Charles R. Cantor, Natasha Broude, Ronald Yaar, and Cassandra L. Smith
Center for Advanced Biotechnology, Departments of Biomedical Engineering, Pharmacology, and Biology, Boston University, Boston MA 02215
Triplet repeats like (GGC)n are an important class of human genetic markers, and they are also responsible for a number of inherited diseases involving the central nervous system. For both of these reasons it would be very useful to have a way to monitor the status of large numbers of triplet repeats simultaneously. We are developing methods to isolate and profile classes of such repeats.
In one method, genomic DNA is cut with one or more restriction nucleases, and splints are ligated onto the ends of the fragments. Then fragments containing a specific class of repeats are isolated by capture on magnetic microbeads containing an immobilized simple repeating sequence. The desired material is then released, and, if necessary, a selective PCR is done to reduce the complexity of the sample. Otherwise the entire captured sample is amplified by PCR. The spectrum of repeats is then examined by electrophoresis on an automated fluorescent gel reader. In our case the Pharmacia ALF is used because of its excellent quantitative signal accuracy. A very complex spectrum of bands is seen representing hundreds of DNA fragments. We have shown that this spectrum is dramatically different with DNAs from unrelated individuals, and the spectrum is markedly dependent on the choice of restriction enzyme, as expected. Repeated measurements on the same sample are highly reproducible. The ability of the method to detect a specific altered repeat length in a complex DNA sample has been validated by examining several individuals with normal or expanded repeat sequences in the Huntington's disease gene. One very powerful application of this method may be the analysis of potential DNA differences in monozygotic twins discordant for a genetic disease. This method can be used to capture genome subsets containing any interspersed repeat. It will also detect insertions and deletions nearby such repeats. Methylation differences between sensitive methylation samples are also detectable when restriction fragments are used.
Conventional analysis of triplet repeats is very laborious since individual repeats must be analyzed by electrophoresis on DNA sequencing gels. The decrease in effort for such analyses will scale linearly as the number of repeats that can be analyzed simultaneously, so we are potentially looking at something like a factor of 100 improvement if the above scheme under development can be effectively realized.
As an alternative approach, we are developing chip-based methods that can detect the length of a tandemly-repeating sequence without any need for gel electrophoresis. Here the goal is to build an array of all possible repeat sequence lengths flanked by single-copy DNA. When an actual sample is hybridized to such an array, the specific alleles in the sample will produce perfect duplexes at their corresponding points in the array and at mismatched duplexes elsewhere. Thus, the task of scoring the repeat lengths is reduced to the task of distinguishing perfect and imperfect duplexes. Currently we are exploring a number of different enzymatic protocols that offer the promise of making such distinctions reliably.