Vaccine 23 (2005) 49594968SARS coronavirus spike polypeptide DNA vaccine priming withrecombinant spike polypeptide from Escherichia coli as boosterinduces high titer of neutralizing antibody against SARS coronavirusPatrick C.Y. Wooa,b,c, Susanna K.P. Laua,b,c, Hoi-wah Tsoia, Zhi-wei Chend,Beatrice H.L. Wonga, Linqi Zhangd, Jim K.H. Chana, Lei-po Wonga,Wei Hee, Chi Mae, Kwok-hung Chana,b,c, David D. Hod, Kwok-yung Yuena,b,c,aDepartment of Microbiology, The University of Hong Kong, Room 423, University Pathology Building, Queen Mary Hospital, Hong KongbResearch Centre of Infection and Immunology, Faculty of Medicine, Hong KongcState Key Laboratory for Emerging Infectious Diseases, The University of Hong Kong, Hong KongdAaron Diamond AIDS Research Center, NY, USAePeking Union Medical College, Beijing, ChinaReceived 30 December 2004; received in revised form 19 May 2005; accepted 29 May 2005Available online 13 June 2005AbstractDifferent forms of SARS coronavirus (SARS-CoV) spike protein-based vaccines for generation of neutralizing antibody response againstSARS-CoV were compared using a mouse model. High IgG levels were detected in mice immunized with intraperitoneal (i.p.) recombinantspike polypeptide generated by Escherichia coli (S-peptide), mice primed with intramuscular (i.m.) tPA-optimize800 DNA vaccine (tPA-S-DNA) and boosted with i.p. S-peptide, mice primed with i.m. CTLA4HingeSARS800 DNA vaccine (CTLA4-S-DNA) and boosted with i.p.S-peptide, mice primed with oral live-attenuated Salmonella typhimurium (Salmonella-S-DNA-control) and boosted with i.p. S-peptide, miceprimed with oral live-attenuated S. typhimurium that contained tPA-optimize800 DNA vaccine (Salmonella-tPA-S-DNA) and boosted withi.p. S-peptide, and mice primed with oral live-attenuated S. typhimurium that contained CTLA4HingeSARS800 DNA vaccine (Salmonella-tPA-S-DNA) and boosted with i.p. S-peptide. No statistical significant difference was observed among the Th1/Th2 index among these sixgroups of mice with high IgG levels. Sera of all six mice immunized with i.p. S-peptide, i.m. DNA vaccine control and oral Salmonella-S-DNA-control showed no neutralizing antibody against SARS-CoV. Sera of the mice immunized with i.m. tPA-S-DNA, i.m. CTLA4-S-DNA,oral Salmonella-S-DNA-control boosted with i.p. S-peptide, oral Salmonella-tPA-S-DNA, oral Salmonella-tPA-S-DNA boosted with i.p S-peptide,oralSalmonella-CTLA4-S-DNAandoralSalmonella-CTLA4-S-DNAboostedwithi.p.S-peptideshowedneutralizingantibodytitersof 1:201:160. Sera of all the mice immunized with i.m. tPA-S-DNA boosted with i.p. S-peptide and i.m. CTLA4-S-DNA boosted with i.p.S-peptide showed neutralizing antibody titers of 1:1280. The present observation may have major practical value, such as immunization ofcivetcats,sinceproductionofrecombinantproteinsfromE.coliisfarlessexpensivethanproductionofrecombinantproteinsusingeukaryoticsystems. 2005 Elsevier Ltd. All rights reserved.1. IntroductionSevereacuterespiratorysyndrome(SARS)hasaffected30countries in five continents with more than 8000 cases and750 deaths. A novel virus, the SARS coronavirus (SARS-Corresponding author. Tel.: +852 28554892; fax: +852 28551241.E-mail address: hkumicrohkucc.hku.hk (K. Yuen).CoV), has been confirmed to be the etiological agent 17.In addition, we have also reported the isolation of SARS-CoV-like viruses from Himalayan palm civets found in a liveanimal market in the Guangdong Province of China, whichimplied that animals could be the reservoir for the ancestorof SARS-CoV 8.Inanimalcoronavirusinfections,ithasbeenshownthatthespike proteins of coronaviruses were highly immunogenic,0264-410X/$ see front matter 2005 Elsevier Ltd. All rights reserved.doi:10.1016/j.vaccine.2005.05.0234960P.C.Y. Woo et al. / Vaccine 23 (2005) 49594968and immunization of animals using spike protein-basedvaccines were able to produce neutralizing antibodies thatwere effective in prevention of infections caused by the cor-responding coronaviruses. For SARS-CoV infection, it hasbeen shown that nucleotides 9521530 of the spike proteingene of SARS-CoV encoded a 193-amino acid fragmentresponsible for attaching to the receptor for SARS-CoV,angiotensin-converting enzyme 2 9. Furthermore, we, andalso others, have shown that patients with SARS producedantibody response against the spike protein of SARS-CoV3,10,11, and it has been demonstrated that the spike proteinis the major target for passive immunization 12,13. Instudiesthatdeterminetherelativeimportanceofhumoralandcell mediated immunity for protection against SARS-CoVinfection, it was confirmed that neutralizing antibody, whenadministered by passive immunization, was crucial in con-ferring protection 14, whereas T-cell immunity was unableto lead to protection 15. In addition, for vaccine candidatesagainst SARS-CoV, spike protein-based DNA vaccinesappeared to be a promising group of vaccine shown toproduce protective immunity against SARS-CoV infections,whereas recombinant spike protein vaccines produced byEscherichia coli were not efficient in terms of generation ofprotective immunity as compared to those generated fromeukaryotic systems such as transfection of cell lines 1425.However, multiple doses of intramuscularly (i.m.) adminis-tered DNA vaccine or recombinant protein generated fromthe eukaryotic systems are quite expensive, and thereforemay not be practical in developing countries. No data on lessexpensive modalities of immunization, such as DNA vaccinefollowed by boosters of recombinant vaccine producedby E. coli or oral mucosal DNA vaccines 2629, areavailable.In this study, we compared the different forms of SARS-CoV spike protein-based vaccines for generation of neu-tralizing antibody response against SARS-CoV using amousemodel.Therelativeeffectivenessofrecombinantspikepolypeptide vaccine produced by E. coli, two different typesof intramuscular spike polypeptide DNA vaccine with andwithout boosters of recombinant spike polypeptide vaccineproduced by E. coli and two different types of oral mucosalspikepolypeptideDNAvaccinewithandwithoutboostersofrecombinant spike polypeptide vaccine produced by E. coliare compared.2. Materials and methods2.1. AnimalsMale Balb/c (H-2d) mice (68 weeks old, 1822g) wereused in all animal experiments. They were housed in cages,under standard conditions with regulated day length, tem-perature and humidity, and were given pelleted food and tapwater ad libitum.2.2. Recombinant SARS-CoV spike polypeptide vaccinefrom E. coliCloning and purification of the spike polypeptide ofSARS-CoV was reported previously 3. Briefly, to producea plasmid for protein expression, primers (LPW742 5?-CGC-GGATCCGAGTGACCTTGACCGGTGC-3?and LPW9315?-CGGGGTACCTTAACGTAATAAAGAAACTGTATG-3?) were used to amplify the gene encoding amino acidresidues 14667 of the spike protein of the SARS-CoVby RT-PCR. This portion of the spike protein was usedbecause it contains the receptor-binding domain withinthe S1 domain that is highly immunogenic, whereas thecomplete spike protein was not expressible in E. coli.The PCR product was cloned into the BamHI and KpnIsites of vector pQE-31 (Qiagen, Hilden, Germany). Theresultant clone was digested by PstI, and the PstI fragmentwhich contained the gene encoding amino acid residues250667 of the spike protein was cloned into expressionvector pQE-30 (Qiagen, Hilden, Germany) in frame anddownstream of the series of six histidine residues. The(His)6-tagged recombinant spike polypeptide (S-peptide)was expressed and purified using the Ni2+-loaded HiTrapChelating System (Amersham Pharmacia, USA) accordingto the manufacturers instructions.2.3. Human codon usage optimized SARS-CoV DNAvaccinesTo enhance the expression of spike polypeptide inhuman cells, the two SARS-CoV DNA vaccines, tPA-optimize800(tPA-S-DNA)andCTLA4HingeSARS800(CTLA4-S-DNA), were constructed using the concept ofhuman codon usage optimization 30 with QUIKChangeMulti Site-Directed Mutagenesis Kit (Strategene, USA)according to manufacturers instructions. The syntheticpolypeptides were cloned into pcDNA3.1(+).2.4. Oral mucosal tPA-optimize800 andCTLA4HingeSARS800 DNA vaccinesTheoralmucosaltPA-optimize800andCTLA4HingeSARS800 DNA vaccines (Salmonella-tPA-S-DNA and Salmonella-CTLA4-S-DNA) were preparedaccording to our published protocol 26,29. tPA-S-DNAand CTLA4-S-DNA were transformed into auxotrophic S.typhimurium aroA strain SL7207 (S. typhimurium 2337-65derivative hisG46, DEL 407 aroA:Tn10Tc-s, a giftfrom Dr Bruce Stocker) 31 by electroporation.2.5. Transfection of 293 cells with tPA-optimize800 andCTLA4HingeSARS800Transfection of 293 cells with tPA-S-DNA and CTLA4-S-DNA was performed according to our published pro-tocol 26,29. Two hundred and ninety-three cells wereP.C.Y. Woo et al. / Vaccine 23 (2005) 495949684961plated at 1107cells per well in Dulbeccos modifiedEagles medium (GibcoBRL, USA) with 10% fetal calfserum (FCS) in a six-well plate on the day before transfec-tion. On the day of transfection, each well was transfectedwith1?gplasmidencodingeukaryoticallyexpressedSARS-CoV spike polypeptide (tPA-S-DNA or CTLA4-S-DNA) orpcDNA3.1(+) (S-DNA-control) with FuGENE 6 Reagent(Boehringer Mannhein, Germany) according to manufac-turers instructions. Forty-eight hours after transfection, thecells were harvested and lysed by freezing and thawing threetimes. After centrifugation at 14000rpm, the supernatantwas used for the detection of SARS-CoV spike polypeptideby Western blot assay using pre-immune rabbit serum andhyperimmune polyclonal serum from rabbit immunized withS-peptide.2.6. Western blot analysisWestern blot analysis was performed according to ourpublished protocol 29. Briefly, 10?l of supernatant of293 cell lysates obtained from 293 cells transfected withtPA-S-DNA, CTLA4-S-DNA or S-DNA-control was loadedinto each well of a sodium dodecyl sulfate (SDS)8%polyacrylamide gel and subsequently electroblotted onto anitrocellulose membrane (Bio-Rad, Hercules, CA, USA).The blot was incubated separately with 1:1000 dilution ofpre-immune rabbit serum or hyperimmune polyclonal serumfrom rabbit immunized with S-peptide. AntigenantibodyinteractionwasdetectedwithanECLfluorescencesystem(AmershamLifeScience,Buckinghamshire,UK).2.7. Immunization scheduleSeventy-two Balb/c mice were used for the immunizationexperiments. The immunization schedule is summarized inTable 1. On days 0, 14 and 28, six mice were immunizedintraperitoneally (i.p.) with S-peptide 0.5?g per mouse(Group 1, Table 1). On day 0, six mice were immunizedi.m. (tibialis anterior muscle) with S-DNA-control 100?gper mouse (Group 2, Table 1) and 12 mice each wereimmunized i.m. with tPA-S-DNA 100?g per mouse (Group3, Table 1) or CTLA4-S-DNA 100?g per mouse (Group5, Table 1). On days 28 and 42, 6 of the 12 mice in thetwo DNA vaccine groups were boosted with i.p. S-peptide0.5?g per mouse (Groups 4 and 6, Table 1). On day 0, 12mice each were immunized orally with S. typhimurium aroAstrain (Salmonella-S-DNA-control) 6109bacterial cellsper mouse (Group 7, Table 1), Salmonella-tPA-S-DNA6109bacterial cells per mouse (Group 9, Table 1)orSalmonella-CTLA4-S-DNA6109bacterialcellsper mouse (Group 11, Table 1). On days 28 and 42, 6of the 12 mice in the three groups were boosted withi.p. S-peptide 0.5?g per mouse (Groups 8, 10 and 12,Table 1).Table 1Immunization schedule for different forms of spike polypeptide-based vaccines against SARS-CoVGroupsFirst dose (day 0)Second doseThird doseVaccinesRoutes ofadministrationDose permouseVaccinesRoutes (days) ofadministrationDose permouseVaccinesRoutes (days) ofadministrationDose permouse1Spike polypeptideIntraperitoneal50?gSpike polypeptideIntraperitoneal (14)50?gSpike polypeptideIntraperitoneal (28)50?g2pcDNA3.1(+)Intramuscular100?g3tPA-optimize800 DNA vaccineIntramuscular100?g4tPA-optimize800 DNA vaccineIntramuscular100?gSpike polypeptideIntraperitoneal (28)50?gSpike polypeptideIntraperitoneal (42)50?g5CTLA4HingeSA RS800 DNAvaccineIntramuscular100?g6CTLA4HingeSA RS800 DNAvaccineIntramuscular100?gSpike polypeptideIntraperitoneal (28)50?gSpike polypeptideIntraperitoneal (42)50?g7S. typhimurium aroA strainOral6109bacterial cells8S. typhimurium aroA strainOral6109bacterial cellsSpike polypeptideIntraperitoneal (28)50?gSpike polypeptideIntraperitoneal (42)50?g9Mucosal tPA-optimize800 DNAvaccineOral6109bacterial cells10Mucosal tPA-optimize800 DNAvaccineOral6109bacterial cellsSpike polypeptideIntraperitoneal (28)50?gSpike polypeptideIntraperitoneal (42)50?g11Mucosal CTLA4HingeSARS800DNA vaccineOral6109bacterial cells12Mucosal CTLA4HingeSA RS800DNA vaccineOral6109bacterial cellsSpike polypeptideIntraperitoneal (28)50?gSpike polypeptideIntraperitoneal (42)50?g4962P.C.Y. Woo et al. / Vaccine 23 (2005) 495949682.8. Measurement of serum antibodies againstSARS-CoV spike polypeptideMice from each group were bled on the day before immu-nization and 42 days after the last dose of vaccine in thecorresponding group. The blood was centrifuged at 2700gfor 20min and the supernatant (serum) was stored at 70Cbefore antibody measurement.Antibodies against SARS-CoV spike polypeptide weremeasured using the enzyme-linked immunosorbent assay(ELISA) according to our published protocol with modifi-cations 3,4. Mouse sera (diluted with PBS-2% skim milk,1:10 for IgM, 1:80 for IgG, 1:1280 for IgG1, 1:40 for IgG2a,1:10 for IgG2b and 1:320 for IgG3) were added to ELISAplatesprecoatedwithS-peptide(80ngperwellforIgM,IgG,IgG2a, IgG2b and IgG3 and 10ng per well for IgG1). Theplates were incubated at 37C for 1h. After washing withwashingbufferfivetimes,100?lperoxidase-conjugatedgoatanti-mouse IgM and IgG, rabbit anti-mouse IgG1, IgG2aand IgG3 and rat anti-mouse IgG2b antibody (Zymed Lab-oratories Inc., USA) diluted according to manufacturersinstructions using PBS-2% skim milk were added to the cor-responding wells accordingly and incubated at 37C for 1h.IgM and total IgG levels were assayed to assess the primaryand secondary immune response, while the IgG subtypeswere used to determine whether the humoral response wasinclined towards the Th1 (IgG2a and IgG2b) or Th2 (IgG1and IgG3) pattern. After washing with washing buffer fivetimes, 100?l diluted 3,3?,5,5?-tetramethylbenzidine (ZymedLaboratories Inc.) was added to each well and incubated atroom temperature for 15min. One hundred microliters of0.3M H2SO4was added and the absorbance at 450nm ofeach well was measured. Each sample was tested in dupli-cateandthemeanabsorbanceforeachserumwascalculated.Theserumantibodylevelofaparticularmousewasdefinedasthe absorbance obtained from the serum taken 42 days afterthe last dose of the vaccine minus that of the correspond-ing mouse taken the day before immunization. The Th1/Th2index of each mouse is calculated by the following formula:IgG2aIgG2b/IgG1IgG3.2.9. Neutralizing antibody assayThe neutralizing antibody assay was modified from ourpublished protocol 32. All work with infectious virus wasperformed inside a type II Biosafety Cabinet, in a BiosafetyContainment level III facility, and the personnel wore pow-ered air-purifying respirators (HEPA Airmate, 3M, St. Paul,MN). Initial screening of mouse sera against the proto-type SARS-CoV strain no. 39849 was performed in 96-wellmicrotiterplatesseededwithfetalrhesuskidney-4cells.Two-fold dilutions of mouse sera (from 1:20 to 1:1280) weretestedinduplicateagainst100TCID50ofSARS-CoV.Acor-responding set of cell controls with sera but without virusinoculation was used as controls. The cells were scored forthe inhibition of the cytopathic effect (CPE) at 48h. The titerof neutralizing antibody is defined as the maximum dilutionof serum at which the percentage of CPE is less than or equalto 50%.2.10. Measurement of lymphocyte proliferation index(LPI)LPI was measured according to our published protocol29. On day 60, single-cell suspensions of spleen cells fromthe six mice of each group were depleted of erythrocytesby adding freshly prepared Geys solution. The cells wereresuspended in RPMI 1640 medium (Gibco BRL, Rockville,MD)supplementedwith15%fetalcalfserumandinoculatedinto microwell plates at 5105cells per well in triplicate.Cells were stimulated with phytohaemagglutinin at 5?gper well (positive control), S-peptide at 0.1?g per wellor RPMI medium (negative control). Cells were culturedat 37C 5% CO2for 3 days, and3H-labelled thymidine(Amersham Pharmacia, Little Chalfont, UK) was added at1?Ci per well for the last 18h. Cells were harvested ontoglass microfibre filter (Whatman International Ltd., UK)using a Model CH1 cell harvester (Insel, Hampshire, UK)and radioactivity was measured by a liquid scintillationcounter (Beckman, Fullerton, CA). The S-peptide-specificLPI of a particular sample is defined as the ratio of thedifference of radioactivity between the sample and thenegative control and that between the positive and negativecontrols.2.11. Interleukin-4 (IL-4) and interferon- (IFN-)assaysIL-4 and IFN-? were assayed according to our publishedprotocol 29. On day 60, spleens from the six mice in eachgroupwereharvested.Single-cellsuspensionswerepreparedand cells from mice within the same group were pooled.2106cells were cultured in 1ml RPMI 1640 mediumsupplemented with 10% fetal calf serum and 5105M 2-mercaptoethanol in 24-well plates. S-peptide was added atfinal concentrations of 0.5, 1.0 and 2.5?g/ml. Supernatant(200?l)fromeachsamplewascollectedat24,48and72hforcytokine measurement. Monoclonal antibodies against IL-4or IFN-? were coated onto wells in 96-well microtiter plates(OptEIA, PharM3ingen, Becton Dickinson, USA) at 1:250dilutionsaccordingtomanufacturersinstructions.Theplateswere incubated at RT for 24h. After washing with wash-ing buffer three times, the plates were blocked with assaydiluent at RT for 1h. After washing with washing bufferthree times, 100?l of supernatant from each sample wasadded to the wells in duplicate. The plates were incubatedat RT for 2h. After washing with washing buffer five times,100?l diluted biotinylated antibody against IL-4 or IFN-?and avidin-horseradish peroxidase conjugate were added tothe wells and incubated at RT for 1h. After washing withwashingbuffereighttimes,100?lTMBsubstratewasaddedto each well and incubated at RT for 30min. Hundred micro-P.C.Y. Woo et al. / Vaccine 23 (2005) 495949684963liters of 0.3M H2SO4was added and the absorbance ofeach well was measured at 450nm, using TMB buffer asa blank. Each pooled sample was tested in triplicate andthe mean absorbance for each pooled sample was calcu-lated.2.12. Statistical analysisComparison was made among the serum antibody levelsand LPI of the various groups of mice using one-wayANOVA. P0.05 was regarded as statistically signi-ficant.3. Results3.1. SARS-CoV spike polypeptide expression in 293 cellstransfected with tPA-optimize800 andCTLA4HingeSARS800The supernatant of 293 cell lysates obtained from 293cells transfected with tPA-S-DNA, CTLA4-S-DNA or S-DNA-control were separated on SDSpolyacrylamide gelsfollowedbyWesternblotanalysiswithserafrompre-immunerabbit serum or hyperimmune polyclonal serum from rab-bit immunized with S-peptide. Prominent immunoreactiveprotein bands of about 90 and 110kDa were visible on theWestern blot that used cell lysates obtained from 293 cellstransfected with tPA-S-DNA and CTLA4-S-DNA, respec-tively, as the antigen and hyperimmune polyclonal serumfrom rabbit immunized with S-peptide as the source of anti-body(Fig.1,lanes1and2).Thesesizeswereconsistentwiththeexpectedsizeof91.4and108.1kDaforthecorrespondingspike polypeptides.3.2. Antibody responseThe antibody levels of the 12 groups of mice on day 42were summarized in Fig. 2. No IgG was detected in miceof Groups 2, 3, 5, 7, 9 and 11, whereas high IgG levelswere detected in mice of Groups 1, 4, 6, 8, 10 and 12. Nostatistical significant difference was observed among theTh1/Th2indexamongthesesixgroupsofmicewithhighIgGlevels.3.3. Neutralizing antibody assayThe number of mice with different neutralizing anti-body titers immunized with different forms of spikepolypeptide-based vaccines against SARS-CoV was shownin Table 2. Sera of all the six mice immunized withi.p. S-peptide, i.m. S-DNA-control and oral Salmonella-S-DN A-control (Groups, 1, 2 and 7) showed no neu-tralizing antibody against SARS-CoV. Sera of the miceimmunized with i.m. tPA-S-DNA, i.m. CTLA4-S-DNA,oralSalmone-lla-S-DNA-controlboostedwithi.p.S-peptide,Fig. 1. Prominent immunoreactive protein bands of about 90 and 110kDawerevisibleontheWesternblot(lanes1and2),indicatingantigenantibodyinteractions between the 293 cell lysates obtained from 293 cells transfectedwithtPA-optimize800andCTLA4HingeSARS800,respectively,andhyper-immunepolyclonalserumfromrabbitimmunizedwith(His)6-taggedrecom-binant spike polypeptide. No antigenantibody interactions were observedbetween the 293 cell lysates obtained from 293 cells transfected with tPA-optimize800 or CTLA4HingeSARS800 and the pre-immune rabbit serum(lanes 3 and 4).oral Salmonella-tPA-S-DNA, oral Salmonella-tPA-S-DNAboosted with i.p S-peptide, oral Salmonella-CTLA4-S-DNAand oral Salmonella-CTLA4-S-DNA boosted with i.p. S-peptide(Groups,3,5and812)showedneutralizingantibodytiters of 6000pg/ml (data not shown).4964P.C.Y. Woo et al. / Vaccine 23 (2005) 49594968Fig. 2. Serum antibody levels (O.D. 450) at day 42 in the 12 groups of Balb/c mice immunized with the various vaccines. The 12 groups correspond to the 12groups of mice described in Table 1 (bar=average of six mice, error bar=1 standard deviation).4. DiscussionAmong all the combinations of vaccines examined in thisstudy, mice primed with SARS-CoV human codon usageoptimizedspikepolypeptideDNAvaccinesandboostedwithS-peptide produced by E. coli generated the highest titerof neutralizing antibody against SARS-CoV. It has beenobserved,andisconfirmedinthepresentstudy,thatS-peptideproduced by E. coli did not induce neutralizing antibodyagainstSARS-CoVinfection(Table2,Group1).Ontheotherhand,recombinantspikepolypeptidegeneratedbyeukaryoticsystems such as transfection of COS7 and BHK21 cells orDNA vaccine was able to elicit high neutralizing antibodytiteragainstSARS-CoVinfection15,21,24.Thiswasprob-ably because when S-peptide produced by E. coli was used,the three dimensional folding and/or the glycosylation of theS-peptide was not optimal for the generation of neutraliz-ing antibodies. In this study, we documented that althoughrecombinant S-peptide produced by E. coli itself was notable to generate neutralizing antibody against SARS-CoVP.C.Y. Woo et al. / Vaccine 23 (2005) 495949684965Table 2Neutralizing antibody titers for different forms of spike polypeptide-based vaccines against SARS-CoVGroupsNeutralizing antibody titers (no. of mice)1:201:201:401:801:1601:3201:6401:12801 (S-peptide)600000002 (S-DNA-control)600000003 (tPA-S-DNA)020040004 (tPA-S-DNA boosted with S-peptide)000000065 (CTLA4-S-DNA)042000006 (CTLA4-S-DNA boosted with S-peptide)000000067 (Salmonella-S-DNA-control)600000008 (Salmonella-S-DNA-control boosted with S-peptide)020400009 (Salmonella-tPA-S-DNA)4200000010 (Salmonella-tPA-S-DNA boosted with S-peptide)2121000011 (Salmonella-CTLA4-S-DNA)5100000012 (Salmonella-CTLA4-S-DNA boosted with S-peptide)22101000infection, mice primed with spike polypeptide DNA vac-cine and boosted with S-peptide from E. coli were able togenerate high titer of neutralizing antibody against SARS-CoV (Table 2, Groups 4 and 6). This indicates that the typeof vaccine used for priming is crucial in determining thetype of immune response developed. Subsequent doses willbooster the immune response generated by the first dose ofvaccine. Of note is that the humoral immune response devel-opedinmiceprimedwithspikepolypeptideDNAvaccineandboostedwithS-peptidefromE.coliwasnotparticularlyoftheTh1 type as compared to that developed in mice immunizedwith S-peptide from E. coli alone. This indicates that a Th1typeimmuneresponsemaynotbeessentialforthegenerationof neutralizing antibodies against SARS-CoV. Although ourresults suggest that priming with DNA vaccines and boost-ing with S-peptide produced by E. coli was successful inthe generation of neutralizing antibody against SARS-CoV,furtherexperimentsusinginfectionmodelstoevaluateitspro-tective immunity are warranted, since anti-spike antibodieshave been shown to enhance the infectivity of coronavirusesin some cell culture systems, as occurred with SARS-CoVand feline infectious peritonitis virus 33,34.The present observation may have major practical value,such as immunization of civet cats, as production of recom-binant proteins from E. coli is far less expensive than pro-duction of recombinant proteins using eukaryotic systemssuch as transfection of cell lines or DNA vaccines. Althoughit has been shown that DNA vaccines are able to generateboth humoral and cellular immunity successfully for variouspathogensinmice,oneofthemajorlimitationsforitsclinicaluse is its ineffectiveness when it is used in humans, unlessa large amount of DNA is used for immunization 35,36.As for the production of recombinant spike polypeptide gen-erated by eukaryotic systems such as transfection of COS7and BHK21 cells 24 or using the baculovirus system 25,although the conformation and/or glycosylation of the spikepolypeptide produced can theoretically be more similar tothe native viral spike protein, it is not easy to scale up theproduction of such recombinant proteins to industrial lev-els. In contrast to recombinant spike polypeptide generatedFig. 3. SARS-CoV spike polypeptide-specific lymphocyte proliferation index of Balb/c mice immunized with the various vaccines. The 12 groups correspondto the 12 groups of mice described in Table 1 (bar=average of six mice, error bar=1 standard deviation).4966P.C.Y. Woo et al. / Vaccine 23 (2005) 49594968Fig. 4. IL-4 (at 72h) and Interferon-? (at 24h) levels of splenic cell culture supernatant in Balb/c mice immunized with the various vaccines. The 12 groupscorrespond to the 12 groups of mice described in Table 1.by eukaryotic systems, a large amount of S-peptide can beproduced by E. coli in a relatively inexpensive way, andsuch S-peptide can be used successfully as boosters. Fur-ther studies on the effectiveness of this mode of vaccinationfor generation of protective immunity against SARS-CoV inother animals could be performed. This principle can also beexamined in vaccination for other pathogens, where “moreeffective” modalities of vaccination, such as DNA vaccine,can be used for priming, and the “less expensive” recom-binant protein produced by E. coli, instead of eukaryoticsystems, can be used as boosters.Spike polypeptide DNA vaccines delivered by the live-attenuated Salmonella system did not induce good neu-tralizing antibody against SARS-CoV infection. We havepreviously shown that hepatitis B virus DNA vaccine pre-sented by the live-attenuated Salmonella system generatedgood cytotoxic T lymphocyte response, but minimal anti-body response, against hepatitis B virus in a mouse model26,27. Furthermore, we found that this immune responsewas able to down-regulate transgene expression in hepatitisB virus surface antigen transgenic mice 28. Subsequently,we reported a comparison of the efficacy of DNA vaccine,DNA vaccine delivered by the live-attenuated Salmonellasystem and recombinant protein vaccine for generation ofprotective immune response against Penicillium marneffei,a thermal dimorphic fungus infecting 10% of HIV positivepatients in China and Southeast Asia, in a mouse model29. Results showed that, similar to hepatitis B virus DNAvaccine presented by the live-attenuated Salmonella system,P. marneffei DNA vaccine delivered by the live-attenuatedSalmonella system did not generate good antibody response,whereas intramuscular DNA vaccine generated the best pro-tectiveimmunityagainstP.marneffeiinfection,implyingthatbothcellularandhumoralimmuneresponseareimportantforprotection against P. marneffei infection 29. In the presentstudy,itwasobservedthat,inlinewiththeresultsofhepatitisBvirusDNAvaccineandP.marneffeiDNAvaccinedeliveredby the live-attenuated Salmonella system, spike polypeptideDNA vaccines delivered by the live-attenuated Salmonellasystem did not induce good antibody response (Fig. 2 andTable2,Groups9and11).Althoughthemicedevelopedhighantibody levels against the spike protein after boosting withP.C.Y. Woo et al. / Vaccine 23 (2005) 495949684967S-peptide(Fig.2,Groups10and12),theantibodieswerenotneutralizing in our cell culture system (Table 2, Groups 10and 12). This may be due to the ineffectiveness of the DNAvaccine delivered by the live-attenuated Salmonella systemin priming the development of neutralizing antibodies in thecorrect configuration, while the “non-neutralizing” antibod-ies against the spike protein were only elicited in response tothe subsequent recombinant S-peptide.AcknowledgementsThis work was partly supported by the Research Fund forthe Control of Infectious Diseases of the Health, Welfare andFoodBureauoftheHongKongSARGovernment01030282,SARS Research Fund, University SARS Donation Fund andthe Research Grant Council Grant.References1 Peiris JS, Lai ST, Poon LL, Guan Y, Yam LY, Lim W, et al. Coro-navirus as a possible cause of severe acute respiratory syndrome.Lancet 2003;361(9366):131925.2 Peiris JS, Chu CM, Cheng VC, Chan KS, Hung IF, Poon LL, etal. Clinical progression and viral load in a community outbreakof coronavirus-associated SARS pneumonia: a prospective study.Lancet 2003;361(9371):176772.3 Woo PC, Lau SK, Tsoi HW, Chan KH, Wong BH, Che XY, et al.Relative rates of non-pneumonic SARS coronavirus infection andSARS coronavirus pneumonia. Lancet 2004;363(9412):8415.4 Woo PC, Lau SK, Wong BH, Tsoi HW, Fung AM, Chan KH, et al.Detection of specific antibodies to severe acute respiratory syndrome(SARS) coronavirus nucleocapsid protein for serodiagnosis of SARScoronavirus pneumonia. J Clin Microbiol 2004;42(5):23069.5 Lau SK, Woo PC, Wong BH, Woo GK, Poon RW, Tsoi HW, et al.Detection of severe acute respiratory syndrome (SARS) coronavirusnucleoapsid protein in SARS patients by enzyme-linked immunosor-bent assay. J Clin Microbiol 2004;42(7):28849.6 Woo PC, Lau SK, Wong BH, Chan KH, Chu CM, Tsoi HW, etal. Longitudinal profile of immunoglobulin G (IgG), IgM, and IgAantibodies against the severe acute respiratory syndrome (SARS)coronavirus nucleocapsid protein in patients with pneumonia due tothe SARS coronavirus. Clin Diagn Lab Immunol 2004;11(4):6658.7 Woo PC, Lau SK, Wong BH, Chan KH, Hui WT, Kwan GS,et al. False-positive resultss in recombinant severe acute respi-ratory syndrome-associated coronavirus (SARS-CoV) nucleocapsidenzyme-linked immunosorbent assay due to HCoV-OC43 and HCoV-229E rectified by Western blotting with recombinant SARS-CoVspike polypeptide. J Clin Microbiol 2004;42(12):58858.8 Guan Y, Zheng BJ, He YQ, Liu XL, Zhuang ZX, CheungCL, et al. Isolation and characterization of viruses related tothe SARS coronavirus from animals in southern China. Science2003;302(5643):2768.9 Wong SK, Li W, Moore MJ, Choe H, Farzan M. A 193-amino acid fragment of the SARS coronavirus S protein effi-ciently binds angiotensin-converting enzyme 2. J Biol Chem2004;279(5):3197201.10 Huang LR, Chiu CM, Yeh SH, Huang WH, Hsueh PR, Yang WZ,et al. Evaluation of antibody responses against SARS coronaviralnucleocapsid or spike proteins by immunoblotting of ELISA. J MedVirol 2004;73(3):33846.11 Hsueh PR, Kao CL, Lee CN, Chen LK, Ho MS, Sia C, etal. SARS antibody test for serosurveillance. Emerg Infect Dis2004;10(9):155862.12 Traggiai E, Becker S, Subbarao K, Kolesnikova L, Uematsu Y, Gis-mondo MR, et al. An efficient method to make human monoclonalantibodies from memory B cells: potent neutralization of SARS coro-navirus. Nat Med 2004;10(8):8715.13 ter Meulen J, Bakker AB, van den Brink EN, Weverling GJ,Martina BE, Haagmans BL, et al. Human monoclonal antibodyas prophylaxis for SARS coronavirus infection in ferrets. Lancet2004;363(9427):213941.14 Bisht H, Roberts A, Vogel L, Bukreyev A, Collins PL, Murphy BR,et al. Severe acute respiratory syndrome coronavirus spike proteinexpressed by attenuated vaccinia virus protectively immunizes mice.Proc Natl Acad Sci USA 2004;101(17):66416.15 Yang ZY, Kong WP, Huang Y, Roberts A, Murphy BR, SubbaraoK, et al. A DNA vaccine induces SARS coronavirus neutralizationand protective immunity in mice. Nature 2004;428(6982):5614.16 Buchholz UJ, Bukreyev A, Yang L, Lamirande EW, Murphy BR,Subbarao K, et al. Contributions of the structural proteins of severeacute respiratory syndrome coronavirus to protective immunity. ProcNatl Acad Sci USA 2004;101(26):98049.17 Bukreyev A, Lamirande EW, Buchholz UJ, Vogel LN, Elkins WR,St Claire M, et al. Mucosal immunisation of African green mon-keys (Cercopithecus aethiops) with an attenuated parainfluenza virusexpressing the SARS coronavirus spike protein for the prevention ofSARS. Lancet 2004;363(9427):21227.18 Gao W, Tamin A, Sol off A, DAiuto L, Nwanegbo E, Robbins PD,et al. Effects of a SARS-associated coronavirus vaccine in monkeys.Lancet 2003;362(9399):18956.19 Kim TW, Lee JH, Hung CF, Peng S, Roden R, Wang MC, et al.Generation and characterization of DNA vaccines targeting the nucle-ocapsid protein of severe acute respiratory syndrome coronavirus. JVirol 2004;78(9):463845.20 Tang L, Zhu Q, Qin E, Yu M, Ding Z, Shi H, et al. Inacti-vated SARS-CoV vaccine prepared from whole virus induces a highlevel of neutralizing antibodies in BALB/c mice. DNA Cell Biol2004;23(6):3914.21 Zeng F, Chow KY, Hon CC, Law KM, Yip CW, Chan KH, et al.Characterization of humoral responses in mice immunized with plas-mid DNAs encoding SARS- CoV spike gene fragments. BiochemBiophys Res Commun 2004;315(4):11349.22 Zhu MS, Pan Y, Chen HQ, Shen Y, Wang XC, Sun YJ, et al. Induc-tion of SARS- nucleoprotein-specific immune response by use ofDNA vaccine. Immunol Lett 2004;92(3):23743.23 Zhang H, Wang G, Li J, Nie Y, Shi X, Lian G, et al. Identifica-tion of an antigenic determinant on the S2 domain of the severeacute respiratory syndrome coronavirus spike glycoprotein capableof inducing neutralizing antibodies. J Virol 2004;78(13):693845.24 Song HC, Seo MY, Stadler K, Yoo BJ, Choo QL, Coates SR, etal. Synthesis and characterization of a native, oligomeric form ofrecombinant severe acute respiratory syndrome coronavirus spikeglycoprotein. J Virol 2004;78(19):1032835.25 Mortola E, Roy P. Efficient assembly release of SARS coronavirus-like particles by a heterologous expression system. FEBS Lett2004;576(12):1748.26 Woo PC, Wong LP, Zheng BJ, Yuen KY. Unique immunogenic-ity of hepatitis B virus DNA vaccine presented by live-attenuatedSalmonella typhimurium. Vaccine 2001;19(2022):294554.27 Zheng BJ, Woo PCY, Ng MH, Tsoi HW, Wong LP, Yuen KY. Acrucial role of macrophages in the immune responses to oral DNAvaccination against hepatitis B virus in a murine model. Vaccine2001;20(12):1407.28 Zheng BJ, Ng MH, Chan KW, Tam S, Woo PCY, Ng SP, et al.A single dose of oral DNA immunization delivered by attenu-ated Salmonella typhimurium down-regulates transgenic expressionin HBsAg transgenic mice. Eur J Immunol 2002;32(11):3294304.4968P.C.Y. Woo et al. / Vaccine 23 (2005) 4959496829 Wong LP, Woo PC, Wu AY, Yuen KY. DNA immunizationusing a secreted cell wall antigen Mp1p is protective againstPenicilliummarneffeiinfection.Vaccine2002;20(2324):287886.30 Deml L, Bojak A, Steck S, Graf M, Wild J, Schirmbeck R,et al. Multiple effects of codon usage optimization on expres-sion and immunogenicity of DNA candidate vaccines encodingthe human immunodeficiency virus type 1 Gag protein. J Virol2001;75(22):109911001.31 HoisethSK,StockerBA.Aromatic-dependentSalmonellatyphimurium are non-virulent and effective as live vaccines. Nature1981;291(5812):2389.32 Chen F, Chan KH, Jiang Y, Kao RY, Lu HT, Fan KW, et al. In vitrosusceptibility of 10 clinical isolates of SARS coronavirus to selectedantiviral compounds. J Clin Virol 2004;31(1):6975.33 Olsen CW, Corapi WV, Ngichabe CK, Baines JD, Scott FW. Mono-clonal antibodies to the spike protein of feline infectious peritonitisvirus mediate antibody-dependent enhancement of infection of felinemacrophages. J Virol 1992;66(2):95665.34 Yang ZY, Werner HC, Kong WP, Leung K, Traggiai E, Lanzavec-chia A, et al. Evasion of antibody neutralization in emerging severeacute respiratory syndrome coronaviruses. Proc Natl Acad Sci USA2005;102(3):797801.35 Manoj S, Babiuk LA, van Drunen Littel-van den Hurk S. Approachesto enhance the efficacy of DNA vaccines. Crit Rev Clin Lab Sci2004;41(1):139.36 Dupuis M, Denis-Mize K, Woo C, Goldbeck C, Selby MJ,Chen M, et al. Distribution of DNA vaccines determines theirimmunogenicity after intramuscular injection in mice. J Immunol2000;165(5):28508.