Negative Feedback Regulation

Two recent studies have indicated that editing can be regulated by negative feedback (Cheng et al. 2003; Sato et al. 2004). In HDV genotype I mutants that overproduce L-HDAg, levels of L-HDAg plateau as replication is shut down—

by L-HDAg (Sato et al. 2004). This inhibition of L-HDAg production occurs because of the location of editing in the HDV replication scheme (Fig. 1). Editing occurs on the antigenome-the replication intermediate-and does not result in L-HDAg synthesisuntil theeditedantigenomefirstservesastemplate for the synthesis of genomes, which then serve as templates for transcription of mRNAs encoding L-HDAg. Thus, L-HDAg can, under these circumstances, limit its own production.

In addition to the plateau in L-HDAg production, Sato et al. observed a plateau in editing at the amber/W site, which is not explained by the above model (Sato et al. 2004). They hypothesize that editing occurs only on newly transcribed antigenomic HDV RNA, perhaps before HDAg has a chance to bind and form a 'mature' RNP, and that L-HDAg indirectly inhibits editing by blocking new RNA synthesis. One important consideration of this model is the mechanism whereby L-HDAg prevents further antigenome synthesis. Other reports have indicated that L-HDAg does not inhibit antigenome RNA synthesis (Macnaughton and Lai 2002; Modahl and Lai 2000). Does inhibition of antigenome synthesis occur indirectly via shutdown of genomic RNA synthesis, or does L-HDAg produced as a result of editing inhibit antigenome RNA synthesis in a manner that is not obvious when L-HDAg is produced in trans? Another consideration is that the observed control of editing by L-HDAg was only observed when editing levels were artificially elevated—either by mutation or by ADAR overexpression. As mentioned above in Sect. 4, L-HDAg does not appear to limit replication of HDV genotype I RNA in transfected Huh-7 cells; rather, as yet undetermined factors limit RNA levels. Perhaps, editing in HDV genotype I is calibrated to these limitations, such that appropriate levels of editing are achieved just before replication is restricted; alternatively, L-HDAg may play a more important role limiting replication in infected hepatocytes than in Huh-7 cells.

Cheng et al. recently showed that L-HDAg is a much better inhibitor of editing in HDV genotype III than is S-HDAg, which is a very poor inhibitor (see Sect. 5.2.1). This observation led to the suggestion that genotype III L-HDAg might directly inhibit its own production by directly inhibiting amber/W site editing. However, preliminary data from our laboratory (R. Chen and J.L. Casey, unpublished results) suggest that this inhibitory activity, which is unrelated to replication, might not be the predominant factor involved in a negative feedback loop to limit editing levels (Cheng et al. 2003). Although L-HDAg is a potent inhibitor of genotype III editing, mixtures of genotype III S-HDAg and L-HDAg at ratios similar to those found in cells replicating genotype III RNA exhibit inhibitory activities similar to S-HDAg. Hence, it appears that levels of L-HDAg achieved during replication might not be sufficient to directly affect editing.

In contrast to the behavior of HDV genotype I, wild-type HDV genotype III replication is affected by levels of L-HDAg production. Genotype III mutants that do not produce L-HDAg replicate at significantly higher levels, and accumulate higher levels of editing (Casey 2002; Cheng et al. 2003). Thus, it seems likely that the predominant factor in the control of maximal editing levels in HDV genotype III is the ability of L-HDAg to inhibit replication. As discussed in Sect. 5.2.1, in the current model for genotype III, editing occurs on a metastable structure that is formed co-transcriptionally; cessation of transcription would prevent further editing because the structure required for editing would not be formed. Thus, while it remains to be determined whether L-HDAg functions to control editing of wild-type HDV genotype I RNA, it seems likely that it does in genotype III. A remaining question for HDV genotype III is why do S-HDAg and L-HDAg inhibit editing at such different levels (Cheng et al. 2003)? One possibility is that the more important aspect for the virus is that S-HDAg is not a good inhibitor. Because editing is already modulated by the conformational dynamics of the RNA, further inhibition by S-HDAg could lead to insufficient levels of editing.

0 0

Post a comment