A strategy to identify differentially expressed genes using representational difference analysis and cDNA arrays

Representational difference analysis (RDA) combined with cDNA arrays is an effective approach to identify differentially expressed genes. To identify differentially expressed genes in c-Myc transgenic mouse liver, we compared the virtues of probing commercially available cDNA arrays with either radi olabeled cDNA pools or radiolabeled difference products (DP2) derived from RDA using c-Myc transgenic and normal mouse liver. Probing commercial and c ustom arrays with DP2 products led to the identification of transcripts of low abundance that were missed when the arrays were initially probed with P CR-amplified cDNA pools, Although DP2 probes also detected abundant transcr ipts that are highly differentially expressed, they failed to identify abun dant transcripts with low differential expression that were detected with c DNA pools. The combined use of radiolabeled cDNA and DP2 products to probe arrays allows a more comprehensive identification of differentially express ed transcripts that are abundant or rare. Our method has the additional ben efit of eliminating false-positive transcripts that lack true differential expression and frequently contaminate DP2 pools. Using this method we ident ified 16 differentially expressed genes in c-Myc transgenic liver, one of w hich is novel. (C) 2001 Academic Press.