A new trick for an old dog: TraY binding to a homopurine-homopyrimidine run attenuates DNA replication.

The effects of the d(GA)(n).d(TC)(n) repeat on plasmid replication in Escherichia coli cells were analyzed using electrophoretic analysis of replication intermediates. This repeat appeared to stall the replication fork progression in E. coli strains carrying F' episomes. The potency of replication stalling increased with the repeat's length but did not depend on its orientation relative to the replication origin, or transcription through the repeat. Treatment of E. coli cells with the protein synthesis inhibitor chloramphenicol abolished replication blockage, indicating that protein binding might be responsible for the repeat-caused replication blockage. Concordantly, dimethylsulfate footprinting in vivo revealed methylation protection of all guanine residues within the d(GA)(n).d(TC)(n). Gel retardation assays with crude cell extracts confirmed the presence of a d(GA)(n).d(TC)(n) -binding activity in F', but not F(-), strains. Further, strains cured from the F' episome lost this activity, while F(-) strains that acquired the F' factor via conjugation, acquired d(GA)(n).d(TC)(n)-binding activity as well. Thus, this d(GA)(n).d(TC)(n)-binding protein is encoded by the F' factor. Purification of this protein by affinity chromatography revealed a single polypeptide with an apparent molecular mass of 15.2 kDa. Microsequencing of its two tryptic peptides revealed two perfect matches with the TraY protein, which is encoded by the F factor. Overexpression of an individual TraY protein in the F(-) E. coli strain conveyed d(GA)(n).d(TC)(n)-binding activity in vitro and replication stalling at d(GA)(n).d(TC)(n) repeats in vivo. We conclude that TraY binding to a homopurine-homopyrimidine repeat is responsible for stalling DNA replication. Biological applications of this phenomenon are discussed.

Transcription through a simple DNA repeat blocks replication elongation.

The influence of d(G)n.d(C)n repeats on plasmid replication in Escherichia coli cells was analyzed using electrophoretic analysis of replication intermediates. These repeats impeded the replication fork in a length- and orientation-dependent manner. Unexpectedly, the replication arrest relied primarily on the repeats' transcription. When the d(C)n sequence served as the transcriptional template, both transcription and replication were blocked. This was true for transcription driven by either bacterial or phage RNA polymerases. We hypothesize that the replication fork halts after it encounters a stalled ternary complex of the RNA polymerase, the DNA template and the r(G)n transcript. This constitutes a novel mechanism for the regulation of replication elongation. The effects of this mechanism on repeat length polymorphism and genome rearrangements are discussed.

Kinetic trapping of H-DNA by oligonucleotide binding.

Homopurine-homopyrimidine mirror repeats are known to adopt the H form under acidic pH and/or negative supercoiling. In H-DNA, one half of the purine strand enters the triplex whereas the second half is unstructured and can form duplex with complementary oligonucleotide. However, because the same oligonucleotide can form triplex with the homopurine-homopyrimidine insert, one could expect that oligonucleotide would make H-DNA thermodynamically less favorable, as was claimed by Lyamichev et al. Nucl. Acids Res. 16, 2165-2178 (1988). Now we show that complex between oligonucleotide and H-DNA, formed under conditions favorable for the H-form extrusion, is kinetically trapped in superhelical DNA and remains stable up much higher pH values than H-DNA alone. Experiments on chemical probing show that such complex exists for a plasmid with native superhelical density at pH7. We have also used this approach to demonstrate a pH-dependent structural transition in yeast telomeric sequence, d(CACACCCA)16.