International Mammalian Genome Society

The 15th International Mouse Genome Conference (2001)

Table of Contents * Complex Genetics * Developmental Genetics * Gene Annotation * Gene Discovery * Genome Sequencing * Functional Genomics * Mutagenesis * Presentations * Verne Chapman Memorial Lecture

Dr Piero Carninci
RIKEN-Genome Science
RIKEN Tsukuba Institute
3-1-1 Koyadai
Tsukuba, Ibaraki
305-0074
Japan

Co-Authors:  1&2)Shibata K, 1&2)Itoh M, 1)Arawaka T, 1)Ishii Y, 1)Konno H, 1)Sato K, 1&2)Hayatsu N, 1&2)Shiraki T, 1&2)Hirozane T, 1&2)Aizawa K, 1)Bono H, 1)Kondo S, 1&2)Kawai J, 3)Yoshiki A, 1&2)Okazaki Y, 1&2)Hayashizaki Y
Institutions:   1)Laboratory for Genome Exploration Research Group, RIKEN Genomic Sciences Center, 2)Genome Science Laboratory, 3)Experimental Animal Facility, RIKEN Tsukuba Institute

Strategies to select, clone and sequence of high quality full-length cdna from rarely transcribed mrnas

We have been preparing full-length enriched, subtracted/normalized cDNA libraries containing long inserts by cap-trapping from >250 different tissues/stages by using the threhalose-thermoactivated RT and the lambda-FLC vectors to clone bulk excise long cDNA in functional vectors. We have produced >1,200,000 3' end ESTs, 115,000 5' ESTs and fully sequenced and functionally annotated >21,000 full-length cDNAs (FANTOM, the largest existing set).

Clustering these sequences gives >140,000 clusters covering most genes, candidates to be different transcriptional units. Remaining clusters are being fully sequenced (http://genome.gsc.riken.go.jp).

Discrepancies with current gene number prediction (35-40,000 only), derives from: (1) false clusters due to sequencing errors, (2) clustering strategies, (3) polymorphisms of transcription initiation/termination sites, but also (4) non-protein-coding RNAs and (5) genes not found by genome annotation. Our data should change the estimation of number/concept of gene.

To select remaining new, rare/very long cDNAs, we are using (1) lambda cDNA libraries subtraction, (2) size selection (~7Kb), (3) stabilization of long cDNAs, (4) improved plasmid preparation and (5) selection of special interest cDNA (preimplantation embryos, cancer, tissues of immunological and neurobiological interest). We also use cytoplasmic mRNA: deriving cDNAs -no unspliced introns- always produce functional proteins.