AT A GLANCE Received his PhD in molecular biology from Aix-Marseille University in 1984. Postdoctoral research at Harvard School of Public Health working on transcriptional regulation.
Served as associate director of R&D at Xenometrix, and scientific attaché at the French embassy in Tokyo.
Joined the new functional genomics laboratory at France’s Atomic Energy Commission (CEA) in 2000.
Interests: Scientific junctions such as the interface between math and biology in modeling genetic networks, and the interface between living cells and glass in biochips.
QWhat role do microarrays play in research at CEA?
AWe are helping researchers study responses to oxidants and heavy metals in yeast, but also more basic aspects such as the transcription machinery.
Our laboratory also has its own research activities. We study the human transcriptome’s dynamic in genotoxicity and carcinogenicity, as well as in stem cell differentiation. Also, with other departments, we work on new biochip concepts.
QHow is the microarray facility set up within the Institute?
AArrays are available at cost for both CEA and Ile de France (Paris area) scientists. We manufacture human, mouse, and yeast arrays, and deliver them to the investigators. They have to isolate and label the RNA, and perform hybridizations, but we can assist them during the process. We then scan arrays in our facility, and provide a list of genes with corresponding fold changes. We face an increasing demand for these services.
QWhat types of microarrays do you use and why?
AWe manufacture our own arrays on poly-L-Lysine and silane slides. Today we routinely use PCR products, but we are testing oligo deposition as well. The glass format is flexible, cheap, and more reliable. But most importantly, by manufacturing our own chips, our team has learned how to design new types of biochips, such as cell arrays, on glass.
QWhat unique types of arrays do you have at CEA?
AWe manufactured a yeast array with 6,000 open reading frames, and a mouse array with 4,000 genes. More recently we started manufacturing a 16K human gene array using a re-sequenced, non-redundant cDNA library obtained from the National Sequencing Center, another member of the Genopole.
QWhat kind of arrayers do you use to make your own arrays?
AWe have two GMS417s from Genetic MicroSystems (now part of Affymetrix) and two MicroGrid II’s from BioRobotics. We also have a GMS418 scanner, an Axon 4000B, a Packard 5000, and an Agilent scanner.
QWhat methods do you use to analyze microarray data?
AWe use our own software as well as GenePix. Each type of software has its own fans among biologists in the lab. For expression profile analysis, our bioinformatics group is developing a database and clustering algorithms. We also intend to get GeneSpring software soon.
QWhat is the biggest challenge you face in working with microarrays?
ALike everyone else in the field, we would like this technology to become as reliable as possible. Molecular biologists seem to be asking for more reliability from microarrays than they ever did from other technologies. It’s a good evolution, as it shows that molecular biology is maturing by being more quantitative and reliable. That’s going to be useful if one wants to map out genetic networks from it, and perform in silico biology one day.
QHow do you tackle this challenge of reliability?
AWe try to optimize each step by paying attention to every detail. You know the wings of butterfly theory, that the flapping of a butterfly’s wings on one side of the planet can turn into a storm on the other side.
QWhat bits of advice would you give to a colleague who is setting up a core microarray facility?
AThere are so many steps in the technology that it’s hard for one person to perfectly handle the entire process. Every step should be performed by dedicated people, who know all the details that count. Pay a lot of attention to the quality of your RNA, optimizing it for each biological model. Finally, never forget that even when you generate beautiful images you are not quite there yet.