ga and gb Glycoproteins of Herpes Simplex Virus Type 1: Two Fonns of a Single Polypeptide

Transcription

1 JOURNAL OF VIROLOGY, Dec. 1980, p X/80/ /11$02.00/0 Vol. 36, No. 3 ga and gb Glycoproteins of Herpes Simplex Virus Type 1: Two Fonns of a Single Polypeptide R. EBERLE AND RICHARD J. COURTNEY* Department ofmicrobiology, University of Tennessee, Knoxville, Tennessee Utilizing a combination of preparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis and sodium dodecyl sulfate-hydroxylapatite column chromatography, we have separated and purified the ga and gb glycoproteins of the major virus-specific glycoprotein region from herpes simplex virus type 1-infected cells. By using purified antigen preparations, antisera specific to each of these glycoproteins were produced. Immunoprecipitation from detergent extracts of infected cells and radioimmune precipitation of the purified antigens have shown that the anti-ga and anti-gb sera each recognize both the ga and the gb glycoproteins. The anti-ga serum was also shown to neutralize virus despite the presence of only minute quantities of the ga glycoprotein in virions. Pulse-chase studies have indicated that the ga and gb glycoproteins are synthesized from a common precursor polypeptide. Together, these data demonstrate that the ga and gb glycoproteins of herpes simplex virus type 1 are antigenically similar but not identical and probably represent two different forms of the same polypeptide which differ in their degree of glycosylation. The virus-specific antigenic components present on the surface of both herpes simplex virus cific glycoprotein (gb), one of the glycoprotein the actual isolation of an individual HSV-spe- (HSV)-infected cells and virions are the viral components which comprises the major glycoprotein region. This same glycoprotein was also components which are exposed to and therefore most probably recognized by both the humoral shown to be the virion glycoprotein which mediates the penetration of adsorbed virions into and cellular immune systems of the infected host. The elucidation and analysis of these surface antigens thus represents an important facet 17). Similar conclusions were reached in our the host cell via its ability to promote fusion (6, in studies directed toward providing a better laboratory with antisera produced to the individual purified glycoprotein components of the ma- understanding of both the pathogenesis and resolution of herpetic infections. The viral glycoproteins have been shown to represent the major HSV-2 (Eberle and Courtney, submitted for jor glycoprotein region of both HSV-1 (2) and viral antigens expressed on the surface of both publication). Although functions have been correlated with two of the three glycoproteins (gb HSV virions (12) and infected cells (4, 13). The viral glycoproteins have also been shown to represent the target antigens in such important region of HSV-1, the function of the third gly- and gc) which comprise the major glycoprotein immune reactions as virus neutralization (2, 14, coprotein (ga) has not been elucidated as yet. 18), antibody-dependent cellular cytotoxicity However, it is clear that this glycoprotein is not (9), and T-cell-mediated cellular cytotoxicity incorporated into virions (21). In this communication we examine the relationship of the ga (Lawman et al., submitted for publication). With the realization of their importance, the and gb glycoproteins to one another and demonstrate that the ga and gb glycoproteins of glycoproteins of HSV have become the subject of intensive investigation during recent years. HSV-1 are antigenically similar and probably Studies by Spear (20, 21) employing sodium represent two different forms of the same polypeptide. dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) were the first to resolve and MATERIALS AND METHODS definitively identify the individual glycoproteins specified by HSV types 1 and 2 (HSV-1 and Viruses, cell cultures, and media. -The KOS strain of HSV-1 was used HSV-2). Using crossed primarily throughout these immunoelectrophoresis, studies. Other HSV-1 strains employed included G, Norrild and Vestergaard identified essentially Ga, HFEM, Hi, ME, Se, and XIII, all of which were the same glycoproteins in HSV-1- and HSV-2- obtained from William Rawls (McMaster University, infected cells as were reported by Spear (10, 22). Hamilton, Ontario, Canada), and BK and CA, which Sarmiento and Spear (18) were the first to report were isolated in this laboratory from a recurrent lip 665

2 666 EBERLE AND COURTNEY lesion and from a skin biopsy of a suspected herpetic dermatitis case, respectively. Production and titration of viral stocks have been described previously (1). HEp-2 cells were propagated in Eagle minimal medium supplemented with 10% donor calf serum (DCS) and 0.075% NaHCO3 and maintained in Eagle minimal medium containing 2% DCS and 0.025% NaHCO3. In experiments where '4C-amino acids ('4C-AA) were used, cells were labeled in Eagle minimal medium with 0.1X AA supplemented with lx arginine and glutamine, 2% DCS, and 0.225% NaHCO3. Where [35S]methionine ([3S]Met) was used, Eagle medium supplemented with lx amino acids without Met, 0.lx Met, 2% DCS, and 0.225% NaHCO3 was used unless cells were pulse-labeled, in which case no unlabeled Met was present in the medium. Chemicals and radioisotopes. All chemicals used for preparative and analytical gel electrophoresis and hydroxylapatite chromatography were purchased from Bio-Rad, Richmond, Calif. D-[6-3H]glucosamine (specific activity, 10 to 25 Ci/mmol) ([3H]GluN), D-[1-4C]GluN (specific activity, 61 mci/mmol), U-'4C-AA (specific activity, 57 Ci/mAtom) and [3S]Met (specific activity, 1,200 Ci/mmol) were obtained from Amersham, Arlington Heights, Ill. Infection, labeling, and detergent extraction. HEp-2 cells were infected at high multiplicity (10 to 20 PFU/cell) and labeled from 4 to 24 h postinfection (p.i.) unless specifically noted otherwise. Harvesting of infected cells and preparation of detergent extracts with 1% sodium deoxycholate and 1% Tween 40 have been described previously (2). Disc SDS-PAGE. Details of methods employed for SDS-PAGE have been described previously (15). Unless noted otherwise, all preparative gels were crosslinked with N,N'-methylenebisacrylamide, whereas slab gels were cross-linked with N,N'-diallyltartardiamide (DATD). Pooled fractions from preparative gels were concentrated and dialyzed with an Amicon ultrafiltration cell equipped with a PM10 membrane (Amicon Corp., Lexington, Mass.). Isotopically labeled proteins in slab gels were detected by autoradiography with Kodak No-Screen NS2T X-ray film. SDS-HTP. The methodology employed for SDShydroxylapatite (SDS-HTP) chromatography was essentially that of Moss and Rosenblum (7) and has been described in detail elsewhere (2). Briefly, HTP was washed with 0.01 M sodium phosphate buffer, ph 6.4, containing 0.1% SDS to remove fines, poured into a column, and equilibrated with the same buffer containing 1 mm dithiothreitol. Samples for SDS-HTP chromatography were denatured in 0.1% SDS-1 mm dithiothreitol at 100 C for 30 s before chromatography. Polypeptides were eluted from the columns by using an increasing linear gradient of phosphate molarity. Purification of glycoproteins and preparation of antisera. The purification of the major glycoprotein region, the gc glycoprotein, and the ga/gb glycoprotein doublet has been described previously (2). Briefly, detergent extracts of HSV-1-infected cells were electrophoresed through preparative gels, and the fractions containing the major glycoprotein region were pooled, concentrated, and re-electrophoresed on a second preparative gel. The isolated glycoprotein region was then separated into the gc glycoprotein and the ga/gb component by chromatography on SDS-HTP columns. To separate the ga and gb glycoproteins from one another, the purified ga/gb component was electrophoresed on 8.6% DATD-crosslinked preparative gels. The ga and gb peaks were pooled, concentrated, and re-electrophoresed on 8.6% preparative gels cross-linked with DATD until single, electrophoretically homogeneous peaks were obtained as judged by autoradiography and Coomassie blue staining of SDS-PAGE analyses. The methods employed for use of purified glycoprotein preparations as immunogens have been described in detail previously (2). Briefly, the purified antigen preparations were electrophoresed into cylindrical gels, and segments containing the banded antigen were cut out from the gels. These gel segments were then emulsified with Freund adjuvant and injected into rabbits to produce antisera. Purification of HSV-1 virions. Details of the methodology used to obtain partially purified HSV-1 virions have been described elsewhere (3). Briefly, 5 x 10' HEp-2 cells were infected at a multiplicity of infection of 10 PFU/cell and labeled from 4 to 24 h p.i. with [3H]GluN (5 ACi/ml) and 14C-AA (1,uCi/ml). At 24 h p.i. the extracellular medium was collected and centrifuged at 14,700 x g for 30 min to pellet virions. The resulting pellet was resuspended in TNE (10 mm Tris (ph 7.4), 100 mm NaCl, and 1 mm EDTA) and layered onto two continuous 20 to 60% sucrose gradients in TNE. Gradients were centrifuged in a Beckman SW27 rotor for 1 h at 107,000 x g. The banded virus was collected and resuspended in 30 ml of TNE. Diluted virions were pelleted for 15 min at 36,000 x g and resuspended in 0.5 ml of TNE. Virions partially purified by this procedure were greater than 90% enveloped, and less than 30% contained empty capsids. Radioimmune precipitation assays. The details of methods used in J. VIROL. radioimmunoassays have been given elsewhere (2). Briefly, 25,Il of antiserum was reacted with 10 to 30 pl of antigen (either purified glycoproteins or detergent extracts of infected cells) in a final volume of 150,ul for 1 h at 370C and an additional 12 h at 40C. Staphylococcus A (20 pl) was then added to adsorb antigen-antibody complexes from solution. The resulting precipitates were washed three times and solubilized in 30 to 200 pl of 2% SDS, 1 M urea, and 2% fi-mercaptoethanol. Endpoint neutralization test. Approximately 2 x 105 PFU of HSV-1 (KOS) grown in MRC5 cells were reacted at 370C for 1 h with 8 U of fresh or heatinactivated guinea pig serum (as a complement source) and serial dilutions of the anti-glycoprotein sera in a final volume of 0.8 ml. All samples were then immediately diluted in ice-cold Eagle medium containing 2% DCS, and the remaining PFU were assayed on Vero cell monolayers. Plaquing, staining of plates, and calculation of the percent plaque reduction all have been described previously (2). Neutralization kinetics. Equal volumes of HSV- 1 (KOS) grown in MRC5 cells (4 x 106 PFU) and antisera (1:5 dilutions) were mixed in a final volume of 0.8 ml in an ice bath. Samples were then placed in a 370C water bath with constant shaking. At specified time intervals, 100-,l aliquots were withdrawn and

3 VOL. 36, 1980 immediately diluted 1:100 in ice-cold diluent. All samples were then further diluted 1:10, and 100 pl was assayed for remaining PFU. Neutralization rate constants were determined by the following formula: log (V/Vo) = kt, where Vo = PFU at zero time, V = PFU remaining at time t, t = time at 370C in minutes, and k = the rate constant for viral inactivation. Pulse-chase labeling of infected cells. For studies on the synthesis of the various glycoproteins from precursor polypeptides, cells were infected and incubated in Met-free medium. At 6 h p.i., cells were pulsed with [35S]Met (30,iCi/ml) for 15 min. Pulse samples were rapidly washed three times in ice-cold phosphatebuffered saline and lysed in ice-cold distilled water. Samples to be chased were rapidly washed three times with warm (370C) phosphate-buffered saline and medium containing 1Ox the normal concentration of unlabeled methionine was added. At varying times after the pulse-labeling period, chase samples were harvested in ice-cold distilled water. RESULTS We recently reported on the separation of the gc glycoprotein from the ga and gb glycoproteins (2). In these studies, we employed an antiserum prepared to a component highly purified by SDS-HTP chromatography which contained both the ga and gb glycoproteins (anti-ga/gb). Based on the results of experiments with the anti-ga/gb serum, the data were interpreted as if this antiserum reacted with a single antigen, i.e., that the ga and gb glycoproteins were antigenically related to one another. Further experiments to verify this premise are described below. Purification of the ga and gb glycoproteins. By using a combination of preparative SDS-PAGE and SDS-HTP chromatography, the pgc(105), gc, and ga/gb components of the major glycoprotein region of HSV-1 were purified from approximately 109 infected HEp-2 cells as previously described (2) (Fig. 1). The ga and gb glycoproteins were then separated from one another on 8.6% preparative polyacrylamide gels cross-linked with DATD (Fig. 10). By pooling the ga and gb peaks and re-electrophoresing these preparations two to three additional times on preparative DATD-cross-linked gels, we were able to obtain electrophoretically pure preparations of the ga and gb glycoproteins (Fig. 1H and I). Analysis of the final preparations of the glycoproteins on slab SDS-PAGE revealed a single polypeptide band in each preparation as determined by both autoradiography (shown in Fig. 1) and staining for protein with Coomassie brilliant blue (not shown). These antigen preparations were then used to produce specific antisera to the ga, gb, gc, and pgc(105) glycoproteins as has been described previously (2). Immunoprecipitation from infected cell MULTIPLE FORMS OF AN HSV-1 GLYCOPROTEIN 667 extracts. The specificity of the ga and gb antisera was initially investigated by determining their ability to selectively precipitate the antigens used in their production from detergent extracts of infected cells which contain a multitude of host cell and virus-specific antigens. Infected cells were (i) pulse-labeled for 15 min with [35S]Met (30,uCi/ml), (ii) pulse-labeled and chased for 2 h in medium containing an excess of unlabeled methionine, or (iii) labeled with [35S]Met (1,uCi/ml) continuously from 4 to 24 h p.i. Detergent extracts were prepared from each sample, and immunoprecipitation was performed as described in Materials and Methods. SDS-PAGE analysis of the polypeptides precipitated by both the ga antiserum and the gb antiserum from each sample is shown in Fig. 2. As is evident from the results presented here, both of these antisera specifically recognize the same polypeptide (pga) in the pulse-labeled extract which has a molecular weight intermediate between those of the ga and gb glycoproteins. Similarly, both of these antisera reacted with the ga and gb glycoproteins in the 2-h chase sample and in the sample labeled from 4 to 24 h p.i. The synthetic interrelationships of the pga, ga, and gb glycoproteins will be dealt with in greater detail below. At this point, it is sufficient to note the recognition of identical polypeptides in each sample by the two antisera prepared to the ga and gb glycoproteins. These data strongly suggest that an antigenic relationship exists between the ga and gb glycoproteins. Radioimmune precipitation of purified glycoprotein antigens. From the immunoprecipitation tests described above, it was evident that the anti-ga and anti-gb sera were each capable of reacting with both the ga and gb glycoproteins in detergent extracts. To verify this observation, it was necessary to test each antiserum for its ability to precipitate known antigens, i.e., the purified ga and gb glycoproteins. The purity of the antigens used in these experiments was comparable to that of the antigen preparations shown in Fig. 1 which were used to produce the antisera. The reactivity of the anti-glycoprotein sera with each of the purified glycoprotein antigens is presented in Table 1. In agreement with the results obtained in the immunoprecipitation tests described above, the reactivity of the anti-ga and anti-gb sera in radioimmune precipitation tests paralleled one another-both sera reacted in a similar manner with all of the antigens tested. Neither of these antisera reacted with either the gc glycoprotein or the major capsid protein of HSV-1 (VP154). Similarly, both the anti-ga and anti-gb sera specifically precipitated the ga and the gb gly-

4 668 EBERLE AND COURTNEY J. VIROL. FIG. 1. Purification of HSV-1 glycoproteins. HSV-1-infected HEp-2 cells were labeled fr-om 4 to 24 h p.i. with [3HJGluN (10 ~ici/ml) and '4C-AA (3 itci/ml) and harvested at 24 h p.i. The major glycoprotein was purified bypreparative SDS-PAGE on 7% gels cross-linked with N,N-methylenebisacrylamide (A) and further fractionated by chromatography on an SDS-HTP column (B). Material eluting fr-om the column which represented the ga/gb component was then mnitially separated into the ga and gb components by electrophoresis on a preparative 8.6% gel cross-lined with DATD (C). The ga glycoprotein was further purified by electrophoresis on an additional preparative 8.6% DATD-cross-linked gel (I). The ge peak fr-om (C) was chromatographed on additional SDS-HTP columns to separate the gb glycoprotein fr-om contaminating gc glycoprotein (E) and finally electrophoresed on an 8.6% DATD-cross-linkedpreparative gel (H). The materials from steps (B), (C), and (E) which represented the gcglycoprotein were combined and purified by electrophoresis onpreparative DATD-cross-linked gels (D, G). The finalpreparations ofeachglycoprotein were analyzed by electrophoresis on 8.6% DATD-cross-linked slab gels. The autoradiographic image of such analysis is shown at the bottom of the figure. An identicalpattern was obtained when this gel was stained with Coomassie brilliant blue to detect protein (not shown). coprotein antigens. The precipitation of both the ga and gb antigens by the anti-ga and anti-gb sera confirms the presence of antigenic sites on the ga and gb glycoproteins which are shared by these two antigens. Neutralization by the anti-ga and antigb sera. Analysis of SDS-PAGE of HSV-1 virions partially purified from the extracellular fluids of infected cells labeled with [35S]Met indicated that only small amounts of the ga glycoprotein were present in mature virions (Fig. 3); the gb and gc glycoproteins thus comprise the major glycoprotein region of virions as previously described by Spear (21). The almost total absence of the ga glycoprotein in virions led us to examine the relative ability of the antiga and anti-gb sera to neutralize virus. Should any accessible antigenic determinants of the gb glycoprotein be shared with the ga glycoprotein, it would be expected that the ga antiserum should neutralize virus despite the absence of the ga glycoprotein in virions.

5 VOL. 36, 1980 MULTIPLE FORMS OF AN HSV-1 GLYCOPROTEIN 669 Anti-gA P P C 4-24 EXT Anti-gB P P/C 4-24 EXT pga-. *-... v"gc *-gb \ga pga-. ogc s-gb ga FIG. 2. Immunoprecipitation from HSV-1-infected cell extracts with anti-ga and anti-gb sera. HSV-1- infected HEp-2 cells were labeled with [35S]Met (1 ACi/ml) from 4 to 24 h p.i. (4-24) or were pulsed with [36S]Met (30 1Ci/ml) from 6.0X to 6.25 h p.i. and either harvested immediately (P) or chased for 2 h in medium containing excess non-radioactive methionine (P/C). Detergent extracts wereprepared from all three samples, and immunoprecipitation was performed by using the antisera prepared to the ga glycoprotein and the gb glycoprotein. Immune precipitates were solubilized and analyzed by SDS-PAGE along with the extract of cells labeled from 4 to 24 h p.i. TABLE 1. Radioimmune precipitation ofpurified glycoprotein antigens by anti-glycoprotein sera Antigen Antiserum ga gb gc VP154 Preimmune 1.6a Anti-gA Anti-gB Anti-gC a Percent precipitation of [3H]GluN counts per minute present in input antigen preparation. VIRIONS VP 123 _aftno _ ga INF CELL EXT VP123 U,ij! To examine this possibility, both the ga and the gb antisera were tested for their ability to neutralize HSV-1 in the presence and absence of active complement. For comparison, an antiserum prepared to the ga/gb component was also tested. The titers of each of the antisera (expressed as 50% or greater reduction of PFU) are shown in Table 2. As is evident, both the antiga and the anti-gb sera were equally capable of neutralizing infectivity of HSV-1 virions despite the absence of significant amounts of the ga glycoprotein in virions. The 50% reduction titers were not significantly different when active or inactive complement was present during the reaction. These data are consistent with the results obtained by radioimmune precipitation tests in that the ga antiserum is capable of recognizing FIG. 3. Comparison ofthe ga glycoprotein content of partially purified virions and infected cells. Infected cels were labeled with 14C-AA (3 yci/ml) and [3HJGluN (5 yci/ml) from 4 to 24 h p.i. and virions purified from the extracellular fluids as described in Materials and Methods. The major glycoprotein region (VP123) was purified from both virions and infected cells by preparative SDS-PAGE to further demonstrate the greatly reduced amounts of the ga glycoprotein which are present in virions as compared with infected cells.

6 670 EBERLE AND COURTNEY TABLE 2. Neutralization of HSV-1 by antiglycoprotein sera Active complementa Antiserum +_ Anti-gA/gB 640b 640 Anti-gA Anti-gB a Fresh guinea pig serum (+) and heat-inactivated guinea pig serum (-) served as the active and inactive complement sources, respectively. b Titers are expressed as the reciprocal of the highest antiserum dilution which gave 50% or greater plaque reduction of 2 x 105 PFU of HSV-1 (KOS). J. VIROL. the gb glycoprotein which is present on the external surface of virions. To further investigate the neutralizing activity of the ga and gb antisera, we examined the kinetics of neutralization of HSV-1 by these two antisera. With equivalent dilutions of the two antisera under conditions of antibody excess, both the anti-ga and anti-gb sera exhibited neutralization kinetic curves consistent with one-hit kinetics (Fig. 4). However, the rate at which the two sera neutralized HSV-1 was found to be significantly different. Neutralization by both the ga and gb antisera consistently resulted in viral neutralization rate constants of approximately for the ga antiserum and for the gb antiserum. The difference in the rate at which these two antisera neutralized virus clearly demonstrates that the gb antiserum was much more effective at neutralizing viral infectivity than was the ga antiserum. These results suggest the existence of some antigenic determinants which are unique to the gb glycoprotein in addition to the common determinants of the ga and gb glycoproteins. Synthesis of the ga and gb glycoproteins. If, as the data presented above suggest, the ga and gb glycoproteins are antigenically related, it may be expected that these two glycoproteins represent different forms of the same polypeptide. If so, it is not at all improbable that one of these glycoproteins may serve as a precursor to the other. To investigate this possibility, HSV-1-infected HEp-2 cell cultures were pulsed with [35S]Met (30,uCi/ml) for 15 min at 6 h p.i. and either harvested immediately or washed and chased for varying lengths of time in media containing an excess of unlabeled methionine. Detergent extracts were prepared from all samples as well as from an infected cell culture labeled continuously from 4 to 24 h p.i. with [35S]Met (1,uCi/ml). Immunoprecipitations were then performed by using the anti-ga/gb serum and the anti-gc serum. SDS-PAGE analysis of the resulting immune precipitates is shown in Fig. 5. With the gc antiserum, any precursors, products, or synthetic intermediates occurring during the pulse-chase studies which are antigenically related to the gc glycoprotein rather than the ga or gb glycoproteins would be detected. As seen in Fig. 5, only two polypeptides were detected with the anti-gc serum. These correspond to the gc and pgc(105) polypeptides described previously by Spear (21), the pgc(105) serving as the sole detectable precursor to the gc glycoprotein. No additional intermediates or products antigenically related to the gc glycoprotein were ever detected. By using the anti-ga/gb serum to immunoprecipitate from the same samples, four distinct polypeptide species were detected (Fig. 5). In the sample harvested immediately after the pulse-labeling period, a single polypeptide designated pga was detected. This polypeptide had a relative electrophoretic mobility slightly less than that of the ga glycoprotein. Within 0.5 h, this polypeptide was quantitatively chased into the ga glycoprotein. After 0.75 h of chase, the gb glycoprotein began to reach detectable levels, its presence being distinct within 2 h after the labeling period. A fourth polypeptide, designated pgb, appeared before the accumulation of detectable levels of the gb glycoprotein. This polypeptide migrated on SDS-PAGE immediately below the ga glycoprotein and was visible in the 0.5- to 2.0-h chase samples. The appearance of this polypeptide only after substantial amounts of the ga glycoprotein were present, but before the appearance of detectable quantities of the gb glycoprotein, suggests that this polypeptide represents an intermediate in the synthesis of the gb glycoprotein from the ga glycoprotein. This conversion, however, appears to be an inefficient process in that total turnover of the ga glycoprotein into the gb glycoprotein was never observed to occur; there were always substantial levels of the ga glycoprotein to be found in infected HEp-2 cells. By using the antiga and anti-gb sera, results identical to these were obtained (see Fig. 2). Since the ga and gb glycoproteins apparently represent two different forms of a single polypeptide, the source of the observed difference in electrophoretic mobility exhibited by the ga and gb glycoproteins was of interest. Although other possibilities do exist, one of the most plausible explanations may be that these two forms differ in the degree to which the common polypeptide moiety is glycosylated, such as differences in the number and/or complexity of the oligosaccharide chains. To superfically examine this prob-

7 VOL. 36, 1980 MULTIPLE FORMS OF AN HSV-1 GLYCOPROTEIN V FIG Th Imlni Kinetics ofneutralization ofhsv-1 by anti-ga and anti-gb sera. Equal volumes ofantisera (diluted 1:5) and virus (4 x 106 PFU) were mixed and placed at 37C with constant shaking. At 4-min intervals, aliquots were removed and assayed for residual infectious virus. Results are expressed as surviving infectious centers/input infectious centers at time t. Antisera used were normal rabbit serum (U), anti-ga serum (0), and anti-gb serum (0). lem, we determined the ratio of [3H]GluN to '4C- AA incorporated into the ga and gb glycoproteins. These data are presented in Table 3. The ga, gb, gc and pgc(105) antigens were all purified as described in Fig. 1. The gc and pgc(105) glycoproteins exhibited a high level of incorporation of GluN relative to the ga and gb glycoprotein antigens, indicating that the gc glycoprotein is highly glycosylated, as has been previously suggested (21). The ratio of GluN to AA incorporated into the ga and gb glycoproteins was found not only to differ from the gc glycoprotein but to be dissimilar from one another as well, the gb form having a higher glucosamine content than the ga glycoprotein. Since the gb form migrates as a polypeptide of greater appar-

8 set1 '5 ' f 20C 4 0 r. *J81. 2.,,_ h. f,,,, d it.. i._.. * *~~~~~~~~~,: 672 EBERLE AND COURTNEY J. VIROL. ;... pgapgboo _<W1, FIG. 5. Immunoprecipitation of HSV-1 glycoproteins synthesized during pulse and pulse-chase labeling conditions. HEp-2 cells were infected with HSV-1 at an input multiplicity of infection of 20 PFU/cell and pulse-labeled for 15 min with [35S]Met (30,uCi/ml) at 6hp.i. After the 15-min pulse, one culture was harvested immediately, whereas the remaining cultures were washed and incubated in medium containing an excess of unlabeled methionine for various time intervals before harvesting. Detergent extracts were prepared from all samples, and immunoprecipitation was performed with antisera prepared to the gc glycoprotein (left panel) and the ga/gb component (right panel). Immunoprecipitates from extracts of cells labeled continuously from 4 to 24 hp.i. with [35S]Met (1,tCi/ml) are included for reference. Time intervals (in hours) between the end of the 15-min pulse-labeling period and harvest are indicated. TABLE 3. Comparison of the relative GluN content of HSV-1 glycoproteins Glycopro- GluN/AA ratioa tein Expt 1 Expt 2 Expt 3 ga 1.32 (1.0)b 2.16 (1.0) 2.11 (1.0) gb 2.17 (1.6) 3.80 (1.8) 3.22 (1.5) gc 7.15 (5.4) 7.17 (3.3) (4.7) pgc 2.77 (2.1) 5.59 (2.6) 4.29 (2.0) a Ratio of GluN to AA determined by comparison of counts per minute per 10 1u of purified glycoprotein preparations. bnumbers in parentheses represent the ratio of GluN to AA in the glycoprotein relative to the GluN/ AA ratio of the ga glycoprotein which was arbitrarily assigned a value of 1.0. ent molecular weight than the ga glycoprotein, cleavage of the polypeptide moiety during conversion of the ga into the gb glycoprotein to thus yield a higher GluN/AA ratio by removal of some AA label is unlikely. These data further suggest but do not prove that the gb glycoprotein may represent a product arising via additional glycosylation of the ga glycoprotein. However, final proof of the presence of identical polypeptide moieties in the two polypeptides will require biochemical analysis of the two glycoproteins such as AA analysis. Variation among HSV-1 strains of the ga and gb glycoproteins. If, as has been proposed by Sarmiento et al. (17), the gb glycoprotein is essential for imparting infectivity to the virion by mediating fusion of the host cell membrane and the virion envelope, then this glycoprotein would be expected to represent the product of a highly conserved gene. Therefore, it would be predicted that an antigenically similar gb glycoprotein (and therefore the ga glycoprotein as well) should be present in cells infected with all HSV-1 isolates. To investigate this premise in a rather limited way, 10 strains of HSV-1 were examined for their ability to induce the synthesis of polypeptides antigenically related to the ga and gb glycoproteins of HSV-1 (KOS). The HSV-1 strains used represented laboratory strains with a high-passage history in tissue culture (KOS, XIII), clinical oral isolates passaged 10 to 30 times in various cell lines (G, Ga, Hi, ME, Se), two clinical isolates from a recurrent lip lesion (BK) and a skin biopsy from a suspected case of herpetic dermatitis (CA), both of which were passaged once in tissue culture after their primary isolation, and a high-passage laboratory strain which induces syncytia in Vero cell cultures (HFEM). HEp-2 cells were infected with the various HSV-1 strains and labeled from 4 to 24 h p.i. with [14C]GluN (5,uCi/ml), and detergent extracts were prepared from the infected cell cultures. Immunoprecipitation with an antiga/gb serum was carried out, and the immune precipitates were analyzed by SDS-PAGE. As shown in Fig. 6, every HSV-1 strain examined produced both the ga and gb glycoproteins. With the exception of HSV-1 (HFEM), low

9 VOL. 36, 1980 MULTIPLE FORMS OF AN HSV-1 GLYCOPROTEIN 673 XIII BK G Ga HFEM Hi KOS ME S. * ga gb FIG. 6. Presence and variation of antigenically related ga and gb glycoproteins in different strains of HSV-1. HEp-2 cell cultures were infected and labeled with ["4C]GluN (3 lici/ml) from 4 to 24 h p.i. Detergent extracts were prepared from all cultures, and immunoprecipitates were solubilized and analyzed by SDS- PAGE. The virus strains used are described in Materials and Methods. levels of the gb glycoprotein were produced relative to the amount of the ga glycoprotein present in the infected HEp-2 cells in this experiment. It is also of interest to note that although slight differences in the electrophoretic mobility of the ga glycoproteins exist between the various strans, in every case this difference was reflected exactly by the gb glycoproteins. For example, the difference in electrophoretic mobility between the ga glycoproteins of HSV-1 (Hi) and HSV-1 (HFEM) was the same as the difference between the gb glycoproteins of these two HSV- 1 strains. Such parallel variation would be expected if the gb glycoprotein represents an alternate form of the ga glycoprotein polypeptide. Thus, these data are also consistent with a relationship between the ga and gb glycoproteins in the form of a common polypeptide moiety as well as indirectly supporting the functional role assigned the glycoprotein by Sarmiento et al. (17). DISCUSSION In previous communications, we reported on the isolation of the major glycoprotein region of HSV-1 and HSV-2, the further fractionation of these glycoprotein regions into subcomponents (gc and ga/gb), and the production of specific antisera to each of these components (2; Eberle and Courtney, submitted for publication). During the course of these studies, we consistently observed the presence of two polypeptide bands within the chromatographically pure SDS-HTP ga/gb components of HSV-1 and HSV-2. Should these polypeptides represent two distinct antigenic moieties, interpretation of many of the results obtained through the use of the anti-ga/ gb sera would be difficult. Consequently, it was highly desirable that the ga and gb glycoproteins be separated from one another and analyzed individually. This was accomplished by employing preparative SDS-PAGE in DATDcross-linked gels, thereby enabling us to produce specific antisera to each of these two glycoproteins. Initial testing of the anti-ga and anti-gb sera by immunoprecipitation from detergent extracts of infected cells unexpectedly revealed precipitation of both the ga and gb glycoproteins by each of the two antisera. Subsequent radioimmune precipitation tests employing purified glycoprotein antigens confirmed these results in that the anti-ga and anti-gb sera were each found to react with both the ga and gb glycoproteins. Similarly, both of these antisera neutralized virus despite the almost total absence of any ga glycoprotein in the virion. These results all support the contention that the ga and gb glycoproteins possess common antigenic sites and that they may represent different forms of a single polypeptide. Since the ga and gb glycoproteins appeared to be antigenically related, the ga/gb antiserum was considered specific for a single polypeptide antigen. Using the anti-ga/gb, anti-ga, and anti-gb sera in pulse-chase studies, we were able to identify four polypeptides which are antigenically related to the ga and gb glycoproteins. The sequential appearance of four antigenically related polypeptides, only two of which were

10 674 EBERLE AND COURTNEY stable (ga and gb) and one of which (pgb) appeared only after the accumulation of the first stable polypeptide (ga), provides strong evidence for the synthesis of both the ga and gb glycoproteins from a single polypeptide. In support of the synthesis of the ga and gb glycoproteins from a common precursor is the immunoprecipitation of the pga polypeptide from pulselabeled infected cell extracts by both the antiga and anti-gb sera. That both of these antisera can recognize not only the "heterologous" glycoprotein but the precursor pga polypeptide as well lends support to the contention that the ga and gb glycoproteins are derived from a common precursor (pga) and share at least a portion of their polypeptide moiety with one another. Using an antiserum prepared to virion glycoproteins, Spear was able to identify a similar set of polypeptides in pulse-labeled infected cells (21). Spear identified the ga, pgb, and gb glycoprotein species and postulated the synthesis of the gb glycoprotein from the pgb species. However, perhaps due to the presence of the pgb glycoprotein in the pulse-labeled sample, it was concluded that the ga and gb glycoproteins represented distinct polypeptides which were unrelated to one another. Aside from the different temporal appearance of the pgb polypeptide and the minor difference in the electrophoretic mobility of the ga and pga glycoproteins, our results essentially parallel those of Spear. In our hands the pgb glycoprotein is not detectable before the accumulation of substantial amounts of the ga glycoprotein. Therefore, it appears that the pgb intermediate cannot be synthesized directly from the pga polypeptide but rather must be derived from the ga glycoprotein. The gb and pgb glycoproteins are probably both derived from the ga glycoprotein via an additional glycosylation step(s). This conversion, however, does appear to be a relatively inefficient process in HEp-2 cells infected with HSV- 1 (KOS) in that the ga glycoprotein was never observed to be quantitatively converted into the gb form of this polypeptide. This process is apparently more efficient in MRC5 cells as opposed to HEp-2 cells and in HSV-2-infected cells, where the amount of the gb glycoprotein formed is greater than the amount of the ga glycoprotein present in infection (Eberle and Courtney, unpublished data). The rapid replication of HSV-1 and HSV-2 in MRC5 cells and of HSV-2 in HEp-2 cells suggests that the efficiency of formation of the gb glycoprotein from the ga glycoprotein may be related to the overall efficiency of the HSV replication process. However, more data are required to substantiate this hypothesis. J. VIROL. From neutralization kinetic tests employing the anti-ga and anti-gb sera, it was readily evident that the anti-gb serum was much more effective at neutralizing HSV-1 than was the anti-ga serum despite the nearly identical endpoint neutralization titers of the two antisera. This suggests that rather than having a greater number of antibody molecules directed to the gb glycoprotein, the gb antiserum recognizes certain antigenic determinants on the gb glycoprotein which the ga antiserum either cannot recognize or for which there is a reduced affinity. Although these two sera may both interact with the same region of the gb glycoprotein, this region may be sufficiently altered (antigenically) from the corresponding region of the ga form such that antibodies synthesized in response to the corresponding region of the ga glycoprotein exhibit a reduced affinity for this region of the gb glycoprotein. In any case, these results suggest that although they are antigenically very similar to one another, the ga and gb glycoproteins may not be antigenically identical despite their common polypeptide moiety. The cause of this antigenic diversity between the two forms is most probably due to the additional carbohydrate which is present on the gb form of this polypeptide. Evidence for such carbohydrate-induced antigenic variation has recently been reported with Semliki Forest virus glycoproteins (5) Ȧlong a similar line, the corresponding ga and gb glycoproteins of HSV-2 also appear to possess antigenic determinants in common with one another. Using an antiserum prepared to the HSV-2 ga/gb antigen, a temporal order of synthesis of the HSV-2 ga and gb glycoproteins from a common pga polypeptide has also been demonstrated (Eberle and Courtney, unpublished data). In addition, neutralization studies employing an anti-hsv-2 ga/gb serum suggest that this antigen is closely associated with virion infectivity, much as has been shown in the HSV- 1 system (17; Eberle and Courtney, submitted for publication). Thus, the ga and gb glycoproteins of HSV-2 appear to be analogous to the ga and gb glycoproteins of HSV-1 in many respects. In further support of this conclusion is the fact that the ga/gb antigens of HSV-1 and HSV-2 exhibit extensive cross-reactivity (8, 11; Eberle and Courtney, unpublished data). Related to both the efficiency of formation of the gb glycoprotein and the cross-reactivity of the ga/gb glycoproteins of HSV-1 and HSV-2 is the function which the gb glycoprotein serves in virion infectivity. As shown by Sarmiento et al. (17), the gb glycoprotein of HSV-1 appears to be responsible for mediating penetration of

11 VOL. 36, 1980 adsorbed virions into the host cell by promoting fusion of the virion envelope with the host cell plasma membrane. Any viral antigen providing a function which is mandatory for virion infectivity such as the ga/gb antigen appears to do would be expected to represent the product of a highly conserved gene since mutations within these genes could easily prove "lethal" to the virus. In exaining a variety of HSV-1 strains, the ga/gb antigen was found to be produced by all of the virus strains, regardless of their passage history or anatomical site of isolation. The antigenic relatedness of the ga and gb glycoproteins of all HSV-1 isolates examined was also demonstrated by the ability of the antiserum prepared to the ga/gb glycoproteins of HSV-1 (KOS) to immunoprecipitate the ga and gb glycoproteins of all the other HSV-1 strains. These observations, taken together with the colinear mapping of the ga/gb genes of HSV-1 and HSV-2 as reported by Ruyechan et al. (16) and the presence of type-common antigenic determinants on the ga and gb glycoproteins of HSV-1 and HSV-2, lends further credence to the notion that the gene encoding the ga/gb polypeptide is vital for the viability of the HSVs and consequently has remained relatively unchanged throughout the evolutionary divergence of HSV- 1 and HSV-2. ACKNOWLEDGMENT This investigation was supported by Public Health Service grant CA 24,564 from National Cancer Institute. LITERATURE CITED 1. Bone, D. R., and R. J. Courtney A temperaturesensitive mutant of herpes simplex virus type 1 defective in the synthesis of the major capsid polypeptide. J. Gen. Virol. 24: Eberle, R., and R. J. Courtney Preparation and characterization of specific antisera to individual glycoprotein antigens comprising the major glycoprotein region of herpes simplex virus type 1. J. Virol. 35: Farber, F. E., and R. Eberle Effects of cytochalasin and alkaloid drugs on the biological expression of herpes simplex virus type 2 DNA. Exp. Cell Res. 103: Glorioso, J. C., and J. W. Smith Immune interactions with cells infected with herpes simplex virus: antibodies to radioiodinated surface antigens. J. Immunol. 118: Kaluza, G., R. Rott, and R. T. Schwarz Carbohydrate-induced conformational changes of Semliki Forest virus glycoproteins determine antigenicity. Virology 102: Manservigi, R., P. G. Spear, and A. Buchan Cell fusion induced by herpes simplex virus is promoted and suppressed by different glycoproteins. Proc. Natl. Acad. Sci. U.S.A. 74: Moss, B., and E. N. Rosenblum Hydroxylapatite MULTIPLE FORMS OF AN HSV-1 GLYCOPROTEIN 675 chromatography of protein-sodium dodecyl sulfate complexes. A new method for the separation of polypeptide subunits. J. Biol. Chem. 247: Norrild, B., H. Ludwig, and R. Rott Identification of a common antigen of herpes simplex virus, bovine herpes mammalitis virus, and B virus. J. Virol. 26: Norrild, B., S. L. Shore, and A. J. Nahmias Herpes simplex virus glycoproteins: participation of individual herpes simplex virus type 1 glycoprotein antigens in immune cytolysis and their correlation with previously identified glycopolypeptides. J. Virol. 32: Norrild, B., and B. F. Vestergaard Polyacrylamide gel electrophoretic analysis of herpes simplex virus type 1 immunoprecipitates obtained by quantitative immunoelectrophoresis in antibody-containing agarose gel. J. Virol. 22: Norrild, B., and B. F. Vestergaard Immunoelectrophoretic identification and purification of herpes simplex virus antigens released from infected cells in tissue cultures. Intervirology 11: Olshevsky, U., and Y. Becker Surface glycopeptides in the envelope of herpes simplex virions. Virology 50: Pauli, G., and H. Ludwig Immunoprecipitation of herpes simplex type 1 antigens with different antisera and human cerebrospinal fluids. Arch. Virol. 53: Powell, K. L, A. Buchan, C. Sim, and D. H. Watson Type-specific protein in herpes simplex virus envelope reacts with neutralizing antibodies. Nature (London) 249: Powell, K. L, and R. J. Courtney Polypeptides synthesized in herpes simplex virus type 2-infected HEp-2 cells. Virology 66: Ruyechan, W. T., L. S. Morse, D. M. Knipe, and B. Roizman Molecular genetics of herpes simplex virus. fl. Mapping of the major viral glycoproteins and of the genetic loci specifying the social behavior of infected cells. J. Virol. 29: Sarmiento, M., M. Haffey, and P. G. Spear Membrane proteins specified by herpes simplex viruses. Im. Role of glycoproteins VP7 (B2) in virion infectivity. J. Virol. 29: Sarmiento, M., and P. G. Spear Membrane proteins specified by herpes simplex viruses. IV. Conformation of the virion glycoprotein designated VP7 (B2). J. Virol. 29: Sim, C., and D. H. Watson The role of type specific and cross reacting structural antigens in the neutralization of herpes simplex virus types 1 and 2. J. Gen. Virol. 19: Spear, P. G Glycoproteins specified by herpes simplex virus type 1: their synthesis, processing and antigenic relatedness to HSV-2 glycoproteins, p In G. dethe, M. A. Epstein and H. zurhausen (ed.), Qncogenesis and herpes viruses II, part 2. International Agency for Research on Cancer Scientific Publications no. 11. Intemational Agency for Research on Cancer, Lyon, France. 21. Spear, P. G Membrane proteins specified by herpes simplex virus type 1. I. Identification of four glycoprotein precursors and their products in type 1 infected cells. J. Virol. 17: Vestergaard, B. F., and B. Norrild Crossed immunoelectrophoresis of a herpes simplex virus type 1- specific antigen: immunological and biochemical purification. J. Infect. Dis. 138: