Never-ending annotation of human chromosome 21: alternative splicing, antisense transcripts and mouse conservation

Katheleen Gardiner, Ph.D.
Eleanor Roosevelt Institute
1899 Gaylord Street
Denver, CO 80206-1210
USA
telephone: 303-336-5652
fax: 303-333-8423
email: gardiner@eri.uchsc.edu
prestype: Platform
presenter: Katheleen Gardiner, Ph.D.

Katheleen Gardiner and Muriel Davisson
Eleanor Roosevelt Institute, Denver Colorado and The Jackson Laboratory, Bar Harbor Maine

The 34 Mb finished sequence of human chromosome 21 was published in May 2000 with 225 genes/models reported. With hand curation, the gene number is now 250, however, much annotation of the genomic sequence remains to be added. We will present information on alternative splicing within coding regions, putative antisense transcripts, and conservation with homologous mouse sequences.

i) Alternative splicing: To determine the number of protein isoforms encoded by chromosome 21, information from dbEST entries was used to detect potential splice variants within coding regions. Of the 200 genes for which sufficient EST data were available, >40% showed two or more splice variants. These included novel splice variants of well studied genes (e.g. APP and HMG14), and prediction of possibly >30 forms for the Intersectin gene.

ii) Antisense transcripts: Information from spliced ESTs was also used to identify putative antisense transcripts to known protein coding genes. Eight examples were found where spliced opposite-strand transcripts contained exons complementary to one or more coding exons of the sense strand gene. Biological functions of the sense genes are diverse; preliminary experimental analysis of the antisense genes suggests that transcription is very low in the tissues tested.

iii) Comparison with mouse genomic sequence: Draft genomic sequences from the public sector and the Celera databases were searched for all segments with homology to human chromosome 21. Comparison of >30 Mb shows that most, but not all, “known” genes and complete cDNAs are present in both species; gene order and orientation is generally, but not always, conserved; a 7 Mb “gene desert” is conserved; intergenic and intronic regions, largely only within non-GC-rich regions, show numerous segments of significant conservation, but unknown function; and a large number of spliced EST sequences appear to be species specific.