AT A GLANCE
Molecular biologist at Spectral Genomics in Houston, Texas
Research focuses on developing cDNA, oligo, and genomic profiling microarrays; as well as using microarrays to compare gene expression profiles in prostate cancer and breast cancer.
PhD, Molecular Biology from Thomas Jefferson University, Philadelphia
Postdoctoral fellowships at Stanford in cancer research, and at the National Cancer Institute working with Jeffrey Green
QHow did you get into doing microarrays?
AWhen I was at Stanford doing cancer-related work, we began to look at lung cancer patients using microarrays. At the NCI they had a fantastic up-and-coming microarray facility. Then Spectral Genomics needed someone to develop a new cancer chip using their chemistry, which is quite different than anything out in the market.
QHow are the arrays different from, say, the typical cDNA arrays you would spot down on glass slides at Stanford?
AIn a conventional array, you don’t need to modify the DNA or plasmid or oligo. The chemistry is on the surface of the glass, and sometimes you do non-covalent attachment, and sometimes amine attachment. Our chemistry is different. We put linkers onto the DNA. The linker allows a tighter covalent bond to the naked glass. Allan Bradley and Wei-wen Cai from Baylor invented the technology — they’re scientific advisors to Spectral Genomics — and Spectral has patented the processes. The first platform we are developing here is using that chemistry with BAC clones, so you can do genomic chips instead of expression chips. A genomic microarray can detect chromosomal changes, including deletions and amplifications. I got on board to develop the same chemistry in the expression arena.
QWhy is this chemistry better than traditional surface chemistry?
AFor genomic arrays the advantage has to be that you should be able to detect two-fold changes with confidence. It means there is duplication at a chromosomal level. If you can detect two-fold changes with confidence, and translate this [ability] to expression arrays where usually two-fold and three-fold changes are often the background noise you get from conventional chemistry, this is going to expand the number of genes that you see that are modulated. Whatever system you are using, this type of chemistry, printing the DNA on the naked glass has no background at all.
QWho are your competitors with chromosomal arrays?
AThe only competitor out there is Vysis, and they have an array coated with some kind of metal. You do spot the clones on it, but our technology allows extremely low background as well as high sensitivity. We believe it’s an extremely versatile system.
QHow did you develop the cancer array?
AI went through NCBI databases, including OMIM, and that’s where I found a lot of genes related to cancer. The microarray includes about 1,700 unique elements, including typical tumor suppressors, oncogenes, cell cycle genes, and cyclins that are very relevant to cancer. So I compiled a list of genes that I would like to pick, then ordered 70-mers from the vendor. The reason we chose chose 70-mers is that we went through a series of studies with different length oligos, and got the best intensity with longer ones. We thought that 70-mers would be a happy medium. We used informatics from Operon, as well as oligos from their company. So far we have been very happy with them.
We just finished a validation of the chip and we can detect almost all of the tumors including lymphomas, breast cancer, and prostate cancer. We look at this chip as a starting point, if you are in cancer research: You get the information that you need without having enormous excess of data.
QWhat if there are other genes involved in cancers that have not previously been seen? This chip would miss them.
AGene discovery is a big part of array work. We are planning to launch the cancer chip this quarter, but this is not the end of the product line. We’re going to continually make progress, and we would like to follow up with chips that have unknown ESTs and known genes. We also have human and mouse arrays, including a mouse chip that represents the mouse genome to a resolution of three megabases. This was followed by a three megabase human, then a one megabase human. A one megabase mouse is coming in March. That means that every one megabase we have coverage in the chromosome. We chose areas that were cytogenetically important.