9th Annual Workshop, October 28-31, 1999
Co-sponsored by the U.S. Department of Energy
Approaches towards global profiling of cDNA mutations
Pan, X. and Weissman, S.
Yale University School of Medicine, Department of Genetics, New Haven, Connecticut, USA
The ability to detect a large fraction of sequence variations between two complex genomes or pools of genomic DNA from two sources would have a wide range of applications. The approach we are taking towards this is to initially focus on comparing variations in cDNA sequence between two sources. Several types of chemical or enzymatic approaches have been used to detect sequence variation. Comparison of cDNA sequences, in principle, could be particularly valuable for identifying mutations associated with specific malignancies, or for more rapidly identifying mutations in simple Mendelian disorders including severe sporadic dominant lesions that could not be mapped by genetic means. We find that certain DNA-glycosylases are quite promising for our purposes. The glycosylases recognize specific types of lesions and have the great advantage that they bind relatively tightly to the abasic DNA that they produce enzymatically. Also, in at least some cases, borohydride reduction can be used to generate covalent links between the enzyme and its product. These features are useful to specifically separate mismatch and perfect match DNA duplexes from the same reaction. Glycosylases exist that would, in combination, detect a large fraction of single base sequence variation; although in certain cases it may be necessary or preferable to use specific nucleases.
Gel display of mismatch fragments in principle is an attractive approach because prior knowledge of the existence of sequence variation is not needed. Various array hybridization procedures would also require considerable development for this purpose because mutations might occur anywhere within a cDNA and also it would be desirable to the extent that it is possible to distinguish new mutations from common polymorphisms in the same mRNA. To achieve the above goal, several technical issues must be dealt with. For accurate comparison, it would be desirable to be able to form heteroduplexes in a single denaturation-renaturation step, then use protocols that permit selective examination of either homoduplexes re-formed from one of the two input cDNAs, or only heteroduplexes in which one strand is derived from each of the two input pools. For display purposes it would be necessary to divide cDNA fragments into non-overlapping, exhaustive subsets, such that each subset would contain a limited number of fragments in the display range. Display conditions should require only stringent PCR steps, and the conditions for PCR should not strongly bias for sequences of a specific range of GC content. Annealing conditions and fragment complexity should be optimized for efficient but stringent annealing of low abundance sequences. We will describe our experimental designs and progress along the above lines.