EPITOPE-BASED SARS VACCINE

This application claims the benefit of U.S. Serial No. 60/638,188, filed December 23, 2004, the contents of which are incorporated herein in its entirety by reference.

Throughout this application, various references are cited and disclosures of these publications in their entireties are hereby- incorporated by reference into this application in order to more fully describe the state of the art to which this invention pertains .

BACKGROUND OF THE INVENTION

Severe acute respiratory syndrome (SARS) emerged as a new infectious disease that claimed 8098 victims including 774 lives in the last outbreak that ended in July 2003 (Summary of probable SARS cases with onset of illness from 1 November 2002 to 31 July 2003. 2003) . A novel coronavirus (SARS CoV) was identified as the aetiological agent (Peiris et al . 2003; Ksiazek et al . 2003; Poutanen et al . 2003; Drosten, C. et al .

2003) . The virus particle most likely transmits through droplets and causes high fever, malaise, rigor, headache, nonproductive cough or dyspnea and may precipitate interstitial infiltrates in the lung and eventual mortality (Peiris et al . 2003) . Adaptive antibody generation is most likely required for clearing the virus since serum from the recovered patients has been found to be beneficial to the incumbent infected patients.

Conventionally, the most effective prevention measure against the lethal SARS virus is vaccination. However, a major safety- concern in development of a vaccine against the SARS coronavirus is that it may induce immunopathology seen for feline infectious peritonitis virus (FIPV) (Marshall et al. 2004), a close relative of the pathogen. It has been found that cats vaccinated actively or passively with feline infectious

peritonitis virus (FIPV) frequently develop disease far more rapidly and more severely than cats that have not had previous exposure to the coronavirus (Pedersen et al . 1980; Weiss et al . 1981; Vennema et al . 1990) . Instead of providing protection, the vaccinations actually facilitate and enhance uptake and spread of the virus, causing an antibody-dependent enhancement (ADE) of infectivity (Vennema et al . 1990; Weiss et al . 1981; Porterfield et al. 1986) was later found that this ADE effect is associated with antibodies targeting many epitopes on the major envelope spike protein(S) (Corapi et al . 1992; Olsen et al . 1992; Olsen et al . 1993; Corapi et al . 1995) that there likely exist viral antigenic peptide sequences capable of inducing neutralizing antibodies without the ADE effect (Corapi et al . 1995) .

Little is known about the antigenicity of the causative agent in natural human immune responses, which is important in design of diagnostic and prophylactic measures. Antibodies targeting the nucleocapsid and the surface spike proteins were found in the serological samples from SARS patients. For examples, relative amount of anti SARS-CoV antibodies targeting the spike protein and nucleocapsid protein was measured (Tan et al. 2004) . In addition, antigenic regions and a short peptide epitope on the nucleocapsid were determined for serological antibodies from SARS patients (Chen et al. 2004; Lin et al. 2003) . However, no detailed epitope map has been obtained and none of the immunodominants that induced protective antibodies in most patients was found. Moreover, none of these investigations was able to determine the viral fragments that would lead to virus- neutralizing antibodies that protected patients against the invading virus.

There are currently three approaches to develop a SARS CoV vaccine that have been reported in the mass media or scientific journals. The first is to use the deactivated SARS CoV as the

antigen, championed by Sinovac Biotech of Beijing, that has been entered into clinical trials (Marshall et al . 2004) . The second is to use adenovirus as a vector to deliver codon-optimized Sl domain of the SARS CoV spike protein and nucleocapsid gene (Gao et al. 2003) . This second experimental vaccine has been tested in four rhesus macaques and found to induce virus neutralizing antibodies and CD8+ cytotoxic T-lymphocyte responses. The third experimental DNA vaccine uses the codon-optimized spike gene in three forms (full-length, without cytoplasmic C-terminus, without the transmembrane domain and the cytoplamic C-terminus) as the antigen. This vaccine has been tested in a mouse model that showed both humoral and cellular immune protection against the SARS CoV (Subbarao et al. 2004; Yang et al. 2004) . However, all three approaches fail to address the problem of potential ADE. Moreover, the animal models used to evaluate the effectiveness and safety of the vaccines are not valid models for SARS disease because neither rhesus macaques or mice developed sickness or disease symptoms similar to SARS patients in the last outbreak, except that SARS CoV induced lung lesions in monkeys similar to that seen for SARS patients. Therefore, although the anti-SARS CoV antibodies in the mouse model have been shown to be immune-protective and not to cause antidoby- dependent enhancement (Subbarao eϋ al. 2004; Yang et al. 2004), ADE is still possible in humans and remains the biggest safety concern in vaccine development against the viral pathogen.

In the last SARS outbreak in 2003, plasma from convalescent SARS patients was shown to be safe and in many cases provide immune protection in passive immunization of infected patients (Wong et al. 2003; Zhou et al. 2003) . Beneficial effects of the plasma humoral antibodies facilitated identifying their viral targets for inclusion in a vaccine, thereby excluding potential ADE- inducing viral components. Through phage-display epitope mapping, the immunodominants regions of the pathogenic virus were determined and these viral protein fragments were included

in the formulation of an epitope-based vaccine for SARS CoV.

This vaccine is superior to other vaccines in various stages of testing as listed above. It is not to induce antibodies with

ADE effect because they will contain only the antigenic fragments that lead to the convalescent phase IgG antibodies from SARS patients, which have been shown to confer immune protection in humans to SARS CoV infection by passive immunization. On the other hand, other vaccines are still overshadowed by the likelihood to induce ADE in humans, risking a disastrous situation if it turns out to be true.

SUMMARY OF THE INVENTION

This invention provides epitope-based vaccine for severe acute respiratory syndrome. A major safety concern in development of a vaccine against the pathogen is that it may induce immunopathology seen for feline infectious peritonitis virus (FIPV) , a close relative of the pathogen. The immunodominant sites on the causative virus were identified by determination of continuous viral epitopes to complementary antibodies in the plasma of convalescent SARS patients. These immunodominants consist of short peptide fragments distributed on various viral proteins, namely, the spike protein, the nucleocapsid protein, the replicase Ia, and the unknown proteins 3a and 9b. A number of epitopes that sparsely distribute on the various viral proteins outside the immunodominant were also identified. Complementary antibodies targeting the immunodominant site on the spike protein effectively neutralize the coronavirus in vitro and are the major immunoglobulins directed against the entire viral envelope in the plasma of convalescent SARS patients. These viral antigenic protein fragments are used to formulate an epitope-based vaccine that will avoid the potential immnuopathologic effects found in vaccines for the feline infectious peritonitis virus.

DESCR I PTI O N OF THE FIGURES

Figure 1. Alignment of the dodecapeptide inserts of the enriched M13 phage clones from biopanning with convalescent phase plasma from SARS patients A-H. The sequences are aligned with Clustal Xl.81.

Figure 2. Alignment of the confirmed viral epitopes with dodecapeptide inserts of enriched M13 phage clones from the biopanning. Epitopes are named in the format of ^λ viral protein- patient-sequential number' . Abbreviations for viral proteins are S, N, R, M, 3a, 3b, 7a, and 9b, representing spike, nucleocapsid, replicases, membrane glycoprotein, protein 3a, 3b, 7a, and 9b, respectively. A-T denote 20 SARS patients whose convalescent phase plasma was used in the phage panning to derive the epitopes. Only the dodecapetides, among a total of 80 from the enriched phage clones, which contain the matching conserved sequence are listed in the alignment. Shaded sequences are the identified convergent antigenic sites. The viral fragment used in confirmation of the epitope is underlined.

Figure 3. Identified antigenic sites on the SARS coronavirus. A vertical red line denotes a confirmed epitope,- a red shaded area is a convergent antigenic site consisting of four or more identified epitopes within a consecutive viral protein sequence; one horizontal blue bar stands for one epitope within a convergent antigenic site which was determined from a different convalescent-phase serum sample. Alignment of the viral epitopes and the matching dodecapeptide inserts of the enriched phage clones from the biopanning is provided in Fig. 2. A box denotes a translated viral protein. S: spike glycoprotein; E: small enveloped glycoprotein,- M: matrix glycoprotein,- N: nucleocapsid protein,- Ia and Ib: replication polyproteins; protein 3a, 3b, 6, 7a, 7b, 8a, 8b, and 9b are unknown proteins.

^' Figure 4. Promiscuous reaction of spike Sl protein with IgG antibodies in normal and SARS patient plasma. Sl denotes the use of denatured spike Sl protein (300 ng/dot) as the antigen for the dot-blotting. A-T denotes plasma samples from 20 convalescent SARS patients. Nl-NlO denotes plasma samples from 10 normal patients. The signals for normal plasma antibodies are not due to impurities from E. coli (please see controls in Fig. 1.) . Similar positive signals were obtained in Western blotting of the spike Sl protein using either normal or SARS convalescent phase plasma, after SDS-PAGE separation (data not shown) .

Figure 5. Analysis of antibodies targeting the immunodominant site on the spike S2 protein. Blocking of the anti-spike S2 protein antibodies in plasma of convalescent SARS patients (A-T and 1-20) by a peptide SL26. The spike S2 protein (375 ng/dot) and deactivated SARS coronavirus (50 ng/dot) were used as the antigen for the rows labeled S2 and SARS CoV, respectively. S2/SL26 indicates pre-saturation of plasma with the peptide SL26 before blotting and use of the spike S2 protein (375 ng/dot) as the antigen. Nl-NlO denotes plasma samples from 10 normal patients and Dl-DlO the sera withdrawn after 15 days of hospitalization from 10 SARS patients who eventually died of the infection; SL26 was 5 μM and 10 μM in blocking the plasma antibodies from patient A-H and 1-20, respectively; an asterisk (*) indicates an SL26 concentration of 200 μM in the blocking of the complementary antibodies.

Figure 6. Analysis of antibodies discontinuous viral surface antigens in plasma from convalescent SARS patients. Shown are the results of ELISA analysis of convalescent phase antibodies from patient A-J against deactivated SARS coronavirus before

(solid lines) and after (dotted lines) neutralization with nucleocapsid protein and peptides encompassing the epitopes identified in the spike, 3a, 3b, and 9b proteins. Concentration

and identity of the added peptides and protein in the neutralization of plasma are given in Table 1.

Figure 7. Neutralization of the SARS coronavirus by antibodies targeting the immunodominant site on the spike S2 protein. (A) Normal Vero E6 cells in the presence of 4 μM of the peptide SL26 encompassing the immunodominant region. (B) Typical cytopathic effect observed for Vero E6 cells inoculated with SARS coronavirus (BJOl) . (C) Protection of Vero E6 cells from SARS CoV infection by plasma antibodies (1:40) from convalescent patient J. (D) Blocking of the neutralization effect seen in panel C by preincubation of the plasma antibodies with 4 μM peptide SL26. (E and F) No protection of Vero E6 cells from SARS CoV infection was observed for the plasma antibodies (1:10) from uninfected patient Nl (E) or the control plasma (1:10) together with 4 μM peptide SL26 (F) . Cell pictures were taken with a phase-contrast microscope at a magnification of 100 x, after 48 h of inoculation at 37°C. Cell cultures started with 4 x 10 ⁴ Vero E6 cells in a total volume of 300 μL, including 5 x 10 ⁴ TCID50 of SARS coronavirus (BJOl) where appropriate. Peptide SL26 is not cytotoxic to the host Vero E6 cells at concentrations up to 1 mM.

Figure 8. ELISA analysis of plasma samples from infected patients and uninfected donors using 3aN-BSA conjugate (A) , an irrelevant peptide RPl-BSA conjugate (B) and recombinant SARS- CoV nucleocapsid protein (C) as the antigen (1 mg per well) . The value of A450 in the plot was obtained by subtracting the reading of a parallel control experiment using BSA as the blank antigen from the sample value. Among 123 patients who had recovered from SARS, 60 (48.8 %) were positive for 3aN-specific antibodies and 117 (95.1 %) were positive for nucleocapsid- specific antibodies. Among 27 patients who had died from SARS, two (7.4 %) were positive for 3aN-specific antibodies and 25

(92.6 %) were positive for nucleocapsid-specific antibodies.

All 25 uninfected donors were negative for 3aN-specific antibodies, but one was found to be positive for antibodies targeting the nucleocapsid protein.

Figure 9. Antibody responses to the 3aN conjugates in mice and rabbit. (A) Titration curves for 3aN-specific antibodies in rabbit antiserum. The rabbit was immunized with a 3aN-KLH conjugate on days 1, 14, 28 and 42 and test bleeding was conducted on days 0, 21, 35 and 49 for ELISA titration using a 3aN-BSA conjugate. Day 0 serum was collected before immunization as a negative control. In a similar experiment, a titre of 6400 for 3aN-specific antibodies was determined for a combined serum collected on day 35 from three mice immunized with a 3aN-BSA conjugate on days 1, 14 and 28. (B) Western dot- blot analysis of 3aN-specific antibodies in the antisera collected from the immunized mice and rabbit on day 35 and 49 after immunization, respectively.

Figure 10. (A) Recognition of 3a-expressing cells by antibodies in rabbit antiserum and convalescent-phase plasma. Rhodamine staining of Vero E6 cells transfected with a plasmid containing the 3a-EGFP gene was negative for the rabbit pre-immune serum, two normal uninfected sera and two convalescent-phase plasma samples that tested negative for 3aN-specificantibodies, but was positive for the rabbit antiserum and five positive convalescent-phase plasma samples from the ELISA screening. (B) Cytotoxicity of 3aN-specific antibodies towards 3a-expressing cells in the presence of the human complement system. HEK293T cells were transfected with the 3a-EGFP-expressing plasmid, treated with control human serum (heat inactivated) and convalescent-phase plasma, incubated in 10% human serum (not heat inactivated) and imaged with a fluorescent microscope. Fluorescent cells (white arrows) treated with convalescent-phase plasma containing 3aN-specific antibodies deformed and started

to detach from the matrix after incubation in human serum, whereas those treated with serological samples without 3aN- specific antibodies were not affected. The deformed cells were confirmed to be dead by trypan blue staining. All experiments were performed in triplicate.

DETAILED DESCRIPTION OF THE INVENTION

This invention provides a vaccine capable of eliciting immune response against Severe Acute Respiratory Syndrome (SARS) Coronovirus in human, comprising at least one of the antigenic epitopes, cDNA which encodes said epitope, its protein counterpart from SARS coronavirus or in combination thereof, wherein said epitopes are known to react with convalescent antibodies from a subject infected with the SARS coronavirus and wherein said vaccine does not contain the naturally occurring viral proteins of the SARS coronavirus.

Even though this invention does not cover vaccines which only contains the naturally occurring viral protein(s) of the SARS coronavirus. However, it is the intention of this invention to encompass fragment of the viral proteins which contain the epitopes. It further the intention of this invention to cover uses of this vaccines to treat and prevent SARS.

As used herein, the epitopes include the small units which would elicit host immune response. These epitopes may include structures which have equivalent activities as an epitope made of peptide.

In an embodiment, the vaccine comprises a construct of the antigenic epitopes of cDNA or their protein counterparts in tandem repeats. In another embodiment, one or more of the antigenic epitopes in the combination are in a tandem repeat or repeated in any order in the construct .

This invention also provides the above vaccines wherein the antigenic epitopes and their repeats in a combination are linked with any DNA sequences or their protein counterparts from any sources and in any orders .

In an embodiment, the said construct of antigenic peptides is grated to a carrier protein or its cDNA counterpart. In another embodiment, the combination of constructs in protein form as described in the above are in any ratios.

In a separate embodiment, the above described vaccine is a recombinant virus containing the construct of antigenic epitopes in DNA form.

In another embodiment, the said vaccine contains at least one recombinant DNA containing the construct of antigenic epitopes. In a further embodiment, the vaccine comprises a combination of constructs in recombinant DNA forms as above described in appropriate ratios.

In a preferred embodiment, the above vaccine contains the antigenic epitope in protein form which has one or more of the following sequences:

Antigenic peptide 1:

SQILPDPLKPTKRSFIEDLLFNKVT

Antigenic peptide 2 :

LRSITAQPVKIDNASPASTVHATATIPLQ

Antigenic peptide 3 : PNQTNWPPALHLVDPQ

Antigenic peptide 4:

AINSVPWSKILAYV

Antigenic peptide 5:

GANKEGIVWVATEGALNTPKDHIGTRNPNNNAATVLQLPQGTTLPKGFYAE

This invention provides a method for determining epitope useful for preparation of a vaccine against Severe Acute Respiratory- Syndrome Coronavirus comprising steps of:

(a) identification of possible antigenic sites on the Severe Acute Respiratory Syndrome Coronavirus;

(b) synthesis of peptides which mimic said sites;

(c) contacting said infectious agent with convalescent antibodies which is known to react to the Severe Acute Respiratory Syndrome Coronavirus; and

(d) determining if the synthesized peptides from step (b) may interfere with reaction, wherein a positive interference indicates that the synthesized peptide contains an epitope useful for preparation of a vaccine against Severe Acute Respiratory Syndrome Coronavirus.

If the peptides do not react with the serum antibodies, these peptides are classified as negative in epitope determination. Only the peptides which show positive reaction will be selected.

In an embodiment, the identification in step (a) is determined by the genome information. In another embodiment, the identification in step (a) is determined by a computer program.

This invention also provides epitopes identified by the above methods. These epitopes may be useful for generation of both monoclonal and polyclonal antibodies against SARS coronavirus.

The generated antibodies may be used for detection diagnosis, and treatment of SARS virus or the SARS virus infected subject. If the epitope is a peptide, the corresponding nucleic acid sequence will be decoded. Accordingly, this invention also provides a nucleic acid molecule encoding the identified

epitopes. The nucleic acid molecule may be useful in production of large amount of peptide when it is placed in an appropriate vector expression system. Therefore, this invention also provides an efficient expression of the identified epitope.

Further, this invention provides a vaccine capable of eliciting immune response against Severe Acute Respiratory Syndrome (SARS) coronovirus in human, comprising of the above-described epitope of its protein counterpart from nucleic acid molecule encoding a protein comprising said epitope or in combination thereof, wherein said vaccine does not contain the naturally occurring viral proteins of the SARS coronavirus in Severe Acute Respiratory Syndrome Coronavirus.

This invention provides antibodies produced by the above identified epitopes. In an embodiment, the antibody is an anti- idiotypic antibody. In a further embodiment, this anti- idiotypic antibody is used as a vaccine. Accordingly, this invention provides a vaccine comprising the anti-idiotypic antibody or a function portion of said anti-idiotypic antibody.

This invention further provides a composition comprising the above described vaccine and an appropriate carrier. Said carrier may be a pharmaceutical carrier. For the purposes of this invention "pharmaceutically acceptable carriers" means any of the standard pharmaceutical vehicles. Examples of suitable vehicles are well known in the art and may include, but not limited to, any of the standard pharmaceutical vehicles such as a phosphate buffered saline solutions, phosphate buffered saline containing Polysorb 80, water, emulsions such as oil/water emulsion, and various type of wetting agents.

This invention provides the above vaccines wherein the above identified epitope (s), its protein counterpart from the infectious agent, nucleic acid molecule encoding a protein

comprising said epitope, its protein counterpart or the nucleic acid molecule is conjugated. Procedures for conjugation are well known in the vaccine field. In an embodiment, the vaccine further comprises an appropriate adjuvant.

This invention provides a method for protecting a subject from Severe Acute Respiratory Syndrome Coronavirus comprising administering to said subject an effective amount of any of the preceding vaccine. The above described vaccine may be used to prevent the onset of the Syndrome in a subject.

The subject includes human and/or animals which may be infected by the SARS virus. Appropriate dosage may be determined by routine experimentation. In addition appropriate route of administration may also be determined. Such routes of administration are but not limited to: Intravenous (IV) , Intraperitoneal (IP), Intradermal (ID) and Oral. _

1. The antigenic map of SARS CoV A global approach was taken to determine the immunodeterminants on the viral pathogen through biopanning of a random dodecapeptide M13 phage library of 1.9 x 10 ⁹ independent clones, directly using plasma from convalescent SARS patients. Probable epitope sequences were derived from alignment of the dodecapeptide inserts of the enriched phage clones and comparison of the resulting consensus sequences with the total proteins coded by the SARS coronavirus genome (Tor 2 strain, GenBank accession: NC_004718) or the viral proteins deposited in the GenBank, using 'BLAST' program seeking short and nearly exact matches. Plasma samples from a total of 20 convalescent SARS patients (A-T) were subjected to the epitope mapping process, one at a time, and 299 distinct consensus sequences were obtained, as shown in Fig. 1. A significant portion of these consensus sequences (92) matches a protein fragment of the SARS coronavirus, which are defined as probable SARS coronavirus

'epitopes. Mostof the remaining consensus sequences align well with proteins of other viruses that might have infected the patients during the disease course, such as human metapeumovirus or influenza viruses. Another small portion of the probable epitope sequences neither matches a SARS protein fragment nor maps to sequences of other viruses, which may represent the mimotopes to the serological antibodies. Under identical conditions, phage clones from biopanning with control sera also showed sequence convergence in the dodecapeptide inserts but no consensus sequences matching protein fragments in SARS coronavirus was identified. These control sera were from two SARS patients before seroconversion (7 days after hospital admission) who eventually recovered from the disease and two patients after the projected seroconversion period (after 15 days of hospital admission) who eventually died of the infection.

Small viral protein fragments matching the consensus sequences from the biopanning were chemically synthesized and subjected to Western dot-blotting to test the presence of complementary IgG antibodies in the plasma samples from which the probable epitopes were derived. Of the total 97 probable epitopes (a few consensus sequences have more than one matching homologue on the viral proteins) , 63 sequences were confirmed while the other 34 were excluded because of the negative testing results. The confirmed epitopes and their alignment with the viral proteins are shown in Fig. 2. Immunoassays with the short peptides alone or in conjugation with a carrier protein (bovine serum albumin) through an enzyme-linked immunosorbent assay (ELISA) did not give consistent testing results because of high background signal. Interestingly, a significant portion of the confirmed SARS epitopes (29) converge to five short and isolated fragments that belong to different viral proteins (Fig. 3) . These convergent antigenic sites locate in the spike protein between amino acid 787-809, the nucleocapsid protein between 127-173,

theunknown protein 3a between 12-37, the unknown protein 9b between 5-17, and the replication polyprotein Ia between 2109- 2116. Every epitope within a convergent antigenic locus originates from a different convalescent SARS patient. Other 34 iminunodeterminants disperse outside of the identified convergent loci without apparent co-localization. Notably, there is only a very short epitope-rich fragment on the S2 domain of the spike protein in the entire viral envelope comprising of the S, M, and E glycoproteins.

2. Analysis of convalescent phage IgG antibodies targeting spike Sl and S2 proteins and the viral surface envelope. Co-localization of the epitopes to the short viral protein fragments indicates that they are the immunodominant sites on the SARS coronavirus. Due to limitations of the adopted phage panning method, there may be more complementary antibodies than those detected that also target these identified convergent loci, but escape detection in the biopanning because of low abundance or low affinity for short peptides in comparison to the cognate intact protein (23) . To test this, the antibodies complementary to the convergent site identified on the S2 domain of the spike protein were analyzed. The prevalence of antibodies targeting the spike Sl domain due to its promiscuous reaction with normal patient plasma was unable to be determined (Fig. 4) . The spike S2 protein (amino acid 681-1203) without the transmembrane domain and the intraviral segment was expressed in fusion to an W-terminal histidine tag in E. coli as inclusion bodies. Using purified denative protein as the antigen for Western dot-blotting, S2-targeting antibodies were detected in most plasma samples (33 out of 40) , whereas no such antibodies was found in normal plasma (Nl-NlO) and sera (Dl-DlO) from SARS patients who died of the viral infection (Fig. 5) . Most of the antibodies in 23 of the 33 positive serological samples were significantly blocked with a low concentration (5 or 10 μM) of a peptide SL26 (Ac-QILPDPLKPTKRSFIEDLLFNKVT LA-OH,

Ac = acetyl) encompassing the identified convergent antigenic region. Antibodies in 10 other plasma samples unaffected by low concentrations of the peptide SL26 were also significantly- blocked when the peptide concentration was increased to 200 μM, indicating they are also complementary to the identified convergent antigenic region on the spike S2 protein but with a low affinity for the synthetic peptide. Indeed, these results show that the identified epitope-rich region on the spike S2 protein is a humoral immunodominant site that induces antibodies in a majority of recovered SARS patients.

Surface antigens of the SARS coronavirus are most important in vaccine development because of their potential to induce virus- neutralizing humoral immunity. To estimate the amount of antibodies targeting discontinuous epitopes and sugar antigens on the envelope proteins, which could not be determined by the employed phage panning method, we used the partially denatured deactivated SARS coronavirus as the antigen to determine antibody levels in plasma from 10 recovered patients (A-J) through enzyme-linked immunosorbent assay (ELISA) . Only one plasma sample from the recovered patient F was tested negative in the assay (Fig. 5) . The anti-SARS coronavirus antibodies in the remaining 9 serological samples were found to be significantly diminished to very low levels after the plasma antibodies were pre-saturated with soluble nucleocapsid protein and peptides containing the identified epitopes of the spike, 3a, 3b, and 9b proteins (as in Table 1) , indicating that complementary antibodies to the discontinuous or glycopeptide epitopes on the viral envelope are insignificant in comparison to that of the nucleocapsid and the identified linear epitopes.

3. Neutralization of SARS CoV by antibodies targeting the immunodominant site on the spike S2 protein

To determine virus-neutralizing activity of the antibodies complementary to the identified immunodominant site on the spike

S2 protein, the plasma sample from patient J, whose anti-S2 antibodies can be effectively blocked by the synthetic peptide SL26 was tested (Fig. 4) . Diluted plasma antibodies (1:40) from this patient were found to completely protect the host Vero E6 cells from infection by the SARS coronavirus (BJOl) (Fig. 6) . However, this virus-neutralizing ability of the plasma was essentially nullified by pre-incubation of the plasma antibodies with 4 μM of the peptide SL26 (Fig. 6 and Table 2) . Indeed, these results show that antibodies targeting the identified spike immunodominant site can effectively neutralize the infectivity of the viral pathogen. In addition, the efficient blocking of the virus-neutralizing antibodies by the peptide SL26 alone provides further support to the conclusion that no significant amount of antibodies targeting the discontinuous or glycopeptide envelope epitopes is present in the convalescent- phase plasma.

4. Iπununogenicity of the identified immunodominant fragments and vaccine formulations To verify the immunogenicity of the identified antigenic fragments of the viral envelope protein S and the potential viral surface protein 3a, peptides SL26 (Ac- QILPDPLKPTKRSFIEDLLFNKVTLA-OH, Ac = acetyl) and p3a (Ac- SITAQPVKIDNASPASTVHATATIP, AC = acetyl) were separately covalently coupled to the carrier protein bovine serum albumin

(BSA) and used to immunized mice with appropriate adjuvants.

After three injections with a dose of 200 μg BSA-peptide per mouse, sera from bleeding of the mice were collected and SARS protein-specific IgG antibodies were analyzed with ELISA, using S2 protein and p3a-BSA conjugate as the antigens to assay spike- specific and 3a-specific antibodies, respectively. The titer for S2-specific IgG was determined to be 128,000 and that for 3a-specific IgG was 800. The mouse antibodies were then subjected to neutralization experiments and were found to effectively block the SARS CoV from getting entry into

permissive host Vero E6 cells. These experiments confirmed that the identified immunodominant fragments indeed are highly immunogenic and that they can efficiently induce neutralizing antibodies for the SARS CoV pathogen. The following gives some hypothetical examples of how the determined antigenic determinants can be used in an epitope- based vaccine :

Example 1. The antigenic peptides can be prepared separately in the form X1-QILPDPLKPTKRSFIEDLLFNKVTLA-X2 and X3-SITAQPVKIDNASPA

STVHATATIP-X4, respectively, where Xl, X2, X3 , and X4 can be natural peptide or protein sequences of any lengths from any sources or any other chemical entities. The chemical entities containing the antigenic peptides can be used individually or in combination at any ratios in a vaccine that may also contain any other ingredients derived from the SARS CoV or other sources.

Example 2. The antigenic peptides can be linked together with a structure X1-QILPDPLKPTKRSFIEDLLFNKVTLA-X2-SITAQPVKIDNASPASTVHA TATIP-X3, or X1-SITAQPVKIDNASPASTVHATATIP-X2- QILPDPLKPTKR SFIEDLLFNKVTLA-X3, where Xl, X2, and X3 can be natural peptide or protein sequences of any lengths from any sources or any other chemical entities. Either structure can be produced chemically or by other means and used together with other ingredients from any sources as a vaccine for SARS CoV.

Example 3. DNA or RNA coding the structures specified in

Example 1 can be cloned into a virus vector and the recombinant virions can be used for vaccination against the pathogenic SARS CoV.

Example 4. DNA or RNA coding the structures specified in

Example 2 can be cloned into a virus vector and the recombinant virions can be used for vaccination against the pathogenic SARS CoV.

5. Experimental Details: Materials and methods

5.1. Preparation of serological samples from SARS and normal patients Serological samples (plasma or serum) were prepared from 40 recovered SARS patients within one month after discharge, 2 patients after 7 days of hospitalization who had confirmed diagnosis of SARS but eventually recovered from the disease, 10 patients after 15 days of hospitalization who had confirmed diagnosis of SARS but later died of the infection, and ten patients who were confirmed not to be infected by SARS coronavirus . The patients were 20-65 years of old who were hospitalized in Princess Margaret Hospital (PMH) , Hong Kong SAR, China. Collection and preparation of the serological samples were agreed to by patients in written consent and authorized by the Hospital Ethic Review Committee. After deactivation at 56 ⁰ C for 45 min, the serological samples were stored at -20 ⁰ C until use.

5.2. Panning of Ml3 phage dodecapeptide library with convalescent phase plasma

All serological samples were deactivated before use by heating at 56°C for 45 min before use. In a typical panning experiment, 10 μL Ph. D. -12™ M13 phage-displayed dodecapeptide library (~2 x 10 ¹¹ phage particles, New England Biolab) and 5.0 μL of plasma from a convalescent SARS patient were mixed and diluted to a final volume of 400 μl with TBST (20 mM Tris.HCl, 137 mM NaCl, 0.1% Tween-20, pH 7.6) . The mixture was incubated at room temperature for 30 minutes and the IgG-binding phages were captured by 100 μL suspension solution of Protein G Sepharose Fastflow resin (Amersham Biosciences) that was pre-blocked for 30 min at room temperature with low speed rocking in 1.0 mL TBST buffer containing 2% bovine serum albumin and washed thrice in the same buffer. The suspension was incubated for 30 min at room temperature with agitation by pipetting at 2-3 min

intervals. Subsequently, the supernatant was drained and the resin washed ten times, each with 1 mL of TBST buffer. The washed resin was finally suspended in 1 mL of 0.2 M Glycine-HCl solution (pH 2.2) and incubated 10 minutes at room temperature before the captured phages were eluted to a 1.5 mL microfuge tube containing 150 μL of 1 M Tris-HCl (pH 9.1) for tittering and amplification in ER2738 cells (New England Biolab) according to the manufacturer's instructions. The panning process was then repeated twice using the amplified phage sub-library while the concentration of Tween-20 in the TBST buffer was increased from 0.1% to 0.2%. Convergence in the dodecapeptide inserts of the phage DNA was obvious at the end of second round panning and became significant at the end of third round. A total of 80 random phage clones after three rounds of panning were sequenced using a dye-terminator cycle sequencing kit (ABI Biosystems) . Phage DNA used in the sequencing was isolated and purified through polyethylene glycol precipitation. Consensus sequences were obtained from alignment of the dodecapeptide inserts of the randomly selected phage clones with Clustal Xl.81. 5.3. Expression and purification of nucleocapsid protein and spike S2 protein ^J

The SARS coronavirus Sl gene, with 1986 nucleotides from the 5'- end, was amplified by PCR with a pair of primers: GGAATTCCATATGAGTGACCTTQACCGGTGC and CATGCTCGAGTGTATGGTAACTAGCAC. The PCR product was digested with Nde I and Xho I ₁ purified with

Gel Extraction Kit (Qiagen) and ligated to the pET22b vector

(Novagen) digested with the same restriction enzymes. The SARS coronavirus S2 gene, with 1566 nucleotides coding amino acid

681-1202 of the spike protein, was amplified by PCR with a pair of primers: CGCGGATCCTTAGGTGCTGATAGTTCAATTG and CCGCTCGAGTTATTAG AAGCCGAGCCAAACATACC. The PCR product was digested with BaMi I and Xho I, purified with Gel Extraction Kit (Qiagen) and ligated to the pET28a(+) vector (Novagen) digested with the same restriction enzymes. The 1269 nucleotide-nucleocapsid gene was amplified by PCR using another pair of primers:

GGAATTCCATATGTCTGATAATGGACCCCAATC and

CATGGGATCCGCCTGAGTTGAATCAGCAG. The PCR product was digested with Nde I and BamH. I, purified, and ligated to the digested pET22b (Novagen) with the corresponding restriction enzymes. Expression of Sl protein was induced for 4 h at 37°C by addition of 0.4 mM isopropyl-β-D-thiogalactopyranoside (IPTG) when OD _60O of the cell culture reached 0.5 after the plasmid constructs containing the Sl domain of the spike gene, in frame with a C- terminal His ₆ -tag, was transformed into E. coli BL21(DE3) (Novegen) . The plasmid constructs containing the S2 domain of the spike gene in frame with an iV-terminal His ₆ -tag (additional N-terminal sequence: MGSSHHHHHHSSGLVPRGSHMASMTGGQQMGRGS) and the nucleocapsid gene in frame with a C-terminal His ₆ -tag were transformed into E. coli BL21(DE3) (Novagen) and E. coli BL21(DE3) codon plus (Novagen), respectively. Expression of S2 protein was also induced for 4 h at 37 ⁰ C by addition of 1 mM IPTG when OD _60O of the cell culture reached 0.5~0.6. Expressed Sl-HiS ₆ protein and His ₆ -S2 protein were obtained as inclusion bodies that were solubilized by heating in 20 mM Tris.HCl buffer (pH7.9) containing 2% SDS, 20 mM DTT, and 400 mM NaCl. Buffer of the solubilized protein solutions were then changed with a desalting column and purified by metal-chelating affinity column chromatography using 5 mL HiTrap ^® column (Amersham Biosciences) according to manufacturer's instructions. The purified proteins were stored at -20°C in TBS buffer (20 mM Tris. HCl, 137 mM NaCl, pH7.6) supplemented with 1% SDS and 2 mM β-mercaptoethanol . Similarly, the nucleocapsid protein was induced for expression as a soluble protein for 4 h at 30 ⁰ C by addition of 0.4 mM IPTG when OD _6O o of the cell culture reached 0.5 and purified with a 5 mL HiTrap ^® column. Purified nucleocapsid protein was desalted and stored at -20 ⁰ C in TBS buffer containing 2 mM β- mercaptoethanol . Purified proteins were >70% in purity as determined by SDS-PAGE.

5.4. Solid-phase synthesis of peptides

All synthetic reactions were carried out under protection of nitrogen atmosphere. All Fmoc-protected amino acids and Wang resin were from GL Biochem (Shanghai) Ltd. Peptide grade N, N- dimethylformamide (DMF) was from Fisher Scientific and dried with 4A molecular sieves before use. All other chemicals, such as diisopropylethylamine (DIEA), N, iV'-diisopropylcarbodiimine (DIC) , 1-hydroxybenzotriazole (HOBt) , and trifluoroacetic acid (TFA) , were from Acros and used directly. Viral antigenic peptide fragments were synthesized by the Fmoc method of solid- phase peptide synthesis using standard DIC/HOBt chemistry. The antigenic fragments were divided into groups of 10-20 peptides and each group was synthesized by a manual split-and-pool approach, using IRORI's MicroKan reactors (ChemTech) with 30 mg of Wang resin (loading value: 1.3 mmol/g) for each peptide. After the synthesis and removal of Fmoc group of the last amino acid residue, the N-terminus was capped with an acetyl group by reaction with acetic anhydride and - a mixture of 88:5:2:5 trifluoroacetic acid/phenol/ triisopropylsilane/water was used to cleave the peptide from the resin (3 h, room temperature) . The peptide products were precipitated twice with cold ether (4 ⁰ C) after the solid support was removed by filtration. Masses of the peptide products were found to be consistent with the calculated values in mass spectra recorded on a Finnigan TSQ 7000 triple stage quadrupole mass spectrometer using fast atom bombardment (FAB) ionization mode. Yield of the crude products was in the range of 15-45% and the purity was > 50% by HPLC analysis. Short peptides (7-15 residues) dissolved in DMSO or H ₂ O were used in the immunoassays directly without further purification. Longer peptides corresponding to the epitope-rich viral protein fragments were purified by HPLC and obtained as white solids after lyophilization. HPLC analysis and purification were carried out with a Waters 600E system coupled to a Model 2487 dual λ absorbance detector, using reversed-phase column (XTerra™ RPi ₈ , 7 μm, 7.8 x 300 mm column, Waters) and a

linear gradient from 0% to 75% acetonitrile in 0.1% trifluoroacetic acid/water over 45 min.

5.5. Western dot-blotting

For dot-blotting, PVDF membrane circles in a diameter of 6 mm were cut from a Hybond ECL membrane (Amersham Biosciences) with a binder punch, marked with a pencil, immersed in methanol for 5 min, washed in water, and dried in the air. Appropriate amount of peptides, proteins, or deactivated SARS coronavirus in 2 μL volumes in DMSO or a suitable solvent was then spotted on the center of the PVDF membrane circles and dried in the air at room temperature. The blots were immersed in ethanol for 5 min, washed in water for 5 min, and blocked for 1 h in blocking buffer (8% milk, 20 mM Tris.HCl, 137 mM NaCl, 0.2% Tween-20, pH 7.6) . Subsequently, they were rinsed with TBST (20 mM Tris.HCl, 137 mM NaCl, 0.2% Tween-20, pH 7.6) , distributed into wells of a 48-well microtiter plate containing 100 μL of diluted serological samples from SARS patients or normal donors, and incubated at room temperature for 1 h under constant shaking in an orbital shaker (200 rpm) . After the blots were rinsed twice with TBST, washed six times in the same buffer (shaking at 250 rpm for 15 min for the first three washes and for 5 min for the remaining washes between buffer changes) , they were blocked one more time in the blocking buffer. The blots were then incubated for 1 h at room temperature with shaking in diluted solution of Anti- Human IgG HRP conjugate (1:50,000, Sigma) and washed again six times with "TBST buffer. Finally, the blotting signal was detected with ECL plus Western Blotting Detection System (Amersham Biosciences) .

Dilution factors for the serological samples in blotting with small peptides, proteins, or the deactivated SARS coronavirus as the antigen were 1:200, 1:400, and 1:800, respectively. To block antibodies directed against certain antigens, the diluted serological samples in TBST were incubated in the microtiter plate with the peptide or protein at appropriate concentrations

for 1 h at room temperature under gentle shaking, before incubation with the blotted PVDF membrane circles.

5.6. ELISA analysis of anti-SARS coronavirus antibodies

The Diagnostic Kit for Antibody to SARS Virus (ELISA) was purchased from Beijing DaJiBiAi Bio-Technology Company. The plasma from SARS patient was diluted in the supplied dilution buffer in a ratio of 1:20, 1:40, 1:80, and 1:160 and 100 μL/well of each diluted solution was added to the supplied 96-well microtiter plate coated with deactivated SARS coronavirus. After incubation in a humidified box at 37°C for 30 min, the used wells in the microtiter plate were washed five times with the supplied washing buffer. Subsequently, 100 μL of the supplied secondary antibody (anti-human IgG-HRP conjugate) was added into each used well for incubation in a humidified box at 37°C for 30 min. The wells were washed five times again with the washing buffer and sequentially added 50 μL of substrate A and substrate B for incubation at 37°C in dark for 10 min. Finally 50 μL of the supplied stop solution was added into each well and the developed signal was read at 450 nm using a microtiter plate reader (TECAN) . For each experiment, supplied positive and negative controls were run in parallel to the assays according to the manufacturer's instructions. All assays were carried out in duplicates.

To determine the anti-SARS coronavirus antibody levels after blocking with the identified antigens, the diluted plasma from the convalescent SARS patients were incubated for 1 h at room temperature with the nucleocapsid protein and peptides containing the identified epitopes on the spike, 3a, 3b, and 9b proteins at appropriate concentrations, before addition into the supplied 96-well plate coated with deactivated SARS coronavirus. Other experimental steps were identical to that described above.

5.7. SARS coronavirus neutralization

All experiments were carried out at biosafety level 3. Vero E6 cells were grown and maintained in DMEM supplemented with 10% fetal bovine serum, 100 units/mL of penicillin G, and 100 μg/mL of streptomycin. Viral infection was carried out by addition of 100 μL medium containing 5 x 10 ⁴ TCID ₅₀ SARS coronavirus (BJOl) to microtiter wells containing 4 x 10 ⁴ host cells per well in a volume of 100 μL, mixed well, and incubated at 5% CO ₂ and 37°C for observation of cytopathic effect. For cytotoxicity assay, the peptide SL26 was diluted in 100 μL medium in two-fold serial dilution and added to the cells to achieve a final peptide concentration of 1 mM, 500 μM, 250 μM, 125 μM, 62.25 μM, 31.12 μM, 15.56 μM, 7.78 μM, and 3.89 μM. For neutralization of the virus with plasma antibodies, plasma from convalescent patient J was diluted in 100 μL medium in 2-fold series to achieve a dilution factor of 1:10, 1:20, 1:40, 1:80, 1:160, 1:320, 1:640, 1: 1280, 1: 2560, and 1: 5120. Each diluted solution was mixed with the same amount of the virus in 100 μL medium as that used for the viral infection, incubated at 37°C for 1 h, and added to microtiter wells containing 4 x 10 ⁴ Vero E6 cells in 100 μL medium. To block the neutralizing effect of the plasma antibodies, two-fold serial dilutions of the peptide SL26 were prepared in 100 μL medium containing 2.5 μL of convalescent phage plasma from patient J, mixed with 100 μL medium containing 5 x 10 ⁴ TCID ₅₀ SARS coronavirus, and incubated for 1 h at 37°C before addition to wells containing 4 x 10 ⁴ Vero E6 cells in 100 μL medium. All cell culture experiments were performed in duplicate on two separate 96-well microtiter plates. Cytopathic effect was observed and cell pictures were taken with a phase- contrast microscope after the cultures were inoculated at 5% CO ₂ , 37°C for 48 h. After 2 days of incubation, live cells were stained with crystal violet for easy observation of viral infection.

6. Amino Terminus of the 3a protein as SARS vaccine

The 3a protein of Severe Acute Respiratory Syndrome (SARS) - associated coronavirus is expressed and transported to the plasma membrane in tissue cells of infected patients. Its short ^-terminal ectodomain was found to elicit strong humoral responses in a half of the patients who had recovered from SARS. The ectodomain-specific antibodies from the convalescent-phase plasma readily recognize and induce destruction of 3a-expressing cells in the presence of human complement system, demonstrating their potential ability to provide immune protection through recognizing and eliminating the SARS coronavirus-infected cells that express the target protein. In addition, when coupled to a carrier protein, the ectodomain peptide elicits 3a-specific antibodies in mice and rabbits in high titers. These results show that the amino terminus of the 3a protein is highly immunogenic and elicits potentially protective humoral responses in infected patients. Therefore, the short extracellular domain is a valuable immunogen in development of a vaccine for the infectious SARS.

Severe Acute Respiratory Syndrome (SARS) is a new infectious disease that is caused by a new strain of coronavirus (CoV, Drosten et al. 2003; Ksiazek et al. 2003; Peiris et al. 2003; Poutanen et al. 2003) . Immunogenicity of the viral pathogen has been a focal point of interest because of its central importance to design of an efficacious vaccine. Several experimental vaccines have been successfully developed to induce protective humoral responses specific for the spike protein, suggesting that it is a major antigen responsible for the protective humoral immunity generated in infected SARS patients (Gao et al. 2003; Bisht et al. 2004; Buchholz et al. 2004; Johnston et al. 2004; Subbarao et al. 2004; Yang et al . 2004; Zhao et al. 2004) . This is consistent with recent surveys of the convalescent-phase serological samples from patients who had recovered from SARS, in which spike-specific antibodies were implicated to confer

long-term immune protection (Guo et al. 2004; He et al . 2004; Zhong et al. 2005) . ^•

Besides the spike protein, the 3a protein and other viral proteins have also been found to be a target of humoral antibodies from SARS patients (Wang et al . 2003; Shi et al . 2004; Chang et al . 2004; Tan et al . 2004a,- Leung et al . 2004; Liu et al. 2004; Chen et al . 2004; Zhong et al. 2005) . While most of these antibodies are only of diagnostic value, the 3a protein-specific antibodies might offer additional immune protection to the infected patients and attracted our attention. Protein 3a is a predicted 274-residue transmembrane protein. Recently, it has been shown to be expressed and transported to the plasma membrane in Vero E6 cells infected with the SARS coronavirus, with the amino terminus (1-35) exposed to the extracellular environment (Tan et al . 2004b) . Experimental evidence has also been furnished for its in vivo expression in a lung section from a SARS CoV-infected patient (Yu et al. 2004) . In addition, this protein has an intracellular perinuclear localization similar to all coronaviral surface proteins (spike, membrane, and small envelope proteins) and extensively interact with them (Tan, et al . 2004b; Zeng et al. 2004) , providing the rationale for its incorporation into the viral envelope in the replication process (Ito et al . 2005) . The role of the 3a protein as a new structural protein of the SARS CoV and its being a target of immune responses in infected patients suggest that its amino terminus might be a valuable immunogen in vaccine development. In this study, we therefore surveyed the prevalence of the antibodies specific for the amino terminus of the 3a protein (3aN) in the serological samples from patients who had recovered from SARS, determined the capability of the antibodies to recognize and eliminate 3a-expressing cells, and tested the antigenicity of the amino terminal peptide in animals.

^' To survey the prevalence of antibodies complementary to the identified antigenic site at the amino terminus of the 3a protein (Zhong et al . 2005), a peptide with the sequence encompassing this epitope (residue 11-44, Ac- RSITAQPVKIDNASPASTVHA TATIPLQASLPFG-OH, AC = acetyl) was chemically synthesized and coupled to bovine serum albumin (BSA) for use as the antigen in an ELISA screen of serological samples from SARS CoV-infected patients. A total of 123 plasma samples collected from patients who had recovered from SARS (28 days after discharge) and 27 sera collected from patients who eventually died of SARS (28 days after hospitalization) , were analyzed. These serological samples were prepared between March and October, 2003, deactivated at 56°C for 45 min, and stored at -20 ⁰ C until used at the Princess Margaret Hospital, Hong Kong SAR, China. Under the given conditions, plasma samples from 25 uninfected donors collected from the Hong Kong Red Cross Blood Transfusion Service were tested negative for antibodies against the peptide conjugate (Figure 8) . All patient blood samples were tested negative for the BSA carrier protein, while only two of them were tested positive for a BSA conjugate with an irrelevant peptide RPl (Ac-GPNLRNPVEQPLSVQA-OH, Ac = acetyl) . As a positive control, the nucleocapsid protein was found to be targeted by specific IgG antibodies in a high percentage of the serological samples of both the recovered (95.1%) and deceased patients (92.6%), in consistence with the clinical diagnosis of infection by the SARS CoV for the patients and the high antigenicity of the nucleocapsid protein revealed in other investigations (Wang et al. 2003; Shi et al. 2004; Chang et al . 2004; Tan et al . 2004a; Leung et al . 2004; Chen et al . 2004) . These control experiments established the validity of the ELISA screening method.

Among the 123 recovered patients, 60 (48.8%) were tested positive for 3aN-specific antibodies (Figure 8) , whereas only 2 (7.4%) out of the 27 deceased patients developed humoral

responses to the antigenic peptide. This high immunoreactivity of the 3aN peptide is consistent with the high positive rate

(71%) of convalescent-phase SARS sera for the whole recombinant

3a protein (Tan et al. 2004a) and the positive immunoreactivity of SARS patient sera for a different N-terminal peptide (Zeng et al . 2004; Zhong et al. 2005) . Noticeably, both the prevalence and levels of 3aN-specific antibodies are significantly lower for the deceased patients in comparison to the recovered patients, despite that both groups of samples have similar positive rate of the nucleocapsid-specific antibodies. These results indeed showed that a substantial proportion of the recovered patients developed antibodies specific for the amino terminus of the 3a protein.

To test the antigenicity of the amino terminus peptide of the 3a protein in animals, it was coupled to a carrier protein (BSA or KLH) and the resulting conjugates were used to immunize three mice and a rabbit. A 12-week old New Zealand white rabbit was immunized by 1 mL of the peptide-KLH conjugate (0.84 mg) emulsified in an equal volume of Freund's complete adjuvant (Sigma) in more than 20 sites by intradermal injection. Booster injections were made with the same amount of the peptide conjugate emulsified in Freund's incomplete adjuvant (Sigma) at an interval of 14 days. The mice were immunized by intraperitoneal injection using a lower dose (0.2 mg of the peptide-BSA conjugate) . As shown in Figure 9A, antibodies specific for the 3a N-terminal peptide were readily induced and reached a titer of 6400 and 64,000 for the mice and rabbit, respectively. The tittering experiments showed that the induced antibodies can recognize the 3aN peptide. This is further supported by a Western dot-blot analysis of the antiserum antibodies with the pure and unconjugated 3aN peptide absorbed on a PVDF membrane (Figure 9B) . Due to the short length of the peptide that is unlikely to form a stable conformation, the antiserum antibodies most likely target a consecutive amino acid

sequence in the 3aN peptide in the range from residue 12 to 37 as determined in the phage panning experiments (Zhong et al. 2005) . These results show that the short amino terminus of the 3a protein is indeed highly antigenic and is able to elicit humoral responses in animals, in accordance with its being a target of the humoral responses in humans .

The 3a protein is expressed as a plasma transmembrane protein in SARS CoV-infected cells with its short amino terminus exposed to the extracellular environment and its carboxy end in the cytoplasm (Ito et al . 2005; Tan et al . 2004b) . The 3aN-specific antibodies in the plasma of recovered patients should be able to offer immune protection by recognizing the SARS-CoV infected cells for elimination by the complement system. To test this, the 3a protein was fused to an EGFP protein at the carboxy terminus and expressed in Vero E6 cells. Fluorescent microscopic analysis found that the 3a-EGFP fusion protein was located on the perinuclear region and the plasma membrane (Figure 10A) , a subcellular distribution indistinguishable from that found for the viral protein without a fusion (Yu et al. 2004; Tan et al. 2004b) . After staining with positive human plasma samples or the rabbit antiserum as the primary antibody and an appropriate anti-IgG-rodamine conjugate as the secondary antibody, the green fluorescent cells were labeled with the red fluorescent rodamine under a microscope, indicating that the 3a specific antibodies in patient plasma and animal antiserum can indeed recognize the ectodomain of the 3a fusion protein on the cell surface. Under identical conditions, such cell labeling was not found for the preimmune rabbit serum, uninfected plasma samples, or convalescent-phase plasma samples that were tested negative for 3aN-specific antibodies. To assess the ability of 3aN-specific antibodies to induce elimination of 3a-expressing cells, HEK293T cells transiently transfected with the 3a-EGFP plasmid were sub- cultured into 96-well microplates in DMEM medium containing 10% fetal bovine serum. After 24 h, the cells were incubated for 1 h

^' with 200 μl of ^" 1:10 dilution of the convalescent plasma tested for 3aN-specific antibodies in the immunofluorescent experiments. The cells were then washed with phosphate buffer saline and incubated in DMEM media supplemented with 10% normal human serum. After another 24 h, all the 3a-expressing fluorescent cells showed death symptoms with obvious morphology change and detachment from the poly-D-lysine matrix, whereas non-fluorescent cells were not affected (Figure 10B) . When the human serum was heat-inactivated before incubation with the cells treated with 3aN-specific antibodies, the 3a-expressing fluorescent cells were also not affected. These results indicated that the 3aN-specific antibodies are able to activate human complement cascade through the classical pathway, leading to elimination of the 3a-expressing cells. Taken together, these experiments demonstrate that antibodies elicited by the 2V- terminus of the 3a protein in humans can specifically recognize and, in the presence of human complement system, eliminate the cells in which the 3a protein is expressed.

SARS-CoV induces strong humoral responses in infected human patients or animals, targeting various structural and nonstructural proteins. So far, the only identified antibodies that can neutralize the virus and provide immune protection exclusively target the major envelope glycoprotein—spike protein (S) . The other viral surface glycoproteins, matrix protein (M) and small membrane protein (E) , have not been found to elicit antibody responses in infected SARS patients. This study revealed that the amino terminus of the 3a protein, a plasma transmembrane protein expressed in infected cells, elicits strong humoral responses in a high percentage of patients who have recovered from SARS and is highly antigenic in animals. These antibodies were also found to readily bind the cells expressing the 3a protein and induce elimination of these cells in the presence of human complement system. Such antibodies can provide immune protection in vivo through recognizing and

binding the surface 3a protein of SARS-CoV-infected cells for destruction by the host complement system. Although protein 3a is a new structural protein of the SARS CoV (Ito et al . 2005), the 3aN-specific antibodies are unlikely to offer protection through blocking the cellular entry of the pathogenic virus. This can be seen from the inability of the M or E-specific antibodies to neutralize the infectivity of corresponding animal coronaviruses (Rottier 1995; Siddell 1995) . Nevertheless, the high prevalence of 3aN-specific antibodies in the plasma of patients who have recovered from SARS and their ability to induce destruction of the infected cells suggest that such antibodies can confer long-term immune protection.

Current efforts to develop a SARS vaccine rely on the spike protein to elicit protective humoral responses (Gao et al. 2003; Bisht et al. 2004; Buchholz et al. 2004; Johnston et al .2004; Subbarao et al . 2004; Yang et al . 2004; Zhao et al . 2004) . However, evasion of neutralization by the SARS CoV subtypes identified in the latest outbreak has been found for the spike- targeting antibodies, especially those specific for the receptor recognition site (Yang et al . 2005) . This is likely a result of molecular evolution of the pathogen under immune pressure and raises concern about the efficacy of the spike-based vaccines. In contrast with the high mutation rate of the spike protein, the antigenic site at the amino terminus of the 3a protein has a much higher stability; no mutations have been identified at this site in the molecular epidemiological studies of the known SARS CoV genome sequences (Ruan et al. 2003; Chinese SARS Molecular Epidemiology Consortium 2004; Yeh et al . 2004) . The high genetic stability and the potential ability to elicit long-term immunity make the amino terminus of the 3a protein a highly valuable supplementary immunogen in the development of a vaccine, which is urgently needed for the infectious SARS disease with high morbidity and mortality.

7. Animal Studies

Construction of the chimera antigen. The B-cell epitopes whose immunogenicity is ^' tested above and the T-cell epitopes determined from the peripheral blood monocyte cells of recovered patients will be used in the construction of an epitope vaccine in an appropriate vector. At least two forms of the chimera immunogen will be tested. First of all, the B-cell epitopes will be grafted to the CDR loops of an IgG protein of the viral FMPV vector that has been shown to highly immunogenic in animals, while the T-cell epitopes will be cloned into the same vector to form a recombinant virus for vaccine production. Alternatively, the cDNA of both the B and T-cell epitopes will be fused together to form a chimera gene that is linked to part of the nucleocapsid protein for the purpose of inducing more T- cell responses. The resulting gene will then be cloned into an adenovirus vector to construct a recombinant virus for vaccine production. The cDNA of the epitopes or the partial nucleocapsid gene will be synthesized with optimized codons for expression in humans. Cloning and fusion of the cDNA will be accomplished through the use of PCR with overlapping oligonucleotides as primers.

Immunization of mice with the chimera immunogen and monitoring of the immune responses. The recombinant viruses constructed above will be grown and their TCID ₅₀ titer determined according to standard protocols. Two forms of the immuogen will be used in test immunization. One is to use the deactivated (by heat and UV) virus with a set protein concentration. The other is to use the live recombinant virus directly for immunization. For immunization test, 2 groups of mice will be used for each recombinant viral vaccine to test the deactivated and live virus form. Each group will be divided into 2 subgroups to received two doses whose level will be determined from the titer/protein concentration of the viral vaccine. The empty viral vectors

without the epitopes will be used as negative controls. The immunization protocol will include one immunization shot and a booster shot with an interval of 4 weeks. The immune responses will be monitored by analyzing the antisera and peripheral blood mononuclear cells (PBMC) from bleeding the mice. Humoral responses specific for the B-cell epitopes will be assayed by

ELISA using antigens mentioned earlier. The T-cell responses will be determined by ELISPOT using synthetic epitope peptides

(and peptides from nucleocapsid) and PBMC from bleeding the mice.

Neutralization of the SARS-like virus by antisera. Similar to the testing of individual epitopes, the virus-neutralizing capability of the antisera from the immunized mice will be determined with the SARS coronavirus. The formulation and dose giving rise to the highest level of neutralizing antibodies and cellular immune responses will be used for vaccination testing in animal models.

Test of the vaccine in animal models.

The most effective vaccine form and dose will subsequently be used to test immunogenicity in four African green monkeys. Two control animals will be immunised with the same amount of empty adenoviral vector. The animals will receive a second vaccination with the same regimen after 4 weeks. The T-cell and B-cell responses will be monitored similar to that done in the mice. In addition, the live SARS coronavirus challenge will also be carried out in the immunized animals (mice and monkeys) and the protective effects will also be assessed by observation of clinical symptoms as well as anatomy of the sacrificed animals.

8. Clinical Studies

Appropriate clinical protocols will be established to get the efficacy of the epitope based SARS vaccines 'disclosed in this invention.

REFERENCES

Bisht, H. Roberts, A., Vogel, L., Bukreyev, A., Collins, P. L., Murphy, B. R., Subbarao, K. 189 & Moss, B. (2004). Severe acute respiratory syndrome coronavirus spike protein expressed by attenuated vaccinia virus protectively immunizes mice. Proc Natl

Acad Sci USA 101, 6641- 6646.

Buchholz, U. J., Bukreyev, A., Yang, L., Laπtirande, E. W., Murphy, B. R., Subbarao, K.,Collins, P. L. (2004). Contributions of the structural proteins of severe acute respiratory syndrome coronavirus to protective immunity. Proc Natl Acad Sci USA 101,

9804-9809.

Chang, M. -S., Lu, Y. -T., Ho, S. -T. & 11 other authors (2004). Antibody detection of SARS198 CoV spike and nucleocapsid protein. Biochem Biophys Res Coπun 314, 931-936.

Chen, Z., Pei, D., Jiang, L. & 11 other authors (2004).

Antigenicity analysis of different 201 regions of the severe acute respiratory syndrome coronavirus nucleocapsid protein. Clin Chem 50, 988-995.

Chinese SARS Molecular Epidemiology Consortium (2004) . Molecular evolution of the SARS coronavirus during the course of the SARS epidemic in China. Science 303, 1666-1669.

Corapi, W. V.; Darteil, R. J.; Audonnet, J. C; Chappuis, 6. E. (1995) . Localization of antigenic sites of the S glycoprotein of feline infectious peritonitis virus involved in neutralization and antibody-dependent enhancement. J. Virol 69,

2858-2862.

Corapi, W. V.; Olsen, C. W.; Scott. F. W. (1992) . Monoclonal antibody analysis of neutralization and antibody-dependent enhancement of feline infectious peritonitis virus. J ^" . Virol 66,

6695-6705. '

Drosten, C, Gϋnther, S., Preiser, W. & 23 other authors (2003).

Identification of a novel 208 coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 348, 1967-1976.

Gao, W., Tamin, A., Soloff, A., D'Aiuto, L., Nwanegbo, E.,

Robbins, P. D., Bellini, W. J., Barratt-Boyes, S. & Gambotto, A.

(2003) . Effects of a SARS-associated coronavirus vaccine in monkeys. Lancet 362, 1895-1896. Guo, J. P., Petric, M. Campbell, W. & McGeer, P. L. (2004) . SARS coronavirus peptides recognized by antibodies in the sera of convalescent cases. Virology 324, 251-256.

He, Y., Zhou, Y., Wu, H., Luo, B., Chen, J., Li, W. & Jiang, S.

(2004) . Identification of immunodominant sites on the spike protein of severe acute respiratory syndrome (SARS) coronavirus: implication for developing SARS diagnostics and vaccines. J

Immunol 173, 4050-4057.

Ito, N., Mossel, E. C, Narayanan, K., Popov, V. L., Huang, C,

Inoue, T., Peters, C. J. & Makino, S. (2005). Severe acute respiratory syndrome coronavirus 3a protein is a viral structural protein. J Virol 79, 3182-3186.

Johnston, R. E. (2004) . A candidate vaccine for severe acute respiratory syndrome. N Engl J Med 351, 827-828.

Ksiazek, T. G., Erdman, D., Goldsmith, C. S. & 24 other authors (2003) . A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 348, 1953-1966.

Leung, D. T., Tarn, F. C, Ma, C. H., Chan, P. K., Cheung, J. L.,

Niu, H., Tarn, J. S. & Lim, P. L. (2004) . Antibody response of patients with severe acute respiratory syndrome (SARS) targets the viral nucleocapsid. J Infect Dis 190, 379-386.

Lin, Y.; Shen, X.; Yang, R. F.; Li, Y. X.; Ji, Y. Y.; He, Y. Y.;

Shi, M. D.; Lu, W.; Shi, T. L.; Wang, J.; Wang, H. X.; Jiang, H.

L.; Shen, J. H.; Xie, Y. H.; Wang, X.; Pei, G.; Shen, B. F.; Wu,

J. R.; Sun, B. (2003) . Identification of an epitope of SARS- coronavirus nucleocapsid protein. Cell Res 13, 141-145.

Liu, X., Shi, Y., Li, P., Li, L., Yi, Y., Ma, Q. & Cao, C.

(2004) . Profile of antibodies to the nucleocapsid protein of the severe acute respiratory syndrome (SARS) -associated coronavirus in probable SARS patients. Clin Diagn Lab Immunol 11, 227-228.

Marshall, E .; Enserink, M. (2004) . Caution Urged on SARS

Vaccines. Science 303, 944-946.

Olsen, C. W.; Corapi, W. V.; Ngichabe, C. K.; Baines, J. D.;

Scott, F. W. (1992) . Monoclonal antibodies to the spike protein of feline infectious peritonitis virus mediate antibody- dependent enhancement of infection of feline macrophages. J.

Virol 66, 956-965.

Olsen, C. W.; Corapi, W. V.; Jacobson, R. H.; Simkins, R. A.;

Saif, L. J.; Scott, F. W. (1993) . Identification of antigenic sites mediating antibody-dependent enhancement of feline infectious peritonitis virus infectivity. J. Gen. Viol 993, 74,

745-749.

Feiris, J. S. M., Lai, S. T., Pooπ, L. L. M. & 14 other authors

(2003) . Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 361, 1319-1325.

Pedersen, N. C; Boyle, J. F. (1980). Immunologic phenomena in the effusive form of feline infectious peritonitis. Am. J. Vet.

Res 41, 868-876.

Porterfield, J. S. (1986) . Antibody-dependent enhancement of viral infectivity. Adv. Vet. Res 45, 2580-2585.

Poutanen, S. M., Low, D. E., Henry, B. & 18 other authors

(2003) . Identification of severe acute respiratory syndrome in

Canada. N Engl J Med 348, 1995-2005.

Rottier, P. J. M. (1995) . The coronavirus membrane glycoprotein, In The Coronaviridae, pp. 115-139. Eidted by S. G. Siddell. New

York: Plenum Press.

Ruan, Y., Wei, C. L., Ee, L. A. & 17 other authors (2003) .

Comparative full-length genome sequence analysis of 14 SARS coronavirus isolates and common mutations associated with putative origins of infection. Lancet 361, 1779-1785. Shi, Y., Wan, Z., Li, L., Li, P., Li, C, Ma, Q. & Cao, C. (2004) . Antibody responses against SARS-coronavirus and its nucleocaspid in SARS patients. Clin Virol 31, 66-68.

Siddell, S. G. (1995) . The small-membrane protein, In The

Coronaviridae, pp 181-189. Edited by S. G. Siddell. New York:

Plenum Press.

Subbarao, K., McAuliffe, J., Vogel, L., Fahle, G., Fischer, S., Tatti, K., Packard, M., Shieh, W. -J., Zaki, S. & Murphy, B.

(2004) . Prior infection and passive transfer of neutralizing antibody prevent replication of severe acute respiratory syndrome coronavirus in the respiratory tract of mice. J ^" Virol

78, 3572-3577. Summary of probable SAKS cases with onset of illness from 1

November 2002 to 31 July 2003 (2003) . WFfO, released on 26

September 2003.

Tan, Y. -J., Goh, P. -Y., Fielding, B. C. & 9 other authors

(2004a) . Profiles of antibody responses against severe acute respiratory syndrome coronavirus recombinant proteins and their potential use as diagnostic markers. Clin Digan Lab Immunol 11,

362-371.

Tan, Y. J., Teng, E., Shen, S., Tan, T. H., Goh, P. Y.,

Fielding, B. C, Ooi, E. E., Tan, H. C, Lim, S. G. & Hong, W. (2004b) . A novel severe acute respiratory syndrome coronavirus protein, U274, is transported to the cell surface and undergoes endocytosis. J Virol 78, 6723- 6734.

Vennema, H.; De Groot, R. J.; Harbour, D. A.; Dalderup, M.;

Gruffydd-Jones, T.; Horzinek, M. C; Spaan, W. J. M. (1990). Early death after feline infectious peritonitis virus challenge due to recombinant vaccinia virus immunization. J. Virol., 64,

1407-1409.

Wang, J., Wen, J., Li, J. & 30 other authors (2003) . Assessment of immunoreactive synthetic peptides from the structural proteins of severe acute respiratory syndrome coronavirus. Clin

Chem 49, 1989-1996.

Weiss, R. C; Dodds, W. J.; Scott, F. W. (1981) . Antibody- mediated enhancement of disease in feline infectious peritonitis: comparisons with dengue hemorrhagic fever. Comp. Immunol. Microbiol. Infect. Dis 4, 175-189.

Weiss, R. C; Scott, F. W. (1981) Pathogenesis of feline infectious peritonitis: nature and development of viremia. Am. J. Vet. Res 42, 382-390.

Wong, V. W.; Dai, D.; Wu, A. K.; Sung, J. J. (2003). Treatment of severe acute respiratory syndrome with convalescent plasma. Hong Kong Med. J 9, 199-201.

Yang, Z.-y. Kong, W. -p., Huang, Y., Roberts, A., Murphy, B. R., Subbarao, K. & Nabel, G.J. (2004) . A DNA vaccine induces SARS coronavirus neutralization and protective immunity in mice. Nature 428, 561-564.

Yang, Z.-y., Werner, H. C, Kong, W. -p., Leung, K., Traggiai, E., Lanzavecchia, A. & Nabel, G. J. (2005) . Evasion of antibody neutralization in emerging severe acute respiratory syndrome coronaviruses . Proc Natl Acad Sci USA 102, 797-780. Yeh, S. -H., Wang, H. -Y., Tsai, C-Y., Kao, C-L., Yang, J. -Y., Liu, H. -W., Su, I.-J., Tsai, S. -F., Chen, D. -S., Chen, P. -J. (2004) . Characterization of severe acute respiratory syndrome coronavirus genomes in Taiwan: Molecular epidemiology and genome evolution. Proc Natl Acad Sci USA 101, 2542-2547. Yu, C-J., Chen, Y. -C, Hsiao, C-H. & 9 other authors (2004) . Identification of a novel protein 3a from severe acute respiratory syndrome coronavirus. FEBS Lett 565, 111-116. Zeng, R., Yang, R. -F., Shi, M. -D. & 31 other authors (2004) . Characterization of the 3a protein of SARS-associated coronavirus in infected Vero E6 cells and SARS patients. J MoI Biol 341, 271-279.

Zhao, P., Ke, J. -S., Qin, Z. -L., Ren, H., Zhao, L. -J., Yu, J.G., Gao, J., Zhu, S. -Y. & Qi, Z. -T. (2004) . DNA vaccine of SARS-Cov S gene induces antibody response in mice. Acta Biochim et Biophysica Sinica 36, 37-41.

Zhong, X., Yang, H., Guo, Z. -F. & 13 other authors (2005) . B cell responses in patients who have recovered from severe acute respiratory syndrome target a dominant site in the S2 domain of the surface spike glycoprotein, J Virol 79, 3401-3408.

Zhou, x. z. ; Zhao, M.; Wang, F. S.; Jiang, T. J.; Li, Y. G.;

Nie, W. M.; Zhou, Z. P.; Wang, Y.; He, J.; Fan, R.; Zhao, J. M. ; Zhang, X. W. (2003) . Epidemiologic features, clinical diagnosis and therapy of first cluster of patients with severe acute respiratory syndrome in Beijing area. Zhonghua Yi Xue Za Zhi 83, 1018-1022.