of Work: When integrated in to the human genome, the HIV-1 genome is transcribed by the RNA polymerase II to produce both messenger and viral genomic RNAs. Our goal is to examine whether transcription by RNA polymerase II may contribute to the high variation observed in certain portions of the HIV genome. To adress this question we have developed a strategy to measure the fidelity of transcription by RNA polymerases including purified T7 RNA polymerase (RNAP)and RNA pol II in human cell extracts. The results with T7 RNAP demonstrate that the applied strategy allowes a quantitative evaluation of transcription fidelity, providing the first measurements of error rates during in vitro transcription in the presence of all four nucleotides on a natural template. The results indicate that the fidelity of transcription is sensitive to a nucleotide pool imbalance and provide direct evidence that local sequence contexts have a pronounced effect on error production during transcription. We have also applied this strategy to study the fidelity of damage bypass by RNA polymerases. We have examined transcription by T7 RNAP past a specific adenine residue adducted with (+)anti-trans- or (-) anti-trans- benzo[a]pyrene diol epoxide (BPDE). BPDE, a metabolite of a common environmental pollutant, benzo[a]pyrene, is an established mutagen and carcinogen. To measure the fidelity of bypass synthesis, an ochre codon reversion assay was developed. Opposite the T of the ochre codon, the transcribed strand contained the N6-BPDE adducted- or a non-adducted adenine. T7 RNAP bypassed (-)anti-trans-BPDE 32% of the time and (+)anti- trans-BPDE 18% of the time. The bypass synthesis was highly mutagenic, with reversion frequency for bypass of (-)anti-trans-BPDE elevated 12,000-fold and that for (+)anti-trans-BPDE bypass 6,000-fold relative to the reversion frequency of transcription on the unadducted template. Adenine was misinserted preferentially opposite the adduct, followed by misinsertion of guanine.
