INTRODUCTION

Hepatitis B virus (HBV)¹ infection is associated with chronic hepatitis and development of hepatocellular carcinoma (Beasley et al., 1981). Over 300 million individuals are infected with HBV worldwide. HBV is the smallest known human virus, with a 3.2-kilobase genome, and it encodes four major open reading frames, pre-C/C, pre-S2/S, P, and X (reviewed by Nasal and Schaller (1993)). HBV DNA in the virion is partially double-stranded and held in circular conformation. The native molecule has a fixed 5 end, whereas the 3 end of the plus (+) strand has variable termini in each molecule. HBV replicates via a reverse transcription mechanism using longer than genome-length, pregenomic RNA. The reverse transcriptase/RNase H activities are encoded by P open reading frame. P protein is covalently attached to the () strand DNA via its terminal protein (TP) domain (reviewed by Nasal and Schaller (1993)).

One of the open reading frames of the HBV genome encodes a protein termed HBx. It is now generally acknowledged that HBx provided in trans can increase gene expression. The transactivation function of HBx has been tested in the context of a wide variety of viral and cellular promoter/enhancer elements (Twu and Schloemer (1987), Zahm et al. (1988), Spandau and Lee (1988), Siddiqui et al. (1989), Colgrove et al. (1989), Hu et al. (1990), Rossner (1992) and references therein). One of the possible mechanisms by which HBx may function was first shown to be via protein-protein interactions with the transcriptional factors ATF-2/CREB (Maguire et al., 1991). Subsequently, interactions of HBx with several cellular proteins were reported. These include p53 tumor suppressor protein (Wang et al., 1994), RNA polymerase subunit RPB5 (Cheong et al., 1995), universal transcriptional factor, TATA-binding protein (TBP) (Qadri et al., 1995), and a putative DNA repair protein, UV-DDB (Lee et al., 1995). Recently, we have shown that HBx interacts with the DNA helicase components (ERCC2 and ERCC3) of basal transcriptional factor TFIIH and stimulates the DNA helicase activity of TFIIH.² All of the proteins with which HBx has been shown to form heteromeric complexes reside in the nucleus. Although not shown by direct protein-protein interactions, the activities of HBx have been shown to influence the events relevant to signal transduction pathways (Cross et al., 1993; Kekule et al., 1993; Natoli et al., 1994; Benn and Schneider, 1994). However, the cellular targets of HBx in signal transduction pathways have not been identified.

Transcription initiation and promoter clearance by RNA ploymerase II requires the melting of the DNA template (Conaway and Conaway, 1993). HBx interactions with the components of basal transcription factors TBP and TFIIH suggest a possible role of HBx at these stages of transcription. The formation and maintenance of single-stranded DNA within the nucleus requires the action of single-stranded DNA binding proteins (SSBs) (Yagil, 1991). The role of SSBs would be to initiate and maintain localized single-stranded conformations during important cellular functions such as DNA replication, transcription, recombination, and DNA excision repair.

Here, we report that HBx interacts with ssDNA in a manner that is not sequence-specific. This study further details characterization of various biochemical parameters of HBx's interaction with ssDNA. These include the effects of mono- and divalent cations, the effect of cardiolipin and heparin, pH and temperature dependence, and variations in the incubation time. Our results show that HBx-ssDNA interactions are tighter at low concentrations of mono- and divalent salt and weaker at higher salt concentrations, consistent with the predicted salt dependence for protein-DNA interactions. The HBx-ssDNA interaction is stable up to 45 °C. Heparin failed to compete completely with DNA for HBx binding, suggesting that HBx interactions with DNA are stabilized by significant interactions with the DNA bases. The biological relevance of HBx interaction with ssDNA remains to be investigated.

EXPERIMENTAL PROCEDURES

Construction of full-length HBx fusion with glutathione S-transferase gene (pGST-X) has been described previously (Qadri et al., 1995). In the plasmid pSPX, the X-gene was cloned within the polylinker in pSP65 in front of the SP6 promoter for in vitro translation of HBx. HBV DNA was extracted from HBV-positive human serum by phenol extraction and ethanol precipitation procedure.

A 26-nucleotide-long oligonucleotide termed RXR-S containing the sequence 5-AGTAAACAGTACATGAACCTTTACCC-3, corresponding to nucleotides 1125-1150 of enhancer I of HBV genome was used in the mobility shift assay. The competitor oligonucleotides are as follows with the following sequence: AS, antisense 5-GGGTAAAGGTTCATGTACTGTTTACT-3; FV-S, 5-AGTAAACAGTAACATGAACCTTTACCC-3; FV-AS, 5-GGGTAAAGGTTCATGTTACTGTTTACT-3; and X-P, 5-AAGAGATGATTAGGCAGAGGT-3. T7-P/P and SP6-P/P primers were purchased from Promega. Competitor oligonucleotides described above and oligonucleotides of d(A)₂₅, d(C)₂₅, d(G)₂₅, and d(T)₂₅ were synthesized at the core facility of the Department of Microbiology, University of Colorado Health Sciences Center (Denver, CO).

SDS, acrylamide, N,N-methylene-bisacrylamide, Coomassie Brilliant Blue R-250, ammonium persulfate, and TEMED were purchased from Bio-Rad. Hydrocholoric acid (HCl), potassium chloride (KCl), sodium chloride (NaCl), magnesium chloride (MgCl₂), and potassium hydroxide (KOH) were from Fisher. Heparin, cardiolipin, HEPES, phenylmethylsulfonyl fluoride (PMSF), isopropyl-1-thio--D-galactopyranoside, dithiothreitol (DTT), glutathione, Nonidet P-40, Triton X-100, and ATP were purchased from Sigma. Protein assay reagents and SDS-PAGE molecular weight markers were purchased from Bio-Rad. [-³²P]ATP was purchased from ICN. Avian myeloblastosis virus reverse transcriptase and T4 polynucleotide kinase were purchased from Promega.

Glutathione S-transferase affinity beads were purchased from Pharmacia Biotech Inc. Single-stranded DNA agarose was purchased from Life Technologies, Inc.

Single-stranded oligonucleotides were labeled with [-³²P]ATP in the presence of T4 polynucleotide kinase. The radiolabeled probe was gel-purified, and an aliquot of 1 × 10³ cpm (0.1 ng; 0.484 nM) was used for a binding reaction with either GST or GST-X fusion proteins. The ³²P-labeled single-stranded DNA probe and fusion proteins were incubated in a buffer in a 25-µl reaction volume that contained, 100 mM KCl, 20 mM HEPES (pH 7.9), 5 mM MgCl₂, 5% glycerol (v/v), 2 mM DTT, and 0.2 mM PMSF at 30 °C for 30 min. The ssDNA-HBx complexes were resolved in a nondenaturing 6% polyacrylamide gel in 0.5 × TBE running buffer at 4 °C. The gels were dried and exposed to x-ray film (XAR; Kodak) at 80 °C with an intensifying screen. Unless specified otherwise, the GST-HBx concentration used in the gel mobility experiments was 4 × 10⁸ M. The results were analyzed using a PhosphorImager by direct exposer of the gel. To quantitate the DNA represented by the different bands, a box surrounding each band was defined, and the relative densities of the bands were calculated. The fraction of protein-DNA complex (X) was calculated from the ratio the intensity of the shifted band to the total DNA. The binding affinity was estimated from the HBx concentration required to bind half of the DNA (Carey, 1991).

A 26-nucleotide-long (RXR-S) ssDNA probe was labeled at the 5 termini with [-³²P]ATP in the presence of T4 polynucleotide kinase. The probe was incubated with either GST or GST-X proteins for 30 min at 30 °C and irradiated in a UV Stratalinker model 1800 (Stratagene) at 312 nm for 30 min at 4 °C. After incubation and UV-irradiation, the covalently cross-linked proteins with radiolabeled ssDNA were separated on 13% SDS-PAGE and autoradiographed.

The GST fusion proteins were purified as described previously (Smith and Johnson, 1988). To dissociate contaminant Escherichia coli proteins, the binding of bacterial lysates to glutathione S-transferase affinity beads and subsequent washing was conducted in the presence of 1.2 M NaCl to remove nonspecific binding of bacterial proteins to HBx. An additional step of purification was included, which involved passing the purified GST-X fusion protein through a single-stranded DNA agarose column. The protein fractions from ssDNA agarose were first eluted in 700 mM KCl, 25 mM HEPES (pH 7.9), 1 mM DTT, 1 mM EDTA, and 0.2 mM PMSF and then dialyzed in a buffer containing either no monovalent cations or 100 mM KCl and 25 mM HEPES, pH 7.9, 10% glycerol (v/v), 1 mM ATP, 1 mM DTT, and 0.2 mM PMSF.

RESULTS

Using several standard methodologies, we demonstrate the ability of HBx to bind single-stranded nucleic acids. The HBx interaction with single-stranded nucleic acids was first examined by electrophoretic mobility shift assay (EMSA), using a glutathione S-transferase-HBx fusion protein. The results of gel mobility shift assays described in Fig. 1A, show that while ssDNA did not bind GST protein, it bound GST-X. These complexes appeared to be specific for ssDNA, since their formation was competitively prevented by a 100-fold excess of several ssDNA oligonucleotides with unrelated sequences (Fig. 1A, lanes 4-10) but not by a double-stranded (ds) DNA (lane 11). This is in agreement with the previous results of others showing that HBx does not bind dsDNA (Rossner (1992) and references therein). The ssDNA-HBx complex was further retarded in the presence of anti-HBx serum, indicating that HBx is a component of that complex (lane 12). Together, these results suggest that HBx interacts with ssDNA in a manner that is non-sequence-specific.

In the next experiment, we examined the direct binding of HBx to ssDNA by the UV cross-linking method. A ³²P-labeled ssDNA oligonucleotide probe was mixed with either GST or GST-X proteins and UV-cross-linked (Fig. 1B). GST protein was not cross-linked to the probe (lane 2). A radiolabled GST-X protein band of approximately 64 kDa, which represents GST-X protein cross-linked to the ³²P-labeled probe, was fractionated by SDS-PAGE (lane 3). The ³²P-labeled probe migrated as a 20-kDa band in the SDS gel as seen in lane 1 and other lanes. Further competition with unlabeled dsDNA and RNA in the UV cross-linking experiment confirms the specificity of HBx interaction with ssDNA (lanes 4-7). It should be noted here that the E. coli SSB protein migrates at about 19 kDa on SDS-PAGE and does not fall within the range of 44-65-kDa proteins, which allays the concern of a contamination of ssDNA binding activity from bacterial extracts in UV cross-linking experiments. In summary, the results obtained by both mobility shift assays and UV cross-linking method indicate that HBx interacts with ssDNA and RNA. Characterization of HBx's interactions with RNA will be described elsewhere.

To provide an alternate source of HBx for binding studies, the X-protein was synthesized in rabbit reticulocyte lysates in the presence of [³⁵S]methionine. The in vitro synthesized [³⁵S]HBx was allowed to interact with single-stranded DNA-agarose beads (Fig. 1C, lane 6). After extensive washing, the mixture was fractionated by 13% SDS-PAGE. The results show that HBx bound to the ssDNA-agarose beads (lane 6). As a positive control, the basal transcriptional factor TFIIB was included (lane 4). TFIIB has been previously shown to bind single-stranded DNA. Unprogrammed lysates do not show a ssDNA binding protein (lane 5). These results together show that HBx, irrespective of the source, displayed affinity for single-stranded nucleic acids.

The genomic HBV DNA within the virions is partially double-stranded, containing a substantial single-stranded DNA region. Most experimental approaches require the disruption of virions in order to observe HBx-ssDNA interactions. Chaotropic agents will likely lead to the dissociation of HBx-ssDNA complexes. With this caveat in mind, we have used phenol-extracted HBV genomic DNA as a competitor in a mobility shift assay to demonstrate that HBx can associate with the ssDNA region of the genomic HBV DNA (Fig. 1D, lane 4). HBV DNA filled in with reverse transcriptase did not compete in this assay (lane 5). The rationale of using HBV DNA (in native conformation and filled in) is only to indicate the specificity of HBx in binding the single-stranded region of the DNA. These results by no means suggest that HBx is associated with the native HBV DNA in the virions. These studies are currently in progress. Filamentous bacteriophage M13, mp18 ssDNA, and filled in dsDNA were used as controls (lanes 6-7).

In the next analysis, purified E. coli single-stranded DNA binding protein (E. coli SSB) was used to investigate its ability to displace HBx-ssDNA complexes. The E. coli SSB protein is a stable homotetramer and binds to ssDNA nonspecifically (Lohman and Ferrari (1994), and references therein). In a competition analysis, E. coli SSB competed with HBx for binding to ssDNA (Fig. 1E, lanes 4-6). These results were further confirmed by the UV cross-linking experiment shown in Fig. 1F (lanes 4-6). In this respect, HBx differs from other sequence-specific mammalian SSB proteins whose interaction with ssDNA is not displaced by E. coli SSB in competition experiments. These results further support the idea that HBx binds to ssDNA without sequence specificity.

To determine the association time of HBx-ssDNA complexes, incubations were performed at various lengths of time at 30 °C (Fig. 2A). The results of the electrophoretic mobility shift assay revealed that the minimum time period required for HBx to interact with ssDNA was 2.0 min (lane 7). After longer exposure (5 days), faint HBx-ssDNA complexes were observed in lanes 4-6 in contrast to the overnight exposure of the gel shown in Fig. 2A (data not shown). Incubation at 3, 4, 5, 10, 20, 30, and 60 min did not result in a change in the intensity of the HBx-ssDNA complex band (lanes 8-14). GST alone did not interact with the radiolabeled single-stranded DNA probe (lane 2). Together, these results show that most of the HBx-ssDNA complexes are formed approximately within 2 min and are stable for at least up to 60 min.

HBx interactions with ssDNA were examined at varying pH, ranging from 3 to 12. The EMSA results show that HBx did not interact with ssDNA at pH 3 and 3.5 (Fig. 2B, lanes 2 and 3). The affinity for HBx to interact with ssDNA was reduced between pH 4 and 6.5 (lanes 4-9). Maximal interactions occurred between pH 7.5 and 11 (lanes 10-16). At pH higher than 11, the interaction was considerably reduced (lanes 17-18). The observation that the complex is most stable between pH 7.5 and 11 suggests that amino acid side chains such as lysine may be involved in the binding; the pK_a for the lysine side chain is 10.53 in the free amino acid. It is possible that the pH stability of the HBx-ssDNA interactions reflect the requirement for lysine to be protonated for complex formation. The weak HBx-DNA interactions at the pH extremes are likely due to protein denaturation.

The HBx-ssDNA interaction studies were conducted in the absence and presence of various concentrations of KCl, NaCl, and MgCl₂ (Fig. 3). Although MgCl₂ is not required for the HBx-ssDNA interaction, tighter binding was observed at low MgCl₂ concentrations compared with NaCl or KCl. Note the absence of free DNA at low concentrations of MgCl₂ (compare lanes 3-9 in Fig. 3C with lanes 2-4 in Fig. 3, A and B). It is possible that the observed tighter binding in the presence of low MgCl₂ concentrations is due to preferential Mg²⁺ binding by the complex. At higher MgCl₂ concentrations, weaker binding was observed. Above 100 mM MgCl₂ concentration, an abrupt decrease in binding was observed, and a higher molecular weight complex is clearly visible in the wells. Higher concentrations of MgCl₂ may promote protein aggregation.

The data described in Fig. 3 clearly show that the HBx-ssDNA interaction decreased with increasing divalent and monovalent salt concentration, consistent with the predicted behavior for protein-DNA interactions (Record et al., 1976, 1978). A higher [KCl] was required to abolish binding relative to [NaCl]; no binding was detected at 800 mM NaCl compared with 1500 mM KCl. These results suggest that preferential interaction of HBx with K⁺ may accompany binding. In addition, MgCl₂ was effective at disrupting the HBx-ssDNA complexes at much lower concentration (300 mM; Fig. 3C, lane 14) compared with the monovalent salts (1500 mM KCl (Fig. 3A, lane 18) and 800 mM NaCl (Fig. 3, lane 14)). These properties are consistent with the expected salt dependence associated with protein-ssDNA interactions (Record et al., 1976, 1978).

Heparin is a glycosamino glycan containing a negatively charged carboxylate or a sulfate group and has been frequently used as a competitor for DNA in protein-DNA binding studies. Heparin is dissimilar in structure to DNA. It has therefore been used to investigate protein-nucleic acid complex stability (i.e., phosphate backbone versus DNA base interactions). Heparin was tested for its ability to compete with ssDNA for HBx binding. A decrease in binding was observed with increasing heparin concentration up to 200 mM (Fig. 4A, lane 10). However, heparin was not able to completely compete with DNA, suggesting that significant nonelectrostatic interactions contribute to the DNA-HBx complex stability. Cardiolipin is a diphosphatidyl glycerol containing two negatively charged phosphate groups. When cardiolipin was used as a competitor in EMSA, similar results were obtained (Fig. 4B).

The ability of the reducing agent dithiothreiotol (DTT) to affect the stability of HBx-ssDNA interactions was determined by incubating the HBx with ssDNA at increasing concentrations of DTT (0.1-100 mM) (Fig. 4C, lanes 2-10). It was observed that at up to 5 mM DTT, the HBx-ssDNA interactions were not affected (lane 5). However, DTT concentrations of 10 mM and above had an inhibitory effect on these interactions (lanes 6-8). The HBx used in this experiment was purified in the presence of a very low concentration of DTT (0.1 mM); therefore, the observed HBx-ssDNA complexes seem to be weaker in comparison with other experiments (Fig. 4, A and B). It is likely that during the preparation of HBx, some of its DNA binding activity was lost due to oxidation of the protein. We conclude that reduced thiol groups are involved in the binding activity of HBx to ssDNA, probably through stabilization of the correct conformation of the protein.

Effect of temperature on HBx-ssDNA interaction was investigated by incubating the reaction mixture at various temperatures (Fig. 5). Significant binding is observed between 20 and 45 °C. A fraction (~20%) of ssDNA probe did not bind HBx at 4 °C (lane 3). At 65 °C only a small fraction (~5%) of the HBx-ssDNA complex was stable (lane 8).

HBx was added in increasing concentrations (5-100 ng; 0.09-90 nM) to a constant amount of radiolabeled DNA (0.1 ng; 0.484 nM) (Fig. 6). The single-stranded DNA oligonucleotide probe representing the nucleotide 1125-1150 sequences of the enhancer I region of HBV genome was radiolabeled at the 5 end by T4 polynucleotide kinase in the presence of [-³²P]ATP. The HBx concentration required to bind half of the DNA was approximately 90 nM, determined from PhosphorImager analysis from EMSA gels. We emphasize that this value is only an order of magnitude estimate for the dissociation constant. We note that these titrations were performed with the bacterially expressed GST fusion protein and may vary with a purified and unfused HBx. In our experience with the Studier's pET expression system in which the X-gene is not fused to any moiety, HBx formed inclusion bodies (Jameel et al., 1990). Denaturation of these complexes uniformly produced functionally inert HBx molecules, perhaps due to malfolding. GST-X fusion protein on the other hand produced soluble protein and hence has been used extensively for in vitro analyses.

The base specificity of the HBx-ssDNA interaction was investigated by performing EMSA in the presence of varying concentrations (0.1, 1, 5, 10, 50, 100, 500, and 1000 ng) of competitor homopolymer oligonucleotides. These are as follows; d(A)₂₅ (Fig. 7A), d(G)₂₅ (Fig. 7B), d(C)₂₅ (Fig. 7C), and d(T)₂₅ (Fig. 7D). The results described in Fig. 7 show that polypyrimidines d(C)₂₅ and d(T)₂₅ were more effective competitors than polypurines d(A)₂₅ and d(G)₂₅. Even in the presence of a molar excess of cold oligonucleotides (d(A)₂₅ and d(G)₂₅) very little effect of competitor oligonucleotides was seen on the formation of the HBx-ssDNA complex (Fig. 7, A and B, lanes 9 and 10). These results suggest that HBx shows a preference for pyrimidine residues during its binding to ssDNA. Similar base preferences have been observed with other single-stranded DNA binding proteins (Lohman and Ferrari, 1994; Kim et al., 1992).

DISCUSSION

The mechanism(s) by which HBx transactivates gene expression is a subject of intense investigation. Its inability to bind cis-acting regulatory DNA elements has been well documented (reviewed by Rossner (1992) and references therein). In this report, we present evidence that HBx interacts with ssDNA in a manner that is not sequence-specific. The experimental approaches used include mobility shift assay, UV cross-linking of HBx with ssDNA, and direct retention of HBx on a single-stranded DNA agarose column. We have conducted a series of experiments relating to the various biochemical parameters of HBx-ssDNA interactions. These studies indicate that HBx displays characteristics that are similar to those known for other viral and cellular SSB proteins. HBx binds preferentially to single-stranded nucleic acids and fails to bind dsDNA. Although HBx does not bind to ssDNA sequence with any specificity, it does show some base specificity of binding, i.e. d(T)₂₅ and d(C)₂₅ are more effective at competing with ssDNA for binding than d(A)₂₅ and d(G)₂₅. This kind of base specificity has been observed for other SSBs (Lohman and Ferrari, 1994; Kim et al., 1992).

The determination of the heat stability of the HBx binding revealed that HBx-ssDNA interactions are stable at 45 °C. Other proteins that have been shown to remain stable in complex with DNA at elevated temperatures include basal transcription factor TFIIA at 55 °C (Waldschmidt and Seifart, 1992), TFIID at 47 °C (Nakajima et al., 1988), and mammalian transcription factor PBP binding to U6 genes at 43 °C (Wanandi et al., 1993). At pH 11, the HBx-ssDNA interaction was fairly stable, which suggests a potential role of lysine residues in these interactions. In the presence of 10 mM DTT or greater, HBx-ssDNA interactions were abolished. This is likely a result of the disruption of Cys-Cys interactions in the protein necessary to maintain a properly folded structure. Our future work will be directed toward altering these residues by site-directed mutagenesis to identify the regions of HBx critical for its interactions within ssDNA.

The single-stranded DNA- or RNA-binding proteins play significant roles in DNA repair, replication, and recombination (Hoeijimakers (1991), Challberg and Kelly (1989), Lohman and Ferrari (1994), and references therein). SSBs can also act as DNA-binding stimulatory factors. One such protein, a 45-kDa DNA-binding stimulatory factor isolated from yeast, has been shown to promote binding of the purified oestrogen receptor to its responsive element (Mukherjee and Chambon, 1990). FBP is an another example of a sequence-specific, single-stranded DNA binding protein which has been shown to activate the upstream element at the FUSE site of the c-myc gene (Duncan et al., 1994). This activity would be consistent with previous observations relating to HBx's ability to induce the DNA binding specificity of ATF-2/CREB and c-Jun·c-Fos heterodimers to their respective sequence motifs (Maguire et al., 1991; Natoli et al., 1994). Confirming these binding properties, a recent report demonstrates that in the presence of HBx, the binding specificity of CREB (K_D = 10⁷ M) with the HBV CRE motif within the enhancer I element increased by an order of magnitude (K_D = 10⁸ M) (Williams and Andrisani, 1995).

HB virion DNA is partially double-stranded with a variable length of single-stranded region. The unusual and unique nature of native virion DNA is the result of incomplete DNA synthesis in the viral core particles. Whether HBx is bound to the single-stranded region of the native genomic HBV DNA is an interesting question that needs to be investigated. In the context of HBV life cycle, HBx may be implicated in binding to the ssDNA region of the genomic DNA and perhaps pregenomic RNA during precore assembly. The data presented here suggest that HBV transactivator protein, HBx, can bind to single-stranded DNA in vitro. Our future studies will focus on defining in greater detail the exact role of HBx in the context of the HBV life cycle. Several eukaryotic DNA viruses encode single-stranded DNA binding proteins that bind to ssDNA. These include SV40 T antigen (Stahl et al., 1986), herpes simplex virus 1 ICP8 and UL9 (Olivo et al., 1988), adenovirus DBP (p72) (Challberg and Kelly, 1989), and Autographa californica nuclear polyhedrosis virus Lef-3 (Hang et al., 1995). The role of these proteins have been documented in DNA replication (Challberg and Kelly, 1989). The A. californica nuclear polyhedrosis virus Lef-3 protein has been shown to play a role in late and very late gene expression.

A wealth of information is available on the biochemical properties of bacterial single-stranded DNA or RNA binding proteins (SSBs). SSBs play significant roles in transcription, DNA repair, replication, and recombination (Hoeijmakers, 1993; Challberg and Kelly, 1989; Lohman and Ferrari, 1994). In its newly assigned role as a viral SSB, one possible function of HBx could be to bind melted template and prevent reannealing during transcription and perhaps other cellular processes where such a function is required. The single-stranded DNA binding activity of HBx may also have an impact on other important cellular functions such as DNA repair. Support for HBx's role in DNA repair comes from a recent report, in which HBx was shown to interact with UV-damaged DNA repair enzyme (Lee et al., 1995). Another possible physiological relevance of single-stranded DNA binding property of HBx could be envisioned during transcription initiation. We have recently shown that HBx can interact with TBP, a principal component of TFIID (Qadri et al., 1995) and components of TFIIH.² HBx-TBP and HBx-TFIIH interactions may provide access for HBx to the melted template, where it can bind to the denatured DNA strand and contribute to the stability of preinitiation complex assembly process.

Our future investigations will be focused on elucidating the biological and functional relevance of HBx-ssDNA interactions. This report, which defines the biochemical criteria of HBx-ssDNA interactions, constitutes the first step toward that goal. These biochemical parameters would be useful in designing strategies in recovering HBx-genomic DNA complexes from HBV particles and in conducting analyses of interactions between HBx and cellular targets in vitro. Further, these studies have the potential to provide clues to its fundamental role in viral and cellular processes and contribute to the understanding of the exact mechanism(s) by which HBx may accomplish its regulatory functions.