Molecular basis of nitrogen mustard effects on transcription processes: role of depurination

DNA was alkylated with nitrogen mustard (HN2) and the rate of release of the alkylpurines was quantitated by HPLC. The half life of depurination of the major product (7-alkylguanine) was 9.1 h at 37 degrees C. End-labelled DNA was used to show that depurination occurred dominantly at 5'-GA, 5'-GG and 5'-GT sequences. Although extensive alkylation was observed at all 5'-GNC and 5'GNT sequences, no depurination was observed at these sites during a depurination time of 20 h at 37 degrees C. Since these sites are potential interstrand crosslinking sequences (G-adduct-G and G-adduct-A, both spanning an intervening base pair), this suggests that these regions have a greatly enhanced stability or that simultaneous depurination of both ends of the crosslink is necessary before these lesions are removed (with a predicted half-life of approximately 80 h at 37 degrees C). Depurination at the lac UV5 promoter impaired the association of Escherichia coli RNA polymerase with that promoter, while in the elongation phase two distinctly different sequence-specific processes were apparent. At 5'-GNC and 5'-GNT sequences transcriptional blockages were maintained with increasing elongation time, whereas at monoadduct sites, the blockage decreased with elongation time (predominantly at 5'-GG and 5'-GC sequences), with an average half-life of approximately 10.7 h. Collectively, these results suggest that the observed read-through past monoadduct sites is due to depurination of the DNA at those sites. E. coli RNA polymerase is therefore able to transcribe efficiently past apurinic sites and presumably does so by incorporating an incorrect base into the nascent RNA.