Tokyo

Japanese research institutes are not known for their openness towards foreign collaborators. But a growing movement to change this was signalled by a two-week meeting in Tsukuba earlier this month, held to annotate a large set of mouse complementary DNA held by the Institute of Physical and Chemical Research (RIKEN).

Unusually for a Japanese research institute, RIKEN's Genomic Sciences Center has initiated and taken the lead in a collaboration based on the free public distribution of data and research materials.

Embracing change: Hayashizaki, left, and his RIKEN colleague, Yasushi Okazaki.

Yoshihide Hayashizaki, a chief scientist at the centre and leader of the project, invited about 50 bioinformaticists and biologists to the Functional Annotation of Mouse (FANTOM) meeting to annotate a subset of his library of 128,500 mouse complementary DNAs, which he says is the world's largest set of mammalian cDNAs.

Many foreign participants say they were struck not only by the event's scale, but also by its taking place in Japan at all. “Japan has always been so secretive,” says one US researcher. Another attributed the inter-nationalism to Hayashizaki's efforts.

Hayashizaki says he faced a catch-22 in getting government approval for the project. “We are told to get involved in international collaborations and to make a name for Japanese science on the international scene, yet we are expected to keep results paid for by Japanese taxpayers, and criticized if we just give them away for other countries to profit from.”

Unlike an entire genome, in which genes are interspersed with ‘junk’ DNA, cDNA is made from RNA produced in the cell, so each sequence represents an expressed gene.

Annotation of cDNA sequences aims to predict the structure, function, homology and protein–protein interactions of their gene products.

Huddled around computers at some times and karaoke machines at others, the group annotated about 21,000 full-length cDNAs. After removing redundant or previously known sequences, this left just under 16,000 unique cDNAs.

Researchers say they have already discovered a number of new genes, or genes that were previously unknown in mice or mammals. The similarity between the mouse and human genomes means that the mouse will almost certainly provide important clues to understanding the human genome.

The data will no doubt find profitable application. Full-length cDNA are “notoriously hard to get,” says participant Peter Mombaerts, a neuroscientist at the Rockefeller University in New York, as the RNA can easily degrade during collection or form secondary structures that make it difficult to copy into cDNA.

After the results are published, Hayashizaki plans to release the data to two public databases, the DNA DataBase of Japan and the Mouse Genomic Database at the Jackson Laboratory in Bar Harbor, Maine. The cDNA clones will be made available to researchers worldwide.

Hayashizaki believes that his clones will become the standard for future research, and he is already planning FANTOM II. Many emerged from this month's meeting convinced that he has the best mouse cDNA around, and that the meeting marked a significant step for Japanese science in keeping pace with collaborative trends elsewhere.