Ribavirin reveals a lethal threshold of allowable mutation frequency for Hantaan virus

Abstract

The broad spectrum of antiviral activity of ribavirin (RBV) lies in its ability to inhibit IMP dehydrogenase, which lowers cellular GTP. However, RBV can act as a potent mutagen for some RNA viruses. Previously we have shown a lack of correlation between antiviral activity and GTP repression for Hantaan virus (HTNV) and evidence for RBV's ability to promote error-prone replication. To further explore the mechanism of RBV, GTP levels, specific infectivity, and/or mutation frequency was measured in the presence of RBV, mycophenolic acid (MPA), selenazofurin, or tiazofurin. While all four drugs resulted in a decrease in the GTP levels and infectious virus, only RBV increased the mutation frequency of viral RNA (vRNA). MPA, however, could enhance RBV's mutagenic effect, which suggests distinct mechanisms of action for each. Therefore, a simple drop in GTP levels does not drive the observed error-prone replication. To further explore RBV's mechanism of action, we made a comprehensive analysis of the mutation frequency over several RBV concentrations. Of importance, we observed that the viral population reached a threshold after which mutation frequency did not correlate with a dose-dependent decrease in the level of vRNA, PFU, or [RTP]/[GTP] (where RTP is ribavirin-5'-triphosphate) over these same concentrations of RBV. Modeling of the relationship of mutation frequency and drug concentration showed an asymptotic relationship at this point. After this threshold, approximately 57% of the viral cDNA population was identical to the wild type. These studies revealed a lethal threshold, after which we did not observe a complete loss of the quasispecies structure of the wild-type genome, although we observed extinction of HTNV.

FIG. 1.

Structures of RBV, SEL, TIA, and MPA.

FIG. 2.

Dose-response curves showing PFU and relative [GTP]/[ATP] following treatment of HTNV-infected cells. Following infection with HTNV, Vero E6 cells were treated with RBV, SEL, or TIA (A) and MPA (B) at several concentrations over a period of 3 days. Vero cells and supernatants were measured for PFU and GTP/ATP content, respectively.

FIG. 3.

Antiviral effects of RBV and MPA on cellular concentrations of HTNV vRNA (vRNAc). HTNV vRNA levels were measured at various combinations of RBV in the absence or presence of MPA.

FIG. 4.

Analyses of mutation spectrum as a function of [RBV]. (A) Mutation frequency was plotted as a function of [RBV] for one representative experiment. [RTP]/[GTP] was assessed and is also presented. Standard deviations are shown for [RTP]/[GTP]. (B) The percentage of cDNAs as a function of the number of mutations relative to that for the wild-type virus is plotted against [RBV].

FIG. 5.

HTNV S vRNA length profile in the absence or presence of various concentrations of RBV. The copy numbers of cDNA primed with the primers described in the text were measured by using a SYBR green quantitative real-time PCR method.