VIROLOGY 131, 296-307 (1983)

Antigenic Relationships of Murine Coronaviruses : Analysis UsingMonoclonal Antibodies to JHM (MHV-4) Virus

JOHN O. FLEMING,' STEPHEN A . STOHLMAN, RICHARD C . IIARMON,2MICHAEL M. C. LAI, JEFFREY A . FRELINGER,2 AND LESLIE P. WEINER

Depar&me us of Neurology and Microbiology, University of Southern CaliforniaSchool of Medicine, 2025 Lowal Avenue, Los Angeles, California 90033

Received April 18, 1983; accepted August 15, 1988

Monoclonal antibodies were produced to .IHMV-DL, a neurotropic member of the mousehepatitis virus (MHV) or marine coronavirus group . Of 23 antibodies isolated, 10 werespecific for the major envelope glycoprotein, gplSO/90, 10 for the nucleocapsid protein,pp6O, and 3 for the minor envelope glycoprotein, gp25 . Eleven different MHV isolateswere used in antibody binding assays to study antigenic relationships among the viruses .Each MIIV isolate tested had a unique pattern of antibody binding, indicating that eachis a distinct strain . Conservation of JHMV-DL antigenic determinants varied among thethree proteins, with pp60 showing intermediate conservation, gpls0/90 little conservation,and gp2.5 marked conservation in the different MHV strains . Monoclonal antibodies topp60 proved most useful in delineating antigenic relationships among MHV strains. Theseantigenic groups correlated with pathogenic types, indicating that pp60 may be one ofthe gene products which mediates the distinct disease patterns manifested by differentmurine coronaviruses .

INTRODUCTION

Coronaviruses are enveloped virusescontaining a single-stranded RNA genomeof positive polarity (ter Meulen et at, 1982 ;Sturman and Holmes, 1983) . The murinecoronavirus or mouse hepatitis virus(MHV) group is of particular interest be-cause of the wide range of diseases pro-duced by its members. Although the ma-jority of MHV strains usually cause in-apparent or latent gastrointestinalinfections in nature (Gledhill and Niven,1955; Rowe et at, 1963), different MHVshave been reported to spontaneously or ex-perimentally produce hepatitis, enteritis,peritonitis, lower respiratory infections,panencephalitis, demyelination, choroi-

' To whom correspondence and reprint requestsshould he addressed .' Current address : Microbiology Research Depart-ment, Cutter Laboratories, Division of Miles Labo-ratories, Tue.., Berkeley, Calif. 94710

' Current address : Department of Bacteriology andImmunology, University of North Carolina, ChapelHill, N. C. 27514

0042-6822/83 $3.00Copyright C 19W by Academic Press, Inc .All lights of reproduction in any form roevved

296

doependymitis, meningitis, myeloprolifer-ation, and wasting disease in susceptiblehosts (Virelizier et aL, 1975; Ishida et al.,1978b; Hierholzer et at, 1979; Wege et at,1992) .

Despite marked variations in pathoge-nicity, the MHVs have relatively simplestructural features, usually consisting ofa major envelope glycoprotein, a minor en-velope glycoprotein, and a nucleocapsidprotein. The major glycoprotein, gp180/90,forms the projecting spikes or peplomers .These are important for cell attachment(Sturman and Holmes, 1983) and may in-fluence tropism for specific tissues . Theminor glycoprotein, gp25, is largely an in-ternal protein, although a small segmentprojects through the virion envelope(Sturman, 1982). This protein is probablyanalogous to the matrix protein of otherenveloped viruses and is thought to playa role in the control of viral maturationand establishment of virion stability(Holmes et at, 1982). The nucleocapsidprotein, ppGO, is intimately associated withthe viral RNA . The simplicity of the struc-

ture of the virion makes the MHV groupan ideal subject for comparative studies ofpathogenesis, as, in principle, variations indisease patterns can be correlated withchanges in protein structure.

The relatedness of the MHV strains hasbeen examined by a number of methods,including cross-neutralization with hyper-immune serum (Hierholzer, 1979 ; Wege, etat, 1981) and, more recently, kinetic neu-tralization (Childs et at, 1983). Most MHVstrains are antigenically distinct by thesemethods, although the MHV3 and A59strains are very similar. Analysis of theviral polypeptides has shown that someMHVs have proteins with unique migra-tion patterns by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (Bondet al., 1979; Cheley et at, 1981; Lai andStohlman, 1981) . More recently, severalstudies have compared the oligonueleotidemaps of closely related MHVs that causedifferent diseases in order to identify thegenes responsible for different disease pat-terns (Lai et at, 1981; Wege et al., 1981 ;Stohlman et at., 1982) .

JHM virus (JHMV) is a neurotropic MHVwhich may cause encephalitis and demy-elination (Bailey et at, 1949; Weiner, 1973) .Its structural features (Massalski et al.,1982), proteins and genome (Lai and Stohl-man, 1981), are those of a typical MHV . Inanimal models of chronic demyelination,JHMV persistence has been demonstratedfor extended periods (Stohlman and Wei-ner, 1981; Knobler et at, 1982). Recentlytwo plaque-size variants of JHMV havebeen isolated and characterized : a largeplaque isolate, JHMV-DL, with a high pro-pensity to acute encephalitis, and a small-plaque variant, JHMV-DS, which predom-inantly causes chronic demyelination(Stohlman et at, 1982) . In this manuscript,we report the use of a panel of JHMV-DL-specific monoclonal antibodies to study theantigenic relationships of the structuralproteins of 11 separate MHV isolates .

MATERIALS AND METHODS

Viruses and cells . Viruses were propa-gated in DBT cells as described previously(Stohlman and Weiner, 1978), except when

MURINE CORONAVIRUS ANTIGENS

297

virus in serum-free media was desiredfor use in radioimmunoassay (RI A) . In thiscase, infected cells were first culturedin Dulbecco's Modified Eagle's Medium(DMEM, Gibco Laboratories, Grand Island,N. Y

at -70 ° .The derivation and characterization of

the two plaque morphology variants ofJHMV (MIIV-4), DL (large plaque) and DS(small plaque), have been described re-cently (Stohlman et at, 1982). MHV-1,MHV-2, MHV-3, and MHV-S viruses wereplaque-purified from virus stocks obtainedfrom Dr. M. Collins, Microbiological As-sociates, Bethesda, Maryland and K . Fu-jiwara, University of Tokyo, Tokyo, Japan .The MHV-K, MHV-D, and MHV-Nuu vi-ruses were obtained as cloned stocks fromDr. K. Fujiwara. The MHV-M virus, anisolate from nude mice with wasting dis-ease, was obtained from Dr. M. Collins .Vesicular stomatitis virus (VSV) andherpes simplex virus, type I (HSV-1), wereobtained from Dr . P. Brayton and Dr. D .Willey, respectively, both of the Universityof Southern California, School of Medicine .For the production of the monoclonal

antibodies, M5 cells, a horse serum-adaptedline of SP2/0-Ag .14, obtained from Dr . J .Davie, Washington University, St . Louis,Missouri, were grown in DMEM supple-mented with 10% horse serum, 20 mMHEPES, 100 IU/ml penicillin, 100 pg/mlstreptomycin, 2 mM L-glutamine, 10 mMMEM nonessential amino acids (GibcoLaboratories, Grand Island, N . Y.), and 5X 10_6 M 2-mereaptoethanol . During se-lection, HAT was added to the medium(final concentrations: 1 X 10_ , Mhypoxan-thine, 4X10-' M aminopterin, and 1 .6x 10-5 M thymidine) (Littlefield, 1964).

Virus neutralization. The ability of themonoclonal antibodies to neutralize viruswas assayed as previously described(Stohlman and Weiner, 1981) . Briefly, 25-

29 8

µl serial dilutions of antibody (heat inac-tivated at 56° for 30 min) were added towells of a 96-well microtiter plate ; sub-sequently, 25 µl of a virus suspension con-taining 10-TCID6„ doses (50% tissue cultureinfective doses) was added to each well .The plates were incubated at 37° for 1 hr,and then 2 X 10° DBT cells were added toeach well. The neutralizing titer was ex-pressed as the highest dilution of antibodywhich prevented CPE in 50% of the wells .

Antibody binding assay. The solid phaseradioimmunoassay (RIA) for antiviral an-tibodies has been described previously(Fleming et al, 1983). Viral antigens wereused in excess; preliminary assays showedthat each virus stock contained suffi-cient antigen to produce maximal bindingwith a hyperimmune anti-MHV standardserum. After cloning, 25 µl of supernatantfrom established hybridomas was assayedat 10 -1 dilution. Serial dilutions of antibodyshowed that this concentration fell withinthe linear region of the assay . In'eross re-active RIA, the binding of antibodies at10-1 dilution was tested against homolo-gous virus (JHMV-DL) and compared tothe binding to heterologous viruses .Relative binding was expressed by a

modification of the convention of Gerhardet at (1981). The counts per minute (cpm)bound to homologous (JHMV-DL) antigenwere normalized to 100% . Binding ofgreater than 50% of this value was con-sidered strongly positive ; 25-50%, moder-ately positive ; less than 25% but more thantwice background, weakly positive; and lessthan twice background, negative . Resultsshown in the block diagrams are the com-posite of at least three assays, each ofwhich consisted of quadruplicate samples .

Antisera. Hyperimmune antisera toJHMV-DL and A59 were made by serialimmunizations of C57BL/6J (B6) mice ob-tained from Jackson Laboratories, BarHarbor, Maine. Normal mouse serum freeof anti-MHV antibody was obtained fromB6 mice immediately after being received .Control antibodies used included mono-clonal anti-mouse Ia (Harmon et al., 1982),and rabbit anti-HSV-1, kindly supplied byDr. D . Willey of the University of SouthernCalifornia, School of Medicine.

FLEMING ET AL.

Preparation of monoclonal antibodies.The technique of Kohler and Milstein(1975) as modified by Harmon et at, (1982),was followed. B6 mice, 6 weeks of age, wereinoculated intraperitoneally with approx-imately 105 plaque-forming units of JHMV-DL grown in serum-free medium . Second-ary immunization, usually 6 weeks laterby the intravenous route, also consisted of105 plaque-forming units of virus in serum-free medium. Suspensions of immunespleen cells were fused to M5 cells at aratio of 2.5 :1 using 34% polyethylene glycolMW 1500 (Aldrich Chemical Co ., Milwau-kee, Wis.) at 37 ° . The suspension was thenwashed, resuspended in HAT medium, andcultured in 24-well plates at 1 X 10 6 cellsper well, supplemented by 1 X 10 6 feederspleen cells/well prepared from nonim-mune B6 mice . Following incubation for 4days at 37°, half of the medium was re-moved from each well and replaced withfresh HAT medium containing 1 X 106feeder cells . On Day 8 this step was re-peated with HT media (aminopterin-free)and feeder cells. On Day 11, wells werescreened by RIA for antiviral antibody .Cells in positive wells were cloned by visualinspection and limiting dilution in 96-wellplates. Clones were expanded, and super-natants were subsequently harvested andreassayed by RIA . Monoclonal antibodieswere assayed for immunoglobin class andsubclass by Ouchterlony immunodiffusion,using antiserum to murine immunoglob-ulin isotypes (Litton Bionetics, Kensing-ton, Md .) .Radioirnmunopreeipitation . The specific-

ity of the monoclonal antibodies for viralproteins was determined by radioimmu-noprecipitation (RIP) . DBT cells were in-oculated with virus at a multiplicity of in-fection of 1-5 for 1 hr at 37° . After removalof the inoculum, the cultures were in-cubated in the presence of serum-freeDMEM containing 1 jig/ml actinomycin D .The medium was removed and replacedwith prewarmed methionine-free DMEM(MFDMEM) for 15 min . The medium wasagain removed and replaced with pre-warmed MFDMEM containing 20,uCi/ml° 5S-methionine (New England Nuclear,Boston, Mass.). Following incubation for

30-45 min, the cultures were placed oncrushed ice, washed 2 times with ice-coldPBS, and solubilized with a buffer com-posed of 10 mM Tris-HCL pH 7 .4, 150 mMNaCl, 0 .5% NP-40, 1 mg/ml N-p-tosyl-L-lysine-chlormethyl ketone HCL (TLCK),200µg/ml phenylmethylsulphonyl fluoride(PMSF), and 500 u/ml aprotinin . The lysatewas clarified by centrifugation at 500 g for5 min and stored at -70°. For immuno-precipitation, the lysates were adjusted to0 .2% sodium dodecyl sulfate (SDS) and in-cubated with an equal volume of polyvalentantiserum or monoclonal antibody as de-scribed by McMillan et al. (1981). The im-mune complexes were adsorbed to a sus-pension of protein-A bearing Staphylococ-cus aureus as described by Kessler (1981) .The final pellet was resuspended in a buffercontaining 67 mMTris-HCL, pH 6 .8,2 .66%SDS, 5 .0% 2-mercaptoethanol, heated at56° for 2 min, and then centrifuged for 1 .5min at 10,000 g. The supernatants were an-alyzed by electrophoresis on 6-15% dis-continous slab polyacrylamide gels .

RESULTS

Production and characterization ofmonoclonat antibodies to JHM-DL virusTwenty-three cloned hybridomas producedantiviral antibodies as determined by RIA .The specificity of each antibody was de-termined by precipitation of JHMV-DLproteins from a lysate of infected cells .Figure 1 shows examples of the immuno-precipitates analyzed on six 15% gradientpolyacrylamide gels . Lysates harvestedlate in infection contained various quan-tities of the degradation products of thenucleocapsid protein (lanes 7, 8, 9, 10, and11 in Fig. 1 and lanes 6 and 8 in Fig. 5) .In addition, a few monoclonal antibodiesimmunoprecipitated other weak proteinbands (see lane 13 in Fig . 1 and lane 11 inFig. 5) which could be eliminated by in-creasing the SDS to 0 .5% in the reactionmixture. Based on the data obtained byimmunoprecipitation, the 23 monoclonalantibodies were placed in three groups .Table 1 shows. that 10 antibodies reactedwith the major glycoprotein, gpISO/90, 10reacted with the nucleocapsid protein,

MURINE CORONAVIRUS ANTIGENS 299

pp6O, and 3 reacted with the minor enve-lope glycoprotein gp25 . All the antibodieswere tested for their ability to neutralizeJHMV-DL. Six of the ten that reacted withgp18O/90 neutralized JHMV-DL. None ofthe antibodies reactive with either pp6O orgp25 were able to neutralize JHMV-DL(Table 1) . The heavy-chain immunoglob-ulin isotype of each antibody was testedby immunodiffusion . All the antibodies re-acted with only a single anti-isotype an-tiserum. The isotypes for each monoclonalantibody are shown in Table 1 .

Reactivities of antibodies with nucleo-capsid protein pp60. Ten monoclonal an-tibodies reactive with the nucleocapsidprotein of JHMV-DL, pp6O, were used toexamine the antigenic conservation of thisprotein in MIIV strains that induce diversetypes of disease. Antigenic preservationwas determined in an RIA which testedthe 10 antibodies with different MHVstrains serving as antigens . Table 2 showsan example of this type of experiment .Anti-pp60 antibodies were tested for theirability to bind to the homologous antigen(JHMV-DL), the small-plaque variant ofthis virus (JHMV-DS), and a related mu-rine coronavirus (MHV-A59), as well ascontrol antigens (HSV-1, VSV) . Table 2

shows binding of at least 25% of the valuesfor homologous virus in 6 of 10 reactionsof anti-pp60 antibodies with MHV-A59 andnone of 10 with JHMV-DS . Since viral an-tigens were used in excess and binding ofantibodies to the glycoproteins of JHMV-DS was equal to the binding to the gly-coproteins of homologous virus (see below),we do not attribute the very low bindingof antibodies to JHMV-DS nucleocapsidprotein, pp6O, to limited antigenic densityof the viral preparation. Anti-pp6O mono-clonal antibodies did not react with controlviral antigens, and control monoclonal an-tibody did not react with MHV antigens .

The relative binding of antibodies to dif-ferent MHV strains in a larger series ofexperiments was used to construct theblock diagram presented in Fig . 2. The pat-tern of reactivities indicates a moderatedegree of antigenic conservation of ppGOamong MHV strains . By inspection, the vi-ruses tested could be divided into two broad

30 0

groups based on their reactivities . The firstgroup consisted of A59, MHV-3, MHV-D,MHV-K, and MHV-Nuu viruses, all ofwhich did not react with monoclonal an-tibodies J.1.1, J .3.14, J .3.7, and J .3.15. Theantigenic sites recognized by these anti-bodies are apparently conserved in the sec-ond group, composed of MHV-2, MHV-S,and MHV-1 viruses. In addition to thesetwo groups, two individual viruses, JHMV-DS and MHV-M, were notable for theirlack of marked binding (less than 25% rpmversus homologous virus) with any of the10 anti-pp60 monoclonal antibodies tested .

FLEMING ET AL .

1

2

3

4

5 8

1 8 9 10 11 12 13

e

FIG . 1 . Immunoprecipitation of JHMV proteins by monoclonal antibodies from [nS]methionine-labeled infected cell lysate . Lanes 1 and 2 are whole lysates prepared at 5 .5 hr post-infection (lane1) and 7.0 hr postinfection (lane 2) . Lane 3 through 6 are monoclonal antibodies reactive withgplSO/90 (lane 3. J.1 .2; lane 4, J.2 .5 ; lane 5, J .7 .5 ; lane 6, J.7.6) . Lanes 7 through 10 are monoclonalantibodies reactive with pp6O (lane 7, J .2 .1 ; lane 8, J.3.7; lane 9, J.3.3; lane 10, J .1 .1) . Lane 11 (J.1 .3),lane 12 (J .2 .7), and lane 13 (J .3 .11) are monoclonal antibodies reactive with gp23 . Samples wereanalyzed on 6-15% gradient polyacrylamide .slab gels .

Reactivities of antibodies with major gly-coprotein 180/90. The major virion enve-lope glycoprotein, gplSO/90, is capable ofa number of functions, including recog-nition of the host cell receptor (Sturmanand Holmes, 1983). It was therefore of in-terest to investigate the antigenic rela-tionships of gp18O/90 in a number of dif-ferent MHV strains which produce a va-riety of diseases. All monoclonal antibodiesrecognized antigenic determinants of thesmall plaque variant JHMV-DS equallywell as the immunogen, JHMV-DL, indi-cating the close antigenic relationship of

TABLE 1

CHARACTERIZATION OF MONOCLONALANTIBODIES TO JHMV-DL'

MURINE CORONAVIRUS ANTIGENS

301

'Monoclonal antibodies were obtained from hy-bridoma tissue culture supernatants. Antigen spec-ificity was determined by RIP, neutralizing titer bymicroassay, and immunoglobulin isotype by immu-nodiffusion as described under Materials and Methods .° Symbol (-) indicates no neutralization at 1 :4 di-

lution .

these two viruses (Fig . 3) . By contrast, veryfew of the antigenic determinants on theother MHVs were closely related to thoseof JHMV-DL, indicating the lack of con-servation of JHMV major glycoprotein an-tigens among the rest of the MHV strains .Several antibodies, such as J.1 .2, J .7.18,

J.2.2, and J .7.2, recognize only JHMV vari-ants DL and DS and thus may appear tohave identical specificities. Nevertheless,differences in fine specificity were dem-onstrated in further experiments . We firstisolated variant JHM viruses which es-caped neutralization by monoclonal anti-bodies. When these variant viruses were

used as antigens, each of the 10 anti-gp180/90 monoclonal antibodies in our panelshowed a distinctive pattern of reactivityto them, indicating each antibody in thisset recognized a unique antigenic deter-minant (Fleming, unpublished observa-tions) .

Reactivity of antibodies with minor gly-coprotein gp25. The smaller viral glyco-protein, gp25, is embedded within and ex-tends through the viral envelope . It hasbeen suggested that it may function asa matrix protein and is essential in vi-ral budding and in the formation of theviral envelope (Holmes et aL, 1982) .Three monoclonal antibodies were obtainedwhich are specific for gp25 . This proteinappears to be highly conserved among theMHV strains (Fig . 4). Nevertheless, thisfinding must be regarded as tentative, inview of the small numbers of available an-tibodies with specificity for gp25 .

Cross immunoprecipitations. To confirmthe pattern of antibody binding determinedby RIA, anti-JHMV-DL monoclonal anti-bodies were tested for their ability to immunoprecipitate the appropriate viralprotein from a lysate of cells infected withthe related MHV, A59. Figure 5 shows thatthose anti-JHMV-DL antibodies (J.7 .5,J.2.1, J .1 .3, and J .2.7) which react with A59by RIA were also able to immunoprecip-itate the appropriate A59-specified protein,indicating that the shared antigens de-tected by RIA were in fact on the analogousprotein. By contrast, those antibodies thatwere unreactive with A59 virus by RIA(J.2.5, J1.2, J .3.7, and J.1.1) did not im-munoprecipitate detectable radiolabeledA59 virus proteins . Taken together, theimmunoprecipitation data confirm thespecificities of the anti-JHMV-DL anti-bodies as determined by RIA .

DISCUSSION

Although marine coronaviruses have arelatively simple structure, they cause awide variety of diseases . In an attempt tounderstand the basis for their diversepathogenicities and to obtain informationon the antigenic characteristics of theseviruses, we have used 23 monoclonal an-

Antigenspecificity

Monoclonalantibody Isotype

Neutralizingtiter °

gP180/90 J .1 .2 IgG 2bJ.7 .18 IgG 3 1 :256J.2.2 IgG 2b 1:512J.7 .2 IgG 2b 1 :10243 .2.5 IgG 2a -J.2.6 IgG 2b 1:128J.7 .1 IgG 2a -J.7 .5 IgG 2b 1 :512J.7 .6 IgG 2b 1 :256J .1 .16 IgG 2b -

pp60 J2.1 IgG 2bJ .3 .13 IgG 2bJ .3 .5 IgG 2a3 .3 .1 IgG 2aJ .3 .3 IgG 2aJ .3 .4 IgG 2aJ .1 .1 IgG 26J .3 .14 IgG 2bJ .3 .7 IgG 2bJ .3 .15 IgG 2a

gp25 J .1 .3 IgG 2bJ2.7 IgG 2aJ .3.11 IgG 2b

tibodies derived from the immunization ofmice with the neurotropic strain JHMV-DL to examine the relationships of thestructural proteins of the principal MHVstrains .

Assessment by RIA and RIP showed thatthe degree of conservation of JHMV-DLantigens, as judged by this panel of anti-bodies, varied among the three principalMHV structural proteins. The nucleocapsidprotein pp60 showed intermediate conser-vation, the major glycoprotein gp180/90little conservation, and the minor glyco-protein gp25 strong conservation . EachMHV strain had a unique pattern of reac-tivities to the set of monoclonal antibodiesused, confirming previous studies by Childset at (1983) which showed that these MHVisolates are in fact separate and distinctstrains .

Relative binding to anti-pp60 antibodiesallowed grouping of the MHV strains intoseveral antigenic families . In general, thisgroup of antibodies showed substantialbinding to all the murine coronavirusestested, with the exception of JHMV-DS andMHV-M, indicating that the nucleocapsidprotein antigens are substantially con-served in most MHV strains . Analysis ofthe relatedness of the pp6o proteins divided

'Sample RIA: Data are given as counts per minute with mean value and one standard deviation of fourreplicate samples. Cpm for JHMV-DS and A59 are also expressed as a percentage of cpm for JHMV-DL(parenthesis) . Symbol (-) indicates background epm and NT indicates that the interaction was not tested .

All antibodies were tested at 10-' dilution. Control antibodies are anti-mouse la (monoclonal antibodyproduced by an identical protocol) and anti-HSV-1 (hyperimmune rabbit serum) .

the viruses into two groupings. The firstconsisted of A59, MHV-3, MHV-D, MHV-K, and MHV-Nuu viruses, which were on-reactive with monoclonal antibodies J .1.1,J.3.14, J.3.7, and J .3.15. The second group,consisting of MHV-2, MHV-S, and MHV-1 viruses, conserved these antigenic spec-ificities, as evidenced by strong reactionswith the same set of monoclonal antibodies .Although these strains have not all beenpreviously compared in a single study, pre-vious investigations do support this generaldivision. For example, the oligonucleotidemaps of MHV-1 and MHV-S are distinctfrom the other MHVs examined (Lai andStohlman, 1981), as are the peptide mapsof the pp6O from these two strains (Cheleyet at, 1981). In addition, the oligonucleotidemaps of A59, MHV-3, MHV-D, and MHV-K are very similar (Lai and Stohlman, 1981 ;Lai, unpublished data) . MHV-M has an oli-gonucleotide map which is very differentfrom any other MHV examined (Lai, un-published data), and it is therefore notsurprising that there is little recognitionof its antigens .

The finding that none of the 10 anti-pp6Oantibodies showed moderate or strongbinding to JHMV-DS was very unexpected,since JHMV-DL and JHMV-DS have a

302 FLEMING ET AL .

TABLE 2

REACTIONS OF ANTI-pp6O MONOCLONAL ANTIBODIES TO HOMOLOGOUS AND HETEROLOGOUS VIRUSES °

Antibodies' JHMV-DL Viruses JHMV-DS A59 VSV HSV-1

J.2.1 14,570 ± 750 910 ± 14 (6 .2%) 12,777 ± 180 (87 .6%)J .3.13 18,145 ± 169 2150 ± 91 (11 .8%) 12,420 ± 367 (68 .4%)J .3.5 5,831 ± 510 43 ± 40 (.7%) 5,176 ± 245 (88 .8%)J .3.1 18,227 ± 95 2535 ± 77 (13 .9%) 12,267 ± 1092 (67 .3%)J .3.3 23,125 ± 1011 2715 ± 558 (11 .7%) 21,122 ± 1665 (91.3%)J.3.4 11,886 ± 1731 1863 ± 286 (15 .7%) 4,896 ± 470 (41 .2%)J.1 .1 3,675 t 90 -J.3 .14 4,051 ± 258J.3.7 347 ± 31J.3.15 1,721 ± 27anti-1aanti-HSV-1 NT NT NT

NTNT

NT34,902 ± 1672

MURINE CORONAVIRUS ANTIGENS

303

FIG. 2 . Reactivities of monoclonal antibodies with the MHV nueleocapsid protein, pp60 . Thebinding of anti-pp6O monoclonal antibodies in RIA with different MHV antigens is expressed asa percentage of mean cpm relative to homologous virus (JHMV-DL) . The blocks represent >50%(U), 25-50% (E), twice assay background-25% (]1) and less than twice background epm (C7) . Alldeterminations represent an average of three assays .

MONOCLONALANTIBODIES

TOgo 180190

J .1 .2

J .7 .18

J. 2 .2

J.7 .2

J .2 .5

J .2 .6

J .7 .1

J.73

J .7 .6

J .1 .16

FIG . 3 . Reactivities of monoclonal antibodies with the MHV major glyeoprotein, gpiSO/90 . Thebinding of anti-gp1SO/90 monoclonal antibodies in RIA to different MHV antigens is expressed asdescribed for Fig. 2.

304

J.1.3

J .2 .7∎ •

∎-SE

J .3-1

common passage history, and their ge-nomes share all but two oligonucleotidespots (Stohlman et aL, 1982). Recently, oli-gonucleotide fingerprints of mRNA No. 7,which encodes pp60, were studied inJHMV-DL and JHMV-DS. The fingerprintsof the two variants were indistinguishable(Lai, unpublished observations) . We pres-ently have no firm explanation for the dif-ference between the data obtained withmonoclonal antibodies and oligonucleotidefingerprinting . Very likely, the monoclonalantibodies detected antigenic sites whichare represented by genetic sequences notdetectable by T 1-oligonucleotide finger-printing. This striking result implies thatthere is a significant antigenic change inthe nucleocapsid protein of the smallplaque variant, JHMV-DS. The lack of an-tigenic differences among JHMV-DL andJHMV-DS in the other two virion proteinssuggests that the change in pp60 may becorrelated with differences in pathogenic-ity between the two viruses (Stohlman, etat, 1982). However, this conclusion is pro-visional and should be supported by furtherstudies, such as the analysis of nonstruc-tural proteins and the complete sequencingof the genomes of the viruses .

Gp180/90 is an external protein whichserves as the target for neutralizing an-tibody, induces cell fusion, and may playa key role in determining host cell range(Holmes et at, 1982; Collins et at, 1982) .None of the anti-gp180/90 monoclonals in-dicated any major antigenic conservationof this protein among the wide range ofmurine coronaviruses, except for JHMV-DS. The fact that none of the gp180/90

FLEMING ET AL .

FIG. 4. Reactivities of monoclonal antibodies with the MHV minor glycoprotein, gp25 . The bindingof anti-gp25 monoclonal antibodies in RIA to different MHV antigens is expressed as describedfor Fig. 2 .

antigens of the other MHVs shared manydeterminants with JHMV is not surprising,as marked variability of external proteins,usually in the context of conservation ofinternal proteins, has been well-docu-mented in viruses such as vesicular sto-matitis virus (Doel and Brown, 1978), po-liovirus (Nottay et aL, 1981), and the mu-rine retroviruses (Niman and Elder, 1982) .This phenomenon may relate to the highmutation frequency of RNA viral genomesin general and the selective pressure thathost immune systems may preferentiallyexert on neutralization-reactive surfacecomponents of viruses (Holland et at, 1982) .Antigenic determinants of the minor

glycoprotein gp25 were highly conservedamong the MHV strains tested. This resultwas not unexpected, since gp25 is thoughtto serve as a matrix protein (Holmes et aL,1982; Sturman, 1982), and therefore it ispossible that many mutations involvinggp25 might be lethal to the virus. Thisfinding is consistent with .previous studiesshowing that the matrix proteins of ve-sicular stomatitis virus (Doel and Brown,1978) and influenza virus (Laver andDownie, 1976) are highly conserved amongdifferent strains .

Studies of viral pathogenicity are alsoconsistent with the divisions proposedbased on the reactivity with the anti-pp60monoclonal antibodies. Most MHV strainscause hepatitis under natural or appro-priate experimental conditions . Neverthe-less, the members of the first group char-acteristically produce other diseases inaddition to hepatitis: A59 can cause de-myelination (Robb et aL, 1979), MHV-3

MONOCLONAL

VIRUSESANTIBODIESTO

JHM JHM MHV MHV MHV MHV MHV MHV MHV MHVgp 25

DL DS A59 3 D K Nuu 2I

S M VSV IHSV

o

FIG. 5. Cross immunopreoipitations: Monoclonal antibodies specific for the three JHMV-DL-specified proteins were tested for their ability to react with A59 virus proteins by immunoprecipitationusing a [ 8]methionine-labeled lysate of A59-infected cells as antigen . Lane 1 is a lysate of JHMV-DL-infected cells . Lane 2 is a lysate of A59 virus-infected cells showing the faster migration ofpp611 . Lane 3 (3 .7 .5), lane 4 (3.2.5), and lane 5 (3.1 .2) are monoclonal antibodies reactive with theJHMV-DL gplSO/90 protein. Lane 6 (J .2 .1), lane 7 (3.3.7), lane 8 (3.3.3), and lane 9 (3 .1 .1) aremonoclonal antibodies reactive with JHMV-DL pp6O . Lane 10 (3.1 .3) and lane 11 (3.2 .7) are monoclonalantibodies reactive with JHMV-DL gp23 . Samples were analyzed on 6-15% gradient polyaerylamideslab gels .

choroidoependymitis (Virelizier et at, 1975)MHV-D enteritis (Ishida et at, 1978a), andMHV-K myeloproliferation in nude mice(Ishida et at, 197Sb). By contrast, themembers of the second group (MHV-2,MHV-S, and MHV-1) have been reportedto produce primarily hepatitis ; MHV-1 andMHV-S have also been considered to be ofrelatively low pathogenicity (Cheley et at,1981; Wege et at, 1982).Thus using the present panel of mono-

clonal antibodies to JHMV-DL, antibodies

MURINE CORONAVIRUS ANTIGENS

1 2 3

s

e

7

a 9 10 11

qMW a

305

to pp60, rather than gp180/90 or gp25, aremost useful in establishing antigenic,re-lationships both among MHV strains andbetween plaque variants of one MHVstrain, JHMV . The fact that these antigenicdivisions to some extent correlate withpathogenicity raises the possibility thatpp60 may play an important role in diseasepotential. However, genetic regions codingfor other gene products may contain al-terations which would result in variationsnot detectable in this study, as shown pre-

306

FLEMING ET AL_

viously by oligonucleotide fingerprinting of-HMV-DL and JHMV-DS (Stohlman et at,1982). Therefore, these conclusions shouldbe considered tentative . Also, the majorglycoprotein, gp180/90,is highly variableand may be expected to influence viralpathogenicity, e.g., by controlling tropismfor specific cell types . This has already beensuggested in one study in which relativehepatotropism among MHV strains maybe correlated with changes in the gene forgp180/90 (Lai et at, 1981) .

Taken together, these findings imply thatdeterminants on both gp180/90 and pp60of MHV may play critical differential rolesin pathogenesis. Such a conclusion wouldbe consistent with the detailed informationavailable concerning the genetic and mo-lecular bases of pathogenicity of influenza(Rott, 1979) and reoviruses (Fields andGreene, 1982). For these viruses, virulencehas been shown to be multigenic, althoughindividual genes are responsible for dif-ferent aspects of pathogenesis. Furtherstudies, including an evaluation of the roleof nonstructural proteins, will be neededto extend these conclusions to the murinecoronaviruses .

ACKNOWLEDGMENTS

We thank Chris Patton, Natalie Stein, Todd Kennell,and Gabriele Olivka for excellent technical assistanceand Raymond Mitchell, Alisa Young, and ClaudiaTroesch for editorial assistance with the manuscript.Alicia McDonough and Scott Linthicum generouslysupplied osI-labeled protein A, and Minnie McMillankindly assisted in the preparation of figures. Thiswork was supported by grants from the National In-stitutes of Health, Grants CA 22662 and NS 18146,the Multiple Sclerosis Society, Grant 1449 Al, andthe National Science Foundation, Grant PCM 10372 .R. Harmon is a recipient of National Research ServiceAward NS-07149, and J . Frelinger is a recipient ofan American Cancer Society Faculty Research Award .

REFERENCES

BAILEY, O. T., PAPPENHEIMER, A. M., CHEEVER, F. S.,and DANIELS, J . B . (1949). A murine virus (JHM)causing disseminated encephalomyelitis with ex-tensive destruction of myelin . J. Exp. Med 90, 195-212.

BOND, C. W., LEIBOWITZ, J . L ., and ROBB, J. A. (1979).Pathogenic murine coronaviruses : II, Character-ization of virus-specific proteins of coronavirusJHMV and A59V. Virology 94, 371-384 .

CHELEY, S., MORRIS, V . L ., CuPPLEs, M. J., and AN-DERSON, R. (1981). RNA and polypeptide homologyamong murine coronaviruses . Virology 115, 310-321 .

CHILDS, J. C., STOHLMAN, S . A., KINGSFORD, L ., andRUSSELL, R. (1983). Antigenic relationships of mu-rine coronavirusea . Arch. ViroL, in press.

COLLINS, A. R., KNOnLER, R. L., PowN.LL, IL, andBUCHMEIER, M . J . (1982) . Monoclonal antibodies tomurine hepatitis virus-4 (strain JHM) define theviral glycoprotein responsible for attachment andcell-cell fusion . Virology 119, 358-371.

DOEL, T. R., and BROWN, F. (1978) . Tryptic peptideanalysis of the structural proteins of vesicular sto-matitis virus . J Gcn Viro& 38, 351-361.

FLEMING, J . 0 ., TING, J . Y . P., STOHLMAN, S . A ., andWEINER, L . P. (1983). Improvements in obtainingand characterizing mouse cerebrospinal fluid : ap-plication to mouse hepatitis virus induced enceph-alomyelitis . J. Neuroimmunot 4,129-140 .

FIELDS, B. N., and GREENE, M . 1. (1982) . Genetic andmolecular mechanisms of viral pathogenesis: Im-plications for prevention and treatment. Nature(London) 300, 19-23.

GERHARD, W., YEWDELL, J ., FRANKEL, M . E., andWEBSTER, R. (1981) . Antigenic structure of influenzavirus haemagglutinin defined by hybridoma anti-bodies . Nature (I,cndnn) 290, 713-717,

GLEDHIia, A. W., and NIvEN, J. S. F. (1955) . Latentvirus as exemplified by mouse hepatitis virus(MHV) . Vet Rev. Annot 1, 82-90.

HARMON, R. C., SHIGEKAWA, B., STEIN, N . M., andFRELINGER, J . A. (1982) . Problems associated withdetermining the specificity of monoclonal antibodiesto major histocompatibility complex antigens . In"Hybridomas in Cancer Diagnosis and Treatment"(M. S. Mitchell and H . F. Oettgen, eds.), pp . 21-30 .Raven Press, New York .

HIERHOLZER, J. C., BRODERSON, J . R., and MURPHY,F. A . (1979) . New strain of mouse hepatitis virusas the cause of lethal enteritis in infant mice . Infect.Im:mun 24, 508-522.

HOLLAND, J ., SPINDLER, K ., HORODYSKI, F, GRABAU,E., NICHOL, S ., and V ANDEPOI, S . (1982) . Rapid evo-lution of RNA genomes. Science 215, 1577-1585 .

HoLMES, K. V., DULLER, E . W., and BEHNKE, J . N.(1982) . Analysis of the functions of coronavirusglycoproteins by differential inhibition of synthesiswith tunicamycin. In`Biochemistry and Biology ofCoronaviruses" (V . ter Meulen, S. Siddel, and H.Wege, eds.) pp. 133-142 . Plenum, New York.

ISHIDA, T ., TAOUCHI, F ., LEE, Y. -S., YAMADA, A ., TA-MnRA, T., and FUJIWARA, K . (1978a). Isolation ofmouse hepatitis virus from infant mice with fataldiarrhea. Lab Anim. Sci 28, 269-276 .

ImmJA, T ., TAMURA, T., UEDA, K., and FHJIWARA, K.(1978b) . Hepatosplenic myelosis in naturally oc-curring mouse hepatitis infection in the nude mouse .Japan J. Vet Sci., 40, 739-743 .

KESSLER, S. W. (1981) . Use of protein A-bearing

staphylococci for the immunoprecipitation and iso-lation of antigens from cells . Methods EnzymoL 73,442-459 .

KNOBLER, R. L., LAMPERT, P. W., and OLDSTONE,M. B. A . (1982) . Virus persistence and recurringdemyelination produced by a temperature-sensitivemutant of MHV-4. Nature (London) 298, 279-283 .

KoHt.FR, G ., and MmSTEIN, C . (1975) . Continuous cul-tures of fusedd cells secreting antibody of predefinedspecificity. Nature (London) 256, 495-497.

LAI, M. M . C., and STOHLMAN, S. A . (1981). Compar-ative analysis of RNA genomes of mouse hepatitisviruses . J. Virol. 38, 661-670.

LAI, M . M. C., BRAYTON, P . R., ARnEN, R. C ., Px1TON,C. D ., PLGH, C., and SToHLMAN, S. A . (1981) . Mousehepatitis virus A59: mRNA structure and geneticlocalization of the sequence divergence from he-patotropic strain MHV-3, J. ViroL 39, 823-834 .

LAVER, W. G., and DOWNIE, J . C . (1976) . Influenzavirus recombination . I . Matrix protein markers andsegregation during mixed infections . Virology 70,105-117 .

LITTLEFIELD, J. W . (1964) . Selection of hybrids frommatings of fibroblasts in vitro and their presumedrecombinants . Science 145, 709-710 .

MASSALSKI, A ., COUI.TER-MACKIE, M ., KNOBLER, R . L .,Brc tIMr,mR, M . J., and DALES, S. (1982) . In vitroand in vivo models of demyelinating diseases : V,Comparison of the assembly of mouse hepatitis vi-rus, strain JHM, in two murine cell lines. Inter-virology 18, 135-146 .

MCMILLAN, M, FRELINGER, J . A ., JONES, P. P., MUR-PHY, D. B., McDEviTT, H. 0., and HOOD, L . (1981) .Structure of murine Is antigens : two-dimensionalelectrophoretic analysis and high-pressure liquidchromatography Cryptic peptide maps of productsof the I-A and I-E subregions andd of an associatedinvariant polypeptide . J. Exp. Med. 153, 936-950 .

NIMAN, H. L., and ELDER, J. H. (1982) . Structuralanalysis of Rauscher virus gp7O using monoclonalantibodies : sites of antigenicity and p15(E) linkage .Virology 123, 187-205 .

NOTTAY, B. K.,IL KEW, 0 . M., HATCH, M. H., HEYWARD,J. T., and OBLIESEI, J . F . (1981). Molecular variationof type I vaccine-related and wild poliovirusea dur-ing replication in humans. Virology 108, 405-423 .

ROBB, J. A., BOND, C . W., and LEiBowIm, J. L . (1979).

MURINE CORONAVIRUS ANTIGENS

Pathogenic murine coronaviruses :I11, Biological andbiochemical characterization of temperature sen-sitive mutants of JHMV . Viroiog-y 94, 385-399 .

RoTT, R. (1979) . Molecular basis of infectivity andpathogenicity of myxovirus: Brief review . Arch. Vi-roL 59, 285-298 .

ROWE, W . P ., HARTLEY, J . W., and CAPPS, W . I. (1963) .Mouse hepatitis virus infection as a highly con-tagious, prevalent, coterie infection of mice . Proc.SOC. Exp Med. Riot 112,161-165.

STOHLMAN, S. A., and WEINER, L. P . (1978), Stabilityof neurotropic mouse hepatitis virus (JHM stain)during chronic infection of neuroblastoma cells .Arch. ViroL 57, 53-61 .

STOH,MMe, S. A., and WEINER, L . P. (1981) . Chroniccentral nervous system demyelination in mice after.JHM virus infection . NeunAogy 31, 38-44 .

STOHLMAN, S . A., BRAYTON, P. R ., FLEMING, J. 0, WEI-NER, L, P ., and LAI, M . M . C . (1982). Murine corona-viruses: isolation and characterization of two plaque.morphology variants of the JHM neurotropic strain .J. Gem. ViroL 63, 265-275 .

STURMAN, L. S . (1982) . The structure and behavior ofcoronavirus A59 glycoproteins . In "Biochemistryand Biology of Coronavirusos" (V. ter Meulen, S.Siddell, and H. Wage, eds .), pp . 1-17. Plenum, NewYork.

STURMAN, L. S., and HOi.MES, K . V . (1983) . "The Mo-lecular Biology of Coronaviruses" Academic Press,New York (in press) .

TER MEULEN, V., SIDDELI, S ., and WEGE, H., eds. (1982) ."Biochemistry and Biology of Coronaviruses,"Plenum, New York.

WEGE, H., STEPHENSON, J. R, KOLA, M, WEGE, H .,and TER MEDIAN, V . (1981) . Genetic variation ofneurotropic and non-neurotropic murine corona-viruses . J Gen. Viral 54, 67-74.

WEGE, H ., SIDDELL, S ., and TER MEULEN, V . (1982) .The biology and pathogenesis of coronaviruses .Cum. Top. Micro. Biota ImmunoL 99, 165-200.

WEINER, L_ P. (1973) . Pathogenesis of demyelinationinduced by a mouse hepatitis virus (JHM :virus) .Arch. NeuroL 28, 298-303.

VmELIZIEE, J . L., DAYAN, A . D., and ALLISON, A. C .(1975) . Neuropathological effects of persistent in-fection of mice by mouse hepatitis virus. InfectImmure. 12,1127-1140 .