and not the cell mediated immunity was responsible for reducing the viral load. The passive transfer of immune serum to naïve mice following SARS-CoV challenge allowed the mice to clear the infection. However, the depletion of CD8 + T cells did not affect the ability of the vaccinated mice to clear the infection. Furthermore, the adoptive transfer of CD4 and CD8 T cells from immunized mice did not provide protection to naïve mice that are challenged with SARS-CoV. Studies in patients during the outbreak have shown that administration of convalescent patient serum protected individuals from acquiring the disease with no undesirable effects. This demonstrates that passive immunotherapy with antibodies more specifically with monoclonal neutralizing antibodies could be a safe and effective treatmentstrategyforthesars-cov. Several groups have produced neutralizing antibodies, against SARS-CoV, that target the spike (S) protein on the virion surface. The S protein is responsible for binding to the angiotensin converting enzyme (ACE2) on the target cells thereby mediating infection. The S protein is divided into two domains, the S1 and the S2. The S1 mediates binding to the ACE2 via the receptor binding domain (RBD, ) while the S2 domain is responsible for the fusion of the viral and host cell membranes. Most of the described neutralizing antibodies to date bind to the RBD within the S1 domain. Previously, using Xenomouse (Amgen British Columbia Inc), our group produced a panel of neutralizing xv

16 Human monoclonal antibodies (HmAbs) that could specifically bind to the ectodmain of the SARS-CoV S glycoprotein. Some of the HmAbs were S1 domain specific, while some were not. In the first part of this study, we describe non-s1 binding neutralizing HmAbs that can specifically bind to the conserved S2 domain of the S protein. However, unlike the S1 specific HmAbs, the S2 specific HmAbs can neutralize pseudotyped viruses expressing different S proteins containing receptor binding domain sequences of various clinical isolates. These data indicate that HmAbs which bind to conserved regions of the S protein are more suitable for conferring protection against a wide range of SARS-CoV variants and have implications for generating therapeutic antibodies or subunit vaccines against other enveloped viruses. Developing a broad spectrum antiviral drug that can block SARS-CoV, EBOV, HeV, and NiV infections would represent a first but an important step for the protection of the war fighter and the public. SARS-CoV, EBOV, HeV, and NiV are enveloped viruses that critically require cathepsin L (catl), a host intracellular lysosomal protease, for their glycoprotein processing and cleavage allowing for virus fusion and entry into the host cells. SARS-CoV and EBOV infect the target cells after cleavage of their fusion glycoproteins by catl in the endocytic vesicles. In case of HeV and NiV, the fusion (F) protein is translocated to the membrane and then internalized allowing for catl mediated cleavage required for fusion to occur. The F protein is then incorporated into the viral particle. In the second part of this study, we have identified catl cleavage sites in the glycoproteins of SARS-CoV, EBOV, NiV, and HeV zoonotic viruses as conserved elements. Thus, the conserved catl cleavage sites can be potential targets for developing broad spectrum anti-viral drugs. There have been several attempts to xvi

17 discover potent inhibitors that can act broadly against several different viruses. For example, novel inhibitors of papain-like protease can inhibit live SARS-CoV infection. Oxocarbazate and pentapeptide amide can block human catl and inhibit SARS and Ebola pseudotyped virus infection. However, blocking host enzymatic activity is likely to have undesirable side effects on human health. Based on the common dependence of SARS-CoV, EBOV, and Henipaviruses on catl for their glycoprotein processing which is necessary for their entry into the target cells, we have carried out a High Throughput Screening Assay (HTSA) for chemical compounds, from Chembridge Corporation, that can block the cleavage of the envelope glycoproteins required for their entry into host cells without inhibiting catl itself so as to preserve its critical host protease function. The HTSA used peptides derived from aforementioned viruses, which contain the catl cleavage sites. We have confirmed that catl can cleave these peptides at the expected sites by mass spectroscopy. We have made use of the HTSA, optimized in our lab, to identify and characterize candidate compounds from libraries of small molecules provided by Chembridge Croporation. The HTSA is a Fluorescence Resonance Energy Transfer (FRET) based assay that is dependent on viral and host pro-neuropeptide (pro-npy) derived peptides labeled on the N-terminus with 5-Carboxytetramethylrhodamine (TAMRA) as a quencher and on the C-terminus with 5-Carboxyfluorescein (5-FAM) as an emitter. We identified 50 compounds out of 5000 compounds screened against SARS-CoV derived peptide that inhibited the catl mediated cleavage at 61% cutoff. The screening of those compounds against the other viral peptides derived from EBOV-GP, HeV and NiV-F 0 glycoproteins identified 12 compounds that specifically inhibited the cleavage of viral peptides and not xvii

18 the host pro-npy derived peptide. The candidate compounds were tested for their inhibitory effect on SARS-CoV and Ebola pseudotyped viruses. One compound (I.D ) was identified as a potential inhibitor of the entry of EBOV and SARS-CoV by 40% and 65% respectively in an in vitro pseudovirus inhibition assay. The compound derivative (I.D ) showed enhanced inhibition of pseudotyped viruses (54% to EBOV pseudotyped virus and 68% to SARS-CoV pseudotyped virus). The IC50 was found to be 15µM and 10 µm for and respectively which confirm the higher potency of The two compounds were confirmed to be specific in action as no inhibitory effect was detected against VSVG-pseudotyped virus (catl independent entry) even at concentration 100µM. The was found to be a mixed inhibitor of catl through Michaelis-Menten dependent kinetics assay. Based on our studies, we believe that the use of HmAbs and antiviral drugs identified in our assay could be optimized to yield broad spectrum therapies against lethal infections caused by those viruses. xviii

19 Chapter I Introduction A. RNA Viruses and Acute Infections RNA viruses are a large group of viruses that possess ribonucleic acid (RNA) as their genetic material and are involved in a wide range of human infections either acute or chronic. This nucleic acid is usually single-stranded RNA (ssrna), but may be doublestranded RNA (dsrna) (1). RNA viruses can be further classified according to the sense or polarity of their RNA into negative sense and positive sense. Positive sense viral RNA is similar to mrna and thus can be efficiently translated by the host cell into viral proteins. Negative-sense viral RNA is complementary to mrna and thus must be converted to positive-sense RNA by RNA dependent RNA polymerase before being translated into viral proteins. Accordingly, purified RNA of a positive-sense virus can directly cause infection though it may be less infectious than the whole virus particle while that of a negative-sense virus is not infectious by itself as it needs to be transcribed into positive-sense RNA. Several RNA viruses were described as the causative agents of acute lethal infections of which Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV), Ebola and Henipaviruses are considered to be a serious health threat (2-5). There is no effective therapy against these life threatening viruses to date. The ideal therapy for acute infections would be passive immunotherapy using monoclonal antibodies or antiviral drugs using small molecules which provide instantaneous protection. Because of the antigenic variants that readily arise during infection due to copying errors commonly found in RNA viruses, the RNA virus may 1

20 2 escape host immune responses (6). Therefore, a clinically useful passive immunotherapy for those RNA viruses should be effective against a broad spectrum of viral variants that may arise during infection. Additionally, developing a broad spectrum antiviral drug that can target the above mentioned viruses would be useful in treating these life threatening diseases. 1. Coronaviruses a. Taxonomy and Morphology The first coronavirus (CoV) reported to cause a disease was the avian infectious bronchitis virus (IBV). IBV was described 30 years prior to the definition of the genus Coronavirus in the 1960s following the discovery of the human CoVs (hcov) causing mild respiratory disease (7-9). The morphological and genomic similarities between the human cold viruses, IBV, and murine hepatitis virus (MHV), another previously described CoV, have led to the grouping of these viruses into the family Coronaviridae under the order Nidovirales, derived from the Latin word nidus, meaning nest, as all viruses in this order produce a 3' co-terminal nested set of subgenomic mrna's during infection (9) (Figure 1). The CoVs are round pleomorphic viruses that range in size from approximately nm, although viruses that are pictured in infected cells may be smaller in size measuring about 85nm. The CoVs are enveloped viruses with a positivesense ssrna genome and with a nucleocapsid of helical symmetry. The genomic size

21 3 Order Nidovirales Family Coronaviridae Group 3 IBV-B TCoV IBV-L Genus Coronavirus (Helical nucleocapsid) BCoV-Lun hcov OC43 MHV-A59 EToV Genus Torovirus (Toroid nucleocapsid) TGEV SARS-CoV PEDV Group 1 Group 2 hcov-229e Figure 1 Taxonomic classification of coronaviruses. Coronaviruses are classified into three groups 1, 2, and 3. Human coronaviruses are classified under group 1 and 2. Group 2 includes animal coronaviruses and SARS-CoV is classified under this group. Group 3 includes avian coronaviruses.

22 4 of coronaviruses ranges from approximately 26 to 32 kilobases, which is the largest among RNA viruses. The name "coronavirus" is derived from the Latin word corona, meaning crown or halo, and refers to the characteristic appearance of virions under electron microscopy (E.M.) with large bulbous surface projections creating an image reminiscent of the solar corona. This morphology is actually formed by the viral spike (S) peplomers, a club shaped membrane protein associated in trimers and projecting from the viral envelope and is thought to determine the virus host tropism (9) (Figure 2). There are approximately 200 copies of S protein on the virion surface. Within the viral envelope, there are two other proteins named the membrane (M) protein and the Envelope (E) protein (10). The nucleocapsid structure and genome size varies among the members of the Coronaviridae. However, the CoVs helical nucleocapsid is generated by the coating the large RNA genome with the nucleocapsid (N) protein (11). b. Coronaviruses and Infection Coronaviruses are the common cause of respiratory and enteric diseases within a wide host range (9). Until recently, human coronaviruses were thought to cause only flulike symptoms. With the discovery of SARS-CoV and more human coronaviruses, e.g. NL63, this thought has dramatically changed (12, 13). The SARS-CoV genome, which is 29.7 kb long, is distinct from other human and animal coronaviruses (10, 14) and was identified as the etiological agent of Severe Acute Respiratory Syndrome (SARS) (15-17). Coronaviruses are classified into 3 groups, the human CoV229E and CoVOC43 viruses that are known to be a major cause of common cold are respectively classified under group 1 and 2, while group 3 contains only avian coronaviruses (9, 18) (Figure 1).

23 Figure 2 SARS-CoV virion structure. Diagram of typical SARS-CoV virion showing main structural proteins, Spike (S), membrane (M), envelope (E), and nucleocapsid (N). 5

24 6 Though originally SARS-CoV was thought to be the only member of a fourth group of coronaviruses, further genomic study demonstrated that SARS-CoV is more closely related to group 2 coronaviruses (18). SARS-CoV was identified in palm civets and other animals found in live animal markets in Guangdong, China (19). The SARS-CoV that exists in animals does not cause typical SARS-like disease in these animals and is not transmitted to humans. Under certain conditions, the virus may have evolved into the early human SARS-CoV, with the ability to be transmitted from animals to humans or even from humans to humans resulting in localized outbreaks and mild human disease. Under selective pressure in humans, the early human SARS-CoV may have evolved into the late human SARS-CoV, which caused local and even global outbreaks and typical SARS in humans with high death rates. c. SARS-CoV genome: Organization, Replication Strategy, and Transcription The replication strategies of the positive sense RNA genomes of the Nidoviruses are similar, however, these viruses differ in the structure of the viral nucleocapsid, the length of the genome, the number of non-structural proteins and subgenomic RNAs generated (9). The SARS-CoV genome is polyadenylated RNA of approximately 29,727 nucleotides in length (10, 14). All the CoVs share a common genome organization. The SARS-CoV genome contains five major open reading frames (ORFs), which encode the replicase polyproteins which includes the RNA-dependent RNA polymerase (first two thirds of the genome), the spike protein (S), envelope protein (E), membrane (M), and nucleocapsid (N) proteins (10, 14). These proteins are homologous to those of other coronaviruses and are essential for viral replication and maturation. Besides these

25 7 major proteins, SARS-CoV genome also encodes about eight putative accessory proteins, which are interspersed between the above genes, showing little homology with other coronaviruses and are unique to SARS-CoV (20). These accessory proteins are dispensable for virus replication (21), however, emerging evidence has indicated that the accessory proteins could modulate cellular processes and interfere with the virushost interaction, thereby contributing to viral virulence and pathogenesis (22-27). The group 2 CoVs contains an additional structural protein, hemagglutinin esterase (HE) which is inserted between the replicase and the S gene and possibly acquired from the Influenza C virus. The HE is not present in SARS-CoV (28). The sites of genome replication in CoV infected cells are interesting structures of double membrane vesicles (DMVs) whose origins are unclear (29). The replication of CoVs including SARS-CoV begins with the generation of a negative stranded RNA which serves as a template for the production of the positive sense RNA genome (29). The N protein, in addition to being the nucleoprotein of the virus particle, serves as an important factor in the replication of the RNA genome. The N protein colocalizes with the replication complexes within the DMVs. Additionally, the efficient expression of the full length cdna clones of CoVs requires N protein which can be provided in trans on a separate plasmid in the producing cells (29, 30). The CoVs use a complex method in transcription known as template switching to form a nested set of subgenomic mrnas from which viral proteins are produced by translation (29). The SARS-CoV forms eight subgenomic mrnas that share common 5 and 3 sequences derived from the genomic RNA (31). The formation of those subgenomic RNAs is regulated by an intergenic transcription regulatory sequence (TRS) that has a conserved core sequence (CS) (10,

26 8 14, 29). The CS of the SARS-CoV has been identified to be AAACGAAC by Rota et al, and CUAAC by Marra et al. The TRS serves as the signal sequence upstream of the ORFs for template switching i.e. discontinuous transcription. Enjuanes et al, describes the transcription model of CoVs in three steps: 1) The formation of a transcription complex, 2) base pair scanning, and 3) template switch. The formation of the transcription complex brings the 5 and 3 ends of the genome into proximity; the RNAdependent RNA polymerase generates the negative strand subgenomic RNA following the copying of the CS; or may be further transcribed through the TRS to produce longer subgenomic RNAs (29). There are viral proteins that were identified to be necessary for an efficient transcription process which include the RNA-dependent RNA polymerase, the helicase and the N protein (29). The N protein has also been shown to be important for the translation of the subgenomic mrna and is suggested to bind to the subgenomic mrna leader sequence and act as a translation initiation signal (29). Additionally, other non-structural proteins have been implicated in the transcription process including an exonuclease, an endonuclease, NendoU and a methyl transferase. The nsp-9 has also been identified to be critical for transcription by acting as a single stranded RNA binding protein to stabilize the template and nascent RNA strands (29, 32). d. SARS-CoV Life Cycle The life cycle of CoVs begins with the binding of the S protein to its specific cellular receptor on the surface of target cells. SARS-CoV enters target cells in vitro via the endosome in a ph dependent manner and requires the endosomal protease, cathepsin L (catl) (33, 34). Once the S protein is cleaved by the catl, it mediates the fusion of the

27 9 endosomal membrane to the viral envelope (35, 36). Blocking of catl cleavage of S protein could be a possible way of blocking virus infection which would be a part of this study. The fusion of the viral envelope to the endosome leads to the release of the viral genome into the cytoplasm. Following the release of the genome, the cap dependent translation of the first two third of the genome starts producing the replicase polyproteins (pp1a and pp1ab) (37). The replicase polyprotein is cleaved by viral proteases to 16 non structural proteins that function in the replication of the viral genome and transcription of subgenomic RNAs (37). The S protein, once translated, forms a trimer in the endoplasmic reticulum through interaction of the oligomerization domains. The S protein is incorporated into the virion surface as it buds through the pre- Golgi compartment (38). The virus finally accumulates in cytoplasmic vesicles which fuse with the cellular membrane allowing for the egress of the newly produced viruses outside the cell. e. The SARS-CoV S protein The S protein of CoVs and that of SARS-CoV mediates the binding and fusion events that are necessary for the initiation of infection of the target cells and was shown to contain the neutralizing epitopes that induce the production of neutralizing antibodies (39-42). The S protein of SARS-CoV shares little homology with the S proteins of other CoVs (approximately 20-27%) (10). The SARS-CoV S protein is approximately 1255 amino acids in length with 23 putative N-linked glycosylation sites (10, 14). A short signal peptide is present at the N-teminus (12 amino acids), that is cleaved during transport through the E.R. and Golgi compartments, and a short transmembrane

28 10 domain at the C-terminus resulting in the exposure of the majority of the protein on the viral surface (10, 14). Like other CoVs, the SARS-CoV S protein is a type I transmembrane glycoprotein that contains a leader peptide (residues 1-12), an ectodomain (residues ), a transmembrane domain (residues ), and a short intracellular tail (residues ) (10). The S protein is functionally divided into 2 domains, the S1(aa ) and the S2 (aa ) domains (43). In many CoVs, the S protein is cleaved by host furin enzyme and the two subunits are held together non-covalently, however, like hcov229e no typical cleavage motif has been identified in the SARS-CoV S protein (10). The SARS-CoV infection is initiated by the attachment of the globular S1 domain of the S protein to the ACE2 receptor, which triggers a conformational change in the S protein (44-46). A 193-amino acid fragment within the S1 domain (aa ) has been identified as a minimal receptor binding domain (RBD), which is sufficient to associate with ACE2 (47-49). Seven cysteine residues are located within the RBD of which two, cysteine residues (348, 467, and 474) have been shown to participate critically in the formation of the tertiary structure of the RBD and changes to these residues reduced the abiity to bind to ACE2 (43). Three functional glycosylation sites are present within the RBD at amino acids 318, 330, and 357. A minimum of one site is required for expression of the S protein and mutation of two glycosylation sites does not affect ACE2 binding (50). Cocrystallization of recombinant S protein and human ACE2 was able to characterize the specific amino acid interactions between the RBD of the SARS-CoV S protein and human ACE2 receptor. The RBD consists of core structure of five antiparallel β- sheets connected by short helical regions and an extended loop formed by

29 11 two anti-parallel β-sheets to form a slightly concave surface that interacts with the tip of ACE2 (43). The crystal structure of RBD bound to ACE2 revealed that a receptor binding motif (RBM) consisting of 70 amino acids (aa ) are directly interacting with the ACE2 (43). The basic residues in between the amino acids 422 and 463 have been demonstrated to mediate entry of SARS S protein pseudotyped virus into target cells (51). Interestingly, the alignment of 96 SARS-CoV clinical isolates showed that there are no alterations in the basic residues within amino acids including R441 and R453 which confirms their important role in mediating viral entry (51). Two residues that are exposed on the RBD surface contacting the ACE2, R426 and N473, were found to be critical for RBD binding to ACE2 receptor. Residues surrounding the R426 were also found to be important for mediating the binding of RBD to the ACE2 (50). The S2 domain of the SARS-CoV S protein (residues ) is more highly conserved than the S1 domain and contains a putative fusion peptide ( ) followed by two conserved heptad repeat (HR1 and HR2) regions responsible for fusion between viral and target cell membranes (35, 36). The cleavage of the S protein by catl triggers the association of the HR1 and HR2 to form a fusion core (35, 43, 52-54), and facilitate fusion with the endosomal membrane required for the virus entry (55). The HR1 domain of SARS-CoV extends from amino acids while the smaller HR2 domain extends from (43). These two domains are connected by interdomain loop of ~ 170 amino acids long. This loop is suggested to allow flexibility between the HR1 and HR2 regions that is required during the fusion process of the viral envelope to the endosomal membrane (36). The HR1 domain is thought to form the inner helical coiled coil region onto which the HR2 domain associates in an anti-parallel manner to form a

30 12 six helix bundle fusion core (35, 36). The formation of the six helix bundle brings the N- terminus of the HR1 domain containing the fusion peptide in close proximity to the transmembrane domain of the S protein placing the cellular and viral membranes in juxtaposition to each other (35, 36). Accordingly, therapies that can disrupt the association of the HR1 and HR2 domains would likely confer protection against SARS- CoV by inhibiting the fusion step in viral entry. Consequently, synthetic HR2 peptides as well as HR2 specific antibodies have been shown to block SARS-CoV infection (56-58). The S protein has been implicated in the interaction with other cellular proteins. The glycoprotein CD209L (L-SIGN) has been shown to serve as SARS-CoV receptor. The ability of CD209L mediate viral entry was discovered using a human lung cdna library. These constructs were expressed in CHO cells, and the involvement of the expressed proteins in viral entry was confirmed by the presence of viral subgenomic mrnas in the target cells (59). Similar to ACE2 expression pattern, CD209L is found to be highly expressed on type II pneumocytes of the lung. CD209L is also implicated in the entry of Ebola and Sindbis viruses and in the attachment of HIV and HCV (59). The interaction of the SARS-CoV S protein with the dendritic cell specific ICAM-3 grabbing non-integrin (DC-SIGN) has been shown to serve as SARS-CoV entry factor (60). The region of the RBD that interacts with the DC-SIGN has been mapped to the amino acids , which is upstream of the identified RBM region that interacts with the cellular receptor ACE2 (39). DC-SIGN has also been implicated in the ability of dendritic cells to transfer SARS-CoV to the target cells in vitro through the formation of a synapse-like structure without productively infecting the DCs, similar to HIV (60).

31 13 f. Tropism Coronaviruses primarily infect the upper respiratory and gastrointestinal tract of mammals and birds (9). Four to five different currently known strains of coronaviruses infect humans. The SARS-CoV has a unique pathogenesis because it causes both upper and lower respiratory tract infections and can also cause gastroenteritis (2). Coronaviruses are believed to be the major cause of all common colds in human adults. The significance and economic impact of coronaviruses as causative agents of the common cold are hard to assess because, unlike rhinoviruses, human coronaviruses are difficult to grow in the laboratory. Coronaviruses cause a wide range of diseases in animals and domestic pets, which can be a threat to the farming industry. Economically significant coronaviruses of farm animals include porcine coronavirus, the more pathogenic transmissible gastroenteritis coronavirus (TGE) and bovine coronavirus (bcov), which can result in diarrhea in young animals (9). Feline Coronavirus has 2 forms, Feline enteric coronavirus is a pathogen of minor clinical significance, but spontaneous mutation of this virus resulted in feline infectious peritonitis (FIP), a disease associated with high mortality (9). Two types of canine coronavirus (CCoV) were identified, one that causes mild gastrointestinal disease and one that has been found to cause respiratory disease (9). Mouse hepatitis virus (MHV) is a coronavirus that causes an epidemic murine illness with high mortality, especially among colonies of laboratory mice (9). Prior to the discovery of SARS-CoV, MHV had been the best-studied coronavirus both in vivo and in vitro as well as at the molecular level. Some strains of MHV cause progressive demyelinating encephalitis in mice which has been used as a

32 14 murine model for multiple sclerosis. SARS-CoV represents a unique member of the Coronaviridae family as it can infect different species and multiple organs within the same species. Additionally, it can replicate efficiently in mice, hamsters, cats, ferrets, and monkeys (61, 62). g. SARS-CoV Outbreak and Pathology SARS disease in humans remains a credible health threat because of the possible animal reservoirs, including bats as the natural reservoir and palm civets as well as raccoon dogs which served as intermediate hosts (19, 63-65). Upon adaptation of the SAR-CoV to humans, the outbreak of SARS started at the end of 2002 resulting in the first major pandemic of the twenty-first century (2). SARS-CoV managed to establish efficient human to human transmission in 2003 resulting in 774 deaths and 8096 cases world wide in which 29 countries were involved (2). Patients infected with SARS-CoV show symptoms of atypical pneumonia. Severe lung damage results in the transition of the disease to acute respiratory distress syndrome (ARDS) and the need for ventilation support. SARS-CoV initially infects type II pneumocytes, which represent the epithelial cells lining the alveolus of the lung. Lung epithelial cells hyperplasia is often associated with SARS-CoV infection (66). Patients infected with SARS-CoV show infiltration of mononuclear cells into the lung leading to lesions in the lung tissue which appear as opacities in the chest radiographs (67, 68). The progression to a more severe state of the disease occurs within the second week of infection, suggesting a possible immune contribution to SARS pathology (68,

33 15 69). The lung damage is probably not due to increased viral replication, but rather due to other mediators (70). The third phase of infection is ARDS which requires the ventilator support and intensive care, approximately half of the patients that get to that stage will die (2, 70). Other clinical manifestations include: lymphopenia, thrombocytopenia, and hyaline membrane formation (diffuse alveolar damage) (66, 71). A unique feature of SARS infection is that, although the primary site of infection is the lung, the virus causes systemic infection involving intestine, kidney, and liver from which viral RNA can be recovered (68). This provides an explanation for the diarrhea associated with the SARS infection, shedding of virus in the urine, feces, and aberration in liver enzymes (68). h. SARS-CoV Evolution SARS-CoV, like all RNA viruses, is characterized by the high rates of mutation that pose a particular problem for developing effective modalities of intervention (72). Such mutations can result in increased virulence, avoiding host immune responses, and changing tissue tropism (6). Several studies have identified genomic variations among SARS-CoV isolates. The variations were found to be in the S protein of the SARS-CoV as well as in accessory proteins (73). Interestingly, the changes within the S protein were found to be localized to the S1 domain while the S2 domain was found to highly conserved among the different isolates (74). Of the S1 changes, two amino acids (N479K and T487S) were found to be important for the transition from palm civet to humans (75). A study of the viral genome sequences of 61 SARS-CoV isolates during the outbreak ( ) indicated that they may be categorized as early, middle, and

34 16 late stage isolates (73). The genomes of SARS-CoV isolated from civet cats and some human isolates of the early stage epidemic contain a single ORF8ab, while viruses isolated from the middle-late phase of the epidemic contain a 29-nucleotide deletion that results in two separate ORFs, ORF8a and ORF8b (19, 73). Isolates recovered in represent a second event of SARS-CoV evolution (74). These isolates appear to be less efficient in using human ACE2 as a receptor. Infection by these isolates is not inhibited by souble ACE2, where as late phase isolates from are susceptible to inhibition by soluble ACE2 (76). The isolates are quite similar to those of palm civets SARS-like-CoV than to the human adapted late phase isolates (personal communication Michael Farzan). This suggests that either the requirement of ACE2 in infection changed or the interaction with ACE2 changed throughout the outbreak. Since the S protein is involved in viral attachment and fusion and in eliciting protective immune responses against SARS-CoV infection (77), antigenic variation or antigenic drift in this protein is of significance. The amino acid sequence analysis of the S protein from 94 clinical isolates (Genbank) identified mutations in the RBD region (Hatem Elshabrawy, data not published). These facts, combined with the recent discovery of a likely reservoir in bats (78), the potential ADE of SARS-CoV (79, 80), and the identification of autoantibodies against lung epithelial cells during SARS-CoV infection (81), raises a considerable challenge to develop effective interventional strategies including antibody and antiviral drug therapy.

35 17 i. Immune responses to SARS-CoV 1. Innate Immunity Innate immune responses are important for clearing infections including viral infections. The functional components of the immune system include: Natural killer (NK) cells, the protein mannan-binding lectin (MBL), surfactant, type I interferon (IFN) responses as well as phagocytic cells including macrophages and DCs (82). The innate immune responses have been extensively studied in SARS-CoV infected patients, animal models and in vitro. Studies in patients showed decreased peripheral NK cells, correlating with disease severity. The type I interferon responses have been shown to be blocked by SARS-CoV in vitro with no detectable levels of interferon in patients serum (82, 83). However, the MBL is of importance in combating SARS-CoV infections probably by binding to carbohydrate residues on the surface of the SARS-CoV S protein (82). 2. Cell mediated Immunity The cell mediated immune responses are critical in eliminating viral infections. The major effector cells in cell mediated immunity against viruses are CD8 + cytotoxic T lymphocytes (CTL). Those types of cells are capable of killing virus infected cells that are expressing viral epitopes in context with MHC class I. The CD4 + T- helper cells are also important in mounting efficient CD8 + T cell responses. The T helper cells are classified into Th1 and Th2 subsets. Th1 cells are important in developing inflammatory CTL responses; however the Th2 cells are important in developing humoral responses together with B cells.

36 18 Understanding the role of T cell responses in context of SARS-CoV infection has been accomplished through examining vaccine candidates. DNA vaccines expressing different truncations of the SARS-CoV S protein has been shown to induce neutralizing antibody responses as well as cellular immune responses (77, 84-86). DNA vaccines expressing SARS-CoV 3a accessory protein induced humoral and cellular immune responses as well (87). Additionaly, induction of specific Th1 type immune responses has been shown upon vaccination with SARS-CoV N, M, and E expressing DNA vaccines (88). 3. SARS-CoV and Humoral Immunity Humoral immune responses appear to play the major role in clearing SARS-CoV infections (77). Several studies in patients and animals have shown that infection with SARS-CoV leads to induction of neutralizing antibodies (Abs) (89-94). These neutralizing antibodies play a significant role in protection against SARS-CoV infection; however, there is a positive correlation between the antibody response and the disease outcome. It has been demonstrated that passive transfer of immune serum to naive mice protected against SARS-CoV virus challenge (77). However, depletion of CD4 + and CD8 + cells did not abolish protection and the adoptive transfer of SARS-CoV specific CD4 + and CD8 + cells did not protect the mice against SARS-CoV challenge (77). The previous findings show the importance of humoral responses over the cellular responses in protection from the SARS-CoV infection. The primary role of neutralizing antibodies in conferring protection against SARS-CoV suggests that passive

37 19 immunotherapy may be a useful therapeutic measure to provide instantaneous protection from SARS-CoV. a. General Properties of Antibodies Antibodies are the main effector proteins produced during an immune response from activated B cells (Plasma cells). Each antibody is formed by a combination of two light (L) and two heavy (H) chains that are linked by disulfide linkages. The L chain contains one variable (V) and one constant (C) region while the H chain is comprised of one V region and three C regions. The variable region of each the L and H chain form three hypervariable (HV) regions which make up the three complementarity Determining Regions (CDRs) of each of the L and H chain. The CDRs of one L and one H chain combine to form the three CDRs of the antibody. Each antibody molecule will finally have two antigen binding sites where each site consists of three CDRs. The antibody recognizes and binds to an epitope which is a short stretch of amino acids in a target protein or formed by the protein tertiary structure (Conformational epitope). The constant region of the heavy chain (or Fc) determines the effector functions of the antibody. There are five isotypes of antibodies depending on the type of the constant region: IgM (µ constant region), IgG (ɣ constant region), IgA (α constant region), IgE (ε constant region), and IgD (δ constant region). The IgG is the most abundant antibody in serum. The tissue distribution of the Ab is determined by the C region of the Ab. Though most Abs are distributed throughout the body by diffusion, delivery to specific regions requires special transport molecules which needs interaction with specific Fc regions resulting in localization of some isotypes in specific regions of

38 20 the body. IgA antibody is present in mucosal secretions, IgG can be transported through the placenta and IgE is found in skin and mucosal layers. Those distribution specificities must be considered when characterizing an Ab for therapeutic use. The whole Ab molecule is not always necessary for effective function. There is some evidence that Fraction antigen binding (Fab) portion of the Ab may be sufficient in some cases. The Fab portion of the Ab consists of the L and H chain variable region with the first constant region. Additionally, one must consider the importance of the Ab subtype when intended to be used therapeutically. In yellow fever virus, for example, a mab with the IgG1 Fc domain is ineffective in clearing the virus infection in a mouse model of infection, whereas, an Ab with the same specificity but with IgG2a Fc domain is effective. This can be attributed to the poor effector functions of the IgG1 Fc domain. It is already established that all IgG subtypes are efficient in diffusing to extravascular sites in tissues. Additionally, the IgG1 and IgG3 are potent activators of complement and have a high affinity for Fc receptors on macrophages, DCs and some tissues. The differences in the ability of IgG subtypes to activate effector functions are critical in the selection of the IgG subtype for immunotherapy. Ab preparations that can be used for passive therapy of viral infections are monoclonal and polyclonal antibodies. Polyclonal antibodies represent a mixture of antibodies recognizing multiple epitopes on a specific target protein. However, the polyclonal Ab preparation carries the background of the animal from which it is derived, which limits its use in treating human beings on one hand due to hypersensitivity issues and decrease the specificity on the other hand. The use of human polyclonal antibodies from convalescent patients will be limited by the fear of transmission of pathogens

39 21 between individuals. Monoclonal Abs recognize a single epitope and therefore more specificity and efficacy is guaranteed and no risk of pathogens transmission. Polyclonal Ab preparation may vary with different lots while monoclonal Ab production relies on hybridomas which eliminate the variation of Ab amount between different lots (95). Most Ab preparations used in immunotherapy are IgG isotype because of the ability of this Ab class to reach tissues throughout the body. The type of Fc of the IgG isotype must be considered because, although the effector functions are important but may be detrimental to the Ab function by enhancing the antibody dependent enhancement of infection (ADE). The nature of the pathogen has a great influence on the subtype of the IgG used. If the neutralizing property of an antibody is required so an IgG subtype with minimal or no effector functions can be used. However, if opsonization of the antigen leading to its phagocytosis is desired, so a subtype with high ability to activate effector functions will be required. For SARS-CoV infection, it is important that the Ab used as immunotherapeutic agent to be neutralizing so that they can prevent the virus from infecting target cells and allow the immune system to respond and clear the viral infection. It is preferable for monoclonal antibodies used to be human antibodies to minimize side effects like serum sickness which may happen if the antibody is of animal origin. There are several ways by which antibodies can protect against a specific pathogen; 1-Neutralization, 2-Opsonization, and 3- Complement activation. Neutralizing antibodies are important when dealing with viral infections because they prevent entry of the virus into the target cells and subsequently deactivates the virus and renders it harmless (96). Four main hypotheses are describing the mechanism of neutralization of viruses by

40 22 Abs. 1- Neutralization happens extracellularly, 2- The Ab binds to the viral target protein and induces a conformational change in the target protein leading to inability of the viral protein to bind to the target cells, 3- Neutralization may happen post-binding to the receptor on the target cells by interfering with the virus uncoating, and 4- Simple occupancy model, in which neutralization occurs by sufficient amount of the antibody occupying the target protein and after a specific threshold is reached, the Abs prevent attachment of the virus or interfere with another step of the virus entry (96). ADE must be considered when optimizing therapies for treating the SARS-CoV infection. ADE can occur by increased uptake of the virus into the target cells through the Fc receptors that may be expressed on the surface of some cell types as has been demonstrated in case of group 2 CoV feline infectious peritonitis virus and in a ferret model of SARS infection (97, 98). b. Monoclonal Antibody Technology The discovery of the monoclonal antibody (mab) technology paved the way for the use of the expansion of the Abs use as therapeutics. With respect to infectious diseases, the mabs provide the opportunity for the production of reliable diagnostic assays as well as powerful therapeutics with highly specific and defined treatment to the target protein (99, 100). The mab technology was developed by Niels Jerne, Cesar Milstein, and George Kohler for which they shared the Nobel Prize in physiology or medicine in Shortly, a mouse mab targeting human T cell CD3 was approved for therapeutic use to prevent rejection of kidney transplants (101). However, the use of this mouse antihuman CD3 antibody was limited due undesirable immune responses in

41 23 treated individuals against the mouse constant region. The challenge to circumvent this problem has been met with increasing effectiveness by development of chimeric antibodies of which some are used today as therapeutics including Herceptin, Remicade, Humira, and Synagis (101). The next development of mab technology was the production of a transgenic mouse (e.g. Xenomouse, whose antibodies were used in this study) that allowed the production of fully human mabs in the context of an intact immune system. In 2006, Vectibix was produced as the first FDA approved human mab for the treatment of epidermal growth factor receptor (EGFR) expressing colorectal cancer (101). Different approaches have been used to produce Human mabs (HmAbs): 1- Hybridoma fusion, 2- Phage display, and 3- Immortalization of convalescent B cells. All of these strategies have been used to produce human mabs against SARS-CoV (91, ). The use of human immunoglobulin transgenic mice takes the advantage of the utility of hybridoma technology producing high affinity mabs of fully human sequences that in the context of the intact mouse immune system undergo somatic hypermutation and affinity maturation (101, 105, 106). The human immunoglobulin transgenic mice come in several varieties IgG (subtype 1, 2, 3, and 4), IgM, IgGD, IgGA, and IgGE allowing for the selection of a mouse strain encoding the most suitable constant region for downstream applications (101). c. SARS-CoV and Antibodies Neutralizing antibodies against SARS-CoV has been shown to be targeted to the S protein. The majority of the neutralizing Abs is targeted to the receptor binding domain

42 24 (RBD) (107). The mechanism of action of those RBD binding Abs was demonstrated to be through the inhibition of binding of the virus to the ACE2 receptor (108). The majority of neutralizing human monoclonal neutralizing antibodies used in a part of this study is binding to regions within the RBD. Other S1 binding neutralizing Abs have been shown to bind to epitopes N-terminus to the RBD (amino acids , ) as well as C- terminus to the RBD (amino acids and ) (91, 102). Another set of neutralizing Abs have been shown to bind to the S2 domain of the SARS-CoV S protein specifically to the HR2 region as well as the S1-S2 junction (94, 109). The presumed action of the S2 binding antibodies is through hthe inhibition of the viral fusion likely by prevention of conformational changes required for the mediation of fusion. Analyzing the sequences of the clinical isolates available up to date in Genbank demonstrates that the changes in the S protein, within the RBD, are tolerated and the majority of the changes by alanine scanning mutagenesis did not result in significant inhibition of the association of the soluble RBD protein and ACE2 (50). This suggests that the S protein of the SARS-CoV can tolerate changes without affecting its ability to bind ACE2 allowing the virus to escape immunological pressure and infect target cells (50). Recent papers have shown the importance of the mabs targeted to the S protein in neutralizing the SARS-CoV. However, the mabs generated against the late isolates may not be effective in neutralizing the zoonotic isolates. Similarly, Rockx et al, demonstrated that mabs could be grouped based on efficacy against early, middle, and late isolates, proving that all mabs are not equally effective against different isolates (110). Additionally, the passage of the virus in the presence of mabs may lead to the generation of escape mutants (111, 112). Those previous findings impose caution when

43 25 producing mabs for use in neutralizing SARS-CoV. Therefore, a complete understanding of the neutralizing Abs binding and developing broad spectrum mabs becomes a necessity for an effective passive immune therapy which will be a major part of this work. d. SARS-CoV and Passive Immunotherapy Studies in patients infected with SARS-CoV during the outbreak demonstrated that the administration of convalescent patient serum protected uninfected individuals against infection, without any obvious adverse effects (92, 113, 114). This suggests that the passive immunotherapy offers a safe and effective treatment strategy. The peak viral titers in SARS-CoV infection are reached 10 days post-infection which furthermore supports the possibility of passive immunotherapy as a potential treatment (115).

44 26 2. Ebola Virus a. Taxonomy and Morphology Ebola virus (EBOV) and Marburg virus (MARV) belong to the Filoviridae family of viruses that produce enveloped virions having a filamentous morphology (116). The Filoviruses are known to cause hemorrhagic fever in humans and nonhuman primates (3). The name Filoviridae is derived from the Latin word filum, which means filament, referring to the filamentous morphology of the viruses (116). The family currently includes the three virus genera "Cuevavirus", Ebolavirus, and Marburgvirus and is classified in the order Mononegavirales. Within the genus Marburgvirus, there is a single species named Lake Victoria marburgvirus that is designated MARV and contains six strains with the prototype Musoke strain. The genus Ebolavirus consists of four recognized species: Ivory Coast Ebolavirus (ICEBOV), Reston Ebolavirus (REBOV), Sudan Ebolavirus (SEBOV), and Zaire Ebolavirus (ZEBOV). The prototype virus for the genus is ZEBOV strain Mayinga. All the members of the family are considered to be World Health Organization Risk Group 4 Pathogens (requiring Biosafety Level 4-equivalent containment) (US Department of Health and Human Services. "Biosafety in Microbiological and Biomedical Laboratories (BMBL) 5th Edition". Retrieved), National Institutes of Health/National Institute of Allergy and Infectious Diseases Category A Priority Pathogens (US National Institutes of Health (NIH), US National Institute of Allergy and Infectious Diseases (NIAID). "Biodefense NIAID Category A, B, and C Priority Pathogens". Retrieved), and Centers for Disease

45 27 Control and Prevention Category A Bioterrorism Agents (US Centers for Disease Control and Prevention (CDC). "Bioterrorism Agents/Diseases" Retrieved 2011). b. Evolution The phylogenetic classification of filoviruses has been determined depending on the viral glycoprotein (GP) gene sequences ( ). This suggests that the family Filoviridae has gone through a great deal of divergence. At the genomic level, Ebola and Marburg viruses differ by approximately 55%, and at the amino acid level, they differ by up to 67%. There is also a difference in the GP gene organization. A 37-40% difference at the nucleotide level and 34-43% difference at the amino acid level occurs between Ebola viruses (117). This indicates that ICEBOV, REBOV, SEBOV, and ZEBOV are distinct and represent different species. Despite the difference between species of EBOV, there is a remarkable genetic stasis (117, 121, 122). The genetic stasis suggests that there is fitness within a given ecological niche that has not been modified over the last few decades. It has been anticipated that the filoviruses are associated with a particular host and genetic variations are just ways of adapting to the natural host. The polymerase amino acid sequence suggests that filoviruses are more cosely related to paramyxoviruses particularly pneumoviruses than to rhabdoviruses (123).

46 28 c. Genome Organization and Replication Filoviruses genome consists of nonsegmented, single, negative-stranded linear RNA molecule and contributes to 1.1% of the total virion mass. The genome size is 19 kb for MARV and 18.9 kb for EBOV ( ). The viral genome encodes nucleoprotein (NP), virion structural protein (VP) 35, VP40, glycoprotein (GP), VP30, VP24, and RNA-dependent RNA polymerase gene (L). The genes of filoviruses are flanked at their 3 and 5 end by noncoding sequences containing replication and encapsidation signals ( ). Highly conserved motifs have been shown to be associated with filoviruses genes for transcriptional start and stop (130, 131). The extragenic sequences at the 3 and 5 end of the genome are highly conserved and have important role as promoters in transcription and replication of filoviruses. The genes between those promoters are delineated by conserved transcriptional signals beginning at the 3 genome end and terminating with a transcriptional stop (polyadenylation) site. The transcription is terminated when the polymerase encounters a series of five to six Us where the polymerase starts to stutter and adds a long polya tail to the transcripts ( ). The filoviruses are characterized by long noncoding sequences at the 3 or 5 end or at both ends which contributes to the increased length of the genome and may have an influence on the stability of the mrna and hence the expression level transcripts (125, 126).

47 29 d. Ebola Viral Proteins Filoviruses including Ebola virus produce seven structural proteins from the seven genes encoded by the genome. Four of the seven proteins directly associates with the negative RNA genome to form the ribonucleoprotein complex (RNP) which includes nucleoprotein (NP), VP30, VP35, L (132). The other three structural proteins associate with the lipid envelope of the virus which includes the glycoprotein (GP 1,2 ) as the major surface spike protein and VP40 as the major matrix protein, while VP24 is believed to be a minor matrix protein involved in the budding process (133, 134). In addition to the structural proteins mentioned above, a non structural soluble precursor protein is produced named sgp and is subsequently cleaved into sgp and delta ( ) peptide (135). Recent studies have shown that the expression of the VP40 protein induces the filamentous particle formation nearly identical to the wild type virus and when coexpressed with the GP protein, the GP spikes are localized to the surface (136). A previous study suggests that both proteins interact to affect morphogenesis. VP24 has been suggested to play a role in budding and assembly, due to its association with the lipid membranes (133). The synthesis of the GP 1,2 involves the processing by the furin enzyme, a subtilisin/kexin-like convertase localized in the trans-golgi, at a polybasic cleavage site. The mature protein consists of the N-terminus GP 1 (140 kda; EBOV) and the carboxy terminal fragment GP2 (26 kda; EBOV) which are linked by a disulfide bond (137). The GP 1 is heavily glycosylated with 50% molecular weight due to N- and O- glycans (138). The mature protein is present as trimeric heterodimer on the surface

48 30 of virus particles and the trimerization is mediated through the GP 2 domain of the protein (139). e. Ebola Virus Transmission, and Disease Human infections with the Ebola virus has always occurred in the rural parts of Africa or through a contact with infected human primates. While the natural virus reservoir was shown to be fruit bats (140), no common event has been associated with proximity to bats, local travel or painful insect bite. Recent outbreaks in Durba/Watsa in the Democratic Republic of Congo have occurred in or around a subterranean gold mine where the reservoirs are likely present. Subsequent spreading of the virus was to people who cared for infected patients, to individuals who were in close contact with body secretions and blood or to those who prepared bodies for burial (141). The primary mechanism of human filoviral infection is the close contact with skin and secretions of an infected individual. Virus is believed to enter through small skin lesions and mucous membranes after which it gets access to the vascular system (142). There is no evidence for aerosol transmission although it can not be excluded (143). The exact doses and route of infection for Ebola virus is currently unknown, however, nonhuman primates are infected by low doses of virus aerosols (144). The EBOV epidemic has stopped because of the low transmissibility of the virus and because of the prevention of the infection within the medical care facilities as well as quarantine practiced by people once the lethality of the virus was recognized. Filovirus infection results in lethal hemorrhagic disease in humans and nonhuman primates (142). The hemorrhaghic fever (HF) that is caused by filoviruses, unlike other viruses, is the most severe form of all viral hemorrhagic fevers and are typically

49 31 associated with hemorrhagic manifestations, coagulation disorders, shock, and hepatic failure (145). The incubation period for the filovirus HF is approximately 4-10 days which is followed by sudden onset of nospecific symptoms like chills, fever, malaise, and myalgia. With the progression of the disease, more severe symptoms are observed, like gastrointestinal symptoms (vomiting, nausea, abdominal pain, diarrhea, and anorexia), respiratory (chest pain, shortness of breath, and cough), vascular symptoms (postural hypotension, and edema), as well as neurological symptoms characterized by headache, confusion, and coma (145). Death in filovirus infections is usually associated with shock due to increased permeability of blood vessels, hypotension, coagulation problems, and focal tissue destruction (146). A humoral response is typically observed in survivors around day 7-11 which may lead to either death or an improvement in health (147). Filoviruses are known to infect almost all tissues, including skin, mucous membranes, and internal organs. Focal necrosis are seen in many organs but maximally in liver, spleen, kidney and gonads (148). Spleen and lymph nodes show extensive follicular necrosis and necrotic debris. The infected lungs usually show interstitial edema and hemorrhage with alveolar damage. Additionally, the heart shows edema and focal necrosis. The primary site of viral replication is believed to be the macrophages and dendritic cells (DCs) (149, 150). Endothelial dysfunction can lead to a wide range of vascular effects leading to hemorrhage and increase in vascular permeability (151).

50 32 f. Ebola Virus Glycoprotein Structure, and Functional Organization Ebola virus glycoprotein (GP) is a class I viral membrane fusion protein that is similar to the prototypic HIV-1 envelope (Env) protein and Influenza virus hemagglutinin (HA) in organization and function (152, 153). The mature GP consists of GP 1 and GP 2 subunits which are the products of GP 0 protein cleavage by furin host protease during the viral assembly (154). The GP 1 mediates the interaction with the receptor on the host cell and interacts with the GP 2 subunit which carries out the fusion of the viral membrane with the host cell membrane (153). The GP 1 -GP 2 forms a trimer which is mediated by the GP 2 subunit (153). The GP 1 subunit is divided into three domains: 1- The N-terminus highly conserved half forms the base which interacts extensively with GP2 holding it in its prefusion state, and the head which has the receptor binding sequences (RBS) (155, 156), 2- The C-terminus domain constitutes a highly glycosylated region which consists of the glycan cap and the mucin domain (Figure 3). The glycan cap is critical for GP folding and has an important role in entry, while the mucin domain is dispensable for EBOV-GP dependent pseudovirus entry in vitro but is proposed to mediate viral adhesion to specific cell types within the host and may play a role in evasion of immune responses by shielding key neutralization epitopes (157, 158). The base and the glycan cap are connected by a flexible loop. The transmembrane fusion subunit, GP 2 contains a hydrophobic internal fusion peptide at the N terminus in addition to N-terminal and C- terminal helical heptad repeats (HR1 and HR2) (152, 159). The GP2 undergoes a large conformational change during membrane fusion.

51 33 S S N SP S S RBS S S Glycan cap Mucin domain FL HR1 HR2 TM C S S Furin cleavage site GP1 GP2 Figure 3 Structural organization and features of the Ebola virus glycoprotein, GP. Linear diagram of Ebola GP showing the cleavage of GP0 by the cellular Golgi endopeptidase furin into GP1 and GP2. SP, signal peptide. RBS, receptor binding sequence. FL, GP2 fusion loop. HR1, GP2 N-terminal heptad repeat. HR2, GP2 C- terminal heptad repeat. TM, GP2 transmembrane domain. Lines and SS indicate intrasubunit and intersubunit disulfide bonds.

52 34 g. Ebola Virus Entry into Host Cells EBOV enters and infects many cell types with a broad mammalian host cell range (160, 161). The broad host cell range suggests that the EBOV either uses a receptor that is ubiquitously expressed in all cell types or there is a redundancy in the receptor that the virus is using to enter. There is some evidence that suggests the later scenario. The C- type lectins e.g. DC-SIGN, L-SIGN which are expressed in many cell types can mediate the entry of EBOV as also observed for other viruses (162, 163). Some of the cells susceptible to EBOV infection do not express the C- type lectins and their importance in vivo remains to be determined. Recently, the T-cell immunoglobulin and mucin domain protein TIM-1, a T-cell costimulatory molecule and phosphatidyl serine receptor was identified as a candidate cell surface receptor for EBOV (164, 165). TIM-1 was shown to interact with GP 1 and mediate entry into highly permissive cell lines while enhancing entry into weakly permissive cell lines. TIM-1 is highly expressed on human epithelial cells including airway epithelium (164). However, other permissive cells like macrophages and DCs do not express it (164). This suggests that EBOV can enter these cells through receptor other than TIM-1. The role of TIM-1 in EBOV entry in vivo remains to be determined. Following the attachment of the virus to the receptor, internalization of the virion happens via a process resembling macropinocytosis (166, 167). The process is involving formation of actin-based plasma membrane ruffles (168). The large particle size of EBOV makes it logical for the virus to use and exploit the macropinocytosis for internalization. However, the use of this route is dependent on the GP interactions rather than the virion size. Potential inducers of macropinocytosis are TIM-1 and other

53 35 lectins both of which have been implicated in macropinocytosis (164, 169). Studies have shown that GP pseudotyped viruses localize with late endosomes at later time posttransduction and a dominant negative inhibitor of Rab7 reduces infection (166). This suggests that delivery to late endosomes is necessary for viral entry. It is not yet known whether the virus delivers its genome to the cytoplasm directly from late endosomes or it should pass through another compartment like lysosomes before entering the cytoplasm. Generally, the class I viral membrane proteins are primed for fusion by a host protease within the virus producer cell which typically involves cleavage at a single cleavage site (170). This cleavage is necessary because it liberates a fusion peptide making it ready for fusion upon virus binding to receptor on the target cell. The cleavage of GP into GP1 and GP2 is dispensable both in vitro and in vivo. This has led to the conclusion that the GP protein is not primed for fusion by furin in virus producing cells but by host endosomal cysteine proteases in the target cell (171, 172). The priming of EBOV GP requires both cathepsin B (catb) and cathepsin L (catl) cysteine proteases (172). Structural and biochemical evidences suggest that the cleavage in the loop connecting the base and the head of the GP 1 subunit relieves a restraint on the underlying GP 2 fusion loop allowing it to emerge and function in the fusion process (153). Another explanation for the importance of cleavage in triggering fusion is that it unmasks the binding site for an unknown receptor with a role in promoting fusion (153). The support for the later speculation comes from the fact that catl treated virions bind better to the target cells. The host protein Niemann-Pick C1 (NPC1) was recently identified as a factor involved in the EBOV virus entry (173, 174). NPC1 is largely

54 36 expressed by all cells localized to late endosomes and lysosomes and function in lysosomal efflux of low density lipoprotein-derived cholesterol. A small molecule that inhibits NPC1 blocks EBOV entry in vitro (173, 174). Hamster cell lines which are deficient in NPC1 are shown to be refractory to infection by EBOV (173, 174). This suggests that the requirement for NPC1 is absolute for EBOV entry. NPC1 appears to be required only for the viral escape to the cytoplasm and its function in EBOV virus entry is completely unrelated to its role in cholesterol metabolism (173, 174). Cleaved but not uncleaved GP can associate with late endosomes in NPC1 dependent manner suggesting that cleavage of GP by cysteine proteases primes the GP for binding to NPC1 (174). h. Prevention of Ebola Virus Infection The first attempt to intervene with EBOV lethal infection was through development of vaccine soon after the outbreak in 1976 and used formalin-fixed or heat inactivated virus to protect guinea pigs and nonhuman primates (151, 175). However, the protection achieved in both studies was inconsistent and the inactivated virus did not mount sufficient immune response to protect baboons against lethal doses of virus (176). There has been a great effort to develop subunit vaccines since 1990 for EBOV. These studies on vaccine development have led to the conclusion that the mice are most protected species while nonhuman primates need more effort. Developing a protective measure against EBOV is required and to date there is no effective therapy against EBOV. Being an acute infection, EBOV hemorrhagic fever needs instantaneous protective measure that is safe and effective. Screening for small molecules inhibitors of

55 37 viral replication or entry would be essential to characterize an inhibitor that interferes with the virus life cycle. The importance of fusion step of the viral membrane with the host membrane in the viral entry, which is mediated by the GP protein, makes it an ideal target for developing inhibitors for the viral entry. Developing potent inhibitors aiming to block the cleavage of the GP protein, leading to fusion inhibition, will be an important part of the current study.

56 38 3. Henipaviruses a. Taxonomy and Morphology The Paramyxoviridae family of viruses has been divided into two subfamilies: Paramyxovirinae and Pneumovirinae. The Paramyxovirinae subfamily comprises five genera: Respovirus, Rubulavirus, Avulavirus, Morbillivirus, and Henipavirus. The genus Henipavirus is represented by the species Hendra (HeV) and Nipah (NiV) viruses. The two viruses are zoonotic viral pathogens responsible for repeated outbreaks with high morbidity and mortality in Australia, Southeast Asia, India, and Bangladesh which lead to the classification of them as Biosafety Level 4 (BSL4) agents (4, 5, 177, 178). The outbreaks began as zoonotic respiratory infections in either horses (HeV) or swine (NiV) that were transmitted to humans in close contact with infected animals (179). The symptoms start 1-2 weeks post infection with fever, chills, headache, and myalgia which may proceed to encephalitis or severe respiratory distress (178, ). HeV and NiV have a broad host tropism that includes pigs, horses, cats, dogs, guinea pigs, hamsters, ferrets, monkeys, and humans (183). HeV and NiV have been classified into a new genus because of their genomic lengths and protein homology is different from other genera of paramyxoviruses (5). b. Genome, Viral proteins, and Entry into Target Cells HeV and NiV are enveloped; negative sense single stranded RNA viruses. As most paramyxoviruses, HeV and NiV genome encodes for two surface glycoproteins which are required for virus entry into the target cells (184). The two proteins are: 1- An

57 39 attachment glycoprotein (G) and 2- Fusion glycoprotein (F). The G protein is required for receptor binding and virion attachment to the host cell, while the F protein is involved in mediating the fusion of the viral membrane with the host cell membrane. The G and F proteins of both HeV and NiV share a high degree of similarity (approximately 83% and 89% amino acid identity respectively) which makes them very similar functionally (184). The G protein of henipaviruses is a type II transmembrane protein that consists of an N- terminus cytoplasmic tail, a transmembrane domain, a stalk domain, and a globular head; however, unlike the other paramyxoviruses, the G protein possesses neither hemagglutinin nor neuramidinase activities. The globular head of the G protein consists of a β-propeller with a central cavity surrounded by six blades, which are composed of four anti-parallel beta sheets ( ). The globular head has been shown to be N- glycosylated at five positions (187). The G protein forms a tetramer which is a dimer of dimers (187, 188). The oligomerization of the protein is mediated by the stalk region (187). While the G glycoprotein mediates viral attachment of the henipaviruses to the target cells, the F glycoprotein is responsible for the fusion of the viral envelope with the host cell membrane. The F protein is a class I viral membrane fusion protein and is initially expressed as a 546 amino acid precursor (F 0 ) which is cleaved into two subunits (F 1 and F 2 ) by the endosomal protease catl during the viral assembly (189). The cleaved F protein forms trimer (189). Unlike other viruses, the processing of the F 0 into the biologically active form is a multistep process which requires the internalization of the F 0 from the cell surface, during the viral assembly, to the endosome mediated by the internalization motif present in the cytoplasmic tail (190, 191). Following cleavage of

58 40 the F protein in the endosomal compartment, the cleaved protein is recycled back to the cell surface to be packaged into the budding viral particle (192). The F 1 subunit contains an N-terminus hydrophobic fusion peptide, two heptad repeats (HR1 and HR2), a transmembrane domain, and a cytoplasmic tail. The HR domains of the henipaviruses are the shortest among the paramyxoviruses (193). The G and F proteins are interacting together regardless of the receptor engagement by G protein, however, the nature and the domains involved in this interaction are not well defined (188, ). The G protein once bound to the receptor changes conformation which induces a change in the F protein conformation triggering fusion (197, 198). Recently, ephrin B2 (EB2) has been shown to serve as HeV and NiV receptor (199). EB2 is a transmembrane anchored ligand of the receptor tyrosine kinases EphB2, EphB3, and EphB4 which play a role in embryonic patterning, axon guidance and angiogenesis (200). EB2 is expressed mainly on endothelial cells and neurons which is consistent with the cellular tropism of HeV and NiV (201). c. Current Targeted Therapeutics for Henipaviruses Licensed and effective antiviral therapeutics for henipaviruses are currently not available. There has been a number of monoclonal antibodies (mabs) that were produced against NiV-G and HeV-G proteins with a range of in vitro neutralization activities (IC50 ~ ng/ml) ( ). Of the mabs developed is the human mab, m102.4, which engages the receptor binding site in NiV or HeV G proteins, and protects ferrets from a lethal virus dose when injected I.V. 10 hrs post-infection (202). Anti-F mabs have been developed to inhibit the fusion step of viral entry (205). Two anti-f

59 41 mabs have been reported to neutralize NiV and HeV in vitro (205). Soluble EB2 or EB3 as well as soluble G protein has also been shown to block virus entry and cell to cell fusion (199, 206, 207), although the possible interference of these inhibitors with the ephrinb function and the antigenicity of the G protein may limit the use of these antivirals. Blocking the viral entry by inhibiting one of the fusion intermediates of the F protein has been thought of as a therapeutic approach. Peptides that mimic the HR2 region of the class I fusion proteins in general have been used to inhibit HIV-1 entry ( ). The same approach has been tried for henipavirus entry inhibition (196). The henipaviruses HR2 peptide has been shown to inhibit the henipavirus pseudovirus entry (211). Moreover, small molecule inhibitors (Quinolone derivatives) have been tested for inhibition of henipaviruses with two compounds shown to be active (212). Calcium infux inhibitors has been shown to inhibit henipavirus replication at micromolar range (213), however, the mechanism is unknown. Specific inhibition of the cathl has been envisioned to inhibit the henipavirus F protein cleavage and fusion step in the virus entry. Recently a small oxocarbazate molecule has been shown to be a specific cathl inhibitor and an effective blocker of SARS-CoV and Ebola viruses that also use cathl for entry (214). This compound can also prove useful in treating henipaviruses infections by preventing generation of mature F protein.

60 42 B. Current Antiviral Strategies for SARS-CoV, Ebola, and Henipaviruses There are no approved effective therapies for SARS-CoV, Ebola, and Henipaviruses to date. These viruses remain a critical threat to human population because of the zoonotic nature of the diseases and the high morbidity and mortality associated with the previous outbreaks. There is a necessity to develop an effective and safe therapeutics against the above mentioned viruses to control any outbreaks that may occur in the furure. The best therapies for those viruses would be those providing instantaneous protection, because of the acute nature of the diseases. Those therapies will include passive immunotherapy with monoclonal antibodies and antiviral drugs with small molecules. Studies performed up to date have used different strategies (listed below) to block those viral infections in vitro and in vivo using animal models. 1. Monoclonal Antibodies (mabs) Antibodies specific for the attachment surface glycoproteins of SARS-CoV, Ebola, and henipaviruses have been tested in neutralizing the virus in vitro and some of them have been tested to block infection in vivo using animal models (91, 102, 108, 196, , 215, 216). The use of mabs to neutralize RNA viral infections will be challenged by their high capability of acquiring mutations which leads to the emergence of escape mutants. Therefore, the mab developed to prevent those viral infections should be of broad spectrum capable of neutralizing a wide range of clinical isolates which would be a part of this study. However, the use of these antibodies in humans still needs more extensive studies to prove efficacy and safety.

61 43 2. HR2 Specific Peptides The use of peptides that mimic the HR2 region of the fusion domain of the glycoproteins in SARS-CoV, and henipaviruses have been taken as an approach to block these viral infection in vitro (56, 211). However, the stability and efficacy of the synthetic HR2 peptides in vivo as well as the bioavailability may limit the use of this approach to protect from these lethal infections. Additionally, the time frame of the effective use of these inhibitors may be narrow because these inhibitors are only effective if present prior to the six helix bundle formation of the fusion core. Another problem that may arise is the mutations that the viruses may acquire in the HR sequences of the fusion domain to prevent those inhibitors from binding. 3. Antiviral Drugs Most current antiviral drugs target differences between the host and the viruses, such as specific viral protein moieties important for viral entry, replication, assembly, budding, etc. However, targeting a viral protein is not the best solution since the virus may escape by mutagenesis. Accordingly, the drugs that target conserved regions of the viral protein or non protein determinants of the viral life cycle for a broad range of viruses, like targeting membrane fluidity necessary for viral entry or exit, would be ideal.

62 44 C. Current Study Our study focuses on the development of broad spectrum therapies, including passive immunotherapy with human monoclonal antibodies (HmAbs) and antiviral small molecule based drugs, to provide instantaneous protection against SARS-CoV, Ebola, and henipaviruses. The study is divided into two parts: 1- The characterization of broad spectrum neutralizing HmAbs against a wide range of SARS-CoV clinical isolates. To identify broadly neutralizing HmAbs, the previously characterized SARS-CoV HmAbs, produced by Amgen incorporation, using Xenomouse (mouse immunoglobulin genes were replaced by human immunoglobulin genes) immunized with SARS-CoV S protein ectodomain, were tested against a relatively large panel of SARS-CoV variants. The RBD amino acid sequences available from 94 SARS-CoV late clinical isolates were first aligned and found mutations in the RBD region of only four clinical isolates relative to the SARS-CoV-Urbani RBD sequence. The HmAbs were tested against pseudoviruses that express S proteins harboring the mutant RBD sequences. The pseudoviruses served as RBD surrogate clinical isolates. The in vitro pseudovirus neutralization assay identified broad spectrum neutralizing HmAbs that can neutralize a wide range of variants by binding to conserved HR1 and HR2 regions of the SARS-CoV S protein. The detailed study will be discussed below. 2- Development of broad spectrum antiviral drugs against SARS-CoV, EBOV, and henipaviruses which are all known to be dependent on cathepsin L (catl) as a host protease for processing their glycoproteins. The assay used for development of those drugs makes use of the catl cleavage sites in the glycoproteins of SARS-CoV, EBOV, and henipaviruses. This study resulted in

63 45 identification of a broad spectrum drug that can block SARS-CoV and EBOV entry in an in vitro pseudovirus based assay and will be discussed in details below.

64 Chapter II Materials and Methods A. Characterization of broad spectrum neutralizing human monoclonal antibodies (HmAbs) against a wide range of SARS-CoV clinical isolates 1. Purpose and Rationale Previous studies have shown that the antibodies form an essential part of the immune response against SARS-CoV. The presence of animal reservoirs for the virus and the likelihood of the reemergence of the disease that will be sporadic in nature suggests that therapies that provide instantaneous protection either antibodies or antiviral drugs will be important in controlling the future outbreaks. Although vaccines represent an important strategy in protecting high risk populations such as laboratory workers and health care professionals, they are unlikely to be used as a protective measure for SARS-CoV because the virus did not cause a pandemic event or a seasonal infection and the disease represents an acute infection that can be best controlled by instantaneous therapies. Previous data from SARS-CoV and other RNA viruses demonstrated that the nature of these viruses makes them a formidable therapeutic target. Due to the high ability of the RNA viruses to exist as quasispecies of different viral variants and their ability to form escape mutants, it is important to understand the properties of the mab therapy and their ability to neutralize a wide range of viral variants when used as passive immunotherapy. Therefore, this part of the study focuses on testing the previously generated SARS-CoV S protein specific HmAbs, produced by 46

65 47 Amgen Incorporation, against a wide range of clinical isolates, that are sequenced to date. The sequences of these isolates demonstrated mutations that are localized to the RBD of the SARS-CoV S protein. The study sought to identify HmAbs that can neutralize a wide range of isolates, as discussed below. The antibodies were produced by immunization of Xenomouse, a transgenic mouse having the immunoglobulin genes knocked out and replaced by fully human immunoglobulin genes. The mouse allows for production of human antibodies and because of the strain chosen for immunization, the HmAbs are of IgG2 subtype. The advantage of using this IgG2 subtype in passive immunotherapy of SARS-CoV is that it is unlikely to cause ADE. Additionally, it has high availability in tissues like all IgGs, however, it is not a potent activator of the effector functions because the IgG2 Fc is not highly recognized by the Fc receptors so the chance of getting enhancement of infection, when using this antibody, is low. 2. Materials and Methods a. Cells 293FT cells were grown in DMEM (Dulbecco s Modified Eagles Medium) medium (Cell gro) supplemented with L-Glutamine (Invitrogen), Sodium Pyruvate (Invitrogen), Non essential amino acids (Invitrogen), and 10% Fetal Bovine Serum (FBS). The cells were used for the protein expression experiments and preparation of SARS-CoV Urbani and mutants pseudotyped viruses. The human embryonic kidney cells (293) were grown up in DMEM + 10%FBS with the same supplementation as the 293FT and used for the establishment of the 293/ACE2 stable cell line as described below.

66 48 b. Construction of expression plasmids for SARS-CoV S1-IgG and full length spike (S) protein mutants The expression plasmid encoding S1 fragment of SARS-CoV Urbani Spike (S) protein with an N terminal C5 signal sequence and a C-terminal human IgG Fc (48), was used as a template in site directed mutagenesis PCRs using QuikChange Lightning Site-Directed Mutagenesis Kit (Stratagene). To generate the Sin845 mutant (S353F), we used the forward primer 5 GCGTGGCCGACTACTTCGTGCTGTACAACTCCACC3 and reverse primer 5 GGTGGAGTTGTACAGCACGAAGTAGTCGGCCACGC3. For GZ-C mutant (Y436H), we used the forward primer 5 GCCACGAGCACCGGCAACCAT AACTACAAGTATCGC3 and reverse primer 5 GCGATACTTGTAGTTATGGTTGCCG GTGCTCGTGGC3, while for GD01 mutant (K344R, F501Y), we used the forward primers 5 GCCTGGGAGCGCAAGAGGATCAGCAACTGCG3 and 5 GCGTGGTCGTG CTGAGCTACGAGCTCCTGAACGCG3 and the reverse primers 5 CGCAGTTGCTGA TCCTCTTGCGCTCCCAGGC3, 5 CGCGTTCAGGAGCTCGTAGCTCAGCACGACCA CGC3 respectively in two independent PCR reactions. The GZ0402 mutant (K344R, F360S, T425I, L472P, D480G and T487S) was generated through five independent PCRs using the following primers: For K344R, the forward primer 5 GCCTGGGAGCGCAAGAGGATCAGCAACTGCG3 and reverse primer 5 CG CAGTTGCTGATCCTCTTGCGCTCCCAGGC3. For F360S, we used the forward primer 5 GCTGTACAACTCCACCTCGTTTAGCACCTTCAAGTGC3 and the reverse primer 5 GCACTTGAAGGTGCTAAACGAGGTGGAGTTGTACAGC3. For T425I, we used the forward primer 5 GCGTGCTGGCCTGGAACATCCGCAACATCGA CGCC3 and the reverse primer 5 GGCGTCGATGTTGCGGATGTTCCAGGCCAGCAC C3. For

67 49 the L472P mutation, we used the forward primer 5 CCCTGCACGCCGCCCGC CCCGAACTGCTACTGGCCG3 and the reverse primer 5 CGGCCAGTAGCAGTTCGG GGCGGGCGGCGTGCAGGG3. Finally for the D480G and T487S mutations, we used the forward primer 5 CCGCTGAACGGCTATGGCTTCTACACTACGTCCGGTATCGG CTACCAG3 and the reverse primer 5 CTGGTAGCCGATACCGGACGTAGTGTAGAA GCCATAGCCGTTCAGCGG3 in a multisite directed mutagenesis PCR using the QuikChange Multi-Site-Directed Mutagenesis Kit (Stratagene). The PCR products were transformed into the Ultracompetent E.coli MC1061/P3 (Invitrogen) and the mutations were confirmed by sequencing (DNA sequencing facility, UIC). The same procedure and primers were used for the generation of the full length S protein mutant constructs using the pcdna3.1- S, coding for the full length SARS-CoV S protein with a C-terminal (C9) tag derived from human rhodopsin protein, as a template. c. Construction of S-ectodomain, S2, HR1 and HR2 domains expression plasmids The pcdna3.1 S encoding the full length S protein of SARS-CoV was used as a template in a PCR reaction to amplify the S-ectodomain (residues ), S2 (residues ) and HR2 (residues ) domains using the forward primers: 5 TATGCTAGCCAGCGGCAGCGACCTGGACCGC3, 5 TATGCTAGCCTTCTCCATCA GCATCACGACCGAG3 and 5 ATCGCTAGCCCACACCAGCCCGGACGTGGACCTG 3 respectively and the reverse primer 5 TATGGATCCCTCTCCTGGAGGTCGATCAG GCTCTCGTTCAG3. The HR1 domain (residues ) was amplified using the forward primer 5 ATCGCTAGCCAACCAGAAGCAGATCGCCAACCAG3 and the reverse primer 5 ATAGGATCCTCGTGCGGCGCGGCCTGCGGGAAGCT3. All the

68 50 forward primers were designed with a 5 NheI site while the reverse primers were designed with a 5 BamHI site. The PCR products were then digested with NheI and BamHI restriction enzymes (New England Biolabs, Ipswich, MA) in a step wise manner and cloned into the C-terminus IgG tag mammalian expression vector (48), which was digested with the same enzymes as well. The ligation was performed using the Quik Ligation Kit (New England Biolabs, Ipswich, MA). The ligation mixtures were transformed into the Ultracompetent E.coli MC1061/P3 (Invitrogen). The inserts were verified by sequencing (DNA Sequencing Facility, UIC). d. Expression and purification of SARS-CoV S1-IgG Urbani and mutant proteins as well as S-ectodomain, S1, S2, HR1 and HR2 domain proteins The plasmids coding for S1-IgG proteins as well as the S protein truncations (Sectodomain, S2, HR1 and HR2 domains) were used to transfect 293FT cells by calcium phosphate transfection method and the proteins were purified using protein A agarose beads as described previously (102). The purified proteins were concentrated through Centricon filters (Millipore, Bedford, MA) then detected by Coomassie blue staining (BioRad, Hercules, CA) following separation on a 4-15% SDS/PAGE gel. The expression was further confirmed by western blot using polyclonal goat anti-human IgG Fc HRP antibody (Promega). e. Purification of the non S1 binding neutralizing human mabs Hybridomas of 56 neutralizing non S1 binding HmAbs were cloned by limiting dilution and the clones were cultured in DMEM medium supplemented with 10% Fetal clone

69 51 (Hyclone laboratories, Logan, Utah) to produce large quantities of HmAbs. Supernatants from these hybridomas were harvested and HmAbs were purified using protein-a agarose beads. Thirty nine HmAbs were successfully purified and the Ab production was confirmed by 4-15% SDS/PAGE followed by Coomassie blue staining. The antibody concentration was measured at 280nm using the Biomate3S UV-Visible Spectrophotometer (ThermoScientific). All HmAbs were diluted to a final concentration of 50µg/ml. f. Enzyme Linked Immunosorbent Assay (ELISA) Medisorp ELISA plates (Nunc, Roskilde, Denmark) were coated with 100ng/well of S1-IgG Urbani protein as well as mutant proteins (Sin845, GZ-C, GD01 and GZ0402) overnight at 4 C. The binding of the 18 HmAbs were tested by ELISA as described previously (102). The anti-human IgG2 HRP mouse monoclonal antibody was used as the secondary antibody (SouthernBiotech, Birmingham, AL). The same procedure was followed for testing the binding of the 39 non S1 neutralizing HmAbs against S protein ectodomain, S2, HR1, HR2 and S1 domain proteins. g. Competitive ELISA inhibition assay Different protein concentrations (0, 25, 50, 100 and 200 nanograms) of Urbani or GZ-C protein were pre-incubated with 0.5µg/ml of 5A7 antibody, O/N at 4 C. The protein/ab mixtures were then added to the wells of the ELISA plate, previously coated with 100ng of the other protein and blocked with 3% Bovine Serum Albumin (BSA) in Phosphate Buffered Saline with 0.05% Tween-20 (PBST), and incubated for 1hr at RT. The plates

70 52 were washed and 50µl/well of HRP conjugated mouse anti-human IgG2 monoclonal antibody (Southern Biotechnology, Birmingham, AL) was added and incubated for 1hr at room temperature. Following washing, the antibody binding was detected using 50µl/well of TMB substrate (BD OptEIA, BD Biosciences Pharmigen, San Diego, CA) and the reaction stopped using 25µl of 10% HCl. The absorbance was then read using a microplate reader (BioRad, Hercules, CA) at 450nm. h. ELISA using Rabbit anti-sars CoV S protein immune serum ELISA plates were coated with 100ngs of the Urbani or mutant SARS-CoV IgG recombinant proteins. The rabbit immune serum was used in serial dilutions as the primary antibody and incubated for 1hr at RT with the Urbani and mutant proteins. After washing, 50µl/well of HRP conjugated donkey anti-rabbit polyclonal antibody (GE health care Biosciences, Pittsburgh, PA) was added and incubated for 1hr at room temperature. The rest of the ELISA procedure was as described before. i. Production of SARS S- Urbani and mutant pseudotyped viruses Pseudotyped viruses were generated by co-transfection of 293FT producer cells (grown in DMEM with 10% FBS) with phiv-gfp-luc expression vector, pgagpol HIV vector, phiv-rev and phiv-tat (217), along with the pcdna3.1-s coding for the SARS-CoV S protein using calcium phosphate transfection according to the previously described protocol (102). For the production of HIV/ E, only HIV vectors were tranfected into the cells. The media were changed the following morning and the supernatants were collected 24 and 48hrs later and pooled. The pseudotyped viruses were concentrated

71 53 through a 20% sucrose cushion at 41,000 rpm using Beckman Ultracentrifuge. The virus pellet was re-suspended in PBS and the incorporation of the S proteins in the virus particles was confirmed by western blot using 1D4 anti-human rhodopsin mouse monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) which recognizes the C9 tag at the C-terminus of the S proteins, while the virus p24ag content was confirmed by mouse anti-hiv1 p24 mouse monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA). j. Establishment of 293/ACE2 stable cell line Human embryonic kidney cells (293) that have been grown in DMEM/10%FBS were seeded at density of 200,000 cells/well in two 6 well plates. The following day, the medium was changed and one of the two plates was transfected with 0.4µg of human ACE2 expression plasmid using Effectene transfection reagent (Qiagen) according to the manufacturer instructions. The other plate was left as a control. The cells in the transfected and untransfected plate were selected with G418 (Cell gro) 24 hrs posttransfection for 7 days at 2.5mg/ml, 1mg/ml, and 0.8mg/ml (2 wells/ concentration). The stable colonies appearing after 7 days in the transfected plate at 1mg/ml G418 concentration, with complete cell death in the untransfected one, were selected and grown up further more under G418 pressure (0.5mg/ml) and used for the pseudovirus antibody entry inhibition experiments.

72 54 k. In vitro pseudotyped virus neutralization assay Entry inhibition of Urbani and mutant pseudoviruses was performed by pre-incubating pseudoviruses, equivalent to 10 nanograms of p24 Ag, quantified by HIV-1 p24 ELISA kit (Express Biotech International, MD), with purified mabs individually or in combinations at 37 C for 1hr. The pseudovirus/mab mixture or pseudovirus alone was added to the target 293/ACE2 stable cell line plated at a density of 60,000 cells/well in 12 well plate, and incubated overnight at 37 C and the medium was replaced the following morning. Forty eight hours later, the cells were lysed and luciferase expression was determined using Luciferase assay kit (Promega, WI)) according to the manufacturer s instructions. The rabbit immune serum was used as a positive control for entry inhibition. The percentage entry inhibition was calculated using the following equation: Luciferase reading of mock treated virus Luciferase reading of Ab treated virus 100 / Luciferase reading of mock treated virus Luciferase reading of HIV E virus The antibody mediated inhibitions of different mutant pseudoviruses were then normalized to HIV/Urbani-S inhibitions.

73 55 B. Development of broad spectrum antiviral drugs against SARS-CoV, Ebola, and Henipaviruses 1. Purpose and Rationale Enveloped viruses enter the target cells by fusion of the viral membrane with the host cell membrane and deliver the viral genome to the cytoplasm. SARS-CoV, Ebola (EBOV), and Henipairuses (Hendra (HeV) and Nipah (NiV) viruses) critically require cathepsin L (catl), a host intracellular lysosomal protease, for their glycoprotein processing and cleavage allowing for virus fusion and entry into the host cells (34, 189, 218, 219). SARS-CoV and EBOV infect the target cells after cleavage of their fusion translocated to the membrane and then internalized followed by catl mediated cleavage required for fusion. The F protein is then incorporated into the viral particle (34, 189, 218, 219). Thus, the conserved catl cleavage sites can be potential targets for developing broad spectrum anti-viral drugs. Antiviral drugs can be broadly divided into 4 major classes. The nucleoside analogs, acyclovir and zidovudine (AZT), are effective against herpes virus and HIV (220, 221). The second class includes viral protease inhibitors such as HIV protease inhibitors, which are effective but exhibit undesirable side effects (222). A third class of antivirals act on virus penetration/un-coating (223, 224), and include Amantadine and Rimantadine used to treat influenza. A fourth class of antivirals consisting of zanamivir (Relenza) and oseltamivir (Tamiflu) block virus release by blocking neuraminidase found on the viral surface (225, 226). Because of development of drug resistance (223, 224), Amantadine (227) and Tamiflu (228) are either not recommended or considered

74 56 ineffective. These drugs act on particular viruses and are not broadly protective. There have been several attempts to discover potent inhibitors that can act broadly against several different viruses. For example, novel inhibitors of papain-like protease can inhibit live SARS-CoV infection (229). Oxocarbazate and pentapeptide amide can block human catl and inhibit SARS and Ebola pseudotyped virus infection (214, 230). However, blocking host enzymatic activity is likely to have undesirable side effects on human health (231). Based on the common dependence of SARS-CoV, EBOV, and Henipaviruses on catl for their glycoprotein processing which is necessary for their entry into the target cells, we have carried out a High Throughput Screening Assay (HTSA) for chemical compounds that can block the cleavage of the envelope glycoproteins required for their entry into host cells without inhibiting the critical host protease function of catl. The HTSA used peptides derived from the glycoproteins of the aforementioned viruses, which contain the catl cleavage sites. We have confirmed that catl can cleave these peptides at the expected sites by mass spectroscopy. We have made use of the HTSA, optimized in our lab, to identify and characterize candidate compounds from libraries of small molecules. The candidate compounds were tested for their inhibitory effect on SARS-CoV and Ebola pseudotyped viruses. The detailed study is presented below.

75 57 2. Materials and Methods a. Cells 293FT cells were grown in DMEM (Dulbecco s Modified Eagles Medium) medium (Cell gro) supplemented with L-Glutamine (Invitrogen), Sodium Pyruvate (Invitrogen), Nonessential amino acids (Invitrogen), and 10% Fetal Bovine Serum (FBS). The cells were used for preparation of SARS-CoV, Ebola, and VSVG pseudotyped viruses and for the Ebola, VSVG pseudotyped viruses entry inhibition experiments. The 293FT transiently transfected with the human ACE2 expression plasmid were used for the SARS-CoV pseudotyped viruses entry inhibition experiments. b. Viral and host derived peptides synthesis Peptides composed of 10 amino acids long derived from glycoproteins of SARS-CoV (S), Ebola (GP), Hendra (F 0 ), Nipah (F 0 ), and host pro-neuropeptide Y (Pro-NPY) were synthesized in the University of Illinois at Chicago (UIC) protein research laboratory. The peptides contain the natural catl cleavage sites in the viral and host pro-npy. The viral and host pro-npy derived peptides were labeled on the N-terminus with 5 Carboxytetramethylrhodamine (Tamra) as a quencher and on the C terminus by 5- Carboxyfluorescein (5-FAM) as an emitter in the protein research laboratory at UIC. The labeled peptides were purified using reversed phase High performance Liquid Chromatography (HPLC) in the UIC protein research laboratory.

76 58 c. Mass spectrometry The SARS-CoV-S, EBOV-GP, HeV-F 0, NiV-F 0, and pro-npy derived labelled and unlabelled peptides (1µM) were incubated with 1µg/ml of human catl purified from human liver (Sigma Aldrich) for 1hr at room temperature in ammonium acetate (NH 4 Ac) buffer ph5.5 containing 4mM EDTA and 8mM dithiothreitol (DTT). The cleavage products were analyzed by MALDI-TOF Mass Spectrometry in the UIC Protein Research Laboratory (PRL). d. Optimization of the High Throughput Screening Assay (HTSA) The HTSA is a Fluorescence Resonance Energy Transfer (FRET) based assay. The labeled SARS-CoV S protein derived peptide was used as a substrate in the primary screen. If the peptide is not cleaved by catl, no fluorescence emission will be detected at 535nm when 5-FAM is excited at 485nm due to the quenching effect of Tamra. In contrast, if the peptide is cleaved, an emission of light at 535nm is detected. The assay was optimized in black 384 well plates (Thermoscientific) using 3µM SARS-CoV-S derived labeled peptide incubated with 1µg/ml human catl (Sigma Aldrich) and further optimized with 1µM SARS-CoV-S derived labeled peptide incubated at room temperature with 0.25, 0.5, and 2µg/ml catl in 50µl total volume of NH 4 Ac buffer ph 5.5 supplemented with 4mM EDTA and 8mM DTT. The fluorescence was measured over time, at 535nm after excitation at 485nm, using fluorescence reader at the UIC HTS facility. The Ebola GP, Hendra and Nipah F 0 derived labeled peptides as well as the host pro-npy derived peptide were tested for cleavage at 1µM concentration by incubation with different concentrations of catl (0.25, 0.5, and 2µg/ml). The quality of

77 59 the screening assay (Z factor) was determined using the following formula: Z-factor = 1 (3(σ p + σ n ) / µ p - µ n ), where σ p = standard deviation of the positive signal, σ n = standard deviation of the negative signal, µ p = mean of the positive signal, and µ n = mean of the negative signal. e. Screening for inhibitors of cathepsin L mediated cleavage of viral glycoprotein derived labeled peptides Primarily, a library of 5000 compounds, at 40 µm concentration, from ChemBridge Corporation was tested in duplicates (in black 384 well plates) for inhibitory effect on catl mediated cleavage of labeled SARS-CoV-S derived peptide in 50µl total reaction volume using the same buffer as mentioned above. The assay was performed with 1µM of the peptide incubated with 0.25µg/ml catl at ph 5.5, simulating the endosomal ph, for 45 minutes at room temperature after which the reaction was stopped by 10µl of 0.5M acetic acid. The fluorescence was read using fluoroscence reader at the UIC HTS facility. The percentage inhibition of the catl mediated cleavage by the screened compounds was calculated using the following formula: positive fluorescence signal in absence of compounds fluorescence signal in presence of compounds 100 / positive fluorescence signal in absence of compounds negative fluorescence signal in absence of the enzyme. The top 50 hits that inhibited the catl cleavage of SARS-CoV peptide at a cutoff of 60% inhibition were screened in duplicates for the inhibition of cleavage of EBOV-GP, Hendra, and Nipah-F 0 as well as cleavage of pro-npy derived labeled peptides as mentioned before.

78 60 f. Preparation of pseudotyped viruses Pseudotyped viruses (EBOV-GP, SARS-CoV-S, VSV-G) were generated by cotransfection of cells of 293FT (grown in DMEM with 10% FBS) with phiv-gfpluc expression vector (18µg), pgagpol HIV vector (1.8µg), phiv-rev (360ng) and phiv- TAT (360ng) (217), along with the pcdna3.1-s (10µg) coding for the SARS-CoV-S glycoprotein or pcdna3.1-gp (10µg) coding for the Ebola virus GP glycoprotein, or pcdna3.1-vsvg (1µg) coding for the VSV-G glycoprotein using calcium phosphate transfection according to the previously described protocol (102). For the production of HIV/ E, only HIV vectors were tranfected into the cells. The media were changed the following morning and the supernatants were collected 24 and 48hrs later and pooled. The virus stocks were frozen at -80 C till used. g. In vitro pseudovirus inhibition assay Different pseudotyped viruses (Ebola-GP, SARS-CoV-S, and VSVG as a negative control), normalized for equal infectivity using HIV-1 p24 Elisa kit (Express Biotech International, MD), were mixed with 10µM of the candidate inhibitory compounds, identified in the screening assay. The virus or virus/compounds mixtures were added to FT/well seeded in 6 well plates. For the SARS-CoV-S pseudotyped viruses inhibition assays, the 293FT were transiently transfected with human ACE2 expression plasmid (0.4µg/well), using Effectene transfection reagent (Qiagen) according to the manufacturer s instructions. Forty eight hours later, the cells were transduced with the SARS-CoV-S pseudotyped viruses treated or untreated with the candidate compounds. Seventy two hours later, the cells were lysed and the luciferase expression was

79 61 determined using luciferase assay kit (Promega) according to the manufacturer s instructions. The cathepsin L inhibitor at 10µM concentration (Calbiochem) was used as a positive control and DMSO treated viruses as a negative control. The percentage entry inhibition of the candidate compounds on different pseudoviruses was calculated using the following formula: Luciferase reading of mock treated virus Luciferase reading of compound treated virus 100 / Luciferase reading of mock treated virus Luciferase reading of HIV E virus. h. IC50 determination Ebola-GP pseudotyped virus was mixed with different concentrations of the compounds and (1-160µM). The mixture was added to 293FT cells seeded at density of cells/well in 6 well plates. Seventy two hours later, the cells were lysed and the luciferase expression was determined using luciferase assay kit (Promega) according to the manufacturer s instructions. The concentration of each compound was plotted against the percentage inhibition of viral entry and the concentration of the compound that inhibits the viral entry by 50% (IC50) was determined from the curve. i. Cytotoxicity assay (MTT assay) 293FT cells were seeded at density of 10 4 cells/well in 96 well plates. The following day, the cells were treated with different concentrations (10, 30, 50, and 100µM) of the 2 compounds ( , and ) and the cytotoxicity effect of the compounds was assessed using MTT reagent (3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium

80 62 bromide) (Roche) for 3 days according to the manufacturer s instructions. Briefly at 24, 48, and 72 hrs following addition of each compound, the cells were washed once with Phosphate buffered Saline (PBS) and fresh DMEM medium without phenol red was added (100µl/well) after which 10µl of MTT reagent was added to each well and incubated for 4hrs at 37 C. 100µl of 10% SDS solubilized in 0.01M HCl was added to each well with vigorous mixing and further incubated for 4hrs at 37 C. The OD was measured at 595nm and percentage viability was calculated relative to the control. j. Enzyme kinetics Different concentrations of the SARS-CoV-S derived labeled peptide (2-64µM) were incubated with fixed concentration of catl (0.5µg/ml) in absence or presence of 50µM of compound for 25 minutes at room temperature. The reaction was stopped with 10µl of 0.5M acetic acid after which the fluorescence was read at 535nm after excitation at 485nm with fluorescence reader at UIC HTS facility. The reaction was performed in 50µl total volume in 384 well plates using NH 4 acetate buffer ph5.5 supplemented with 4mM EDTA and 8mM DTT. The Velocity of the reaction at different substrate concentrations was calculated (Fluorescence units/minute) and plotted versus substrate concentration. The inverse velocity was further plotted versus inverse substrate concentration (Lineweaver-Burk plot) from which the K m and V max were calculated.

81 Chapter III Results A. Characterization of broad spectrum neutralizing human monoclonal antibodies (HmAbs) against a wide range of SARS-CoV clinical isolates 1. Expression of SARS-CoV S1 IgG1 Urbani and mutant proteins The SARS-CoV S protein consists of S1 domain in which RBD contains the major neutralizing epitopes, and S2 domain which consists mainly of HR1 and HR2 domains (35, 43, 48, 92, 102) (Figure 4A). To identify broadly neutralizing HmAbs, the previously characterized HmAbs were tested against a relatively large panel of variants. We aligned the RBD amino acid sequences available from 94 SARS-CoV late clinical isolates and found mutations in the RBD region of only four clinical isolates relative to the Urbani RBD sequence (Figure 4B). The clinical isolates with the identified RBD mutations are named Sin845, GZ-C, GD01 and GZ0402 (GenBank accession number: AY , AY , AY and AY respectively). The identified mutations were inserted within the RBD by site directed mutagenesis into the Urbani S1 sequence that is fused to human IgG1 Fc tag at the C-terminus (48). The Urbani and the mutated S1-IgG proteins were expressed in 293FT cells, purified and analyzed by SDS/PAGE (Figure 5A) followed by western blot (Figure 5B). 63

82 64 A S1 domain S2 domain SP RBM FP HR1 HR2 TM CP ,013 1,145 1,184 1,215 1,255 RBD B Sin845 GZ-C Urbani GD01 GZ0402 Sin845 GZ-C Urbani GD01 GZ0402 Sin845 GZ-C Urbani GD01 GZ WERKKISNCVADYFVLYNSTFFSTFKCYGVSATKLNDLCFSNVYADSFVVKGDDVRQIAP WERKKISNCVADYSVLYNSTFFSTFKCYGVSATKLNDLCFSNVYADSFVVKGDDVRQIAP WERKKISNCVADYSVLYNSTFFSTFKCYGVSATKLNDLCFSNVYADSFVVKGDDVRQIAP WERKRISNCVADYSVLYNSTFFSTFKCYGVSATKLNDLCFSNVYADSFVVKGDDVRQIAP WERKRISNCVADYSVLYNSTSFSTFKCYGVSATKLNDLCFSNVYADSFVVKGDDVRQIAP ****:******** ****** *************************************** GQTGVIADYNYKLPDDFMGCVLAWNTRNIDATSTGNYNYKYRYLRHGKLRPFERDISNVP GQTGVIADYNYKLPDDFMGCVLAWNTRNIDATSTGNHNYKYRYLRHGKLRPFERDISNVP GQTGVIADYNYKLPDDFMGCVLAWNTRNIDATSTGNYNYKYRYLRHGKLRPFERDISNVP GQTGVIADYNYKLPDDFMGCVLAWNTRNIDATSTGNYNYKYRYLRHGKLRPFERDISNVP GQTGVIADYNYKLPDDFMGCVLAWNIRNIDATSTGNYNYKYRYLRHGKLRPFERDISNVP ************************* **********:*********************** FSPDGKPCTPPALNCYWPLNDYGFYTTTGIGYQPYRVVVLSF FSPDGKPCTPPALNCYWPLNDYGFYTTTGIGYQPYRVVVLSF FSPDGKPCTPPALNCYWPLNDYGFYTTTGIGYQPYRVVVLSF FSPDGKPCTPPALNCYWPLNDYGFYTTTGIGYQPYRVVVLSY FSPDGKPCTPPAPNCYWPLNGYGFYTTSGIGYQPYRVVVLSF ************ *******.******:*************: Figure 4 Comparative sequence analysis of the receptor binding domain of spike proteins in SARS-CoV clinical isolates. (A) Domain structure of the SARS-CoV spike protein (SP; signal peptide, RBM; receptor binding motif, RBD; receptor binding domain, FP; putative fusion peptide, HR1; heptad repeat 1, HR2; heptad repeat 2, TM; transmembrane domain, CP; cytoplasmic domain). (B) Amino acid sequence alignment of aa within the receptor binding domain (aa ) of Urbani SARS-CoV-S protein, Sin845, GZ-C, GD01, and GZ0402 mutant S proteins. Amino acid differences are shown in bold.

83 65 A kda Urbani Sin845 GZ-C GD01 GZ B Urbani Sin845 GZ-C GD01 GZ0402 Figure 5 Expression and purification of SARS-CoV S1 proteins (aa ). 293FT cells were transiently transfected with either Urbani S1-IgG expression plasmid or each of the mutant S1-IgG plasmids. Recombinant proteins were purified from the supernatants 72 hrs post-transfection using protein-a agarose beads, concentrated and detected by (A) Coomassie blue staining and (B) Western blot using goat polyclonal anti-human IgG antibody.

84 66 2. The S1 proteins containing RBD sequences of Sin845, GD01, and GZ0402 isolates show low binding to S1 specific neutralizing HmAbs, while that of GZ-C isolate shows higher binding Relative binding of HmAbs to different S1 proteins at different concentrations of antibodies was determined. The binding at the highest concentration used (2.5µg/ml) is shown. Interestingly, the Sin845-S1 protein failed to react with 16/18 HmAbs (OD ~ 0.2) when compared to the control OD of ~ However, HmAbs 4D4 and 6B1 showed about 50% binding to Sin845 S1 protein relative to their binding to Urbani S1 protein (Figure 6A). The GD01-S1 protein showed a diminished binding to 16/18 HmAbs and binding of about 40% and 60% to 4D4 and 3C7 HmAbs respectively (Figure 6B). The GZ0402-S1 protein showed minimal binding to 15/18 HmAbs, and 57%, 52% and 69% binding to HmAbs 4D4, 6B1 and 3C7 respectively (Figure 6C). Surprisingly, the GZ-C S1 protein showed an increased binding to all 18 HmAbs (Figure 6D). The diminished binding to the Sin845, GD01 and GZ0402 mutants was further confirmed by the minimal to no binding of the HmAbs 5A5, 5D6 and 4G2, even when the wells were coated with an excessive amount of mutant S1 proteins (i.e. 600ng) relative to their significant binding to only 100ngs of the Urbani-S1 protein (data not shown). The validity of these findings was confirmed when we found that an anti-sars- CoV-S Urbani polyclonal serum showed strong reactivity (OD ~ 0.4) against the GZ-C- S1 mutant even at a high dilution (1/1280) while it showed much lower binding to Sin845-S1, GD01-S1 and GZ0402-S1 proteins relative to its binding to the Urbani-S1

85 67 A B C D Figure 6 Reactivity of the 18 neutralizing HmAbs with SARS-CoV S1 proteins. Medisorp ELISA plates were coated with 100ng/well of Urbani and RBD mutant S1-IgG proteins and 2.5 µg/ml of each HmAb was used as the primary antibody. Anti-human IgG2 HRP mouse monoclonal antibody was used as secondary antibody. OD was measured at 450nm. Error bars represent SD of a representative experiment performed in triplicates. (A) Urbani versus Sin845 mutant. (B) Urbani versus GD01 mutant. (C) Urbani versus GZ0402 mutant. (D) Urbani versus GZ-C mutant.

86 68 protein (Figure 7A). Enhanced binding to GZ-C-S1 protein was further validated when we found that as little as 25ng of the GZ-C protein could block HmAb 5A7 binding to the Urbani-S1 protein while as much as 200ng of Urbani protein was significantly less efficient in blocking the HmAb 5A7 binding to GZ-C-S1 protein (Figure 7B). 3. The S proteins containing RBD sequences of Sin845, GD01, GZ0402 and GZ-C isolates do not affect pseudovirus entry Pseudoviruses expressing S proteins containing RBD sequences of Sin845, GD01, GZ0402 and GZ-C isolates were prepared to serve as RBD surrogates for those clinical isolates. 293FT cells were co-tranfected with the RBD containing S protein expression vector and the HIV coding vectors. The pseudoviruses were concentrated and the S protein and the HIV p24 Ag incorporation into the viral particles were confirmed by western blot (Figure 8A). In HIV/ E, as expected, no surface glycoprotein was detected. Pseudoviruses expressing the mutant S proteins entered 293 cells, stably expressing ACE2, with equal efficiency when compared to the HIV/S positive control (Figure 8B). 4. Pseudoviruses containing S proteins with RBD sequences of Sin845, GD01 and GZ0402 isolates escape neutralization while GZ-C shows enhanced neutralization by S1 specific HmAbs Consistent with the binding data shown above, entry inhibition of Sin845-S, GD01-S and GZ0402-S pseudoviruses ranged from 10-45%, except for the HmAb, 4D4, which showed % inhibition, relative to that seen with Urbani-S pseudovirus by the

87 69 A B Figure 7 Reactivity of Urbani SARS-CoV-S protein antibodies with Urbani and mutant S1 proteins. (A) Different dilutions of a rabbit anti-urbani SARS-CoV-S protein immune serum were tested in an ELISA against Urbani as well as mutant S1-IgG proteins. Anti-rabbit donkey polyclonal HRP antibody was used as the secondary antibody. (B) Competitive ELISA assay: Different protein concentrations of Urbani or GZ-C proteins were pre-incubated with 5A7 antibody then the protein/ab mixtures were tested for binding to the other protein by ELISA. OD was measured at 450nm.

88 70 A Urbani Sin845 GZ-C GD01 GZ0402 HIV/ E S protein P24 B Figure 8 Pseudoviruses expressing the spike glycoprotein of clinical isolates can enter cells with equal efficiency as HIV/S. (A) Pseudoviruses, produced by co-transfecting 293FT cells with HIV viral vectors and pcdna3.1-s encoding the SARS Urbani-S protein or its mutants (i.e. Sin845, GZ-C, GD01 and GZ0402), were concentrated and confirmed for S protein and HIVp24 protein content by western blot. (B) Different pseudoviruses were tested for entry into stable 293/ACE2 cells by measuring the relative luciferase expression (RLU) 72hrs posttransduction. The HIV/VSVG pseudovirus was used as a positive control and HIV/ E as a negative control. Error bars represent SD of representative experiment performed in triplicates.

89 71 corresponding antibodies (Figure 9A, 9B). In contrast, these antibodies showed more efficient inhibition of GZ-C mutant (Figure 9A). The HmAbs did not show significant inhibition of VSV-G pseudotyped virus which ensures the specificity of the HmAbs (data not shown). 5. Differential reactivity of non-s1 binding HmAbs with S ectodomain, S2 domain, HR1 and HR2 regions suggest multiple mechanisms of virus neutralization The recombinant S protein ectodomain, S2 domain, HR1 and HR2 proteins were expressed in 293FT cells and purified using protein-a agarose beads (Figure 10). Thirty nine non-s1 binding but Urbani strain S-ectodomain binding and neutralizing HmAbs (102), were successfully purified and tested for binding to different regions of the S protein, including S1 domain as a negative control and full-length S-ectodomain as a positive control. OD which is 3x negative control (control OD ~ 0.13) was considered positive. Twenty two HmAbs bound to S2 domain out of which nine and thirteen bound specifically to the HR1 and the HR2 regions respectively (Table I). Interestingly, seventeen HmAbs bound to S-ectodomain but failed to bind to HR1 and HR2 regions of the S2 domain. This suggested that these HmAbs are binding to conformational epitopes that are only available in the S protein ectodomain and not in its subdomains. Inhibition of different pseudoviruses entry by HR1 and HR2 binding HmAbs ranged from 60 to 110% of the Urbani-S pseudovirus inhibition at an antibody concentration of 25µg/ml (Table II). In contrast, the S-ectodomain binding HmAbs were less effective and showed entry inhibition ranging from 10-45% of Urbani-S inhibition, except for the HmAb 4G10 which showed ~76% neutralization of Sin845-S virus, and

90 72 A B Figure 9 In vitro pseudovirus neutralization assay. Eighteen neutralizing HmAbs were tested against different mutant as well as Urbani pseudoviruses. Pseudoviruses equivalent to 10ng of HIVp24 were incubated for 1hr with 25µg/ml of each of the HmAbs at 37 C. The virus/ab mixtures were then added to 293/ACE2 stable cell line. Seventy two hours later, the virus entry was determined by luciferase expression. The percentage entry inhibitions obtained with Abs were calculated and normalized to HIV/Urbani-S inhibitions (A) HIV/GZ-C and HIV/Sin845 inhibitions (B) HIV/GZ0402 and HIV/GD01 inhibitions. Polyclonal rabbit immune serum (PolyAb) was used as a positive control. Error bars represent SD of representative experiment performed in triplicates.

91 73 A B kda kda S1 domain S2 domain S-ectodomain C kda HR1 HR Figure 10 Expression and purification of SARS-CoV-S protein domains. 293FT cells were transfected with the plasmids coding for each of the S protein domains and the proteins were purified from the supernatants 72hrs post-transfection using protein-a agarose beads, concentrated and detected by Coomassie blue staining of 4-15% SDS/PAGE. (A) S glycoprotein ectodomain, (B) S1 and S2 domain of the S protein, and (C) HR1 and HR2 domains.

92 74 Ab S- ect. a,b S2b HR1 b HR2 b S1 b BR c 1F S-ect a 3F S-ect a 4E S-ect a 6C S-ect a 4G S-ect a 3F S-ect a 6D S-ect a 2C S-ect a 2G S-ect a 1D S-ect a 4 E S-ect a 1C S-ect a 2B S-ect a 2E S-ect a 1G S-ect a 6H S-ect a 1D S-ect a 1F HR1 4A HR1 1D HR1 2A HR1 5C HR1 2B HR1 6H HR1 6C HR1 4F HR1 5G HR2 5B HR2 3A HR2 5 E HR2 6H HR2 1E HR2 3H HR2 5B HR2 5D HR2 2D HR2 3E HR2 5G HR2 2D HR2 a S glycoprotein ectodomain b Values in the table are average OD of three experiments c Likely binding region TABLE I DIFFERENTIAL REACTIVITY OF 39 SARS-CoV NON-S1 NEUTRALIZING HUMAN MONOCLONAL ANTIBODIES TO SPIKE PROTEIN FRAGMENTS. The 39 non S1 binding neutralizing HmAbs were tested for binding to different SARS- CoV-S protein fragments by ELISA, the values in the table are average ODs of representative experiment in triplicates measured at 450 nm. The likely binding region is shown in the table.

93 75 Virus Sin845 GZ-C GD01 GZ0402 Ab Percentage entry inhibition (normalized to HIV/Urbani-S inhibition) BR 1F S-ect 3F S-ect 4 E S-ect 6C S-ect 4G S-ect 3F S-ect 6D S-ect 2C S-ect 2G S-ect 1D S-ect 4 E S-ect 1C S-ect 2B S-ect 2 E S-ect 1G S-ect 6H S-ect 1D S-ect 1F HR1 4A HR1 1D HR1 2A HR1 5C HR1 2B HR1 6H HR1 6C HR1 4F HR1 5G HR2 5B HR2 3A HR2 5E HR2 6H , HR2 1 E HR2 3H HR2 5B HR2 5D HR2 2D HR2 3 E HR2 5G HR2 2D HR2 PolyAb TABLE II HUMAN MONOCLONAL ANTIBODIES TO HR1 AND HR2 CAN EFFICIENTLY NEUTRALIZE SURROGATE CLINICAL ISOLATES. The 39 non S1 binding HmAbs were tested for neutralizing the different pseudoviruses. Pseudoviruses equivalent to 10 nanograms of HIVp24 were incubated for 1hr with 25µg/ml of each of the HmAbs at 37 C. The virus/ab mixture was then added to 293/ACE2 stable cell line for 72 hrs after which the virus entry was determined by luciferase expression. Polyclonal Rabbit immune serum (PolyAb) was used as positive control. BR; likely binding region, S-ect; S glycoprotein ectodomain

94 76 the HmAbs 3F1 and 2G11 which showed 92% and 98.4% neutralization of the GZ-C-S virus (Table II). Collectively, the above results showed that the HR1 and HR2 binding HmAbs are more effective in inhibiting the entry of the RBD surrogate clinical isolates. Those HmAbs did not inhibit the entry of VSV-G pseudotyped virus (data not shown). 6. Combinations of SARS-CoV HmAbs targeted to different regions of the S glycoprotein more efficiently inhibit the entry of RBD surrogate clinical isolates Combinations of 4D4 (binds to S1, N-terminal of RBD), 1F8 (binds to HR1) and 5E9 (binds to HR2) HmAbs were tested to see if they can more effectively inhibit viral entry. The combinations of 4D4/1F8, 4D4/5E9 and 1F8/5E9 HmAbs were more effective in blocking Urbani pseudovirus entry compared to the individual antibodies (p value < 0.05). The same pattern of inhibition was seen with the Sin845-S, GZ-C-S and GZ0402- S pseudoviruses (p values= ). However, these HmAb combinations exhibited similar levels of GD01 pseudovirus blocking as seen with the 1F8 or 5E9 HmAbs when used individually. Maximum inhibition of 90-95% (p values = ) was noted when a combination of 4D4/1F8/5E9 HmAbs was used (Figure 11). These results indicated that a cocktail of HmAbs targeting different conserved regions of the S protein is likely to be more effective in neutralizing different SARS-CoV clinical isolates than individual antibodies with specificity to those regions.

95 Figure 11 Combinations of HmAbs more efficiently inhibit the entry of SARS-CoV RBD surrogate clinical isolates. Neutralizing HmAbs binding to different regions of S protein 4D4 (S1), 1F8 (HR1), 5E9 (HR2) were tested for their ability to neutralize pseudoviruses in different combinations as well as individually at a concentration of 6.25µg/ml each. The virus/ab mixture was incubated for 1 hr at 37 C then added to 293/ACE2 stable cell line. Seventy two hours later, the virus entry was determined by luciferase expression. The percentage entry inhibitions by individual antibodies as well as combinations of antibodies were calculated. Error bars represent SD of representative experiment performed in triplicates. Statistical analysis was done using Student-t test, significant differences are indicated by asterisks,* p <

96 78 B. Development of broad spectrum antiviral drugs against SARS-CoV, Ebola, and Henipaviruses 1. Synthesis of viral and host proteins derived peptides that contain the natural cathepsin L cleavage sites CatL cleavage sites in the glycoproteins of SARS-CoV, EBOV, NiV and HeV zoonotic viruses were identified as conserved elements. Peptides (10 amino acids long), derived from the glycoproteins of SARS-CoV, EBOV, HeV and NiV and the host proneuropeptide Y (pro-npy) that contain the naturally conserved catl cleavage sites, were synthesized in the protein research laboratory at UIC (Figure 12). The peptides derived from each of the aforementioned viruses were labeled on the N-terminal amino acid with 5-Carboxytetramethylrhodamine (Tamra) as a quencher and on the C terminus with 5-Carboxyfluorescein (5-FAM) as an emitter. If the peptide is not cleaved, no fluorescence emission is detected at 535nm when FAM is excited at 485nm due to the quenching effect of Tamra. In contrast, if the peptide is cleaved by catl, an emission of light at 535nm is detected. 2. The synthesized peptides are efficiently cleaved at the expected sites by cathepsin L The unlabeled peptides derived from the viral glycoproteins as well as the host pro-npy derived peptide were tested for cleavage, at 1µM concentration, by catl (1µg/ml). The

97 79 A B I V A Y T M S L G A S H P L R E P V N A C D V G D V R L A G V I V G D V K L A G V V E R Q R Y G K R S S P Figure 12 Synthesis of host and viral derived peptides. Peptides derived from glycoproteins of viruses and host pro-neuropeptide Y and composed of 10 amino acids were synthesized. (A) SARS-CoV Spike (S) protein derived peptide, (B) Ebola virus GP protein derived peptide, (C) Nipah virus fusion protein (F 0 ) derived peptide, (D) Hendra virus fusion protein (F 0 ) derived peptide, and (E) Human pro-neuropeptide Y (Pro-NPY) derived peptide. Arrows indicate the cathepsin L cleavage sites.

98 80 cleavage was detected using Matrix Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI-MS) compared to the catl untreated control. As expected, the four viral derived peptides were found to be cleaved at the expected sites as shown by MS (Figure 13, 14). All the labeled peptides were similarly cleaved by catl confirming that the fluorophores did not affect the catl mediated cleavage (Figure 15, 16). 3. Optimization of the High Throughput Screening Assay (HTSA) The HTSA used labeled peptides derived from the aforementioned viruses, which contain the natural catl cleavage sites. The HTSA is a Fluorescence Resonance Energy Transfer (FRET) based assay. If the peptide is not cleaved, no fluorescence emission is detected at 535nm when FAM is excited at 485nm due to the quenching effect of Tamra. In contrast, if the peptide is cleaved, an emission of light at 535nm is detected. The SARS-CoV derived labeled peptide was used initially to optimize the HTSA. We hypothesized that our FRET based assay can be used for HTSA of small molecules to identify compound/s that will inhibit conserved catl cleavage sites of diverse viral glycoproteins while minimally interfering with host substrate/s cleavage thus paving the way for developing safe and effective broad spectrum anti-virals. The assay was optimized in black 384 well plates using 3µM SARS-CoV-S derived labeled peptide incubated with 1µg/ml human catl and further optimized with 1µM SARS-CoV-S derived labeled peptide incubated at room temperature with 0.25, 0.5, and 2µg/ml catl in 50µl total volume of NH 4 Ac buffer ph 5.5 supplemented with 4mM EDTA and 8mM DTT. The fluorescence was measured over time, at 535nm after

99 81 Voyager Spec #1=>BC=>NF0.7[BP = , A 90 B % Intensity % Intensity Mass (m/z) Mass (m/z) C D % Intensity % Intensity Mass (m/z) Mass (m/z) Figure 13 SARS-CoV and EBOV derived peptides are cleaved efficiently by cathepsin L (catl) at the expected sites. (A) SARS-CoV untreated peptide ( Da), (B) SARS-CoV catl treated peptide (500, and Da), (C) EBOV untreated peptide ( Da), and (D) EBOV catl treated peptide (630.57, and Da).

100 A 90 B % Intensity % In te n s ity % Intensity C Mass (m/z) % In ten sity D Mass (m /z) Mass (m/z) Mass (m /z) Figure 14 Nipah and Hendra virus derived peptides are cleaved by catl. (A) Nipah virus untreated peptide ( Da), (B) Nipah virus catl treated peptide (544.66, and 457 Da), (C) Hendra virus untreated peptide ( Da), and (D) Hendra virus catl treated peptide (538, and 477 Da).

101 A B % Intensity % Intensity Mass (m/z) Mass (m/z) C D 1.3E % Intensity % Intensity Mass (m/z) Mass (m/z) Figure 15 Labeled SARS-CoV and EBOV derived peptides are cleaved efficiently by catl at the expected sites. (A) SARS-CoV untreated peptide ( Da), (B) SARS-CoV catl treated peptide (978, and 964 Da), (C) EBOV untreated peptide (2017 Da), and (D) EBOV catl treated peptide (1022, and 1016 Da)

102 E 90 A 90 B % Intensity % Intensity Mass (m/z) Mass (m/z) E+4 90 C 90 D % Intensity 50 % Intensity Mass (m/z) Figure 16 Labeled Nipah and pro-npy derived peptides are cleaved efficiently by catl at the expected sites. The labeled Nipah and pro-npy derived peptides were incubated with catl for 1hr at room temperature after which cleavage was detected by MS. (A) Nipah untreated peptide ( Da), (B) Nipah catl treated peptide (957, and 958 Da), (C) Pro-NPY untreated peptide (2133 Da), and (D) Pro-NPY catl treated peptide ( , and Da). Mass (m/z)

103 85 excitation at 485nm, using fluorescence reader at the UIC HTS facility. The labeled SARS-CoV derived peptide at 3µM was found to be cleaved by catl at concentration of 1µg/ml and the cleavage was seen in the form of increased fluorescence over time with no increase in fluorescence in catl untreated peptide (Figure 17A). Next, the SARS- CoV derived peptide was used at a concentration of 1µM incubated with 0.5µg/ml catl, and the peptide was found to be efficiently cleaved as shown by rapid increase in fluorescence over time (Figure 17B). The rate of reaction was found to be dose dependent with higher rate at catl concentration of 2µg/ml compared to concentration of 0.5µg/ml (Figure 17C). The validity of the assay was determined based on the Z- factor calculation. The Z-factor is a measure of statistical effect size. It has been proposed for use in HTS to judge whether the response in a particular assay is large enough to be reliable. Zhang and colleagues (232), defined Z-factor based on the following calculations: 1- Compute the threshold value for negative controls as the mean signal of the negative controls plus three times their standard deviation. 2- Compute the threshold value for positive controls as the mean signal of the positive controls minus three times their standard deviation. 3- Compute the difference between the two thresholds and call it the 'separation band' of the assay, S. If the threshold computed in step 1 is less than the one computed in step 2, then this difference is positive. Otherwise, this difference will have a negative value. 4- Compute the absolute value of the difference between the two means and call it the dynamic range of the assay, R. 5- Compute the Z as S/R. A Z-factor of 1 means an ideal assay, however, an assay can never have a Z-factor of 1. This value is approached when you have a huge dynamic range with tiny standard deviations.

104 86 A B C 2µg/ml Fluorescence units 0.5µg/ml Time (min) Figure 17 Optimization of HTSA. (A) 3µM of SARS-CoV derived peptide was incubated for 25 minutes with or without 1µg/ml catl. (B) 1µM of SARS-CoV peptide incubated for 25 minutes with 0.5µg/ml catl. (C) Dose dependent cleavage of SARS-CoV derived peptide with 0.5, and 2 µg/ml catl. Cleavage was measured as fluorescence units over time.