implicated a mechanism by which the ISDR sequence influences response to interferon, and seemed to imply that inhibition of PKR was vital for HCV persistence. However, subsequent studies showed that the ISDR is not essential to the ability of NS5A protein to confer interferon resistance. Using a virus-rescue model system, Polyak and colleagues9 showed that when provided in trans form, the NS5A protein could rescue interferonsensitive viruses from the antiviral effects of interferon in cell culture. Importantly, NS5A-deletion mutants lacking the ISDR domain retained this interferon rescue bioactivity. In another study, the ability of NS5A to confer interferon resistance in cell culture was not related to the phenotype of HCV from which the NS5A protein was derived. 10 Then Podevin and colleagues11 confirmed the findings that NS5A can rescue other viruses from the antiviral effects of interferon, but found no evidence that NS5A interacts with PKR in cells derived from human liver. In a more recent study, Polyak and colleagues12 reported that NS5A induces the expression of interleukin-8, an inflammatory cytokine that inhibits the antiviral action of interferon. These investigators showed that increased intracellular NS5A levels correlated directly with increased interleukin-8 concentrations, and in an in-vitro assay NS5A directly transactivated the promoter for the interleukin-8 gene. Furthermore, interleukin-8 alone mimicked the antiviral effects of NS5A in the trans rescue culture system. Finally, in human beings treated with interferon alone, raised interleukin-8 concentrations in serum were associated with treatment failure. 13 These data offer compelling evidence for a novel mechanism of interferon resistance mediated by the HCV NS5A protein—ie, induction of interleukin-8.

Overall, these studies show that NS5A can interact with PKR under certain conditions, and that this interaction is ISDR dependent. Such findings give a molecular basis for understanding the association between ISDR sequence and clinical response to interferon in vivo. However, in-vitro models show that NS5A can also confer interferon resistance independent of its interaction with PKR, and irrespective of ISDR sequence. This latter mechanism, possibly interleukin-8 dependent, may help explain the high rate of HCV persistence during natural infection, and may also help account for clinical resistance to exogenous interferon seen in many genotype 1 infections (some of which lack a consensus ISDR sequence). It is noteworthy that the cytoplasmic (PKR modulation) and nuclear (interleukin-8 induction) bioactivities of NS5A are not mutually exclusive, yet need not be independent because of possible cross-talk pathways in the cytokine network.