Sensitivity of mitochondrial transcription and resistance of RNA polymerase II dependent nuclear transcription to antiviral ribonucleosides

Abstract

Ribonucleoside analogues have potential utility as anti-viral, -parasitic, -bacterial and -cancer agents. However, their clinical applications have been limited by off target effects. Development of antiviral ribonucleosides for treatment of hepatitis C virus (HCV) infection has been hampered by appearance of toxicity during clinical trials that evaded detection during preclinical studies. It is well established that the human mitochondrial DNA polymerase is an off target for deoxyribonucleoside reverse transcriptase inhibitors. Here we test the hypothesis that triphosphorylated metabolites of therapeutic ribonucleoside analogues are substrates for cellular RNA polymerases. We have used ribonucleoside analogues with activity against HCV as model compounds for therapeutic ribonucleosides. We have included ribonucleoside analogues containing 2'-C-methyl, 4'-methyl and 4'-azido substituents that are non-obligate chain terminators of the HCV RNA polymerase. We show that all of the anti-HCV ribonucleoside analogues are substrates for human mitochondrial RNA polymerase (POLRMT) and eukaryotic core RNA polymerase II (Pol II) in vitro. Unexpectedly, analogues containing 2'-C-methyl, 4'-methyl and 4'-azido substituents were inhibitors of POLRMT and Pol II. Importantly, the proofreading activity of TFIIS was capable of excising these analogues from Pol II transcripts. Evaluation of transcription in cells confirmed sensitivity of POLRMT to antiviral ribonucleosides, while Pol II remained predominantly refractory. We introduce a parameter termed the mitovir (mitochondrial dysfunction caused by antiviral ribonucleoside) score that can be readily obtained during preclinical studies that quantifies the mitochondrial toxicity potential of compounds. We suggest the possibility that patients exhibiting adverse effects during clinical trials may be more susceptible to damage by nucleoside analogs because of defects in mitochondrial or nuclear transcription. The paradigm reported here should facilitate development of ribonucleosides with a lower potential for toxicity.

Figure 1. Antiviral ribonucleoside triphosphates are substrates for POLRMT.

(A) Nucleoside analogs used in this study. (B) RNA-primed DNA templates. The first templating base is underlined. (C) Nucleoside analog incorporation catalyzed by POLRMT. Reaction products from POLRMT-catalyzed nucleotide incorporation using the indicated nucleoside analog triphosphate and RNA primer/DNA template nucleic acid scaffolds shown in panel b. Reactions proceeded for 30 s. RNA primer was extended to n+1 in the presence of each nucleoside triphosphate. (D) Percentage of RNA product relative to correct nucleotide (ATP, CTP, GTP or UTP) is shown. Error bars represent s.e.m. n = 3. Both 2′-deoxy-2′-fluoro-2′-C-methyl-UTP and ribavirin triphosphate were the least incorporated nucleoside analogs. (E) Inhibition of mitochondrial transcription by ethidium bromide (EtBr) and the recovery of mitochondrial transcription. Northern blots of mitochondrial transcripts ND1, ND5 and nuclear transcript GAPDH 24 h post EtBr treatment and 1, 2 and 3 day(s) post wash. Mitochondrial transcripts ND1 and ND5 were inhibited by EtBr treatment and recovered post wash; there was no effect on GAPDH. (F) Effect of 6-methlypurine on POLRMT-catalyzed transcription in vivo. Northern blots of ND1, ND5 and GAPDH after EtBr treatment and recovery in the presence of 6-methylpurine.

Figure 2. Non-obligate chain terminators inhibit RNA elongation by POLRMT.

(A) Non-obligate chain termination of RNA synthesis in vitro. Products from POLRMT-catalyzed nucleotide incorporation in the presence of the next correct nucleotide substrate, UTP or ATP. Reactions proceeded for 10 min. Reactions containing ATP, 7-deaza-ATP, 3-deaza-ATP, 6-methylpurine-TP, ribavirin-TP, CTP,2′-deoxy-2′-fluoro-CTP and GTP were readily extended to n+2 product by POLRMT. Reactions containing 2′-C-methyl-ATP, 2′-C-methyl-CTP, 4′-methyl-CTP, 4′-azido-CTP and 2′-C-methyl-GTP were unable to be extended to n+2, demonstrating the ability of these nucleoside analogs to be non-obligate chain terminators for POLRMT once incorporated in nascent RNA. 3′-dATP and 3′-dCTP were used as positive controls. (B–D) Production of full-length mitochondrial RNA transcripts in cells is impaired in the presence of 2′-C-methyladenosine and 4′-azidocytidine. (B) Experimental design. Huh-7 cells were treated with EtBr for 24 h to deplete mitochondrial transcripts from cells, washed, treated with 2′-C-methyladenosine or 4′-azidocytidine for 1, 2 and 3 days, total RNA isolated and Northern blots performed. Northern blots of ND1, ND5 and GAPDH after EtBr treatment and recovery in the presence of (C) 2′-C-methyladenosine and (D) 4′-azidocytidine. Cells treated with a minimum of 50 µM 2′-C-methyladenosine showed specific inhibition of mitochondrial transcription and the inability to produce both ND1 and ND5 transcripts, whereas a minimum of 50 µM 4′-azidocytidine only inhibited production of ND5; GAPDH was unaffected by treatment with 50 µM 2′-C-methyladenosine or 4′-azidocytidine. At higher concentrations of 4′-azidocytidine GAPDH showed some sensitivity.

Figure 3. TFIIS prevents accumulation of antiviral nucleotides in Pol II transcripts.

(A) Schematic of synthetic nucleic scaffolds for transcription elongation complex (TEC) assembly with calf thymus Pol II. The first templating base is underlined. The TEC with 11-nt RNA (TEC-A11) was assembled using TDS50, NDS50 and RNA9 (see Table S2 for the complete oligonucleotide sequences) in the presence of 10 µM each GTP and ATP (TEC-A11) and purified from the unincorporated DNA, RNA and NTPs; TEC-C12 was obtained by addition of 10 µM CTP to TEC-A11. (B) Reaction products from Pol II-catalyzed nucleotide incorporation in the absence and presence of TFIIS. The concentration of unmodified substrate NTP and analogs were 500 µM; TFIIS was added at 10 µM. Reactions proceeded for 2 min. Reactions with ribavirin-TP proceeded for 10 min. (C,D) Reaction products from Pol II-catalyzed nucleotide incorporation in the presence of the next correct nucleotide substrate. The concentration of the unmodified substrate NTP and analogs were 500 µM; TFIIS was added at 10 µM. Reactions proceeded for 2 min. Reactions with ribavirin-TP, 4′-methyl-CTP and 4′-azido-CTP proceeded for 10 min. (E) Percent inhibition by TFIIS on Pol II nucleoside analog incorporation.

Figure 4. Predicting adverse effects of antiviral ribonucleosides during preclinical development: The mitovir score.

Correlations between (A) cytotoxicity in Huh-7 cells and MT4 cells ( Table 1 , CC₅₀), (B) cytotoxicity in MT4 cells ( Table 1 , CC₅₀) and the efficiency of nucleotide incorporation ( Table 2 , k _pol/ K _d,app), (C) cytotoxicity in MT4 cells ( Table 1 , CC₅₀) and mitovir score for MT4 cells ( Table 2 ), (D) cytotoxicity in MT4 cells ( Table 1 , CC₅₀) and the mitovir score for each analogue corrected to account for the presence of the nucleotide with which the analogue competes, ATP or CTP, and (E) cytotoxicity in Huh-7 cells ( Table 1 , CC₅₀) and mitovir score for Huh-7 cells ( Table 2 ). Error bars represent s.d. Nonparametric (Spearman) correlations with r values shown. In parentheses are one-tailed P-values calculated from Spearman coefficients to provide a measure of statistical significance of correlation.