Assembly of Hepatitis Delta Virus Particles: Package of Multimeric Hepatitis Delta Virus Genomic RNA and Role of Phosphorylation

VIROLOGY

249, 12±20 (1998) VY989310

ARTICLE NO.

Assembly of Hepatitis Delta Virus Particles: Package of Multimeric Hepatitis Delta Virus Genomic RNA and Role of Phosphorylation Tien-Shun Yeh and Yan-Hwa Wu Lee1 Institute of Biochemistry, School of Life Sciences, National Yang-Ming University, Taipei, Taiwan, Republic of China Received November 24, 1997; returned to author for revision December 31, 1997; accepted June 25, 1998 We previously demonstrated that both casein kinase II (CKII) and protein kinase C (PKC) positively modulate the hepatitis delta virus (HDV) RNA replication but not the assembly of the empty hepatitis delta antigen (HDAg) particle. In this study, we investigated whether phosphorylation of HDAg by these two kinases plays a role in assembly of the HDV virion. As demonstrated by in vivo labeling and kinase inhibitor experiments, the phosphorylation level of large HDAg but not small HDAg in HDAg-expressing HuH-7 cells was diminished by CKII inhibitor (DRB), whereas no effect was observed for the phosphorylation level of two HDAgs when treated with protein kinase A (PKA) inhibitor (HA1004) or PKC inhibitor (H7). Cotransfection experiment also demonstrated that packaging of HDV genomic RNA was not affected by the kinase inhibitor DRB or H7 and mutation at the putative CKII phosphorylation sites (serine-2, serine-123, or both), and the putative PKC site (serine-210) of HDAg did not elicit any significant effect on the HDV virion assembly. Therefore, based on the previous work and the present study, it seems that the status and biological significance of phosphorylation of HDAg vary depending on the HDV life cycle. Although in the HDV RNA replication cycle, phosphorylation of small HDAg by CKII or PKC plays important role in HDV replication, phosphorylation of the same HDAg by these two kinases does not occur during the HDV RNA virion assembly, and phosphorylation of the large HDAg by CKII does not confer any regulatory role in the assembly of HDV virion and empty viral particles. Our study also showed that the large HDAg without the small HDAg could efficiently assemble both monomeric and dimeric HDV genomic RNAs into secreted HBV-enveloped virus-like particles. Increasing the transfected small HDAg-expressing plasmid led to an enhancement of the packaging efficiency for the monomeric HDV genomic RNA with little effect on the packaging of dimeric HDV RNA. Similarly, HDAgs could package the trimeric HDV genomic RNA, albeit less efficiently. CsCl density gradient centrifugation confirmed that HDAgs and the monomeric and multimeric (dimer and trimer) HDV genomic RNAs formed an HBV-enveloped virus-like particle at a density of 1.23±1.25 g/ml. Thus, the assembly of the HDV virion seems to not impose much restriction on the size of HDV RNA for packaging. © 1998 Academic Press

inhibits HDV RNA replication and is required for HDV assembly (Chao et al., 1990; Chang et al., 1991; Glenn and White, 1991; Ryu et al., 1992). It has also been noted that the formation of HDV viral particles requires the isoprenylation and specific amino acid sequence at the carboxyl-terminal of the large HDAg (Glenn et al., 1992; Lee et al., 1994, 1995). Moreover, studies on the role of HBV envelope proteins in the assembly of HDV particles have indicated that although the small HBsAg is sufficient for assembly of the HDV virion, the large HBsAg is necessary for infectivity of the virion (Ryu et al., 1992; Sureau et al., 1993; Wang et al., 1991). The HDAg has multiple functional domains, which is relevant to their biological properties, such as oligomerization, RNA-binding ability, and nuclear transport (for reviews, see Lazinski and Taylor, 1993; Lai, 1995 and references therein). The two forms of the HDAg bind to HDV RNA with equal affinity, forming a ribonucleoprotein particle (RNP), and they recognize the truncated HDV RNA, which is capable of folding into a rod-like structure (Chang et al., 1988, 1995; Lin et al., 1990; Chao et al., 1991; Ryu et al., 1993; Lazinski and Taylor, 1994). Notably, previous studies have also indicated that the large HDAg

INTRODUCTION Hepatitis delta virus (HDV) contains a 1.7-kb singlestranded circular RNA genome, delta antigen (HDAg), and a surrounding lipid envelope embedded with hepatitis B virus (HBV) surface antigen proteins (HBsAg) (Rizzetto, 1983; Bonino et al., 1984; Wang et al., 1986, 1987; Makino et al., 1987; Kuo et al., 1988). HDAg is encoded by an antigenomic strand of HDV RNA and exists as two protein species: a small 24-kDa form and a large 27-kDa form (Wang et al., 1986, 1987; Makino et al., 1987; Kuo et al., 1988). The large HDAg contains the entire small HDAg open reading frame (ORF) plus 19 additional amino acids at the carboxyl terminus. This carboxyl-terminal 19-amino-acid extension in the large HDAg is generated by an RNA editing event during the process of HDV replication (Weiner et al., 1988; Luo et al., 1990). These two isoforms of HDAg have opposing biological functions. Although the small HDAg is required for genome replication (Kuo et al., 1989), the large HDAg

1 To whom reprint requests should be addressed. Fax: 886±2-2826± 4843. E-mail: [email protected].

0042-6822/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

12

ASSEMBLY OF HDV PARTICLES

13

FIG. 1. Analysis of HDV RNA package after the cotransfection. HuH-7 cells were cotransfected with a plasmid expressing the mutant HDV RNA (pCDm2G) (20 mg), a plasmid expressing HBsAg (pMTS) (20 mg), and a plasmid expressing large HDAg (pECE-D, designated DWT) or small HDAg (pECE-d, designated dWT) at the indicated amounts. Total cellular RNA (A) and the RNA of viral particles (B) from day 3 post-transfected cells were prepared and analyzed by Northern blotting using HDV strand-specific riboprobe (see Materials and Methods). The intracellular (A) and secreted HDAgs (B) were detected by Western blotting using a human anti-HDAg antibody (see Materials and Methods). The positions for the monomeric or dimeric HDV RNA, as well as the positions for the large (L) or small HDAg (S), are indicated. (M) Mock without transfection. (G RNA) Genomic HDV RNA. (AG RNA) Antigenomic HDV RNA.

alone is capable of packaging monomeric HDV genomic RNA into virus particles, whereas the incorporation of small HDAg increases the HDV RNA packaging efficiency (Wang et al., 1994). Both HDAgs have been detected as nuclear phosphoproteins (Chang et al., 1988; Hwang et al., 1992; Yeh et al., 1996), although certain data suggest that the large HDAg but not the small HDAg is phosphorylated (Bichko et al., 1997). Because it is well known that phosphorylation can regulate the activities of protein at multiple levels, including nuclear transport, dimerization, nucleic acid binding, and transcriptional activation (Hunter and Karin, 1992), similarly, phosphorylation may affect these properties of the HDAg. We previously investigated the functional role of HDAg phosphorylation in HDV RNA replication, packaging, and nuclear transport (Yeh et al., 1996). Our results suggest that although phosphorylation of serine-2 by casein kinase II (CKII) is important for the replication function of small HDAg, phosphorylation of three conservative serine residues (serine-2, serine-123, and serine-210) by CKII or protein kinase C (PKC) seems functionally irrelevant to the assembly of the empty viral particle or in the transdominant inhibitory activity of large HDAg on HDV replication (Yeh et al., 1996). Furthermore, they are not a prerequisite for either form of HDAg entrance into the nucleus. The remaining issue of whether phosphorylation of these serines plays a role in the RNA-binding ability during the virion assembly is addressed in this study. Interestingly, we found that contrary to HDV RNA replication, phosphorylation by both CKII and PKC does not elicit any effect on the HDV RNA package. Moreover, we demonstrate here that HDAgs together with HBsAg can assemble the multimeric (dimer and trimer) form of HDV RNA with almost indistinguishable property as the HDV virion-containing monomeric form of HDV genomic RNA.

RESULTS HDAgs are capable of copackaging the dimeric form of HDV RNA in the presence of HBsAg To investigate the role of phosphorylation of HDAgs in HDV RNA packaging in vitro without the complication generated from RNA replication, we adapted the fourplasmid cotransfection system developed by Wang et al. (1994). In this system, each component of the HDV virion was supplied by an individual plasmid, including the expression plasmids for the HBsAg (pMTS), the large HDAg (pECE-D), the small HDAg (pECE-d), and the HDV RNA template (pCDm2G). The latter construct pCDm2G contains a tandem dimer of the full-length, mutated HDV cDNA harboring a frameshift mutation that prevents the production of both species of HDAg (Wang et al., 1994). This plasmid presumably can supply the HDV genomic RNA for assay of the HDV genome package by cotransfecting with the expression constructs of both HDAgs and HBsAg (Wang et al., 1994). To assess the feasibility of this system for assay of HDV virion assembly, HDAgs and HDV RNAs from within cells (cellular), as well as those secreted in the presence of HBsAg (medium) after day 3 post-transfection, were evaluated by immunoblotting and Northern blotting. As demonstrated in Figure 1A, in HuH-7 cells transfected with plasmid pCDm2G, the intracellular 3.4-kb primary HDV transcript could undergo self-processing to generate a full-length, monomeric HDV RNA (Fig. 1A, lane 2). However, without the small HDAg expression, this mutant HDV RNA could not replicate because no antigenomic HDV RNA was found (Fig. 1A, lane 2 in cellular G RNA and AG RNA sections). If HuH-7 cells were cotransfected with four plasmids under the conditions in which the amounts of transfected plasmids for HBsAg and pCDm2G were fixed (20 mg each)

14

YEH AND LEE

but varying the plasmid ratio for the small to large HDAg, results from Northern blot analysis indicated that both forms of HDAgs could enhance the yield and the ratio of monomeric to dimeric HDV genomic RNAs compared with that without the HDAg-expressing plasmid (Fig. 1A, compare lane 2 with lanes 3±7 in cellular G RNA). This suggested that the HDAg could stabilize the HDV RNA and enhance the ribozyme activity as well, which is in accord with previous reports (Lazinski and Taylor, 1994; Jeng et al., 1996). It is also noted that the ratio of intracellular monomeric to the dimeric HDV RNA varied depending on the different combination of both forms of HDAg introduced. In particular, the dimeric HDV RNA was present predominantly if the transfected small HDAg-expressing plasmid was excluded or the transfected large HDAg-expressing plasmid was in great excess compared with the small HDAg-expressing plasmid (Fig. 1A, lanes 4 and 7 in cellular G RNA). However, as judged by the presence or lack of antigenomic RNA, it appears that as long as the large HDAg-expressing plasmids are present in equal (lane 6) or excess (lanes 4 and 7) amount compared with the small HDAg, no detectable HDV replication occurred. When examining the secreted HDV-like particle, in the absence of large HDAg expression, no HDV-like particle was secreted in the culture medium (Fig. 1B, lane 3). This result supports the essential role of large HDAg in HDV viral particle assembly as previously noted elsewhere (Chang et al., 1991). Interestingly, exclusion of the small HDAg expression plasmid while leaving the cells transfected with only the large HDAg expression plasmid and the other two plasmids (pCDm2G and pMTS) could lead to a copackage of the large HDAg with HDV RNA in the culture medium (Fig. 1B, lane 4). A similar observation was reported by other investigators (Wang et al., 1994). However, it should be noted that in this particular experiment, both the dimeric and monomeric forms of HDV RNAs were packaged into the HDV viral particle with almost equal efficiency (Fig. 1B, lane 4). Increasing the transfected small HDAg plasmid from 2 to 20 mg (smallto-large HDAg plasmid ratio, 2:20 or 20:20; lanes 6 and 7) or increasing the plasmid ratio of transfected small HDAg to large HDAg (20:2 ratio; lane 5), however, resulted in a significant increase in the packaging of the monomeric form of HDV RNA, which is consistent with a previous report (Wang et al., 1994). The amount of dimeric HDV RNA packaged into the virus-like particle was largely unaltered except in the transfected cells containing equal amounts (20 mg each) of transfected large and small HDAg plasmids, whereas a slight increase in secreted dimeric form of HDV RNA was noted (Fig. 1B, lane 6). These results reveal that the large HDAg could package the HDV dimeric RNA independent of the small HDAg. On the contrary, the packaging efficiency of the monomeric form of HDV RNA was highly dependent

on introduction of the small HDAg into the transfection experiment. Characterization of HDV virus-like particles containing the multimeric form of HDV RNA To further demonstrate that multimeric HDV RNAs were actually packaged into virus-like particles, HuH-7 cells were cotransfected with 20 mg each of four plasmids. In this series of experiments, the HDV RNA templates for package were provided by construct pCDm2G or pCDm3G. The construct pCDm3G contains a mutated tandem trimer of HDV cDNA; therefore, it can yield the unprocessed trimeric and self-processed dimeric or monomeric forms of genomic HDV RNAs in the HuH-7 cell line but in varying amounts (Fig. 2A, lane 3). Like the dimeric HDV genomic RNA, the trimeric one also was packaged into the virus particle (Fig. 2B, lanes 2 and 3). To test whether these virus-like particles with multimeric HDV RNA contain the HBsAg as its envelope, the virus particles collected from day 3 post-transfection were immunoprecipitated with mouse anti-HBsAg monoclonal antibody, and the immunoprecipitates were analyzed by Northern blotting using the HDV antigenomic RNA as a probe. Results suggest that all the trimeric, dimeric, and monomeric forms of HDV RNAs can form a virus-like particle with HBV envelope protein (Fig. 2C, lanes 2 and 3). This RNA-packaging activity by HDAgs was rather sequence specific because no package of abundant cellular RNA such as 18S rRNA was detected in the HBsAg-enveloped particle (Fig. 2D, lanes 2 and 3). CsCl gradient centrifugation was then performed to determine the density of the viral particle containing the multimeric forms of HDV RNAs. When the construct pCDm2G was used to express HDV RNA template for package, both the dimeric and monomeric HDV RNAs were banded together in fractions 13±15 (Fig. 3A). These fractions had a density of 1.23±1.25 g/ml, which is similar to the density of the HDV virion reported by others (Chang et al., 1991; Wang et al., 1994). Notably, the peak fractions for HBsAg appeared at fraction 12, which corresponded to a density of empty virus-like particle at 1.21 g/ml. Similarly, when the HDV RNA template for the package was provided by the construct pCDm3G, all three sizes of HDV RNAs were also banded together in a fraction with a density of 1.24 g/ml (Fig. 3B). These studies demonstrated that all the trimeric, dimeric, and monomeric HDV RNAs were indeed packaged into viral particles with a density distinct from that of empty particles. Inhibitors of CKII and PKC do not affect the assembly and secretion of the HDV viral particle Our previous studies have shown that both CKII and PKC positively modulate the HDV RNA replication but not the assembly of the empty HDAg particle (Yeh et al., 1996). Here, we investigated whether phosphorylation of

ASSEMBLY OF HDV PARTICLES

15

FIG. 2. Northern blot analysis of HDV virion RNA after immunoprecipitation with anti-HBsAg antibody. HuH-7 cells were cotransfected with the equal amounts (20 mg each) of four plasmids: pECE-D, pECE-d, pMTS, and pCDm2G (lane 2) or pCDm3G (lane 3). At day 3 post-transfection, virions were prepared from the culture medium (B) and immunoprecipitated with anti-HBsAg antibody (C and D). Cellular RNA (A and D, lane 4) and viral RNA (B±D) were prepared and analyzed by Northern blotting using HDV strand-specific riboprobe (B and C) or using the cDNA fragment of 18S rRNA as a probe (D) (see Materials and Methods). (Lane 1) Mock without transfection. [Lane 4 (D)] Sample from the cellular RNA to serve the positive control. The positions for the trimeric, dimeric, or monomeric HDV RNA are indicated.

HDAg by these two kinases plays a role in HDV virion assembly. When examined for the in vivo phosphorylation of both forms of HDAg in HuH-7 cells transiently transfected with the HDAg expression construct, our results suggested that the CKII inhibitor (DRB) could block the phosphorylation intensity (50% reduction) of large HDAg but not that of the small HDAg (Fig. 4, compare lane 2 with lane 3 and lane 6 with lane 7). Interestingly, both PKC and PKA inhibitors, H7 and HA1004, did not elicit any apparent effect on the phosphorylation intensity of two HDAgs (Fig. 4, compare lanes 4 and 5 with lane 2 and lanes 8 and 9 with lane 6). Next, we examined the effect of kinase inhibitor on the HDV virion assembly. To this end, HuH-7 cells cotransfected with HDV dimeric cDNA plasmid (pCDm2G) and the expression constructs for the large HDAg (pECE-D), small HDAg (pECE-d), and HBsAg were treated with the kinase inhibitor. As shown in Figure 5, on treatment with DRB or H7, the same level of HDAgs was detected in both the intracellular and extracellular fractions of the control and inhibitor-treated cells. Furthermore, the levels of dimeric and monomeric forms of HDV genomic RNA present in the intracellular and extracellular fractions of the treated cells, as detected by the strandspecific riboprobe, were similar to those of transfected cells without the inhibitor treatment. Apparently, these two kinase inhibitors do not exert any significant inhibition of HDV RNA packaging. Mutation at serine-2, serine-123, or serine-210 of the HDAgs does not block the HDV viral RNA package According to our previous study, serine-2, serine-123, and serine-210 are the three most conserved serine residues among HDAg variants; they also are the recog-

nition sites for CKII or PKC kinases (Yeh et al., 1996). To further examine whether phosphorylation of serine-2, serine-123, or serine-210 residues in two species of HDAg is required for the formation of the virus-like particle, the wild-type, or mutant form, of the HDAg expression construct harboring serine-to-alanine substitution mutation at these three serine residues, together with the HBsAg expression construct and HDV cDNA plasmid pCDm2G, were cotransfected into HuH-7 cells. Analysis of the secreted HDV viral particles by immunoblotting and Northern blotting indicated that all HDAg mutant variants, including the double mutant harboring substitution mutations at both serine-2 and serine-123, essentially retained the wild-type ability for the formation of HDV viral particles (Fig. 6). Thus, phosphorylation of HDAg at serine-2, serine-123, and serine-210 by CKII or PKC, if it occurs in vivo, does not appear to play any significant role in HDV viral RNA particle assembly. This result is consistent with the finding obtained from the inhibitor experiment (Fig. 5) and further confirms that PKC and CKII kinases are not involved in the assembly of HDV RNA particle. DISCUSSION Our previous studies demonstrated that CKII and PKC modulate the HDV RNA replication but not the assembly of the empty viral particle. To further assess whether CKII and PKC regulate the package of HDAg for HDV RNA without the complication of HDV RNA replication, we adopted the four-plasmid cotransfection experiment (Wang et al., 1994) to address this question. In this system, the HDV RNA for the virion package was provided by the construct pCDm2G or pCDm3G, and the amounts of the large and small forms of HDAg-

16

YEH AND LEE

FIG. 3. CsCl density gradient analysis of HDV viral particles. (A) HuH-7 cells were cotransfected with equal amounts (20 mg each) of plasmid pECE-D, pECE-d, pCDm2G, and HBsAg construct pMTS. (B) pCDm3G instead of pCDm2G was used to express the HDV RNA. At day 3 post-transfection, virions were prepared from the culture medium and subjected to isopycnic centrifugation in a CsCl gradient (see Materials and Methods). (Top) Density profile of each fraction. Each fraction was analyzed for HDV genomic RNA by Northern blotting using HDV strand-specific riboprobe and for HDAg or HBsAg (expressed as P:N ratio) by Western blotting or EIA. All symbols are similar to those described in the legend for Figure 1 or 2.

expressing plasmid transfected under these conditions (20 mg each) were such that no HDV RNA replication was allowed (see Fig. 1A). On the basis of in vivo labeling and kinase inhibitor treatment experiments, we have three lines of evidence to support that phosphorylation of HDAg does not confer any regulatory role in the HDV virion assembly. (1) Neither HDAgs is the in vivo substrate for PKC and PKA. (2) Although the phosphorylation intensity of large HDAg was diminished by the CKII inhibitor, the assembly of HDV virion was not affected by the inhibitors of CKII and PKC. (3) Mutation of HDAg at the putative CKII or PKC recognition sites does not influence the packaging of HDV viral particles. Therefore, based on this work and our previous studies (Yeh et al., 1996), it appears that the phosphorylation status of the small HDAg varies depending on the HDV life cycle. In its

RNA replication cycle, the phosphorylation of the small HDAg in the CKII or PKC recognition site occurs and plays an important role in regulation of the HDV RNA replication process (Yeh et al., 1996). Nevertheless, phosphorylation of the same antigen by these two kinases apparently does not occur during packaging of both HDV virion and its empty virus-like particle (Yeh et al., 1996; present work). Intriguingly, phosphorylation of the large HDAg seems dispensable to any of its function known so far (e.g., trans-dominant function, viral and empty particle assembly, nuclear transport) (Yeh et al., 1996; present work). This implies that phosphorylation probably does not affect the oligomerization or the RNA-binding affinity of the HDAg, which presumably are important for assembly of HDV virion. Our finding that HDAgs could package the dimeric or

ASSEMBLY OF HDV PARTICLES

FIG. 4. Effects of kinase inhibitors on the HDAg phosphorylation. HuH-7 cells were transfected with the expression construct pECE-d (dWT) (lanes 2±5) or pECE-D (DWT) (lanes 6±9) (20 mg), and at the appropriate time, cells were treated with or without kinase inhibitor (100 mM) and processed for in vivo 32Pi labeling (see Materials and Methods). The HDAgs were immunoprecipitated from the cell lysates, and half of the aliquots of the protein samples were analyzed by SDS±PAGE and autoradiography. The other half of the aliquots of the samples were subjected to Western blot analysis using a rabbit antiHDAg antibody (see Materials and Methods). The relative phosphorylation level was determined with a PhosphorImager. The relative positions for the large (L) or small (S) HDAgs are indicted. (Lane 1) Mock without transfection. The molecular mass (in kDa) of the protein marker is indicated.

17

trimeric form of HDV RNA into the virus-like particles was rather surprising because in a similar system, no such phenomenon was found (Wang et al., 1994). Although the exact basis for the cause of this phenomenon remains unclear, there are several possible explanations. The amount of DNA used for transfection as well as the transfection efficiency may contribute to this discrepancy. Nevertheless, the question arises of whether the monomeric and dimeric HDV RNAs can be copackaged into the same virus particle. Although we have evidence to show that the viral particle with multimeric form of HDV RNA has the same buoyant density as that of monomeric one and forms an envelope with HBsAg and HDAgs, this does not provide direct evidence to support the notion that these two forms of HDV RNA are copackaged into the same particle. Despite this limitation, it is reasonable to envision that the size of virion with multimeric HDV RNA may be larger than the one packed with the monomeric HDV RNA. If this is the case, it is of relevance to know whether there is any large size limitation for package of HDV RNA. When considering that the truncated HDV genomic RNA with a rod-like structure but a size of ,350 nucleotides is competent for packaging into RNP particles (Lazinski and Taylor, 1994), it seems that the rod-like structure but not the size of HDV RNA governs the assembly of HDV virion by the HDAg. Further examination of the capability of HDAg to assemble the larger size (more than trimer) of HDV RNA may verify this viewpoint. Large HDAg has two important roles in the HDV life cycle (for reviews, see Lazinski and Taylor, 1993; Lai, 1995 and references therein). This antigen can suppress the replication of HDV RNA, but it is necessary for packaging of HDV virion or empty virus-like particles (Chao et al., 1990; Chang et al., 1991; Glenn and White, 1991; Ryu

FIG. 5. Effects of kinase inhibitors on the HDV RNA package. HuH-7 cells were cotransfected with equal amounts (20 mg each) of plasmid pECE-D, pECE-d, pCDm2G, and pMTS (lanes 3±5). At day 2±3 post-transfection, cells were continuously treated with 100 mM DRB (lane 4) or H7 (lane 5) or left untreated (lane 3). Total cellular RNA and the RNA of viral particles from day 3 post-transfected cells were prepared and analyzed by Northern blotting using HDV strand-specific riboprobe (see Materials and Methods). The intracellular and secreted HDAgs were detected by Western blotting using a human anti-HDAg antibody (see Materials and Methods). Samples for analysis has been prepared from the same amount of protein (intracellular HDAg) or the same number of transfected cells (medium fraction). (Lane 1) Mock sample. (Lane 2) All experimental conditions are similar to those for lanes 3±5 except that cells were cotransfected with equal amounts (20 mg each) of plasmid pCDm2G and pMTS but not HDAg-expressing plasmid. All symbols are similar to those described in the legend for Figure 1.

18

YEH AND LEE

FIG. 6. Effects of serine-to-alanine substitution mutations on the assembly of HDV RNA by a small and/or large HDAg with HBsAg. HuH-7 cells were cotransfected with the pCDm2G, pMTS, and the large or small HDAg-expressing construct pECE-D (DWT) or pECE-d (dWT) or its mutant derivatives (20 mg each). Analyses of the HDV genomic RNA, as well as the HDAgs in these transfected cells after day 3 post-transfection, were followed essentially as described in the legend for Figure 1. (M) Without transfection. The designations for HDV mutants are described in Materials and Methods. D2, pECE-D2A; D123, pECE-D123A; D-210, pECE-D210A; d2, pECE-d2A; d123, pECE-d123A; D2/123, pECE-D2/123A; d2/123, pECE-d2/123A. All symbols are similar to those described in the legend for Figure 1.

et al., 1992). Small HDAg can trans-activate the replication of HDV RNA (Kuo et al., 1989). Conceivably, the ratio between large and small HDAgs in its infected cells modulates the HDV life cycle. In the hepatocyte, the ratio of large HDAg to small HDAg is ;0.1 (Bonino et al., 1986), whereas in the virion, this ratio increases to 0.9 (Chang et al., 1991). Interestingly, in this study, we found that the ratio between these two HDAgs modulates the selection of monomeric or dimeric HDV genomic RNA for assembly. When this ratio in the transfected cells is approximately the same as that in liver cells, the packaged HDV RNA is predominantly monomeric (Fig. 1B, lane 5). Notably, when the small HDAg is absent, the large HDAg can package the monomeric and dimeric with apparently equal efficiency. However, increasing the small HDAg expression level, which may reflect the natural condition for infected cells with active HDV replication, the packaging efficiency of monomeric HDV RNA, but not the dimeric one, caused significant enhancement. This explains why the monomeric form of HDV genomic RNA is the predominant form present in the virion. The exact mechanism for the preference of packaging of multimeric HDV RNA by the large HDAg, but not by the small HDAg, is not yet known. Presumably, this may result from the conformational difference between these two anti-

gens as pointed out previously (Hwang and Lai, 1993a, 1994), or it may reside on the attribute that the large HDAg, but not the small HDAg, interacts with HBsAg (Hwang and Lai, 1993b), which is an important factor in the assembly of HDV RNA. Alternatively, this may result from the selectivity of the small HDAg, but not the large HDAg, for packaging of the monomeric, replicative form of HDV genomic RNA. In fact, in transfected or infected cells, most of the HDV RNA has been found to be associated with small HDAg and not with large HDAg (Ryu et al., 1993). MATERIALS AND METHODS Plasmids The HDV sequence used in this study was derived from plasmid pSVLD3 (Kuo et al., 1989). Plasmid pCDm2G (Wang et al., 1994) and pCDm3G contain a tandem dimer or trimer of the full-length, mutated HDV cDNA in which a frameshift mutation in the HDAg ORF destroys the production of both large and small HDAgs and requires the functional small HDAg in trans for replication. Plasmid pGEM3L, which contains a ScaI/EcoRI fragment (1.1 kb) of the HDV cDNA driven by the SP6 or T7 phage promoter (Hu et al., 1996), was used for pro-

ASSEMBLY OF HDV PARTICLES

duction of strand-specific probes by in vitro transcription. Plasmid pMTS contains the major HBV surface antigen (HBsAg) gene under the control of metallothionein promoter (Sheu and Lo, 1992). This plasmid was used in the transfection experiment to provide HBsAg (Yeh et al., 1996). The expression plasmids for the wild-type and mutated small HDAg defective in the CKII recognition site (e.g., pECE-d, pECE-d2A, or pECE-d123A), as well as the plasmids encoding the ORF of wild-type large HDAg and its mutant derivatives defective in CKII or PKC recognition site (e.g., pECE-D, pECE-D2A, pECE-D123A, or pECE-D210A), have been described previously (Yeh et al., 1996). For construction of the HDAg-expressing plasmid harboring substitution mutation at both serine-2 and serine-123 residues, the 0.5-kb BglII/EcoRI fragment harboring the serine-2 mutation in HDAg ORF was excised from pECE-d2A and used to replace the same fragment on pECE-D123A or pECE-d123A. The resulting construct pECE-D2/123A or pECE-d2/123A was also sequenced to confirm their mutated residues. Cell culture and transfection DNA transfection experiments for this study were adopted from the four-plasmid cotransfection system reported by Wang et al. (1994). HuH-7 cells were seeded (6 3 106 cells/15-cm-diameter dish) and cultured at 37°C under 5% CO2 for 20 h in the Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 IU/ml penicillin, 100 mg/ml streptomycin, and 2 mM l-glutamine. Cells were cotransfected with the appropriate amounts of plasmid combinations (60±80 mg/15-cmdiameter dish) by a calcium phosphate coprecipitation method (Graham and van der Eb, 1973). When applicable, the transfected cells were treated with 100 mM kinase inhibitor DRB or H7 (Sigma Chemical, Poole, Dorset, UK) during days 2±3 post-transfection. The cells and media were harvested at day 3 post-transfection. Isolation of HDV particles and characterization of HDV particles by CsCl centrifugation The secreted HDV viral particles were purified according to a method described previously (Wu et al., 1991; Yeh et al., 1996). When applicable, the HDV viral particles were also purified by immunoprecipitation with mouse anti-HBsAg monoclonal antibody (DAKO Japan, Kyoto, Japan) from culture medium. The pellets recovered were further processed for HDAg or HDV RNA analysis (Yeh et al., 1996). For isopycnic centrifugation, the viral particles prepared from a 35-ml sample of pooled medium of day 3 post-transfected HuH-7 cells by high speed centrifugation (Ti 55.2 rotor, 45,000 rpm centrifugation for 5 h) were suspended in culture medium and subjected to centrifugation in a CsCl gradient (average density, 1.205 g/ml; final density, 1.122±1.422 g/ml) at 38,000 rpm in an SW41 rotor for 60 h. The gradients were fractionated into 0.5-ml

19

samples from the top, and aliquots of each fraction were concentrated by centrifugation at 60,000 rpm for 2 hr in a Beckman TLA100.2 rotor and then analyzed for various viral markers. The density of each fraction was determined by the respective refractive index using a refractometer. Northern and Western blotting Total RNA from transfected HuH-7 cells or from viral pellets was isolated as previously described (Yeh et al., 1996). Northern blotting to detect HDV RNA and preparation of strand-specific riboprobe were essentially the same as described previously (Yeh et al., 1996). The HDAg was detected by the immunoblot using human anti-HDAg sera as the primary antibody and horseradish peroxidase-conjugated antibody as the secondary antibody. The blot was detected by enhanced chemiluminescence (ECL Kit, Amersham Japan, Tokyo, Japan). Metabolic labeling of phosphorylated HDAg with and immunoprecipitation of the HDAg

32

Pi

For labeling the large and small HDAgs, HuH-7 cells (7 3 106) were transfected with the expression plasmid pECE-D or pECE-d (40 mg); after 32 h post-transfection, cells were treated with or without kinase inhibitor (DRB, H7, or HA1004; Sigma, 100 mM each) for 6 h and processed for 32Pi labeling of the HDAg as described previously (Yeh et al., 1996). After 4 h of labeling, cells were washed free of labeled medium and processed for immunoprecipitation with Protein A±Sepharose 4B-bound human anti-HDAg antibody (Yeh et al., 1996). Bound proteins were extracted from the Sepharose 4B beads, and half of the aliquots of protein samples were analyzed by SDS±PAGE and autoradiography, and the other half were used for Western blot analysis and probed for HDAg using rabbit anti-HDAg antibody and the ECL system for detection. ACKNOWLEDGMENTS We are grateful to Drs. S. J. Lo and C. Weaver for critical reading of the manuscript. This work was supported by Grants DOH85-HR-502, DOH86-HR-502, and DOH87-HR502 from the National Health Research Institute of the Republic of China (to Y.-H.W.L.).

REFERENCES Bichko, V., Barik, S., and Taylor, J. (1997). Phosphorylation of the hepatitis delta virus antigens. J. Virol. 71, 512±518. Bonino, F., Heermann, K. H., Rizzetto, M., and Gerlich, W. H. (1986). Hepatitis delta virus: protein composition of delta antigen and its hepatitis B virus derived envelope. J. Virol. 58, 945±950. Bonino, F., Hoyer, B., Shih, J. W.-K., Rizzetto, M., Purcell, R. H., and Gerin, J. L. (1984). Delta hepatitis agent: structural and antigenic properties of the delta-associated particle. Infect. Immunol. 43, 1000±1005. Chang, M.-F., Baker, S. C., Soe, L. H., Kamahora, T., Keck, J. G., Makino, S., Govindarajan, S., and Lai, M. M. C. (1988). Human hepatitis delta

20

YEH AND LEE

antigen is a nuclear phosphoprotein with RNA-binding activity. J. Virol. 62, 2403±2410. Chang, M.-F., Chen, C.-H., Lin, S.-L., Chen, C.-J., and Chang, S. C. (1995). Functional domains of delta antigens and viral RNA required for RNA packaging of hepatitis delta virus. J. Virol. 69, 2508±2514. Chang, F.-L., Chen, P.-J., Tu, S.-J., Wang, C.-J., and Chen, D.-S. (1991). The large form of hepatitis antigen is crucial for assembly of hepatitis virus. Proc. Natl. Acad. Sci. USA 88, 8490±8494. Chao, M., Hsieh, S.-Y., and Taylor, J. (1990). Role of two forms of hepatitis delta virus antigen: evidence for a mechanism of self-limiting genome replication. J. Virol. 64, 5066±5069. Chao, M., Hsieh, S.-Y., and Taylor, J. (1991). The antigen of hepatitis delta virus: examination of in vitro RNA-binding specificity. J. Virol. 65, 4057±4062. Glenn, J. S., Watson, J. A., Havel, C. M., and White, J. M. (1992). Identification of a prenylation site in delta virus large antigen. Science 256, 1331±1333. Glenn, J. S., and White, J. M. (1991). trans-Dominant inhibition of human hepatitis delta virus genome replication. J. Virol. 65, 2357±2361. Graham, F. L., and van der Eb, A. J. (1973). A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52, 456± 467. Hu, H.-M., Shih, K.-N., and Lo, S. J. (1996). Disulfide bond formation of the in vitro-translated large antigen of hepatitis D virus. J. Virol. Methods 60, 39±46. Hunter, T., and Karin, M. (1992). The regulation of transcription by phosphorylation. Cell 70, 375±387. Hsieh, S.-Y., and Taylor, J. (1992). Delta virus as a vector for the delivery of biologically-active RNAs: possibly a ribozyme specific for chronic hepatitis B virus infection. In ªInnovations in Antiviral Development and the Detection of Virus Infectionsº (T. M. Block, D. Jungkind, R. L. Crowell, M. Denison, and L. R. Walsh, Eds.) pp. 125±128. Plenum, New York. Hwang, S. B., and Lai, M. M. C. (1993a). A unique conformation at the carboxyl terminus of the small hepatitis delta antigen revealed by a specific monoclonal antibody. Virology 193, 924±931. Hwang, S. B., and Lai, M. M. C. (1993b). Isoprenylation mediates direct protein-protein interactions between hepatitis large delta antigen and hepatitis B surface antigen. J. Virol. 67, 7659±7662. Hwang, S. B., and Lai, M. M. C. (1994). Isoprenylation masks a conformational epitope and enhances trans-dominant inhibitory function of the large hepatitis delta antigen. J. Virol. 68, 2958±2964. Hwang, S. B., Lee, C.-Z., and Lai, M. M. C. (1992). Hepatitis delta antigen expressed by recombinant baculoviruses: comparison of biochemical properties and post-translational modifications between the large and small forms. Virology 190, 413±422. Jeng, K.-S., Su, P.-Y., and Lai, M. M. C. (1996). Hepatitis delta antigens enhance the ribozyme activities of hepatitis delta virus RNA in vivo. J. Virol. 70, 4205±4209. Kuo, M. Y.-P., Chao, M., and Taylor, J. (1989). Initiation of replication of the human hepatitis delta virus genome from cloned DNA: role of delta antigen. J. Virol. 63, 1945±1950. Kuo, M. Y. P., Goldberg, J., Coates, L., Mason, W., Gerin, J., and Taylor, J. (1988). Molecular cloning of hepatitis delta virus RNA from an infected woodchuck liver: sequence, structure, and applications. J. Virol. 62, 1855±1861. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680±685. Lai, M. M. C. (1995). The molecular biology of hepatitis delta virus. Annu. Rev. Biochem. 64, 259±286. Lazinski, D. W., and Taylor, J. M. (1994a). Recent developments in hepatitis delta virus research. Adv. Virus Res. 43, 2672±2680. Lazinski, D. W., and Taylor, J. M. (1994b). Expression of hepatitis delta virus RNA deletions: cis and trans requirements for self-cleavage, ligation, and RNA packaging. J. Virol. 68, 2879±2888.

Lee, C.-Z., Chen, P.-J., and Chen, D.-S. (1995). Large hepatitis delta antigen in packaging and replication inhibition: role of the carboxylterminal 19 amino acids and amino terminal sequences. J. Virol. 69, 5332±5336. Lee, C.-Z., Chen, P. J., Lai, M. M. C., and Chen, D. S. (1994). Isoprenylation of large hepatitis delta antigen is necessary but not sufficient for hepatitis delta virus assembly. Virology 199, 169±175. Lin, J.-H., Chang, M.-F., Baker, S. C., Govindarajan, S., and Lai, M. M. C. (1990). Characterization of hepatitis delta antigen: specific binding to hepatitis delta virus RNA. J. Virol. 64, 4051±4058. Luo, G., Chao, M., Hsieh, S.-Y., Sureau, C., Nishikura, K., and Taylor, J. (1990). A specific base transition occurs on replicating hepatitis delta virus RNA. J. Virol. 64, 1021±1027. Makino, S., Chang, M.-F., Shieh, C.-K., Kamahora, T., Vannier, D. M., Govindarajan, S., and Lai, M. M. C. (1987). Molecular cloning and sequencing of a human hepatitis delta (d) virus RNA. Nature 329, 343±346. Netter, H. J., Hsieh, S.-Y., Lazinski, D., and Taylor, J. (1993). Modified HDV as a vector for the delivery of biologically-activate RNAs. Prog. Clin. Biol. Res. 382, 329±335. Polo, J. M., Lim, B., Govindarajan, S., and Lai, M. M. C. (1995). Replication of hepatitis delta virus RNA in mice after intramuscular injection of plasmid DNA. J. Virol. 69, 5203±5207. Rizzetto, M. (1983). The delta agent. Hepatology 3, 729±737. Ryu, W.-S., Bayer, M., and Taylor, J. (1992). Assembly of hepatitis delta virus particles. J. Virol. 66, 2310±2315. Ryu, W.-S., Netter, H. J., Bayer, M., and Taylor, J. (1993). Ribonucleoprotein complexes of hepatitis delta virus. J. Virol. 67, 3281±3287. Sheu, S. Y., and Lo, S. J. (1992). Preferential ribosomal scanning is involved in the differential synthesis of the hepatitis B viral surface antigens from subgenomic transcripts. Virology 188, 353±357. Sureau, C., Guerra, B., and Lanford, R. E. (1993). Role of the large hepatitis B virus envelope protein in infectivity of the hepattitis delta virion. J. Virol. 67, 366±372. Wang, C.-J., Chen, P.-J., Wu, J.-C., Patel, D., and Chen, D.-S. (1991). Small-form hepatitis B surface antigen is sufficient to help in the assembly of hepatitis delta virus-like particles. J. Virol. 65, 6630± 6636. Wang, H.-W., Chen, P.-J., Lee, C.-Z., Wu, H.-L., and Chen, D.-S. (1994). Packaging of hepatitis delta virus RNA via the RNA-binding domain of hepatitis delta antigens: different roles for the small and large delta antigens. J. Virol. 68, 6363±6371. Wang, K.-S., Choo, Q.-L., Weiner, A. J., Ou, J.-H., Najarian, R. C., Thayer, R. M., Mullenbach, G. T., Denniston, K. J., Gerin, J. L., and Houghton, M. (1986). Structure, sequence and expression of the hepatitis delta (d) viral genome. Nature 323, 508±514. Wang, K.-S., Choo, Q.-L., Weiner, A. J., Ou, J.-H., Najarian, R. C., Thayer, R. M., Mullenbach, G. T., Denniston, K. J., Gerin, J. L., and Houghton, M. (1987). Structure, sequence and expression of the hepatitis delta viral genome. Nature 328, 456. Weiner, A. J., Choo, Q.-L., Wang, K.-S., Govindarajan, S., Redeker, A. G., Gerin, J. L., and Houghton, M. (1988). A single antigenomic open reading frame of the hepatitis delta virus encodes the epitope(s) of both hepatitis delta antigen polypeptides p24d and p27d. J. Virol. 62, 594±599. Wu, J.-C., Chen, P.-J., Kuo, M. Y. P., Lee, S.-D., Chen, D.-S., and Ting, L.-P. (1991). Production of hepatitis delta virus and suppression of helper hepatitis B virus in a human hepatoma cell line. J. Virol. 65, 1099± 1104. Yeh, T.-S., Lo, S. J., Chen, P.-J., and Lee, Y.-H. W. (1996). Casein kinase II and protein kinase C modulate hepatitis delta virus RNA replication but not empty viral particle assembly. J. Virol. 70, 6190±6198.