Phenotypic analysis of mutations at residue 146 provides insights into the relationship between NS5A hyperphosphorylation and hepatitis C virus genome replication

Abstract

The hepatitis C virus genotype 2a isolate, JFH-1, exhibits much more efficient genome replication than other isolates. Although basic replication mechanisms must be conserved, this raises the question of whether the regulation of replication might exhibit isolate- and/or genotype-specific characteristics. Exemplifying this, the phenotype of NS5A hyperphosphorylation is genotype-dependent; in genotype 1b a loss of hyperphosphorylation correlates with an enhancement of replication. In contrast, the replication of JFH-1 is not regulated by hyperphosphorylation. We previously identified a novel phosphorylation site in JFH-1 NS5A: S146. A phosphomimetic substitution (S146D) had no effect on replication but correlated with a loss of hyperphosphorylation. In genotype 1b, residue 146 is alanine and we therefore investigated whether the substitution of A146 with a phosphorylatable (S), or phosphomimetic, residue would recapitulate the JFH-1 phenotype, decoupling hyperphosphorylation from replication. This was not the case, as A146D exhibited both a loss of hyperphosphorylation and a reduction in replication, accompanied by a perinuclear restriction of replication complexes, reductions in lipid droplet and PI4P lipid accumulation, and a disruption of NS5A dimerization. In contrast, the S232I culture-adaptive mutation in the low-complexity sequence I (LCSI) also exhibited a loss of hyperphosphorylation, but was associated with an increase in replication. Taken together, these data imply that hyperphosphorylation does not directly regulate replication. In contrast, the loss of hyperphosphorylation is a consequence of perturbing genome replication and NS5A function. Furthermore, we show that mutations in either domain I or LCSI of NS5A can disrupt hyperphosphorylation, demonstrating that multiple parameters influence the phosphorylation status of NS5A.