The HDV Replication Cycle

Currently, there are no convenient cell culture model systems to study HDV infection. Thus, many of the details of the natural HDV replication cycle are still unclear. To date, most studies of the HDV replication cycle have been carried out using cultured cells transfected with HDV cDNA constructs longer than genome length (1.2x to 3x genomic length) under the control of strong foreign promoters. By this technique, the S-HDAg, essential for initiation of HDV RNA replication, is generated either from an mRNA transcribed directly from the transfected HDV cDNA or, alternatively, provided in trans from a co-transfected plasmid. In either case, this experimental approach introduces an artificial requirement of a DNA-dependent transcription step in ordertogenerateaprecursor HDVRNA, which, inturn, leadstosubsequent RNA replication. Our laboratory has pioneered an alternative approach that circumvents the need for the cDNA step and involves the transfection of the 1.2x genomic-length HDV RNA together with an mRNA encoding HDAg (Modahl and Lai 1998). This method has led to very significant revisions of the previous concepts of HDV RNA established by cDNA transfection.

HDV presumably enters cells through a similar cellular receptor to that used by HBV as infection of both these viruses depends on the large form of HBsAg in the envelope (Sureau et al. 1993). The following steps, from uptake, uncoating to delivery of the viral RNA to the nucleus (where RNA replication takes place), are unknown although the latter is most likely reliant on the combined RNA-binding and nuclear localizing abilities of the HDAg present as part of the infecting ribonucleoprotein complex.

The next step is HDV RNA replication, which is thought to proceed by a double rolling circle model similar to that proposed for viroids (Branch and Robertson 1984; Fig. 3). In this model, input circular genomic HDV RNA serves as a template for synthesis of the complementary antigenomic strand. As HDV RNA synthesis continues, monomers of antigenomic HDV RNA are cleaved from the growing transcript by the ribozyme activity intrinsic to both polarities of HDV RNA. The resultant antigenomic RNA species are then ligated into a circular form. The latter process was originally thought to be dependent on the self-ligating activity of HDV RNA (Sharmeen et al. 1989) but more recent evidence suggests that this may be carried out by a cellular RNA ligase (Reid and Lazinski 2000). The circularized antigenomic HDV RNA then serves as a template for genomic HDV RNA synthesis, with the subsequent replication steps proceeding by a similar mechanism (Fig. 3). Initially, evidence for this model came from the detection of greater-than-unit HDV RNA intermediates in infected or transfected cells (Chen et al. 1986; Kuo et al. 1989; Macnaughton et al. 1990) and from the observation that mutations interfering with ribozyme activity severely inhibited RNA replication (Mac-naughton et al. 1993). More recently, metabolic labeling experiments have provided firm evidence that HDV RNA replication proceeds by this mechanism (Macnaughton et al. 2002) and that the replication intermediates are likely to be very long (at least 10x genome length). Nevertheless, the replication cycle must be asymmetrical as at least 20 times more genomic than antigenomic HDV RNA is synthesized in infected cells (Chen et al. 1986). The



Fig. 3 HDV RNA double rolling circle replication. Adapted from the model for viroid replication originally put forward by Branch and Robertson (1984)

origins of replication on both genomic and antigenomic RNAs have not been unequivocally determined, although antigenomic-sense RNA synthesis may initiate from one end of the rod-like structure of HDV RNA downstream of a putative RNA promoter element (Beard et al. 1996; Fig. 1).

The replication model above can account for the production of full-length genomic and antigenomic HDV RNA. However, since this model is based on viroid replication, not surprisingly, the synthesis of the antigenomic-sense 0.8-kb HDAg-encoding mRNA does not fit easily into this scheme. The 5' end of this transcript starts at position 1631 just upstream of the initiation codon of the ORF for HDAg (Hsieh et al. 1990; Gudima et al. 2000). The 3' end of the transcript is 76 nucleotides downstream from the termination codon of HDAg and 15 nucleotides downstream of an AAUAAA polyadenylation signal (Hsieh et al. 1990; Fig. 1). The HDAg mRNA transcript is the least abundant of the three HDV RNA species, present in infected cells at approximately 1000 times lower frequency than the genomic species.

As HDV RNA replication proceeds, RNA editing occurs at nucleotide 1015 of the genomic sense RNA (Casey et al. 1992) that ultimately leads to the pro duction of L-HDAg. This event initiates the production of new virus particles by promoting the interaction of the HDV ribonuclearprotein complex and HBsAg. How the genomic RNA species is specifically selected for packaging was a puzzle for a long time, especially as HDV RNA replication is restricted to the nucleus and HBsAg occurs only in the cytoplasm. However, the likely mechanism behind this selection process was revealed recently when it was demonstrated that genomic HDV RNA species is specifically and rapidly exported to the cytoplasm soon after transcription (Macnaughton and Lai 2002a; see later). Nevertheless, many aspects ofvirus packaging and secretion are still unclear. The editing process is nonreversible; thus, the extent of RNA editing in the cells must be regulated, so that L-HDAg is not over-produced. Such a feedback inhibition is likely caused by the enhanced accumulation of deleterious mutations particularly near the putative promoter element that is triggered by the editing event (Macnaughton et al. 2003) and/or L-HDAg by itself (Cheng et al. 2003). This mechanism also explains why HDV RNA genomes encoding L-HDAg are noninfectious (Glen and White 1991; Macnaughton et al. 2003).

0 0

Post a comment