This study consists of a novel strategy for the identification of potential cancer exclusively expressed genes, which might lead to the development of valuable diagnostic molecular markers. Normal- and cancer-paired tissues from the same patient were collected and subjected to the construction of primary complementary deoxyribonucleic acid (cDNA) libraries. The cancer cDNA library was generated by subtracting normal cDNAs from the primary cancer library. The remaining clones in the subtracted cancer library were sequenced and Basic Local Alignment Search Tool (BLAST)-analyzed against current nonredundant and est_human databases in the GenBank. The clones that had no matches with any known gene sequences except the human genomic sequences were identified as ideal candidates for potential molecular marker development. The candidate marker sequence and associated primer sequences were identified on the target clone sequence of the region with the least, or no, matched homologs. The specificity of the markers was measured by a polymerase chain reaction test experiment with the DNA templates from different human normal tissues, genomic DNA, and additional patients' tissues. Results from bioinformatical and experimental approaches used suggest that the methodology has a high potential to identify cancer exclusive transcripts. Thus, it might result in the development of diagnostic markers.

No abstract

Synthesis of full-length cDNA libraries is an essential step for the study of gene function. The method for selecting the intact mRNA directly affects the number of full-length transcripts. We have developed a novel method for intact mRNA selection based on the elimination of uncapped mRNAs. A negative-selection strategy that removes both uncapped mRNA and other non-mRNA molecules that present a phosphate at the 5'-end has been applied in the mRNA purification procedures. Briefly, after performing a standard mRNA purification, a biotinylated oligoribonucleotide is ligated to the 5-end phosphate of uncapped mRNAs. Streptavidin extraction is then performed to remove truncated and non-mRNAs from the intact mRNAs. By comparing random sequencing results of mouse brain full-length and standard cDNA libraries, there was a significant increase of full-length clones with the modified procedure. The results showed that the full-length library contained more than 68% full-length clones with the 5'-end positions ranging between -485 to +100 compared to the standard library with 33% of full-length clones and 5'-end positions ranging between -233 to +100. The data were analyzed using the t-test with the significance level set at p<0.05. The statistical results showed that there were significant differences between two libraries in both 5'-end position and mRNA size (p<0.05).