New concepts for genomics
09 March 2007 (Volume 2 Issue 3)
Out of Italy
A senior scientist at both the Discovery Research Institute in Wako and the Genome Sciences Center in Yokohama, Carninci’s interests range from technology development to analyzing genome networks in the brain.
Born in Trieste in northeastern Italy, Carninci gave up a promising career as a soccer player to pursue science instead. In 1989, he obtained his doctorate in biological science at the University of Trieste.
In the same year, he spent six months as a research associate at Trieste’s International Center for Genetic Engineering and Biotechnology, where he developed more efficient way of sequencing DNA. After that came a year of military service, in which he served as a health assistant. It was, Carninci recalls, the most enjoyable time of his career—working in the army during the day and then rushing into his university laboratory to do research until midnight.
With his military service over, Carninci became a researcher at a biotech company called TALENT, spun out of his supervisor’s lab, and developed easy-to-use methods to extract and sequence highquality DNA.
Carninci was also involved Italy’s small genome project, but was suffering from small budgets and insufficient support. So when he met Hayashizaki at a conference in Germany in 1994, they found a mutual interest: Carninci was looking for a place to improve genome analysis techniques; Hayashizaki had just been recruited by RIKEN to launch a genome project and needed technical supports.
An encyclopedia of the mouse
In 1995, Carninci joined the Genome Science Laboratory in Tsukuba just as Hayashizaki was launching the Mouse Encyclopedia Project to collect and sequence full-length mouse cDNAs as a model for human diseases. Carninci’s first mission was to develop technology to clone full-length mouse messenger RNA in the form of cDNAs. Conventional technology often failed to clone the gene from the beginning to end, missing an important ‘cap’ where protein translation is initiated. After eight months during which he often saw the dawn from his lab, Carninci completed his ‘cap trapper’ technology—capturing the cap and picking up full-length genes while eliminating broken ones1.
Carninci also developed a way of stabilizing an enzyme that stimulates the conversion of long messenger RNA into cDNA. A few years later, he improved the technology to enable it to remove copies of highly expressed RNAs and collect only rare ones, helping RIKEN stand out from its competitors.
By 2000, Carninci’s technologies had created a database with 21,000 full-length cDNAs. But it was just the start. “We had a very long list of genes, but two-thirds of them were unknown. And we had piles of printed papers of data, but there was no clue to their biological patterns.”
Massive efforts to interpret genes
Driven by this frustration, the team led by Hayashizaki and Okazaki convened an ambitious project to annotate all the cDNAs. They called the project FANTOM (Functional Annotation of Mammalians) and organized a meeting in the summer of 2000, inviting 50 Japanese and overseas researchers. It was “annotation prison camp”, Carninci jokes, with guests either in a hotel in a rice field in Tsukuba or in RIKEN for the entire fortnight.
The results2 “helped us to understand the first 21,000 cDNAs, yielding the first view of transcriptome functions,” Carninci says. FANTOM-2 in 2002, which looked at a larger transcript of another 39,000 cDNAs, revealed that several RNAs don’t encode proteins but have other functions3.
Carninci continued to improve technologies, and invented methods such as CAGE (Fig.1) and GSC, which were designed to collect the short sequence tags that identify where transcription sites start and end. Those new methods were used in FANTOM-3, which ran from the end of 2002 through 2006. Out of two million mouse cDNAs, researchers analyzed 103,000 in detail. The result? “We completely changed a genetic philosophy,” Carninci says.
For example, it had been believed that only 2% of the genome are important in encoding for human (and mouse) proteins, but the researchers found that about 63% of the genome play a critical role by being actually transcribed into RNAs. They also found the percentage of RNAs that aren’t coding proteins (non-coding RNAs)4 constitutes even more than half of RNAs, and these noncoding RNAs are often transcribed in the opposite orientation of protein coding genes5. They also found the starting point of gene’s activity, called a promoter, was scattered in broader sites than had been predicted and that these sites are poorly conserved between species, suggesting that gene expression became more important than sequence conservation as evolution progressed6.
The series of achievements at FANTOM projects are largely attributed to Hayashizaki’s leadership and ideas, but, says Okazaki, “behind all these was Piero’s ability to develop technologies by looking at both a big picture and details”.
An appetite for discovery…and food
Carninci and other researchers are preparing to start FANTOM-4, partly to deepen analysis of the relationships between promoter elements and transcriptional functions. “Dr Carninci has an incredible amount of energy. He knows what is not written in books,” comments Charles Plessy, a junior colleague of Carninci. And “his great appetite is not only for discovery, but also for food,” says the French scientist, adding he shares with Carninci the culture that a good lunch is important in life.
Over the past few years, Carninci has been drawn into the brain science, so in addition to working for FANTOM projects, he collaborates with neuroscientists at RIKEN and elsewhere to study transcriptional patterns in neurons. Carninci says RIKEN is a good place for non- Japanese young researchers to work, because it is trying to bring Japan’s research system to a new level. “I recommend they consider coming to RIKEN. Here, my dreams have come true,” Carninci says.