Ribonuclease mapping and antisense targeting of hepatitis C virus RNA terminal sequences

Date of Completion

January 2004


Biology, Molecular|Health Sciences, Pharmacology




Terminal sequences of the hepatitis C virus (HCV) positive-strand RNA genome and negative-strand replication intermediate include cis -acting elements essential for viral translation and RNA synthesis. In comparison to the positive-strand 5' untranslated region (UTR), the other HCV terminal sequences remain poorly characterized with respect to RNA structure, and relatively unexplored as targets for antiviral therapy. We investigated RNA folding of the HCV genotype 1a negative-strand terminal sequences by enzymatic probing and thermodynamic modeling of secondary structure. Sites accessible to antisense hybridization within the negative-strand termini and positive-strand 3'-UTR were also mapped by RNase H digestion in the presence of combinatorial 2'-O-methyl RNA/DNA libraries. The nuclease sensitivity data are consistent with structural models featuring a three stem-loop domain in the 5'-terminal 113-nt region and a seven stem-loop domain in the 3'-terminal 365-nt region. Although several of these stem-loops correspond to structures resident in the complementary positive-strand RNA sequence, the structural models do not constitute mirror images of the HCV 5'-UTR and X region. The proposed models may facilitate the identification of structural features which contribute to protein binding, strand stability, and template recognition by the viral replication complex. The hybridization accessibility profile of the negative-strand 3' terminus correlates with inhibitory activity of individual antisense constructs, reported in the literature. Direct evaluation of the X region accessibility map led to the identification of 15-mer oligodeoxynucleotides and a 10-23 deoxyribozyme exhibiting antisense activity against HCV 3'-UTR and X RNA transcripts in vitro. However, transfection of the optimized deoxyribozyme and a panel of small interfering RNA (siRNA) constructs targeting the X region sequence of an HCV 1b-based subgenomic replicon in human hepatoma cells failed to inhibit viral replication in a sequence-specific manner. Supplementary experiments with sequence-modified siRNA constructs targeting the nonstructural 5B coding region suggested a strong strand bias for RNA interference directed against the replicon. The results establish the hepatitis C virus X region and negative-strand sequences as unfavorable targets for siRNA-mediated inhibition, in comparison to the 5-UTR and coding region of the positive strand. ^