Preferential DNA repair in active genes

We have demonstrated that essential genes in rodent cells and in normal human cells are preferentially repaired after UV damage. In rodent cells, some genes are repaired much more efficiently than the bulk of the genome. This may explain the long standing paradox that their overall genome repair is low, but their UV survival is as high as for proficiently repairing human cells. Rodent cells appear to repair efficiently only genomic regions of vital importance. In normal human cells, we have found that genes are repaired faster than the bulk of the genome, but eventually (after 24 hrs) all genomic regions are proficiently repaired. The demonstration of preferential DNA repair mandates caution in interpreting correlations between overall DNA repair capacity and other biological parameters. Changes in preferential DNA repair could have profound effects on such parameters without noticeably altering overall genome repair levels since the vital regions only constitute less than 1% of the genome. We have correlated overall genome repair, repair in the DHFR gene, and UV resistance for three different cell lines: CHO, XPC and normal human. The results further suggest that determinations of DNA repair in specific genomic sequences may be more important than overall DNA repair measurements for correlations to other biological end points such as resistance to UV damage. Although DNA repair heterogeneity has been demonstrated in XPC, we have found that the preferentially repaired regions in these cells do not include the essential DHFR gene. DNA repair may normally be regulated over the genome in a similar manner to that for transcription, and we propose that this regulation is deficient in the human DNA repair deficient syndrome XPC. We have also analyzed the genomic fine structure of DNA repair in and around the DHFR gene in CHO cells. We find a region of preferential DNA repair of approximately 60-80 kb in length with maximal DNA repair efficiency at the 5' end of the gene and in its 5' flanking sequences. This size corresponds very well with proposed and measured lengths for loops or domains of higher order structure in chromatin, and suggests that DNA repair efficiency in genomic regions might reflect aspects of local chromatin structure, and thus provide us with a probe for the detection of chromatin structural changes. We have found considerable differences in the repair efficiency of different genes within the same cell.(ABSTRACT TRUNCATED AT 400 WORDS)