- Email: [email protected]
Characterization of hepatitis A virus capsid proteins with antisera raised to recombinant antigens
Journal of Virological Methods, 32 (1991) 213-220 0 1991 Elsevier Science Publishers B.V. / 0168.8510/91/$03.50
ADONIS0168851091
213
125F
VIRMET 01156
Characterization of hepatitis A virus capsid proteins with antisera raised to recombinant antigens Bruce C. Ross’ and David A. Anderson* ‘Department of Clinical Pathology and 2Macfarlane Burnet Centrefor Medical Research, Fairfield Hospital, Victoria. Australia
(Accepted 13 December 1990)
Summary The capsid proteins of hepatitis A virus (HAV) were expressed as fusion proteins of P-galactosidase in E. coli using the expression vector hgtl 1. Four fusion proteins were stably expressed and used to immunize rabbits to obtain mono-specific antisera. The antisera were unable to neutralize viral infectivity or react with HAV by radioimmunoassay. Three of the antisera were able to recognize HAV antigens in infected BS-C-l cells by immunofluorescence and denatured capsid proteins by immunoblot analysis. The antisera were used to investigate the migration of the capsid proteins in gels by immunoblot analysis using standard SDS-PAGE conditions and in gels containing urea. The migration of VP1 and VP3 correlated with their molecular weights predicted from the nucleotide sequence and was consistent in either the presence or absence of urea. However, VP2 migrated with an apparent molecular weight significantly higher than the predicted value and, in gels containing urea, migrated as a doublet. It is proposed that the upper band of this doublet represents VPO, the proteolytic precursor of VP2 and VP4. The relative molecular mass (MJ of VP4 was estimated to be less than 1 kDa, which is substantially lower than the 2.5 kDa predicted from the nucleotide sequence. Hepatitis A virus; Capsid protein; Expression
Introduction Hepatitis A virus (HAV) is a positive-strand RNA virus which has been classified within the Picornaviridae (Gust et al., 1983). A genome organization for HAV has been proposed from the predicted amino acid sequence using poliovirus as a comparative model (Najarianet al., 1985; Cohenet al., 1987). The location of some of the gene Correspondence
to:
B.C. Ross, Dept. of Clinical Pathology, Fairfield Hospital, Victoria, 3078, Australia.
214
products of HAV have been confirmed directly using antisera raised to synthetic peptides (Weitz et al., 1986; Wheeler et al., 1986) and amino acid sequencing of virion proteins (Linemeyer et al., 1985; Gauss-Muller et al., 1986). Although these studies have determined the primary amino acid sequence of the capsid proteins and VPg, the proteolytic processing of the HAV polyprotein and morphogenesis are poorly understood. The predicted amino acid sequence of HAV suggests that the HAV capsid is composedoffourstructuralproteins VP1 (33.2 kDa), VP2 (24.8 kDa), VP3 (27.8 kDa) and VP4 (2.5 kDa). Studies of the HAV capsid proteins by SDS-PAGE have determined that the observed molecular weight for VP1 corresponds to the predicted value but the estimates of VP2 and VP3 appear variable. Amino acid sequence data obtained from gel-purified VP3 (Linemeyer et al., 1985) and HPLC-purified VP2 (Gauss-Muller et al., 1986) indicated that VP3 was the faster migrating protein with an apparent relative molecular weight (M,) of 29 kDa. Conversely, immunoblot analysis with antisera raised to synthetic peptides determined that VP2 migrated faster with an apparent M, of 27 kDa (Wheeler et al., 1986). This discrepancy was reportedly due to the presence of urea in the gels affecting the migration of the proteins (Wheeler et al., 1986), although this was not directly demonstrated. Several studies have detected a protein migrating slower than VP2 (Hughes et al., 1984; Hughes and Stanton, 1985; GaussMuller et al., 1986) and have proposed that this protein may represent the precursor of VP2 and VP4, VPO. In an effort to clarify the migration of the capsid proteins in gels and obtain antibody reagents to enable a detailed study of morphogenesis, we aimed to express the capsid proteins in E. cc&. This paper describes the expression of the capsid proteins of HAV in E. coli as fusion proteins of P-galactosidase, the characterization of the antisera raised to these proteins and the identification of the capsid proteins of the cytopathic HM 175 strain of HAV. The preparation of the HAV cDNA clones was described previously (Ross et al., 1986). Restriction enzyme fragments from the cDNA clones were inserted into bacteriophage agtl 1 (Young and Davis, 1983) after blunt-ending with Klenow polymerase and attachment of EcoRI linkers using standard techniques (Maniatis et al., 1982). The regions of the HAV genome selected for expression are depicted in Fig. 1. Four recombinant phages were constructed: hFPl containing the entire coding region of VP2, 57 nucleotides of VP4 and 219 nucleotides of VP3; hFP2 containing the coding region of VP 1, 135 nucleotides of VP3 and 387 nucleotides of 2A; kFPVP2 containing 450 nucleotides of VP2; hFPVP3 containing 405 nucleotides of VP3 and 162 nucleotides of VPl. Together, these fragments contain each of the regions coding for the antigenic epitopes identified by Wheeler et al. (1986). Ligated DNA was packaged into phage proteins using an in vitro packaging kit (Amersham, England). Plaques containing recombinant phage with the desired insert were identified by plaque hybridization (Benton and Davis, 1977). Phage containing HAV cDNA inserts were used to infect E. coli Y 1089 cells at a multiplicity of 0.1 and grown at 37°C to an A 650of 0.4. The culture was then incubated at 45°C for 15 min, made up to 10 mM with isopropyl-P-o-thio-galactopyranoside (IPTG) and incubated for 2 h at 37°C. Cells were pelleted by centrifugation at 12000 x
215
3’
=’ 0
1
2
3
4kb
HAV genome Hincll Xbal
Restriction enzyme sites
BstEll
I
I
Bglll
BamHl Avall
Hindlll
I
hFP1
cDNA fragments subcloned for expression
IFPVPS
I
I____
LFPP
hFPVP3
Fig. 1, Fragments of the HAV genome inserted into bacteriophage &et1 I for expression in E. co/i. The locations of the polyprotein cleavage sites have been described previously (Najarian et al., 1985: Cohen et al., 1987) and are labelled using the picomavirus L434 nomenclature (Rueckert and Wimmer, 1984). Restriction enzyme sites used for the construction of recombinant phage are indicated.
g for 5 min and lysed in 2% SDS, 10 mM NaP04 (pH 6.5) and 1 mM dithiothreitol. Lysates were subjected to SDS-PAGE and examined for the presence of a stable fusion protein. Approximately one half to one third of the plaques containing a HAV cDNA insert were able to produce a stable fusion protein and each protein migrated in gels consistent with their predicted molecular weight (results not shown). Fusion proteins were purified from E. coli lysates by preparative SDS-PAGE for animal immunizations. After electrophoresis, gels were stained with Coomassie blue and the protein band representing the fusion protein was excised. The fusion protein was then electroeluted and excess SDS was removed by dialysis against 0.1 M ammonium bicarbonate. Rabbits were immunized intramuscularly with approximately 50 pg per dose of purified fusion protein in 1: 1 protein:Freunds’ complete adjuvant (FCA). The level of response to P-galactosidase was monitored by indirect radioimmunoassay and after 4-5 doses a stationary titre of p-galactosidase specific IgG was obtained. Mouse polyclonal ascites fluid was produced by a method obtained from Dr. W. Maskill, Fairfield Hospital (personal communication). White male Swiss mice were given 0.2 ml intraperitoneal injections of 9:l FCA:virus at weekly intervals. After approximately 3 weeks the mice produced ascites fluid, which was harvested regularly and the injections were continued until ascites fluid ceased to be produced. The response of the immunized rabbits to I)-galactosidase and HAV was determined by radioimmunoassay. For P-galactosidase a commercially available protein preparation was used as antigen (Pharmacia, Sweden) in a standard indirect radioimmunoassay procedure to detect IgG. Anti-HAV antibodies were measured using an
216
antigen capture radioimmunoassay as previously described (Coulepis et al., 1985). Each rabbit immunized with either purified p-galactosidase or fusion protein responded to P-galactosidase with a P/N ratio (post-immune divided by pre-immune) of greater than 15 at a serum dilution of 1: 10 000. None of the animals receiving fusion protein responded significantly to HAV by radioimmunoassay when tested at a serum dilution of 150. All P/N ratios were 1.4 or less whereas sera from a rabbit immunized with HAV virions reacted with a P/N of 17.5. The fusion protein antisera were also tested for their virus neutralizing activity using the radioimmunofocus assay described by Lemon et al. (1983) and modified by Anderson et al. (1987). All antisera raised to recombinant proteins were negative by neutralization while the antiserum raised to HAV showed a 99% reduction in viral infectivity (results not shown). The failure of the rabbit antisera raised to these proteins to react with HAV by radioimmunoassay or neutralize viral infectivity is similar to a previous study in which a fusion protein of HAV VP1 with the TrpE gene of E. coli also failed to elicit a neutralizing antibody response (Johnston et al., 1988). These results are consistent with earlier studies in which the binding of polyclonal and monoclonal antibodies is prevented if the virion is denatured by treatment with SDS (Gerlich and Frosner, 1983; Hughes et al., 1984) and gel-purified capsid proteins have been shown to elicit poor antibody responses to HAV by competitive radioimmunoassay and neutralization (Hughes and Stanton, 1985). This data implies that the major neutralization site of HAV is strictly conformationally dependent and may require the interaction of several capsid proteins for immunogenicity, although Gauss-Muller et al. (1990) have demonstrated priming for a neutralizing response using HAV-P-galactosidase fusion proteins. Indirect immunofluorescence was used to detect antibodies that react with intracellular viral proteins in HAV-infected cells (results not shown). Immunofluorescence with antisera raised against the fusion proteins FPl, FP2 and FPVP2 showed small cytoplasmic and perinuclear foci of fluorescence in infected cells. The pattern of cytoplasmic and perinuclear fluorescence was similar to that observed with polyclonal antibodies raised to HAV. The failure of the antisera to react by radioimmunoassay or neutralization implies that the foci of HAV antigen detected probably represent accumulations of capsids within the cytoplasm which have been partially denatured during fixation of the cells in acetone. Rabbit antisera raised to the fusion proteins were reacted with immunoblotted HAV virion proteins after SDS-PAGE (Fig. 2). Anti-FP2 reacted with a protein migrating at 33 kDa which is consistent with the M, of VP1 . Anti-FPl and anti-FPVP2 appear to react with the same protein migrating with an M, of 29 kDa and presumably represents VP2, because FPVP2 did not contain any component of VP3. However, the virion proteins VP2 and VP3 migrate with similar mobilities in gels and further analysis was necessary to confirm the identity of this protein. A previous study suggested that the migration of VP2 and VP3 varies, depending on whether urea is used in the gel (Wheeler et al., 1986). In an effort to obtain optimal separation of these capsid proteins, gels were run in the presence and absence of urea before immunoblotting with the fusion protein antisera and a mouse polyclonal ascites fluid raised to HAV which reacts with all capsid proteins. In the absence of
217
1234
5
92.5 6946-
30-
Fig. 2. Immunoblot analysis of HAV capsid proteins with the fusion protein antisera after electrophoresis through 12% SDS-polyacrylamide gels. Lanes contain molecular weight markers (lane I), or HAV virions immunoblotted with anti-FP2 (lane 2), anti-FPl (lane 3), anti-FPVP2 (lane 4) and anti-FPVP3 (lane 5).
urea the mouse polyclonal ascites fluid identified three proteins migrating with apparent M,s of 33,29, and 28 kDa (Fig. 3A) which represent VPl, VP2 and VP3. The antisera raised to FPl and FPVP2 both recognized the 29 kDa protein which represents VP2. Immunoblot analysis of HAV virions after SDS-PAGE with 3.5 M urea (Fig. 3B) resolved the capsid proteins into four bands with apparent M,s of 33,30.6,30 and 28 kDa. Under these conditions the antiserum to FPl (Fig. 3B, lane 3) and FPVP2 (Fig. 3B, lane 4) reacted with a protein doublet migrating with apparent M,s of 30.6 and 30 kDa. The M,s of the capsid proteins estimated from these gels and their comparison with the predicted sizes of the capsid proteins are summarized in Table 1. The migration of VP1 and VP3 in SDS-PAGE was consistent with their predicted M,s and was not altered by the presence of urea (Table 1). However, the putative VP2 protein recognized by antisera to FPl and FPVP2 migrated with an apparent M, considerably higher than its predicted value and, in gels containing urea, migrated as a doublet. This migration of the putative VP2 appears to differ from two previous studies of the HAV capsid proteins. In gels without urea VP2 migrates with an apparent molecular weight higher than that of VP3 (Fig. 3A) which is contrary to the finding of Wheeler et al. (1986). In gels containing urea the electrophoretic profile of the capsid proteins detected with the mouse anti-HAV polyclonal ascites fluid (Fig. 3B, lane 2) is different to that reported by Gauss-Muller et al. (1986) for mature HAV particles banding at 1.32 g/ml in CsCl gradients. However, our results appear similar to another HAV study in which virion proteins were purified from acrylamide gels containing
218
92.5
92.5
69
69
46
46
30
30
Fig. 3. ~mmunoblot analysis of HAV virions after electrophoresis through 12% SDS-polyac~lamide gels under either standard conditions (A) or in the presence of 3.5 M urea(B). Nitrocellulose strips were reacted with mouse anti-HAV ascites fluid (lane 2) rabbit anti-FPl (lane 3), rabbit anti-FPVPZ (lane 4) and rabbit anti-FPVPS (lane 5). The molecular weights of protein standards in lane 1are indicated. TABLE 1 Molecular mass estimates of HAV capsid proteins from immunoblot analysis (Fig. 3A,B) compared with their predicted values (Cohen et al., 1987) Protein
Predicted
VPI VP0
.33.2 kDa 27.3 kDa
VP2 VP3
24.8 kDa 27.8 kDa
Observed Non-urea gel
Urea gel
32.6 kDa
33.0 kDa 30.6 kDa 29.1 kDa 30.0 kDa 28.1 kDa
28.3 kDa
urea (Hughes and Stanton, 1985). One explanation for differences in the migration of the HAV proteins in these studies is the varying conditions in SDS-PAGE between laboratories. It is a well established phenomenon that many factors such as the source of SDS can contribute todifferences in protein migration. This is p~i~ularly relevant considering the small difference in migration between VP2 and VP3 under standard SDS-PAGE conditions and the finding that the migration of VP2 is profoundly affected by urea. However,
219
some of the apparent differences are substantial. For example, there is a 2-kDa shift in the migration of VP2 when compared to Wheeleret al. (1986) resulting in an inversion of VP2 and VP3. It is possible that this difference could be variation between strains of HAV which may have altered VPO/VP2 cleavage junctions. Although this possibility is unlikely, it cannot be ruled out until direct comparisons between strains in the same laboratory are made and the contribution of factors affecting the migration of HAV capsid proteins in SDS-PAGE are determined. The doublet identified with the antisera to VP2 probably represents VP0 and VP2. Recent studies in our laboratory on the analysis of sub-viral particles by sucrose gradient ultracentrifugation has demonstrated a precursor/product relationship between these bands (Anderson and Ross, 1990). The upper band of the doublet is present in pentamers, procapsids, and provirions whilst the lower band is present in provirions and virions. This data is consistent with the morphogenesis described for other picomaviruses (Rueckert, 1985) and indicates that the doublet represents VP0 and VP2. Further studies need to be completed to precisely define the morphogenesis of HAV which, on the basis of the data presented here, suggests that the M, difference between VP0 and VP2 is significantly smaller than the predicted value. Even considering the possible trimming of the amino terminus of the HAV polyprotein for processing by myristilation (Chow et al., 1987), the difference between VP0 and VP2 should be approximately 2 kDa. We are attempting to answer these questions by introducing mutations into a complete infectious copy of the genome to determine if cleavage at the proposed VPO/VP2 junction is occurring and continuing our studies on the sub-viral particles of HAV.
Acknowledgements The authors would like to thank Scott Bowden for his invaluable editorial assistance and Mr. H. Weatherley for providing the Maroline M. Weatherley Scholarship which supported BCR during this work. These studies were supported by a grant from the National Health and Medical Research Council. Australia.
References Anderson, D.A., Locamini, S.A., Ross, B.C., Coulepis, A.G., Anderson, B.N. andGust, I.D. (1987) Singlecycle growth kinetics of hepatitis A virus in BSC-1 cells. In: M.A. Brinton and R.R. Rueckert (Eds), Positive Strand RNA Viruses. Alan R. Liss, New York, pp. 497-507. Anderson, D.A. and Ross, B.C. (1990) Morphogenesis ofhepatitis A virus: isolation and characterization of subviral particles. J. Virol. 64,5284-5289. Benton, W.D. and Davis, R.W. (1977) Screening of hgf recombinant clones by hybridization to single plaques in situ. Science 196,180-l 82. Chow, M., Newman, J.F.E., Filman, D., Hogle, J.M., Rowlands, D.J. and Brown, F. (1987) Myristilation of picomavirus capsid protein VP4 and its structural significance. Nature 327,482486. Coulepis, A.G., Veale, M.F., MacGregor A., Komitschuk, M. and Gust, I.D. (1985) Detection of hepatitis
220 A virus and antibody by solid-phase radioimmunoassay and enzyme-linked immunosorbent assay with monoclonal antibodies. J. Clin. Microbial. 22, 119-124. Cohen, J.I., Ticehurst, J.R., Purcell, R.H., Buckler-White, A. and Baroudy, B.M. (1987) Complete nucleotide sequence of wild-type hepatitis A virus: comparison with different strains of hepatitis A virus and other picornaviruses. J. Virol. 61,50-59. Gauss-Muller, V., Lottspeich F. and Deinhardt, F. (1986) Characterization of hepatitis A virus structural proteins. Virology 155,732-736. Gauss-Muller, V., Mingquan, Z., von der Helm, K. and Deinhardt, F. ( 1990) Recombinant proteins VP I and VP3 of hepatitis A virus prime for neutralizing antibody response. J. Med. Virol. 3 I,2777283 Gerlich, W.H. and Frosner, G.G. (1983) Topology and immunoreactivity of capsid proteins in hepatitis A virus. Med. Microbial. Immunol. 172, 101-106. Gust, I.D., Coulepis, A.G., Feinstone, S.M., Locarnini, S.A., Moritsugu, Y., Nejera, R. and Siegl, G. (I 983) Taxonomic classification of hepatitis A virus. Intervirology 20, t-7. Hughes, J.V. and Stanton, L.W. (I 985) Isolation and immunizations with hepatitis A virus structural proteins: induction of antiprotein, antiviral, and neutralizing responses. J. Virol. 55.395401. Hughes, J.V., Stanton, L.W., Tomassini, J.E., Long, W.J. and Scolnick, E.M. (1984) Neutralizing monoclonal antibodies to hepatitis A virus: partial localization of a neutralizing antigenic site. J. Virol. 52, 465473. Johnston, J.M., Harmon, S.A., Binn, L.N., Richards, O.C., Ehrenfeld, E. and Summers, D.F. (1988) Antigenie and immunogenic properties of a hepatitis A virus capsid protein expressed in Escher-ichia co/i. J. Infect. Dis. 157, 1203-121 I. Lemon, S.M., Binn, L.N. and Marchwicki, R.H. (1983) Radioimmunofocus assay forquantitation of hepatitis A virus in cell culture. J. Clin. Microbial. 17,834-839. Linemeyer. D.L., Menke, J.G., Martin-Gallardo, A., Hughes, J.V., Young, A. and Mitra, S.W. (1985) Molecular cloning and partial sequence of hepatitis A viral cDNA. J. Virol. 54,247-255. Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, New York. Najarian, R., Caput. D., Gee, W., Potter, S.J., Renard, A., Merryweather, J., van Nest, G. and Dina, D. (1985) Primary structure and gene organization of human hepatitis A virus. Proc. Natl. Acad. Sci. USA 82, 2627-263 1. Ross, B.C.. Anderson, B.N., Coulepis, A.G., Chenoweth, M.P. and Gust, I.D. (1986) Molecular cloning of cDNA from hepatitis A virus strain HMI 75 after multiple passages in vivo and in vitro. J. Gen. Virol. 67, 1741-1744. Rueckert, R.R. (I 985) Picornaviruses and their replication. In: B.N. Fields, D.M. Knipe, R.M. Chanock, J.L. Melnick, B. Roizman and R.E. Shope (Eds), Virology. Raven Press, New York, pp. 705-738. Rueckert. R.R. and Wimmer, E. (1984) Systematic nomenclature of picomavirus proteins. J. Virol. 50, 957-959. Weitz, M., Baroudy. B.M.. Maloy. W.L., Ticehurst, J.R. and Purcell, R.H. (1986) Detection of a genomelinked protein (VPg) of hepatitis A virus and its comparison with other picomaviral VPgs. J. Virol. 60, 1244130. Wheeler, C.M., Robertson, B.H.. van Nest, G., Dina, D., Bradley, D.W. and Fields, H.A. ( 1986) Structure of the hepatitis A virion: peptide mapping of the capsid region. J. Virol. 58,307-3 13. Young. R.A. and Davis, R.W. (1983) Efficient isolation of genes by using antibody probes. Proc. Natl. Acad. Sci. USA 80,1194-l 198.
ADONIS0168851091
213
125F
VIRMET 01156
Characterization of hepatitis A virus capsid proteins with antisera raised to recombinant antigens Bruce C. Ross’ and David A. Anderson* ‘Department of Clinical Pathology and 2Macfarlane Burnet Centrefor Medical Research, Fairfield Hospital, Victoria. Australia
(Accepted 13 December 1990)
Summary The capsid proteins of hepatitis A virus (HAV) were expressed as fusion proteins of P-galactosidase in E. coli using the expression vector hgtl 1. Four fusion proteins were stably expressed and used to immunize rabbits to obtain mono-specific antisera. The antisera were unable to neutralize viral infectivity or react with HAV by radioimmunoassay. Three of the antisera were able to recognize HAV antigens in infected BS-C-l cells by immunofluorescence and denatured capsid proteins by immunoblot analysis. The antisera were used to investigate the migration of the capsid proteins in gels by immunoblot analysis using standard SDS-PAGE conditions and in gels containing urea. The migration of VP1 and VP3 correlated with their molecular weights predicted from the nucleotide sequence and was consistent in either the presence or absence of urea. However, VP2 migrated with an apparent molecular weight significantly higher than the predicted value and, in gels containing urea, migrated as a doublet. It is proposed that the upper band of this doublet represents VPO, the proteolytic precursor of VP2 and VP4. The relative molecular mass (MJ of VP4 was estimated to be less than 1 kDa, which is substantially lower than the 2.5 kDa predicted from the nucleotide sequence. Hepatitis A virus; Capsid protein; Expression
Introduction Hepatitis A virus (HAV) is a positive-strand RNA virus which has been classified within the Picornaviridae (Gust et al., 1983). A genome organization for HAV has been proposed from the predicted amino acid sequence using poliovirus as a comparative model (Najarianet al., 1985; Cohenet al., 1987). The location of some of the gene Correspondence
to:
B.C. Ross, Dept. of Clinical Pathology, Fairfield Hospital, Victoria, 3078, Australia.
214
products of HAV have been confirmed directly using antisera raised to synthetic peptides (Weitz et al., 1986; Wheeler et al., 1986) and amino acid sequencing of virion proteins (Linemeyer et al., 1985; Gauss-Muller et al., 1986). Although these studies have determined the primary amino acid sequence of the capsid proteins and VPg, the proteolytic processing of the HAV polyprotein and morphogenesis are poorly understood. The predicted amino acid sequence of HAV suggests that the HAV capsid is composedoffourstructuralproteins VP1 (33.2 kDa), VP2 (24.8 kDa), VP3 (27.8 kDa) and VP4 (2.5 kDa). Studies of the HAV capsid proteins by SDS-PAGE have determined that the observed molecular weight for VP1 corresponds to the predicted value but the estimates of VP2 and VP3 appear variable. Amino acid sequence data obtained from gel-purified VP3 (Linemeyer et al., 1985) and HPLC-purified VP2 (Gauss-Muller et al., 1986) indicated that VP3 was the faster migrating protein with an apparent relative molecular weight (M,) of 29 kDa. Conversely, immunoblot analysis with antisera raised to synthetic peptides determined that VP2 migrated faster with an apparent M, of 27 kDa (Wheeler et al., 1986). This discrepancy was reportedly due to the presence of urea in the gels affecting the migration of the proteins (Wheeler et al., 1986), although this was not directly demonstrated. Several studies have detected a protein migrating slower than VP2 (Hughes et al., 1984; Hughes and Stanton, 1985; GaussMuller et al., 1986) and have proposed that this protein may represent the precursor of VP2 and VP4, VPO. In an effort to clarify the migration of the capsid proteins in gels and obtain antibody reagents to enable a detailed study of morphogenesis, we aimed to express the capsid proteins in E. cc&. This paper describes the expression of the capsid proteins of HAV in E. coli as fusion proteins of P-galactosidase, the characterization of the antisera raised to these proteins and the identification of the capsid proteins of the cytopathic HM 175 strain of HAV. The preparation of the HAV cDNA clones was described previously (Ross et al., 1986). Restriction enzyme fragments from the cDNA clones were inserted into bacteriophage agtl 1 (Young and Davis, 1983) after blunt-ending with Klenow polymerase and attachment of EcoRI linkers using standard techniques (Maniatis et al., 1982). The regions of the HAV genome selected for expression are depicted in Fig. 1. Four recombinant phages were constructed: hFPl containing the entire coding region of VP2, 57 nucleotides of VP4 and 219 nucleotides of VP3; hFP2 containing the coding region of VP 1, 135 nucleotides of VP3 and 387 nucleotides of 2A; kFPVP2 containing 450 nucleotides of VP2; hFPVP3 containing 405 nucleotides of VP3 and 162 nucleotides of VPl. Together, these fragments contain each of the regions coding for the antigenic epitopes identified by Wheeler et al. (1986). Ligated DNA was packaged into phage proteins using an in vitro packaging kit (Amersham, England). Plaques containing recombinant phage with the desired insert were identified by plaque hybridization (Benton and Davis, 1977). Phage containing HAV cDNA inserts were used to infect E. coli Y 1089 cells at a multiplicity of 0.1 and grown at 37°C to an A 650of 0.4. The culture was then incubated at 45°C for 15 min, made up to 10 mM with isopropyl-P-o-thio-galactopyranoside (IPTG) and incubated for 2 h at 37°C. Cells were pelleted by centrifugation at 12000 x
215
3’
=’ 0
1
2
3
4kb
HAV genome Hincll Xbal
Restriction enzyme sites
BstEll
I
I
Bglll
BamHl Avall
Hindlll
I
hFP1
cDNA fragments subcloned for expression
IFPVPS
I
I____
LFPP
hFPVP3
Fig. 1, Fragments of the HAV genome inserted into bacteriophage &et1 I for expression in E. co/i. The locations of the polyprotein cleavage sites have been described previously (Najarian et al., 1985: Cohen et al., 1987) and are labelled using the picomavirus L434 nomenclature (Rueckert and Wimmer, 1984). Restriction enzyme sites used for the construction of recombinant phage are indicated.
g for 5 min and lysed in 2% SDS, 10 mM NaP04 (pH 6.5) and 1 mM dithiothreitol. Lysates were subjected to SDS-PAGE and examined for the presence of a stable fusion protein. Approximately one half to one third of the plaques containing a HAV cDNA insert were able to produce a stable fusion protein and each protein migrated in gels consistent with their predicted molecular weight (results not shown). Fusion proteins were purified from E. coli lysates by preparative SDS-PAGE for animal immunizations. After electrophoresis, gels were stained with Coomassie blue and the protein band representing the fusion protein was excised. The fusion protein was then electroeluted and excess SDS was removed by dialysis against 0.1 M ammonium bicarbonate. Rabbits were immunized intramuscularly with approximately 50 pg per dose of purified fusion protein in 1: 1 protein:Freunds’ complete adjuvant (FCA). The level of response to P-galactosidase was monitored by indirect radioimmunoassay and after 4-5 doses a stationary titre of p-galactosidase specific IgG was obtained. Mouse polyclonal ascites fluid was produced by a method obtained from Dr. W. Maskill, Fairfield Hospital (personal communication). White male Swiss mice were given 0.2 ml intraperitoneal injections of 9:l FCA:virus at weekly intervals. After approximately 3 weeks the mice produced ascites fluid, which was harvested regularly and the injections were continued until ascites fluid ceased to be produced. The response of the immunized rabbits to I)-galactosidase and HAV was determined by radioimmunoassay. For P-galactosidase a commercially available protein preparation was used as antigen (Pharmacia, Sweden) in a standard indirect radioimmunoassay procedure to detect IgG. Anti-HAV antibodies were measured using an
216
antigen capture radioimmunoassay as previously described (Coulepis et al., 1985). Each rabbit immunized with either purified p-galactosidase or fusion protein responded to P-galactosidase with a P/N ratio (post-immune divided by pre-immune) of greater than 15 at a serum dilution of 1: 10 000. None of the animals receiving fusion protein responded significantly to HAV by radioimmunoassay when tested at a serum dilution of 150. All P/N ratios were 1.4 or less whereas sera from a rabbit immunized with HAV virions reacted with a P/N of 17.5. The fusion protein antisera were also tested for their virus neutralizing activity using the radioimmunofocus assay described by Lemon et al. (1983) and modified by Anderson et al. (1987). All antisera raised to recombinant proteins were negative by neutralization while the antiserum raised to HAV showed a 99% reduction in viral infectivity (results not shown). The failure of the rabbit antisera raised to these proteins to react with HAV by radioimmunoassay or neutralize viral infectivity is similar to a previous study in which a fusion protein of HAV VP1 with the TrpE gene of E. coli also failed to elicit a neutralizing antibody response (Johnston et al., 1988). These results are consistent with earlier studies in which the binding of polyclonal and monoclonal antibodies is prevented if the virion is denatured by treatment with SDS (Gerlich and Frosner, 1983; Hughes et al., 1984) and gel-purified capsid proteins have been shown to elicit poor antibody responses to HAV by competitive radioimmunoassay and neutralization (Hughes and Stanton, 1985). This data implies that the major neutralization site of HAV is strictly conformationally dependent and may require the interaction of several capsid proteins for immunogenicity, although Gauss-Muller et al. (1990) have demonstrated priming for a neutralizing response using HAV-P-galactosidase fusion proteins. Indirect immunofluorescence was used to detect antibodies that react with intracellular viral proteins in HAV-infected cells (results not shown). Immunofluorescence with antisera raised against the fusion proteins FPl, FP2 and FPVP2 showed small cytoplasmic and perinuclear foci of fluorescence in infected cells. The pattern of cytoplasmic and perinuclear fluorescence was similar to that observed with polyclonal antibodies raised to HAV. The failure of the antisera to react by radioimmunoassay or neutralization implies that the foci of HAV antigen detected probably represent accumulations of capsids within the cytoplasm which have been partially denatured during fixation of the cells in acetone. Rabbit antisera raised to the fusion proteins were reacted with immunoblotted HAV virion proteins after SDS-PAGE (Fig. 2). Anti-FP2 reacted with a protein migrating at 33 kDa which is consistent with the M, of VP1 . Anti-FPl and anti-FPVP2 appear to react with the same protein migrating with an M, of 29 kDa and presumably represents VP2, because FPVP2 did not contain any component of VP3. However, the virion proteins VP2 and VP3 migrate with similar mobilities in gels and further analysis was necessary to confirm the identity of this protein. A previous study suggested that the migration of VP2 and VP3 varies, depending on whether urea is used in the gel (Wheeler et al., 1986). In an effort to obtain optimal separation of these capsid proteins, gels were run in the presence and absence of urea before immunoblotting with the fusion protein antisera and a mouse polyclonal ascites fluid raised to HAV which reacts with all capsid proteins. In the absence of
217
1234
5
92.5 6946-
30-
Fig. 2. Immunoblot analysis of HAV capsid proteins with the fusion protein antisera after electrophoresis through 12% SDS-polyacrylamide gels. Lanes contain molecular weight markers (lane I), or HAV virions immunoblotted with anti-FP2 (lane 2), anti-FPl (lane 3), anti-FPVP2 (lane 4) and anti-FPVP3 (lane 5).
urea the mouse polyclonal ascites fluid identified three proteins migrating with apparent M,s of 33,29, and 28 kDa (Fig. 3A) which represent VPl, VP2 and VP3. The antisera raised to FPl and FPVP2 both recognized the 29 kDa protein which represents VP2. Immunoblot analysis of HAV virions after SDS-PAGE with 3.5 M urea (Fig. 3B) resolved the capsid proteins into four bands with apparent M,s of 33,30.6,30 and 28 kDa. Under these conditions the antiserum to FPl (Fig. 3B, lane 3) and FPVP2 (Fig. 3B, lane 4) reacted with a protein doublet migrating with apparent M,s of 30.6 and 30 kDa. The M,s of the capsid proteins estimated from these gels and their comparison with the predicted sizes of the capsid proteins are summarized in Table 1. The migration of VP1 and VP3 in SDS-PAGE was consistent with their predicted M,s and was not altered by the presence of urea (Table 1). However, the putative VP2 protein recognized by antisera to FPl and FPVP2 migrated with an apparent M, considerably higher than its predicted value and, in gels containing urea, migrated as a doublet. This migration of the putative VP2 appears to differ from two previous studies of the HAV capsid proteins. In gels without urea VP2 migrates with an apparent molecular weight higher than that of VP3 (Fig. 3A) which is contrary to the finding of Wheeler et al. (1986). In gels containing urea the electrophoretic profile of the capsid proteins detected with the mouse anti-HAV polyclonal ascites fluid (Fig. 3B, lane 2) is different to that reported by Gauss-Muller et al. (1986) for mature HAV particles banding at 1.32 g/ml in CsCl gradients. However, our results appear similar to another HAV study in which virion proteins were purified from acrylamide gels containing
218
92.5
92.5
69
69
46
46
30
30
Fig. 3. ~mmunoblot analysis of HAV virions after electrophoresis through 12% SDS-polyac~lamide gels under either standard conditions (A) or in the presence of 3.5 M urea(B). Nitrocellulose strips were reacted with mouse anti-HAV ascites fluid (lane 2) rabbit anti-FPl (lane 3), rabbit anti-FPVPZ (lane 4) and rabbit anti-FPVPS (lane 5). The molecular weights of protein standards in lane 1are indicated. TABLE 1 Molecular mass estimates of HAV capsid proteins from immunoblot analysis (Fig. 3A,B) compared with their predicted values (Cohen et al., 1987) Protein
Predicted
VPI VP0
.33.2 kDa 27.3 kDa
VP2 VP3
24.8 kDa 27.8 kDa
Observed Non-urea gel
Urea gel
32.6 kDa
33.0 kDa 30.6 kDa 29.1 kDa 30.0 kDa 28.1 kDa
28.3 kDa
urea (Hughes and Stanton, 1985). One explanation for differences in the migration of the HAV proteins in these studies is the varying conditions in SDS-PAGE between laboratories. It is a well established phenomenon that many factors such as the source of SDS can contribute todifferences in protein migration. This is p~i~ularly relevant considering the small difference in migration between VP2 and VP3 under standard SDS-PAGE conditions and the finding that the migration of VP2 is profoundly affected by urea. However,
219
some of the apparent differences are substantial. For example, there is a 2-kDa shift in the migration of VP2 when compared to Wheeleret al. (1986) resulting in an inversion of VP2 and VP3. It is possible that this difference could be variation between strains of HAV which may have altered VPO/VP2 cleavage junctions. Although this possibility is unlikely, it cannot be ruled out until direct comparisons between strains in the same laboratory are made and the contribution of factors affecting the migration of HAV capsid proteins in SDS-PAGE are determined. The doublet identified with the antisera to VP2 probably represents VP0 and VP2. Recent studies in our laboratory on the analysis of sub-viral particles by sucrose gradient ultracentrifugation has demonstrated a precursor/product relationship between these bands (Anderson and Ross, 1990). The upper band of the doublet is present in pentamers, procapsids, and provirions whilst the lower band is present in provirions and virions. This data is consistent with the morphogenesis described for other picomaviruses (Rueckert, 1985) and indicates that the doublet represents VP0 and VP2. Further studies need to be completed to precisely define the morphogenesis of HAV which, on the basis of the data presented here, suggests that the M, difference between VP0 and VP2 is significantly smaller than the predicted value. Even considering the possible trimming of the amino terminus of the HAV polyprotein for processing by myristilation (Chow et al., 1987), the difference between VP0 and VP2 should be approximately 2 kDa. We are attempting to answer these questions by introducing mutations into a complete infectious copy of the genome to determine if cleavage at the proposed VPO/VP2 junction is occurring and continuing our studies on the sub-viral particles of HAV.
Acknowledgements The authors would like to thank Scott Bowden for his invaluable editorial assistance and Mr. H. Weatherley for providing the Maroline M. Weatherley Scholarship which supported BCR during this work. These studies were supported by a grant from the National Health and Medical Research Council. Australia.
References Anderson, D.A., Locamini, S.A., Ross, B.C., Coulepis, A.G., Anderson, B.N. andGust, I.D. (1987) Singlecycle growth kinetics of hepatitis A virus in BSC-1 cells. In: M.A. Brinton and R.R. Rueckert (Eds), Positive Strand RNA Viruses. Alan R. Liss, New York, pp. 497-507. Anderson, D.A. and Ross, B.C. (1990) Morphogenesis ofhepatitis A virus: isolation and characterization of subviral particles. J. Virol. 64,5284-5289. Benton, W.D. and Davis, R.W. (1977) Screening of hgf recombinant clones by hybridization to single plaques in situ. Science 196,180-l 82. Chow, M., Newman, J.F.E., Filman, D., Hogle, J.M., Rowlands, D.J. and Brown, F. (1987) Myristilation of picomavirus capsid protein VP4 and its structural significance. Nature 327,482486. Coulepis, A.G., Veale, M.F., MacGregor A., Komitschuk, M. and Gust, I.D. (1985) Detection of hepatitis
220 A virus and antibody by solid-phase radioimmunoassay and enzyme-linked immunosorbent assay with monoclonal antibodies. J. Clin. Microbial. 22, 119-124. Cohen, J.I., Ticehurst, J.R., Purcell, R.H., Buckler-White, A. and Baroudy, B.M. (1987) Complete nucleotide sequence of wild-type hepatitis A virus: comparison with different strains of hepatitis A virus and other picornaviruses. J. Virol. 61,50-59. Gauss-Muller, V., Lottspeich F. and Deinhardt, F. (1986) Characterization of hepatitis A virus structural proteins. Virology 155,732-736. Gauss-Muller, V., Mingquan, Z., von der Helm, K. and Deinhardt, F. ( 1990) Recombinant proteins VP I and VP3 of hepatitis A virus prime for neutralizing antibody response. J. Med. Virol. 3 I,2777283 Gerlich, W.H. and Frosner, G.G. (1983) Topology and immunoreactivity of capsid proteins in hepatitis A virus. Med. Microbial. Immunol. 172, 101-106. Gust, I.D., Coulepis, A.G., Feinstone, S.M., Locarnini, S.A., Moritsugu, Y., Nejera, R. and Siegl, G. (I 983) Taxonomic classification of hepatitis A virus. Intervirology 20, t-7. Hughes, J.V. and Stanton, L.W. (I 985) Isolation and immunizations with hepatitis A virus structural proteins: induction of antiprotein, antiviral, and neutralizing responses. J. Virol. 55.395401. Hughes, J.V., Stanton, L.W., Tomassini, J.E., Long, W.J. and Scolnick, E.M. (1984) Neutralizing monoclonal antibodies to hepatitis A virus: partial localization of a neutralizing antigenic site. J. Virol. 52, 465473. Johnston, J.M., Harmon, S.A., Binn, L.N., Richards, O.C., Ehrenfeld, E. and Summers, D.F. (1988) Antigenie and immunogenic properties of a hepatitis A virus capsid protein expressed in Escher-ichia co/i. J. Infect. Dis. 157, 1203-121 I. Lemon, S.M., Binn, L.N. and Marchwicki, R.H. (1983) Radioimmunofocus assay forquantitation of hepatitis A virus in cell culture. J. Clin. Microbial. 17,834-839. Linemeyer. D.L., Menke, J.G., Martin-Gallardo, A., Hughes, J.V., Young, A. and Mitra, S.W. (1985) Molecular cloning and partial sequence of hepatitis A viral cDNA. J. Virol. 54,247-255. Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, New York. Najarian, R., Caput. D., Gee, W., Potter, S.J., Renard, A., Merryweather, J., van Nest, G. and Dina, D. (1985) Primary structure and gene organization of human hepatitis A virus. Proc. Natl. Acad. Sci. USA 82, 2627-263 1. Ross, B.C.. Anderson, B.N., Coulepis, A.G., Chenoweth, M.P. and Gust, I.D. (1986) Molecular cloning of cDNA from hepatitis A virus strain HMI 75 after multiple passages in vivo and in vitro. J. Gen. Virol. 67, 1741-1744. Rueckert, R.R. (I 985) Picornaviruses and their replication. In: B.N. Fields, D.M. Knipe, R.M. Chanock, J.L. Melnick, B. Roizman and R.E. Shope (Eds), Virology. Raven Press, New York, pp. 705-738. Rueckert. R.R. and Wimmer, E. (1984) Systematic nomenclature of picomavirus proteins. J. Virol. 50, 957-959. Weitz, M., Baroudy. B.M.. Maloy. W.L., Ticehurst, J.R. and Purcell, R.H. (1986) Detection of a genomelinked protein (VPg) of hepatitis A virus and its comparison with other picomaviral VPgs. J. Virol. 60, 1244130. Wheeler, C.M., Robertson, B.H.. van Nest, G., Dina, D., Bradley, D.W. and Fields, H.A. ( 1986) Structure of the hepatitis A virion: peptide mapping of the capsid region. J. Virol. 58,307-3 13. Young. R.A. and Davis, R.W. (1983) Efficient isolation of genes by using antibody probes. Proc. Natl. Acad. Sci. USA 80,1194-l 198.