Identification of antigenic determinants by polyclonal and hybridoma antibodies induced during the course of infection by vaccinia virus

VIROLOGY

148.84-96

(1986)

Identification Antibodies

of Antigenic Determinants Induced during the Course

SHARON Cytobiology

WILTON, Group,

JAMES

Department

Western

Ontario,

Received

June

by Polyclonal and Hybridoma of Infection by Vaccinia Virus’

GORDON,

of Mhobiologg London, Ontario 26, 1985;

accepted

AND and N6A

SAMUEL

Immunology, 5C1, Canada

October

DALES2 University

of

3, 19X5

In order to extend the understanding of determinants involved in the humoral response in the infected host, mice were subjected to an immunization regimen using both active and uv-killed vaccinia virus. The spectrum of antibody specificity in hyperimmune sera was followed by Western blotting. Comparable studies involving Western blotting and immunofluorescence were conducted with a panel of monoclonal antibodies derived from hybridomas selected from similarly immunized animals. Hyperimmune sera contained circulating antibodies primarily against three polypeptides of 28K, 35K, and 62K. These antigens were shown to be located both at the surface and within the virion. The repertoire of monoclonal antibodies included some that reacted with the 28K and 35K antigens and others that recognized a 32K core complex component or a nonvirion cell surface component, corresponding to the viral hemagglutinin. Within the panel of monoclonal antibodies was a large group which reacted with a 3ZK antigen found in the IHD-J virion but absent from the IHD-W strain. This finding correlates with the absence of a 3ZK polypeptide from the IHD-W particle. Overall, the current findings reveal the absence of any particular correlation between the incidence of polyclonal antibodies in the circulation of the immune host and the frequency of selected hybridomas against vaccinia antigens. Application of this type of immunological analysis should provide useful data concerning the detection and mapping of the antigens and their epitopes which are significant for humoral immunity. 0 1986 Academic

Press, Inc.

INTRODUCTION

Immunization with poxviruses evokes both humoral antibody and cytotoxic cellular (CTL) responses (Westwood et aL, 1965; Loh and Riggs, 1961; Mims, 1964; Hutt, 19’75; Doherty et a,!., 1976; Koszinowski and Ertl, 1976; Hapel et aL, 1978; Byrne et ak, 1983; Mallon and Holowczak, 1985). The relative importance of circulating antibodies vs CTLs in the suppression of the infection is not fully understood to date. However, it is quite clear that upon the appearance of circulating antiviral antibodies the infection abates in the animal and human host (Cho and Wenner, 1973; Soekawa et al, 1977). Analysis of convalescent sera by the double diffusion Ouch1 Supported Canada. z To whom

reprint

0042-6822186

$3.00

Copyright All righta

by the

Medical requests

Research should

Q 1986 by Academic Press. Inc. of reproduction in any form reserved.

Council

of

be addressed. 84

terlony and other serological methods reveals that antisera of the immune host contain a spectrum of antibodies reactive with surface or internal virion antigens, as well as with virus-induced antigens which are inserted at the surface of infected cells (Ueda et al, 1969; Weintraub and Dales, 1974; Payne, 1979). These findings reveal that circulating antibodies may be directed against a multiplicity of vaccinia virus antigenic determinants. The potential of vaccinia virus as a recombinant vaccine vector for introducing antigens of herpes simplex (Paoletti et aL, 1984), influenza (Bennink et uL, 1984), VSV (Mackett et ah, 1985), rabies (Wicktor et al, 1984; Kieny et aL, 1984), hepatitis B viruses (Beale, 1984; Moss et UC, 1984) and other infectious agents such as the malarial parasite (Smith et al, 1984), has been demonstrated recently. These develop-

VACCINIA

ANTIGENS

ments draw attention to the antigenic determinants of vaccinia which may be most relevant for immune responsiveness and surveillance. It is likely that the usefulness of this virus as a recombinant vector may be broadened by creation of as many serological subtypes as possible, each carrying a different endowment of foreign genes, thus providing a repertoire of vectors which can be used successively on the same individual. Before vaccinia virus vectors of variable serotype can be selected it is essential to identify the antigenic determinants involved during antibody responses to the normal infection. For this reason we undertook to examine which polyclonal antibodies are induced during the immunization of mice. Furthermore a comparison was made between the polyclonal and monoclonal antibodies (Mab) produced by a panel of hybridomas derived from the immunized mice. The Mabs were then used to characterize some of the viral antigens.

AND

ANTIBODIES

85

phase-contrast microscopy. The suspension was centrifuged at 500g for 5 min to remove free nuclei and intact cells, the supernatant was mixed with sodium dodecyl sulfate (SDS)-dissociating buffer (Laemmli, 1970), adjusted in volume whereby 1 ml contained cytoplasmic components from 10’ cells and the proteins were completely dissociated by heating for 2 min at 100”. Purified virus was prepared and proteins of the viral envelope isolated as described by Stern and Dales (1976). The suspension of purified virions in 10 mM Tris-HCl, pH 7.3, contained 1 mg/ml of protein. To remove detergent-soluble components, the suspension was adjusted to 1% NP-40 by the addition of freshly prepared 10% NP40 and incubated at 37” for 15 min. To effect stripping of envelopes (Easterbrook, 1966; Gold and Dales, 1968), one-half of this material was subsequently incubated in 0.12 M 2-fi-mercaptoethanol for 10 min at 37”. The incubation mixture was loaded onto a sucrose cushion (40% w/v sucrose in 10 mM Tris-HCl, pH 7.3) and centrifuged for 20 min at 100,OOOg. Both the suspended enMATERIALS AND METHODS velope material above the cushion and at the interface, and the pelleted fraction Cells and viruses. Mouse Lz, human consisting of viral core complexes (Gold HeLa, rat myoblast L6H9, and kangaroo and Dales, 1968), were collected and prePtKz cell lines described in Silver et aZ. pared for one-dimensional polyacrylamide (1979) and Dales et al. (1983), were maingel electrophoresis (Laemmli, 1970) (SDStained in monolayer cultures using nu- PAGE) as described above. trient medium consisting of Eagle’s MEM Pol&onal antisera. Mouse polyclonal supplemented with 10% fetal bovine serum. antisera were raised against purified IHDIHD-J, the hemagglutinin (HA+) parent W vaccinia virus, according to the proceand IHD-W, the HA- variant of vaccinia dures described by Dales et aL (1983). Rabvirus (Weintraub et al, 1974) were propabit antisera to IHD-J virus were produced gated and titrated on L cells. Cell monoby immunizing with purified virions aclayers were infected at an m.o.i. of 0.1-5 cording to Gold and Dales (1968). Selection and testing of h$widomas. The PFU/cell as appropriate. Virus purification, which includes sedimentation through library of monoclonal or hybridoma (Mab) a potassium tartrate gradient, followed the antibodies was selected by immunizing of method adopted in Stern and Dales (1974). BALB/c mice with either purified IHD-W Preparation of antigen for immunoelecvirions or lysates of IHD-J infected L cells. Details of the methods employed for imtroblotting. L cell monolayer cultures, either creation of, heretofore, ununinfected or infected, incubated for 18 hr munization, at 37’ were washed with PBS, then the cells characterized hybridomas (HBs) by fusion cell lines, selection of were removed by scraping, pelleted at 2000g with plasmacytoma HB colonies, repeated cloning, screening of and resuspended in PBS: deionized water (1:3). The cells were allowed to swell on ice, clones producing Mab against (>) vaccinia virus antigens and determination of the then Dounce homogenized to effect greater type of each HB have been than 90% cell rupture, as assessed by immunoglobulin

86

WILTON,

GORDON,

described in detail elsewhere (Dales et al., 1983). Procedure for immunoelectroblotting. The dissociated proteins, separated on 11% SDS-polyacrylamide gels were transferred to nitrocellulose sheets by the method of Towbin et aL (1979), termed Western blotting. The modified procedure of Batteiger et al. (1982) was then adopted. Briefly, nitrocellulose strips were soaked in PBS containing 0.05% Tween-20 (PBS-T& for a minimum period of 1 hr at 4”. Specific antibodies in fresh PBS-T, were applied to the strips for 2 hr at 20”. Following extensive washing with PBS-Tz,, the strips were bathed for 2 hr at 20” with conjugates of ‘%I-Protein A or ‘%I-labeled goat > mouse Ig antibodies (3 X lo4 cpm/nitrocellulose strip), again thoroughly washed with PBSTz,,, dried, and exposed to X-ray film in the presence of an intensifying screen. Immune-dot-blotting procedures. Antigens for immuno-dot-blotting were obtained in solubilized lysates of IHD-J infected cellular material from which the nuclei had been removed. The solubilizing solution contained 2% (v/v) NP-40 in 10 mM Tris-HCl, pH 7.4, 0.2 M NaCl, 0.5 M KCl, 2 mM EDTA, 1 mM PMSF (Talbot et aL, 1984). The final suspension contained about 107/ml enucleated cell equivalents. Antigens dissociated in SDS were prepared as outlined for Western Blotting. To apply the antigen, drops of 10 ~1, containing about 10 pg protein, were spotted onto nitrocellulose strips. The strips were dried, fixed with 1% formaldehyde, washed thoroughly with distilled water, and equilibrated to a neutral pH with PBS. After saturation with the blocking buffer PBS-Tzo, the spots were overlaid for 1 hr at 20”, with a solution of the relevant antibodies. Then the strips were washed with PBS-TZO and reacted for 1 hr at 20’ with ‘%I-labeled goat > mouse Ig containing 3 X lo4 cpm per strip. Finally, the strips were washed, dried, and exposed to X-ray film. Immuno~uorescent microscopy. Coverslip cultures of uninfected or vaccinia virus infected Lz cells were acetone fixed and reacted 30-60 min with either specific or nonrelated antibodies. The primary antibodies were tagged for 30 min with goat

AND

DALES

> mouse IgG conjugated to fluorescein or rhodamine, as described in Dales et al. (1983). Reagents. Electrophoresis grade SDS was purchased from Bio-Rad Laboratories, NP-40 from Particle Data Laboratories. Rabbit > mouse K light chain serum was obtained from Miles Laboratories, Inc., rhodamine conjugated to goat > mouse IgG from Cappel Laboratories, ‘?-Protein A (sp act 8.5-9.2 &i/pg) and lzr’I-labeled goat > mouse Ig (8 &i/kg) from New England Nuclear. Molecular weight (MW) markers, purchased from Bio-Rad Laboratories, included phosphorylase B, bovine serum albumin, ovalbumin, carbonic anhydrase, soy bean trypsin inhibitor, and lysozyme, of MW, respectively, 92,000 (92K), 66K, 45K, 31K, 21K, and 14K. RESULTS

IdentQication by Immunoelectroblotting of Anti- Vaccinia Antibodies in Mouse

Antisera Serum collected from BALB/c mice infected with vaccinia virus (strain IHD-W) was used in the identification of viral antigens that may play a role in the host’s immune response to virus infection. To present the antigen in the form of individual polypeptide or glycoprotein bands, purified vaccinia virus (strain IHD-W) was solubilized and the extract separated by SDS-PAGE. Then the viral polypeptides were transferred onto nitrocellulose and reacted with the antisera followed by ‘%Ilabeled goat > mouse immunoglobulin. To ascertain the reproducibility of the responses, antisera from a group of eight mice were tested. The immune sera became bound reproducibly to three prominent viral antigens, with MW of 28K, 35K, and 62K, respectively, and to several minor antigens of MW 25K, 40K, 42K, 60K, 70K, and 94K, as illustrated in Fig. 1, lanes 3-9. None of the above antigens reacted with immunoglobulins in any of the preimmune sera, exemplified by lane 2 of Fig. 1. Some variability in the intensity of labeling of antigens with MW 25K and 28K was evident among different sera, as revealed by comparing lanes 7, 8, and 9 of Fig. 1.

VACCINIA

ANTIGENS

If the presence in hyperimmune mouse sera of antibodies recognizing core and other internal virion components was due to the presence of unintegrated antigens produced during active replication of the virus (Dales and Pogo, 1981), it might be expected that such antigens would not be synthesized or become available to act as immunogens in mice inoculated with uvinactivated virus. Therefore, antibodies to these immunogens should be absent. In fact, antibodies to the major 62K antigen were present in antisera from mice inoculated with killed virus, as shown by the Western blotting test, albeit in reduced amounts (Fig. 1, lane 12). However, >28K antigen antibodies were not detectable in these antisera but antibodies to the minor l

4240-

2

3

4

567

AND

ANTIBODIES

87

antigens of MW 55K, 60K, and 98K as well as to the major envelope 35K antigen were present. Preimmune sera from mice which subsequently provided the antisera did not bind any vaccinia antigens, as illustrated in lanes 2 vs 3 of Fig 1. Evidently, mice immunized with uv-killed IHD-W vaccinia virus produced antibodies against both the envelope and internal core components. As a consequence of viral replication, the core and other antigens, including those which have not been integrated into virions or appear at the cell surface, would be available as immunogens and should evoke specific antibodies. For this reason, the location within the virions of the various antigens identified by the mouse hyperimmune sera was determined. For this pur8

9lOll

12

a

FIG. 1. Characterization of anti-vaccinia virus polyclonal antibodies in hyperimmune mouse antisera by immunoelectroblotting. As the antigen purified vaccinia virus (strain IHD-W) was solubilized, electrophoresed through a 11% SDS-polyacrylamide gel, then transferred to nitrocellulose and reacted with either a preimmune serum (lane 2) or seven IHD-W antisera (lanes 3-11). The preimmune and antiserum used in lanes 2 and 3 are from the same animal. IHD-W virus reacted with antiserum from a mouse immunized with uv-killed pure IHD-W vaccinia (lane 12). For detection by autora?-labeled goat > mouse immunoglobulin was used as the second reagent. To detect diography, nonvirion vaccinia antibodies the hyperimmune antiserum was reacted with lysate from infected cells (lane 10). A lysate of uninfected L cells is included in lane 11; lanes 9-11, the same hyperimmune serum was used. As a standard in aiding the identification of viral antigens, a lysate from vaccinia virus infected L cells which had been pulse-labeled with [%]methionine, was coelectrophoresed (lane 1).

88

WILTON,

GORDON,

pose, polypeptides of entire purified virions of strain IHD-W, those extracted by NP40 alone, those isolated as components of stripped envelopes with 2-P-mercaptoethanol and those sedimented with the fraction of isolated core complexes, were subjected to Western blotting analysis. Antibodies to the major 28K and 62K virion antigens, were also binding components in the NP-40 insoluble and core materials (Fig. 2, lanes 5 and 7), suggesting that these two antigens are present in the virion cores. The bulk of the predominant 35K antigen, marked by antibodies in hyperimmune sera, was detected in the NP-40 soluble and envelope fractions (Fig. 2, lanes 4 and 6). Since, however, residual amounts 1

234567

62-

35-

2a-

FIG. 2. Comparisons between the binding of antibodies in hyperimmune mouse antisera to solubilized vaccinia virus polypeptides present in virions, detergent extracts, stripped envelopes, and core complexes. Extracts of viral material were electrophoresed through 11% SDS-polyacrylamide gels then subjected to Western blotting. Lane 1 contained lysate from vaccinia virus-infected L cells pulse-labeled with [%]methionine; lane 2, lysate of uninfected cells. Other samples consisted of purified virus strain IHDW (lanes 3-7); whole virions (lane 3), NP-40 soluble extract (lane 4), and the NP-40 insoluble fraction (lane 5); stripped envelopes (lane 6) and core complexes (lane 7). Antisera used were from a mouse immunized with active virus (lanes 2-7).

AND

DALES

of 35K protein were present also in the NP40 insoluble and core fractions (Fig. 2, lanes 5 and 7), neither the detergent extraction nor the stripping of envelopes could have been complete. Of the minor antigens, those of 40K, 42K, 55K, and 98K were detected in the NP-40 extractable and envelope fractions (Fig. 2, lanes 4 and 6) while those of 60K and 94K remained with the virion core complexes (Fig. 2, lanes 5 and 7). Although the stripped envelopes possessed greater quantity and number of viral polypeptides than the NP-40 extracts, as demonstrated by silver staining of SDSpolyacrylamide gels (data not shown), there was no difference in the Western blotting patterns of these fractions (Fig. 2, lanes 4 and 6). To ascertain whether mouse hyperimmune sera also contained antibodies to virus-induced nonvirion antigens, Western blotting was carried out with an infected cell lysate. It is evident in lane 11 of Fig. 1 that antisera from mice immunized with active IHD-W virus did not contain any detectable antibodies to nonvirion cellular antigens. The greater labeling of the 94K antigen, weaker labeling of 62K, and absence of label at the 60K position indicates that relative concentrations of these antigens in the lysate were different from those in virions (Fig. 1, lanes 9, 10). Since it is clearly established that 94K is a precursor of the 62K core polypeptide (Moss and Rosenblum, 1973; Stern et al., 1977), it was not surprising that there was less 62K and more 94K antigen in the lysates than within virions. Identity

of Antigen

Binding

Mabs Selected an Active Vaccinia Virus Ir4fectim

from Mice Which Had Experienced

(a) Immuno~uorescent staining. Following immunization of BALB/c mice with either purified IHD-W strain vaccinia virus or lysates of IHD-J infected cells a panel of Mabs specific for virus-induced antigens was selected, as described previously (Dales et al, 1983). HBs between murine plasmacytoma cell lines P3 X 63Ag8 or 6531, and spleen cells were created by eight independent fusions. Out of about 2000

VACCINIA

ANTIGENS

AND

ANTIBODIES

89

wells seeded hybrid cell colonies appeared viral materials by SDS-PAGE in one diin 344 wells. Among the latter, selection mension. The antigen tested was either and repeated subcloning generated 32 sta- suspensions of purified virions or lysates ble Mab-producing clones, of which 23 were of infected cells, depending on the question specific for viral antigens and 9 were au- being asked. Among the 23 Mab tested, only toantibodies, predominantly against com- 13 were reactive in the Western blotting ponents of cellular cytoskeleton (Dales et test with vaccinia virus antigens. It is rea,!, 1933). The autoantibodies will not be markable that 7 of the 13 recognized a 32K considered in this article. polypeptide. Among these ‘7, B5-3.2, C2-3.5, The primary screening of anti-vaccinia G3-2.1, HS-2.2 and 2A-11.10, reacted with virus Mabs was carried out by means of avidity (lanes l-5 of Fig. 4), while 2-5.5 immunofluorescence testing of IHD-J in- and 2A-4.11 gave faint bands and only fected Ls, HeLa, L6H9 and PtKz cells. With when applied at lo-fold greater concentraany individual Mab used the pattern of tion. Whether the same or different epiviral antigen distribution was essentially topes of the 32K antigen are recognized by the same, irrespective of the cell line used this group of ‘7 remains to be determined. as the host. The rhodamine or fluorescein It is, however, certain that at least 5 origtag occurred usually in inclusions within inated from independent fusions and canthe cytoplasm, ranging in size from that of not, therefore, have originated as sister vaccinia virus viroplasmic foci, termed also subclones. The group of 7 also includes “factories” and B-type inclusions, to very various immunoglobulin types (Table 1). small and numerous granules the size of Selection at high frequency of Mabs > 32K individual virions. The most common virus polypeptide suggests that during imstaining pattern, evident with 9 of the 23 munization this antigen effectively and Mabs, was that illustrated in the selected frequently induced clonal expansion of examples in Fig. 3, B-D and F, which shows splenocytes. that Mabs, 2A-11.10, C2-3.5130-1.2, and 2Three other strongly reactive Mabs were 1.5 obtained from four different fusions re- recognized, 2A-1.2 binding a 35K compoacted with antigen within entities the size nent, H3-1.4 an antigen spread over a broad of “factories” and virions. One Mab, H3- band with an approximate MW 85K, and 1.4, was bound uniformly to the surface of 130-1.2 a 28K antigen (lanes 6-8 of Fig. 4). infected cells, accentuating the presence of After reducing the quantity of H3-1.4 Mab, very long and numerous microvilli which it became evident that binding occurred to emanated from the cell surface. Another at least four antigenic bands ranging in Mab, Bll-5.1, described previously (Dales MW from about 70K to 90K (lanes 9 and et a& 1933), which had a similar surface 11 of Fig. 4). This antigen was present in distribution on infected cells was also a lysate of cells infected with IHD-J virus cross-reactive with an epitope of the cy- but absent from purified virions (lane lo), toskeletal protein vimentin. Staining with consistent with it being the vaccinia viboth H3-1.4 and Bll-5.1 was obtained only rus HA. under conditions which allow surface Among the weakly reacting Mabs, not expression of the viral hemagglutinin (HA) illustrated, were 2-5.5 and 2A-4.11 within as a late-late function (Dales et al, 1976), the group of seven binding to a 32K antiproviding suggestive evidence that the two gen, A2-2.1 recognizing a 35K antigen, 2AMabs reacted with epitope(s) on the HA. 29.6 with affinity for an 80K component and A summary of the data on the Ig type, in- Bll-5.1 reacting with a nonvirion 85K antensity and distribution patterns of im- tigen which, as discussed previously, was munofluorescence for our collection of at the cell surface (Dales et al, 1983). A list Mabs is given in Table 1. of Mabs positive in Western blotting is (b) Characterization by Western blotting. shown in Table 1. The location of antigens recognized by Information about the MWs of antigens reacting with our Mabs was obtained by one of the group of seven, C2-3.5 and >28K immunoelectroblotting after separation of Mab 130-1.2 was determined on virions and

90

WILTON,

GORDON,

AND

DALES

FIG. 3. Selected examples of L cells infected with IHD-J virus and reacted with representative Mabs. After fixation with acetone the coverslip cultures were reacted with individual Mabs followed by rhodamine-conjugated goat > mouse IgG and examined under uv optics. (a) Phase contrast image of the cell group shown in (b), reacted with Mab C2-3.5. Note that not all cells contain virus product. In the immunofluorescent positive cell the antigen is present in both discrete, cytoplasmic inclusions and very fine and numerous granules (virions?) concentrated particularly at the surface. In (cf, (d),

VACCINIA 12345676

ANTIGENS 91011

85 K-

35K32 K28 K-

FIG. 4. Autoradiogram showing specificity of Mahs for individual polypeptides of IHD-J vaccinia virus in a lysate of infected cells ascertained by Western blotting. Each lane contained material from 2 X 106 cells. Lanes l-11 immunoelectroblots reacted with Mabs and %Protein A. Lanes l-5 show reactivity with C2-3.5, G3-2.1, H8-2.2, 2A-11.10, and B5-3.2, respectively, which all bind to an antigen 32K in MW. Lane 6, Mab 2A-1.2 identifies antigen of 35K, lane 7, Mab H3-1.4 binds to a broad hand, average MW 85K, antigen; and lane 8, Mab 130-1.2 identifies a 28K antigen. Lane 9, IHD-J lysate reacted with Mab 130-1.2 and H3-1.4; lane 10, IHD-J pure virus reacted with H3-1.4; lane 11, IHD-J lysate reacted with H3-1.4. By reducing the intensity of labeling it is evident in lanes 9 and 11 that H3-1.4 identifies an antigen in four bands of slightly different, MW, corresponding together to the broad band evident in lane 7.

subviral fractions by the procedures used with hyperimmune sera, described in Fig. 2 above. Both Mabs reacted with internal, core complex antigens not released after extraction of virions with NP-40 or solubilization of the viral envelopes (Fig. 5). It

AND

ANTIBODIES

91

should be noted, however, that a minor fraction of the 28K and some of the 32K antigen was released during stripping of the envelopes (lane 6 of Fig. 5). The group of seven Mab which reacted with a 32K antigen in IHD-J lysate (lanes l-5 of Fig. 4), did not identify a comparable polypeptide when tested with IHD-W lysate. This finding correlates with the presence of a 32K band, evident after silver staining of IHD-J material subjected to SDS-PAGE, and the absence of such a band from equivalent preparations of IHDW virus (data not shown). To demonstrate that the 32K protein is, indeed, absent from the latter strain of vaccinia, Western blots were carried out with either polyclonal rabbit antiserum to IHD-J which had been extensively absorbed with a lysate of cells infected with IHD-W virus (Ichihashi and Dales, 1971), or with Mab C2-3.5 belonging to the group of seven. It is evident in lanes 3, 5, and 6 of Fig. 6 that both antibody preparations identified a component in IHD-J virions and lysate which was missing from virions and lysate of IHD-W virus, shown in lanes 4,7, and 8 of Fig. 6. It should be pointed out that, with the exception of Mab B5-3.2, all the Mabs in the group of seven were synthesized by HBs created from splenocytes of mice immunized with the IHD-J strain of virus. Absence of reactivity in Western blotting by 10 of the 23 Mabs, which were positive in the immunofluorescence testing (Table l), implied that during preparation of some antigens for immunoelectroblotting the native configuration of the relevant epitopes was altered. This idea was examined by testing the reactivity of Mabs with antigens undissociated by SDS and 2-@-mercaptoethanol. For this purpose binding of Mabs was tested by dot blots on viral material deposited onto sheets of nitrocellu-

and (f), staining was due to binding of Mabs 1.30-1.2, 2A-11.0, and Z-1.5, respectively. The patterns of antigen distribution are similar to those in (h), although only the Mab used in (d) binds to an antigen of the same MW as that in (b), while the HB which labels cells in (c) reacts with vaccinia antigens of lower MW. Presence of antigen in large inclusions (arrows) and minute granules, concentrated at the surface, is particularly clear in C (arrow heads). In (e), reaction with Mab H3-1.4 reveals uniform extensive surface labeling, including that of prominent, numerous microvilli (arrows). Magnifications (a)-(d), (f), X1350, (e), X1800.

92

WILTON,

GORDON,

AND

TABLE

DALES

1

SUMMARY OF TESTS WITH PANEL OF ANTI-VACCINIA

Mab

k

type’

Pattern of antigen distribution

Immunofluorescenceb

+

A2-2.1 A12-1.2 A17-1.1 B5-3.2

G 2A, 2B, k G2A,k G2A.k

++ ++ ++

Bll-5.1= B12-1.2 C2-3.5

M, k G 1, k G2A.k

+ ++ ++

C12-1.8 Cli’-2.4 G3-2.1 G5-5.3 H3-1.4 HS-2.2

M G2 G 1, k G 1, k G 1, k Gl

++ + ++ ++ ++ ++

19-3.2 130-1.2 l-4.5 l-7.8 2-1.5 2-5.5 2A-1.2 2A-4.11 2A-11.10 24-29.6

G3 G 1, k G 2, G2 G 2A Gl Gl G 2A G2A

++ ++ ++ ++ ++ + ++ + ++ ++

Very small numerous granules Small inclusions Large to small inclusions Very small numerous granules Uniform surface staining Large to small inclusions Large to small inclusions numerous very small granules Medium size inclusions As C2-3.5 As C2-3.5 Medium size inclusions Uniform surface staining Very small numerous granules Large inclusions As C2-3.5 As C2-3.5 As C2-3.5 As C2-3.5 Large granular inclusions Large granular inclusions As C2-3.5 As C2-3.5 Large inclusions

’ Complete typing was not obtained in most cases. b + weak, ++ intense staining. ‘This Mab has cross-reactivity with a cellular cytoskeleton d nd = not done.

lose. The Mabs reactive in Western blots generally also showed binding in the dotblot test (Table 1, lane 3 of Fig. 7). However, only 2-1.5 among the 10 Mabs negative in Western blotting gave a positive dot-blot reaction (lane 2 of Fig. 7). The discrepancy between the immunofluorescence and the other tests, involving reactivity of the remaining 9 Mabs, remains unexplained. DISCUSSION

Immunization of mice with replicating or uv-killed vaccinia virus evokes a predictable spectrum of antibodies reactive by

Mabs

component,

and

Western blotting

Dot blotting

+

+

+

-

+ +

ndd

+ -

+

+

+ -

vimentin.

Western blotting. The most prevalent antigen detected has a MW of 35K, and is the NP-40 extractable component of the envelope, identified previously (Stern and Dales, 1976). With the exception of >28K antibodies, which were absent in hyperimmune sera of mice immunized with killed virus, antibodies to the other abundant and minor components were present, including those binding to internal, core complex polypeptides 94K, 62K, and 60K of the virion. This observation indicates that the immune system can recognize antigens integrated into virions, as well as the unassembled, nascent polypeptide products of

VACCINIA 12

3

4

5

6

ANTIGENS 7

62-

353226-

AND

1.2 and the weakly reactive A2-2.1. Most of the other Mabs reacted with antigens in the core complex. The most frequently represented were seven HBs synthesizing Mab against a 32K antigen which was absent from IHD-W virions. Nor were the seven Mabs reactive by immunofluorescence with antigens in IHD-W infected cells. Since, with one exception, these HBs were derived from mice injected with IHDJ virus, were selected from clones originating from five independent splenocyteplasmocytoma fusions and produced Mabs of different Ig types, it is not possible that all originated from the same parental HB clone. Although there is strong evidence for the binding of all Mabs in the group to the same 32K polypeptide, whether they

FIG. 5. Identification by Western blotting of surface and internal vaccinia virus antigens which react with polyclonal and Mabs. The SDS-PAGE was conducted in 11% gels. Lane 1, lysate of L cells infected with IHDJ vaccinia virus and labeled with [%jmethionine; lanes 2-3, purified IHD-J whole virions; lane 4, NP40 soluble fraction; lane 5, NP-40 insoluble component; lane 6, stripped envelope material; lane 7, core complexes. In lanes 3-‘7 both Mabs C2-3.5 and 130-1.2 were used, while in lane 2 polyclonal mouse antiserum was applied.

synthesis. Judging by data from previous (Moss and Rosenblum, 1973; Stern et aZ., 1977; Silver and Dales, 1982) and current analyses it can be assumed that the 94K polypeptide is the precursor of the 62K component of cores. The data presented here, from Western blotting with hyperimmune polyclonal antisera (Fig. 1) revealed that the 94K component is abundant in the lysate but barely evident in virions, while the 62K is more prominent in virions than in the lysate. Despite an anticipated correlation between the types of anti-vaccinia antibodies in immune sera and the Mabs produced by cloned HBs derived from immunized mice, the most prominent polyclonal antibodies against the detergent-extractable 35K envelope component were represented, in Western blotting, by only two Mabs, 2 A-

93

ANTIBODIES

1234

5

6

7

8

32

FIG. 6. Comparison between IHD-J and -W vaccinia virus strains by Western blotting demonstrating the presence or absence of a 32K antigen. [86S]methioninelabeled lysate of IHD-J infected cells, lane 1; uninfected cell material, lane 2; IHD-J infected lysate, lanes 3 and 5; purified IHD-J vaccinia virus, lane 6; IHD-W infected lysate, lanes 4 and 7; purified IHDW vaccinia virus, lane 8. Reacted with polyclonal rabbit r IHD-J antiserum after adsorption with IHD-W infected lysate, lanes 2-4; reacted with Mab C2-3.5, lanes 5-8.

94

WILTON,

GORDON,

NP-40

FIG. 7. Dot immunoblots made with extracts of uninfected (I) or IHD-J infected cells (II). The lysates were treated with NP-40 or with SDS, prior to addition of Mabs. Material from lOa cells, equivalent to approximately 10 pg protein suspended in 10 pl, was spotted onto nitrocellulose paper and processed as described in the text. The Mabs were added in sufficient amounts to ensure that they were present in antibody excess, as determined by preliminary assays. Lane 1, no antibody; lane 2, Mab 2-1.5, 40 pl; lane 3, Mab B53.2, 10 jd.

have an affinity for the same or different epitopes remains to be established. At present we cannot explain the significance of the high frequency at which these HBs were represented among the group of 23. Among the 23 Mab binding vaccinia antigens in the immunofluorescence test, most reacted with cytoplasmic inclusions ranging widely in size. The larger foci were most likely the “factories” where nascent antigens become concentrated and virion assembly occurs (Dales and Pogo, 1981). A large fraction, 9/23, of the Mabs was reactive with antigens in both the “factories” and very small granules, scattered throughout the cytoplasm, but most numerous at the cell margin. If the tentative identification of these granules as mature virions is confirmed one could conclude that these Mabs are reactive with the same epitopes, whether the antigen occurs in precursor pools or after it has been incorporated into assembled virions. Two of our Mabs reacted with antigen uniformly distributed at the surface of infected cells. One of these, Mab Bll-5.1 studied previously (Dales et cd., 1983), was shown to cross-react with the viral HA and an epitope of the host cytoskeletal protein,

AND

DALES

vimentin. The other Mab, H3-1.4, reacted only with a virus-induced, nonvirion antigen, giving a prominent immunofluorescence of the plasma membrane and long microvilli (Fig. 3E), which become prominent during vaccinia infection (Dales and Siminovitch, 1961; Krempien et cd., 1981). Since the presence or absence of the antigen binding H3-1.4 could be correlated absolutely with circumstances which are permissive or restrictive for expression of HA (Dales et aZ., 1976), it is reasonable to assume that this Mab is specific >HA. Further confirmation of the identity of H31.4 is the matching of its affinity for antigen of a wide apparent MW range, which contains carbohydrate (Weintraub and Dales, 1974; Shida and Dales, 1981), and reacts with polyclonal, monovalent > HA rabbit antibodies (Shida and Dales, 1982). When Mab H3-1.4 was applied at appropriately low concentrations it revealed that the broad HA band consists of several components resolved into at least four distinctive minor bands (lanes 9 and 11 of Fig. 4). This heterogeneity of MWs is very likely due to qualitative and quantitative differences in the carbohydrate chains attached to the HA polypeptide backbone, which bears both the asparagine -O- linked and the serine, threonine -N- linked carbohydrate moieties (Shida and Dales, 1981). Overall, this study reveals that the immune response in mice, immunized with either live or killed vaccinia virus, evokes a spectrum of prominent antibodies such as those against the 35K envelope component, but these antibodies may be represented only rarely or not at all among Mabs produced by a selected HB panel. Conversely HBs which may appear at high frequency produce Mabs against internal antigens, such as the 32K polypeptide of IHD-J vaccinia, which are not detected by antibodies in the hyperimmune, polyclonal sera. Thus, in assessing which virus-induced components are most relevant for evoking humoral antibodies, the hyperimmune antisera are likely to be most useful. On the other hand, for analytical studies concerned with identification of individual antigens as either virion or nonvirion components, the selection of specific

VACCINIA

ANTIGENS

AND ANTIBODIES

95

S., STERN, W., WEINTRAUB, S. B., and HUIMA, T. (1976). Genetically controlled surface modifications by poxviruses influencing cell-cell and cellvirus interactions. In “Miles 9th International Symposium” (Beers, R. F., Jr. and Bassett, E. G., eds.), pp. 253-270. Raven Press, New York. DOHERTY, P. C., BLANDEN, R. V., and ZINKERNAGEL, ACKNOWLEDGMENTS R. M. (1976). Specificity of virus immune effector T cells for H-2K or H-2D compatable interactions: Samuel Dales acknowledges with gratitude the help, implications for H-antigen diversity. Transplant expertise, and use of facilities at the Scripps Clinic Rev. 29,89-124. and Research Foundation, La Jolla, California, proEASTERBROOK,K. B. (1966). Controlled degradation of vided by Michael Oldstone and his colleagues, in parvaccinia virus in vitro: An electron microscopic ticular Michael Buchmeier for selection of the MAbs. study, J. Ultrastruct. Res. 14,484-486. We thank Mitchell Zimmer for excellent technical as- GOLD, P., and DALES, S. (1968). Localization of nusistance and acknowledge the invaluable discussions cleotide phosphohydrolase activity within vaccinia. with Wayne Flintoff and Sven Beushausen concerning Proc Natl Acad. Sci. USA 60,845-852. Western blotting. HAPEL, A. J., BABLANIAN, R., and COLE, G. A. (1978). Induction requirements for the generation of virus specific T-lymphocytes. I. The nature of the host REFERENCES cell-virus interaction that triggers secondary pox BA~EIGER, B., NEWHALL, W. J. V., and JONES, R. B. virus specific cytotoxic T lymphocyte induction. J. (1982). The use of Tween-20 as a blocking agent in Immunol 121,736-743. the immunological detection of proteins transferred Hum, L. M. (1975). The immune response to infection to nitrocellulose membranes. J. Inzmund Methods with vaccinia virus in mice. I. Infection and the 55,297-307. production of antibody neutralizing cell-associated BEALE, J. (1984). Recombinant vaccinia for prevention with cell-free virus. J. Hyg 74,301-314. of hepatitis B. Nature (London) 311,12-13. ICHIHASHI, Y., and DALES, S. (1971). Biogenesis of BENNINK, J. R., YEWDELL, J. W., SMITH, G. L., MOLLER, poxviruses: Interrelationship between hemaggluC., and MOSS, B. (1984). Recombinant vaccinia virus tinin production and polykaryocytosis. virology 46, primes and stimulates influenza haemmagglutinin533-543. specific cytotoxic T cells. Nature &w&n) 311,578KIENY, M. P., LATHE, R., DRILLIEN, R., SPEHNER, D., 579. SKORY,S., SCHMITT, D., WIKTOR, T., KOPROWSKI, H., BYRNE, J. A., SOI.OSKI, M., and HOLOWCZAK, J. A. and LECOCQ,J. P. (1984). Expression of rabies virus (1983). Immune responses of DBA/2 mice bearing glycoprotein from a recombinant vaccinia virus. tumors: Cell-mediated immune responses after Nature (London) 312.163-166. challenge with vaccinia virus. Cancer Immunol. ImKOSZINOWSKI, U., and ERTL, H. (1976). Role of early munother. l&81-87. viral surface antigens in cellular immune response CHO, C. T., and WENNER, H. A. (1973). Monkey pox to vaccinia virus. Eur. J. Immud 6,679-683. vir?)s. Bactiol: Reu. 37,1-18. KREMPIEN, U., SCHNEIDER, L., HILLER, G., WEBER, K., DALES, S., FUJINAMI, R. S., and OLDSTONE, M. B. A. KATZ, E., and JUNGWIRTH, C. (1981). Conditions for (1983). Infection with vaccinia favors the selection pox virus-specific microvilli formation studied durof hybridomas synthesizing autoantibodies against ing synchronized virus assembly. Virology 113,556intermediate filaments, one of them cross-reacting 564. with the virus hemagglutinin. J. Immunol. 131, LAEMMLI, U. K. (1970). Cleavage of structural proteins 1546-1553. during the assembly of the head of bacteriophage DALES, S., and OLDSTONE,M. B. A. (1982). Localization T4. Nature (Lunobn) 227,680-684. at high resolution of antibody-induced mobilization of vaccinia virus hemagglutinin and the major his- LOH, P., and RIGGS, J. L. (1961). Demonstration of the sequential development of vaccinial antigens and tocompatibility antigens on the plasma membrane virus in infected cells: Observations with cytoof infected cells. J. Exp. Med. 156,1435-1447. chemical and differential fluorescent procedures. J. DALES, S., and POGO, B. G. T. (1981). Biology of poxExp. &fed 114,149-160. viruses. In “Virology Monographs” (D. W. KingsMACKETT, M., YILMA, T., ROSE, J. K., and Moss, B. bury and H. zur Hansen, eds.), Vol. 18, Springer(1985). Vaccinia virus recombinant% Expression of Verlag, New York. VSV genes and protective immunization of mice and DALES, S., and SIMINOVITCH, L. (1961). The developcattle. Science (Washington, D. C!) 227,433-435. ment of vaccinia virus in Earle’s L strain cells as examined by electron microscopy. J. Biophys. MALLON, V., and HOLOWCZAK, J. A. (1985). Vaccinia virus antigens on the plasma membrane of infected Biochem Cytol LO,475-503.

mRNAs from immunoprecipitated polyribosomes and mapping of vaccinia genes in which the gene products are recognized after in vitro translation, the Mabs can be usefully employed.

DALES,

96

WILTON,

GORDON.

cells. I. Viral antigens transferred from infecting virus particles and synthesized after infection. Virology 141,201-220. MIMS, C. A. (1964). Aspects of the pathogenesis of virus diseases. Bacterial. Rev. 28,30-71. Moss, B., and ROSENBLUM, E. N. (1973). Protein cleavage and poxvirus morphogenesis: Tryptic peptide analysis of core precursors accumulated by blocking assembly with rifampicin. J. hfol BioL 81,267-269. Moss, B., SMITH, G. L., GERIN, J. L., and PURCEU, R. H. (1984). Live recombinant vaccinia virus protects chimpanzees against hepatitis B. Nature (London) 311,67-68. PAOLETTI,

E., LIPINSKAS,

B. R., SAMSONOFF,

C., MER-

S., and PANICALI, D. (19&i). Construction of live vaccines using genetically engineered poxviruses: Biological activity of vaccinia virus recombinants expressing the hepatitis B virus surface antigen and the herpes simplex virus glycoprotein D. Proc Nat1 Acad Sci USA 81,193-197. PAYNE, L. (1979). Identification of vaccinia hemagglutinin polypeptide from a cell system yielding large amounts of extracellular enveloped virus. J. Viral 31,147-155. SHIDA, H., and DALES, S. (1981). Biogenesis of vaccinia: Carbohydrate of the hemagglutinin molecule. Virology 111,56-72. SHIDA, H., and DALES, S. (1982). Biogenesis of vaccinia: Molecular basis for the hemagglutinin-negative phenotype of the IHD-W strain. firologp 117,219237. SILVER, M., and DALES, S. (1982). Biogenesis of vaccinia: Interrelationship between post-translational cleavage, virus assembly, and maturation. virology 117,341-356. SILVER, M., MCFADDEN, G., WILTON, S., and DALES, S. (1979). Biogenesis of poxviruses: Role for the DNAdependent RNA polymerase II of the host during expression of late functions. Proc. Natl. Acad. Sci. USA76,4122-4125. SMITH, G. L., GODSON, G. N., NUSSENZWEIG, V., NusSENZWEIG, R. S., and Moss, B. (1984). Plasmodium knowlesi sporazoite antigen: Expression by infectious recombinant vaccinia virus. Science (Washington, D. C) 224, 397-399. CER,

AND

DALES

SOEKAWA, M., MORIGUCHI, R., MORITA, C., KITAMURA, T., and TANAKA, Y. (1977). Electron microscopical observations on the development of vaccinia, cowpox and monkey pox viruses in pig skin. Zentralbl. Bakteriol. Midrobiol. Hyg. Abt. 1 Orig. A237,425-443. STERN, W., and DALES, S. (1974). Biogenesis of vaccinia: Concerning the origin of the envelope phospholipids. Virology 62,293-306. STERN, W., and DALES, S. (1976). Biogenesis of vaccinia: Isolation and characterization of a surface component that elicits antibody suppressing infectivity and cell-cell fusion. Virology 75.323341. STERN, W., POGO, B. G. T., and DALES, S. (1977). Biogenesis of poxvirus: Analysis of the morphogenetic sequence using a conditional lethal mutant defective in envelope self-assembly. Proc Natl Acad Sci USA 74,2162-2166. TALBOT, P. J., KNOBLER, R. L., and BUCHMEIER, M. J. (1984). Western and dot immunoblotting analysis of viral antigens and antibodies: Application to murine hepatitis virus. J. Immunol. Methods. 73, 177-188. TOWBIN, H. T., STACHELIN, T., and GORDON, J. (1979). Electrophoretic transfer of proteins from polyacrylamide gels to nitrocelluose sheets: Procedure and some applications. Proc. Natl. Acad Sci USA 76,4350-4354. UEDA, Y., ITO, M., and TAGAYA, J. (1969). A specific surface antigen induced by poxvirus. virolm, 38, 180-182. WEINTRAUB, S., and DALES, S. (1974). Biogenesis of poxviruses: Genetically controlled modifications of structural and functional components of the plasma membrane. Virology 60,96-127. WESTWOOD,

J. C. N.,

ZWARTOUW,

H. T., APPLEYARD,

G., and T. TITMUSS, D. H. J. (1965). Comparison of the soluble antigens and virus particle antigens of vaccinia virus. J. Gen. Microbial 38,47-53. WIKTOR, T. J., MACFARLAN, R. I., REAGAN, K. J., DIETZSCHOLD, KIENY,

B.,

CURTIS,

M. P., LATHE,

P. J., WUNNER,

R., LECOCQ,

W.

H.,

J. P., MACKETT,

M., Moss, B., and KOPROWSKI, H. (1984). Protection from rabies by a vaccinia virus recombinant containing the rabies virus glycoprotein gene. Proc. Natl. Acad. Sci. USA 81.7194-7198.