Transcription arrest at an abasic site in the transcribed strand of template DNA

A dedicated excision repair pathway, termed transcription-coupled repair (TCR), targets the removal of DNA lesions from transcribed strands of expressed genes. Transcription arrest at the site of the lesion has been proposed as the first step for initiation of TCR. In support of this model, a strong correlation between arrest of transcription by a lesion in vitro and TCR of that lesion in vivo has been found in most cases analyzed. TCR has been reported for oxidative DNA damage; however, very little is known about how frequently occurring and spontaneous DNA damage, such as depurination and base deamination, affects progression of the transcription complex. We have previously determined that the oxidative lesion, thymine glycol, is a significant block to transcription by T7 RNA polymerase (T7 RNAP) but has no detectable effect on transcription by RNA polymerase II (RNAP II) in a reconstituted system with all of the required factors. Another oxidative lesion, 8-oxoguanine, only slightly blocked T7 RNAP and caused RNAP II to briefly pause at the lesion before bypassing it. Because an abasic site is an intermediate in the repair of oxidative damage, it was of interest to learn whether it arrested transcription. Using in vitro transcription assays and substrates containing a specifically positioned lesion, we found that an abasic site in the transcribed strand is a 60% block to transcription by T7 RNAP but nearly a complete block to transcription by mammalian RNAP II. An abasic site in the nontranscribed strand did not block either polymerase. Our results clearly indicate that an abasic site is a much stronger block to transcription than either a thymine glycol or an 8-oxoguanine. Because the predominant model for TCR postulates that only lesions that block RNAP will be subject to TCR, our findings suggest that the abasic site may be sufficient to initiate TCR in vivo.