GENERATION AND ANALYSIS OF ESTS FROM MOUSE SPERMATOGENIC CELL AND SERTOLI CELL CDNA LIBRARIES

Over 28,000 expressed sequence tags (ESTs) were generated from seven cDNA libraries prepared from mouse spermatogenic cells at different stages of development and from a mouse Sertoli cell cDNA library. The collection of ESTs was used to identify trends in gene expression during spermatogenesis and to identify novel as well as testisspecific transcripts. The cDNA libraries were arrayed at the Lawrence Livermore National Laboratories and utilized for single-pass 5′ endsequencing at the Washington University Genome Center. Replicates of the cDNA clones were provided to I.M.A.G.E. Consortium sites for distribution and 21,278 ESTs that met quality standards were deposited in the NCBI public database. Our analysis of the ESTs was carried out by downloading the sequences in FASTA format and using accession numbers to assemble the annotation data that was available for approximately 19,000 of these ESTs. The annotations were imported into OmniViz and clustered using a K-means algorithm applied to UniGene cluster identification numbers. We found that 7–15% of ESTs from individual libraries were not included in UniGene Clusters. In addition, over 6% of the ESTs in UniGene Clusters were testis-specific, with the majority coming from the round spermatid cDNA library. No testisspecific UniGene Clusters contained ESTs found in libraries from all three germ cell types (spermatogonium, spermatocyte, and round spermatid), but over 4% of the non-testis-specific UniGene Clusters did contain ESTs found in all three. However, the Washington University Genome Center estimated that less than one-half of the genes expressed in each cell type were identified. These ESTs are useful for identifying novel genes for further study and for assessing where and when genes are expressed in the seminiferous epithelium. These studies were supported by the Intramural Research Program, NIH, National Institute of Environmental Health Sciences. (poster)