Generation of Longer cDNA Fragments from SAGE Tags for Gene Identification

Serial analysis of gene expression (SAGE) is a powerful technique for genomewide analysis of gene expression (1 –14 ). However, almost two-thirds of SAGE tags cannot be used directly for gene identification for two reasons. First, many of SAGE tags match to multiple known expressed sequences because of the short length of SAGE tag sequences (12 –14 ). Second, many SAGE tags do not match any known expressed sequences because the sequences corresponding to these SAGE tags have not been identified yet (2 –14 ). These problems substantially diminish the power of the SAGE technique. The GLGI (Generation of Longer complementary DNA [cDNA] fragments from SAGE tags for Gene Identification) technique was designed to solve these two problems (15 ). The basic principle of GLGI is to use the SAGE tag as the sense primer, and anchored oligo-(dT) primers as the antisense primer to amplify the original 3′ cDNA from which the SAGE tag was derived. The size of 3′ cDNA will be hundreds of bases longer, which is long enough for solving the two problems. In a typical SAGE project, hundreds or thousands of SAGE tags need to be further analyzed. To facilitate such large-scale performance, we developed the GLGI method into a high-throughput procedure (16 ). In this high-throughput GLGI procedure, 3′ cDNAs starting from the last CATG are used as the templates for GLGI amplification, a SAGE tag sequence is used as the sense primer, and a universal sequence located at the 3′ end of all the cDNA templates generated from anchored oligo(dT) primers is used as antisense primer to amplify the original 3′ cDNA template from which the SAGE tag was derived.