Role of Mitochondrial RNA Polymerase in the Toxicity of Nucleotide Inhibitors of Hepatitis C Virus

Abstract

Toxicity has emerged during the clinical development of many but not all nucleotide inhibitors (NI) of hepatitis C virus (HCV). To better understand the mechanism for adverse events, clinically relevant HCV NI were characterized in biochemical and cellular assays, including assays of decreased viability in multiple cell lines and primary cells, interaction with human DNA and RNA polymerases, and inhibition of mitochondrial protein synthesis and respiration. NI that were incorporated by the mitochondrial RNA polymerase (PolRMT) inhibited mitochondrial protein synthesis and showed a corresponding decrease in mitochondrial oxygen consumption in cells. The nucleoside released by the prodrug balapiravir (R1626), 4'-azido cytidine, was a highly selective inhibitor of mitochondrial RNA transcription. The nucleotide prodrug of 2'-C-methyl guanosine, BMS-986094, showed a primary effect on mitochondrial function at submicromolar concentrations, followed by general cytotoxicity. In contrast, NI containing multiple ribose modifications, including the active forms of mericitabine and sofosbuvir, were poor substrates for PolRMT and did not show mitochondrial toxicity in cells. In general, these studies identified the prostate cell line PC-3 as more than an order of magnitude more sensitive to mitochondrial toxicity than the commonly used HepG2 cells. In conclusion, analogous to the role of mitochondrial DNA polymerase gamma in toxicity caused by some 2'-deoxynucleotide analogs, there is an association between HCV NI that interact with PolRMT and the observation of adverse events. More broadly applied, the sensitive methods for detecting mitochondrial toxicity described here may help in the identification of mitochondrial toxicity prior to clinical testing.

FIG 1

Structures of HCV NI and their clinical progress. The names of some corresponding nucleosides for the prodrugs are provided in parentheses.

FIG 2

Homology model of ATP incorporation into human PolMRT based on a crystal structure of T7 RNA polymerase (PDB ID 1S76). The active site is well conserved, with only N926 (shown) and Q992 (not shown, but positioned above the substrate/template bases) being different. A 1′CN substitution shifts H1125 toward 2′. A 2′CMe substitution shifts Y999 toward 1′. When combined, these two substitutions are then in conflict and the analog is a poor substrate.

FIG 3

Effects of compounds on mitochondrial protein synthesis in PC-3 cells after 5-day treatments. The levels of mitochondrial protein synthesis, represented by COX-1 protein (•), nuclear protein synthesis, represented by SDH-A protein (◆), and ATP (■) were curve fitted with solid lines, dashed lines, and dotted lines, respectively. (A) Puromycin showed nonselective inhibitive effects on the COX-1, SDH-A, and ATP levels, while chloramphenicol and ddC specifically inhibited COX-1 synthesis. (B) 4′-AzidoC and BMS-986094 showed selective inhibition of COX-1 synthesis. (C) RBV at a concentration not resulting in any observed toxicity (10 μM) potentiated the mitochondrial and general cellular toxicity associated with BMS-986094. In contrast, at 10 μM, RBV showed no appreciable effect on mitochondrial or nuclear protein synthesis or ATP level.

FIG 4

Effects of compounds on mitochondrial spare respiratory capacity in PC-3 cells (A and B) and primary rat cardiomyocytes (C) after 3-day treatments. The relative levels of spare respiratory capacity (•), DNA (◆), and ATP (■) were curve fitted with solid lines, dashed lines, and dotted lines, respectively. The spare respiratory capacity was normalized by cell numbers. (A) Puromycin showed nonselective inhibition of the spare respiratory capacity, DNA, and ATP levels, while ddC and chloramphenicol specifically inhibited mitochondrial spare respiratory capacity but showed minimal effects on cellular DNA and ATP levels. (B) 4′-AzidoC and BMS-986094 showed selective inhibition of mitochondrial respiration in PC-3 cells. (C) BMS-986094 showed selective inhibition of mitochondrial respiration in primary rat cardiomyocytes.

FIG 5

Effects of BMS-986094 on RNA transcript levels in PC-3 cells during a 1-day treatment. The relative levels of transcripts of preribosome (PolI), myc (PolII), mitochondrial ATP6 (PolRMT), and total RNA are shown in green (•), blue (▲), red (▼), and black (⭘). Cells were treated with BMS-986094 (100 μM) (left) or the positive control RNA Pol1 inhibitor CX-5461 (5 μM) (right), and transcripts in purified total nuclear RNA were quantified by RT-qPCR.