Evidence for two preferred hairpin folding patterns in d(CGG).d(CCG) repeat tracts in vivo.

Unusual DNA secondary structures have been implicated in the expansion of trinucleotide repeat tracts that has been found to be responsible for a growing number of human inherited disorders and folate-sensitive fragile chromosome sites. By inserting trinucleotide repeat sequences into a palindromic clamp in lambda phage we are able to investigate their tendencies to form hairpins in vivo in any particular alignment and with odd or even numbers of repeat units in the hairpin. We previously showed that with d(CAG).d(CTG) repeat tracts there was a markedly greater tendency to form hairpins with even numbers of repeat units than with odd numbers, whereas d(GAC).d(GTC) repeats showed no such alternation despite having the same base composition. We expected that d(CGG).d(CCG) repeats, might show the same pattern as d(CAG).(CTG) repeats since they are also involved in trinucleotide repeat expansion disorders. The pattern was not so clear and we wondered whether this might be because d(CGG).d(CCG) repeats have more than one possible alignment in which they could self-anneal. We now present results for all three alignments, which suggest that while even-membered hairpins are preferred in the frame d(CGG).d(CCG), hairpins with odd numbers of trinucleotides are more stable in the frame d(GGC).d(GCC). In both cases the base-pair predicted to close the terminal loop of unpaired bases is 5'C.3'G which has previously been found to be a favoured loop-closing pair.