See article on page 357
The hepatitis B virus (HBV) X gene encodes a 154 amino acid protein called hepatitis B virus X protein (HBx). 1 This protein is a multifunctional protein exhibiting numerous activities affecting gene transcription, intracellular signal transmission, cell proliferation and apoptotic cell death. Of these activities, the best known is its promiscuous transactivation activity. 2 HBx is capable of up-regulating a wide range of cellular and viral genes, including its own. Although its precise role in the viral replication cycle remains unknown, HBx is required for natural woodchuck hepatitis virus infection. 3 HBx does not directly bind DNA. The transactivation activity is, therefore, subjected to complex mechanisms such as protein–protein interactions, regulation of phosphorylation, mRNA stablization and alteration of nucleocytoplasmic translocation. It is possible that the multiple activities of HBx perturb cell growth and differentiation to a certain extent and contribute to hepatocarcinogenesis. Several transgenic mice experiments indicate that mice harbouring HBx either develop liver cancer or have accelerated development of neoplasms when they are exposed to other carcinogens. 4,5 In this issue of the Journal, Arbuthnot et al. have comprehensively reviewed the possible roles of HBx in hepatocarcinogenesis. 6 They have focused on three major aspects: apoptosis, DNA repair and the mitogen-activated protein kinase (MAPK) and Janus kinase (JAK)/signal transducer and activator of transcription (STAT) pathways.
Apoptosis, or programmed cell death, is necessary for the elimination of damaged or differentially redundant cells. Two apoptotic pathways were discussed in the review; the ‘classical’ caspase-mediated pathway, and the p53-mediated pathway. The caspase-mediated pathway is initiated by Fas ligand–Fas (or tumour necrosis factor (TNF)-α–TNF receptor) interaction, which leads to activation of the receptor. Subsequently, a mediator (e.g. TNF receptor-associated death domain (TRADD) or Fas-associating protein with death domain (FADD)) is activated through the interaction of its own death domain and that on the cytoplasmic portion of the receptor. Thereafter, the downstream caspase proteases are activated, leading to mitochondrial dysfunction and subsequent endonucleolytic cleavage, DNA fragmentation and nuclear disassembly. 7 Many proteins in the Bcl-2 family serve as regulators for the caspase-mediated pathway, while other cellular factors, such as p53, indirectly regulate this pathway. p53 is capable of mediating cell cycle arrest and apoptosis. 8 This protein is a transcription activator. It transcriptionally activates several genes including mdm2, p21Cip1/Waf1, Bax, Gadd45, cyclin G and insulin-like growth factor (IGF)-BP3. Mouse double-minute 2 (mdm2) binds directly to p53 to inhibit transcriptional activation. p21Cip1/Waf1 inactivates cyclin-dependent kinase and thus, the phosphorylation of the retinoblastoma protein (pRB), leading to cell cycle arrest. Other proteins, such as Bax (a member of Bcl-2 family), can modulate the caspase-mediated apoptosis pathway. Although these p53-activated proteins play important regulatory roles, there are lines of evidence suggesting that mechanisms other than transcriptional activation are responsible for p53-mediated apoptosis. 9,10
An early report demonstrated an inhibitory effect of HBx on apoptosis, 11 but was contradicted by a later report that showed a pro-apoptotic effect in experiments using HBx transgenic mice. 12 Similarly, an initial report suggested that HBx sensitized Chang liver cells in TNF-α-induced apoptosis 13 but other researchers demonstrated a resistant effect using Rev2 cells. 14 The interaction between HBx and p53 is better defined than other interactions, largely because the interaction can be shown by co-immunoprecipitation experiment. 15 Although the molecular mechanism for the inhibition of p53 by HBx is still in debate, two interesting findings are noteworthy. First, the subcellular localization of p53 can be altered by coexpression of HBx. 16 Second, the interacting domains of the two proteins overlap with the transactivation domain of HBx and a region in p53 responsible for forming stable tetramers. 17,18
Transcription factor IIH (TFIIH) is a basal transcription factor of all structural genes. This ‘factor’ is a protein complex. Several of its subunits, including two helicases with opposing polarity, xeroderma pigmentosa complementation group (XP)B (excision repair cross complementation group (ERCC)3) and XPD (ERCC2), have been identified in recent years. Additionally, TFIIH exhibits a kinase activity for the carboxy-terminal domain of the large subunit of RNA polymerase II. Interestingly, this kinase is also capable of activating other cyclin-dependent kinases (CDK)–cycline complexes, leading to the argument that a link between TFIIH and cell cycle regulation exists. 19 Because of the bidirectional helicase activities, TFIIH plays an important role in nucleotide excision repair. 20 The XPA protein, and possibly other damaged DNA binding (DDB) proteins, first recognize a damaged site. Then, TFIIH (XPB and XPD), replication protein A (RPA), and two structure-specific endonucleases, XPG and ERCC1, are recruited to this site. The helicases open the helix and the endonucleases make the incision. During the process, p53 interacts with several factors, including XPB and XPD, presumably to inhibit translesion transcription. Accumulating evidence has suggested that HBx may disturb this scenario in at least two ways, by interfering with the p53–XPB/XPD interaction by interacting with p53, 21 and disturbing the recognition of damaged DNA by interacting with a putative DDB. 22 The significance of the latter is still unclear. 22
The MAPK pathway is a general term for a number of cellular signal cascades that involve the sequential activities of protein kinases in a similar way. Such pathways start with activation of cellular receptors with intrinsic protein kinase activities (receptor tyro-sine kinases or RTK), resulting in dimerization/autophosphorylation of multiple tyrosine residues in their own cytoplasmic ports. Thereafter, the Src homology 2 (SH2) domain-containing protein, including Src family protein kinases (Shc and Grb2), are recruited to the docking sites. An ‘adapter’ (e.g. Son-of-sevenless (Sos)), which mediates the interaction between Grb2 and Ras activities, converts Ras/guanosine diphosphate to Ras/guanosine triphosphate (GTP). Ras/GTP then stimulates MAPK through cascades of kinase phosphorylation. Finally, the MAPK (e.g. Jun N-terminal kinase (JNK) or extracellular signal-regulated kinase (ERK)) is activated and thus capable of activating targeted transcription factors such as Jun, Fos and Myc. 23 HBx up-regulates the MAPK pathway in different ways, such as increasing GTP uptake by Ras to increase the Ras signalling over a prolonged period of time or enhancing the association between Shc, Grb2 and Sos. 24 However, other reports suggest that HBx is capable of activating the targeted genes through alternative routes independent of the MAPK pathways. 25,26
The JAK/STAT pathway starts with activation of receptors to generate receptor dimers, leading to auto-phosphorylation of JAK and phosphorylation of the tyrosine residues on the receptors. The phosphorylated receptors allow docking of STAT, which are subsequently phosphorylated. The phosphorylated STAT are released from the receptors and translocated to the nucleus to activate the targeted genes. 27 HBx is capable of activating JAK1-tyrosine kinase, leading to constitutive phosphorylation of STAT and thus constitutive activation of the targeted genes. 28 The exact mechanism for the effect of HBx on JAK activation remains unknown.
In summary, emerging evidence suggests that HBx is capable of interfering with several important cellular signalling pathways. The exact molecular mechanism is, however, largely unknown. Furthermore, contradictory results were reported using different experimental systems. To improve our understanding of the action of HBx, several issues need to be addressed. In the signalling pathways, the roles of some presumably functional molecules should be further defined (e.g. the role of DDB protein in nucleotide excision repair pathway). It would be very helpful if the signalling pathway itself could be further clarified (e.g. the p53-mediated apoptosis pathway). None of the putative roles of HBx in the signalling pathways has yet been convincingly proven by experiments using human tissue. To perform such experiments, the availability of a highly specific antibody against HBx is imperative. The report suggesting that most of the anti-HBx used are not specific, greatly compromises the credibility of currently available data. 29 Finally, to claim that HBx is capable of directly interfering with a particular signalling pathway, speci-fic interaction between HBx and a key player in the pathway should be clearly demonstrated. In lieu of such evidence, the effect of HBx on a signalling pathway can only be indirect and the real mechanism remains elusive.