AT A GLANCE
PhD in cellular and developmental biology, SUNY, Stonybrook, 1996
National Science Foundation International Scientist Training Program in Antarctica, studying receptor proteins related to speciation in sea urchin eggs, 1996
Postdoctoral fellow at Harvard studying population genetics of various sea urchin species from coral reefs in Hawaii, looking at receptors related to speciation
Postdoctoral fellow at Steve Gullans’ Lab at Harvard Medical School and Brigham and Women’s Hospital, 1997 to 2000, studying the effect of growth factors on signaling pathways in cultured human cells
2000-2001, research scientist at Aventis Pharmaceuticals
How did you get into working with microarrays?
I was doing a postdoc with Steve Gullans at Harvard, working on a receptor protein in kidneys. We got a beta version of an arrayer from Cartesian technologies, and a beta version of a microarray scanner from GSI Lumonics, which was later Packard and is now PerkinElmer, and said, “OK, let’s do microarrays.” I picked out my own genes by hand, designed my own experiments, and developed my own surfaces and my own protocols for doing microarrays. I also developed a dendrimer-based DNA detection system for microarrays, and licensed it to Genisphere.
What are dendrimers, and how does this detection system work?
Dendrimers are balls of oligos of a defined size, and the outer shell has a defined number of free ends, to which are attached a defined number of fluors. By modifying the [target sequence] primer to include a capture sequence, you can also include a complementary tag on the dendrimer. You hybridize the cDNAs to the array, then wash out the unbound target, and detect them with the pre-bound primer. In oligo [arrays] the hybridization depends on the transcript length and the sequence. [But with dendrimers] you get rid of the sequence-specific artifacts by binding the dendrimers to the cDNAs or prebinding the cDNAs, then hybridizing them on the arrays. Each cDNA on the array has the same exact number of fluors, so this gives you a better representation of the signal intensities vs. bound material. Dendrimers became a very successful business for Genisphere.
Currently, what microarray-related projects are you working on at TeleChem/arrayit.com?
We are involved in development of protein arrays, as well as DNA arrays. We do a lot of consulting for people in terms of the microarray experiment setup and application, especially the design of microarrays. We do trouble shooting and onsite consultations, mostly for customers but also for anyone who is using microarrays.
What is the number-one issue you deal with when troubleshooting?
Design can be complicated. Sometimes the budget is a limiting factor, when customers want to [design] a complicated array and include a lot of gene families. Also, microarrays are still more of an art than a science. It can [consequently] be difficult to troubleshoot over the telephone because it is very technique-oriented, and not something that you can learn just from a protocol.
Have you gone back to studying receptor proteins with
Not yet, but I am starting to come full circle, because I am bringing my proteomic interests into microarrays. We are developing surfaces for protein arrays, as well as printing buffers, washing buffers, and hybridization buffers. There’s a huge market for pre-spotted protein arrays as well.
In your work, you have worked with both Affymetrix arrays at Aventis, and now self-spotting systems at TeleChem. Which one do you find preferable?
I think they are very complementary. The Affymetrix platform allows the user to get a full genomic view of the human or E. coli genome and for other organisms for which there are Affymetrix chips. But there are a lot of people that are working on cows and dogs and pigs, [who] need a custom spotted array. We are now working with people looking at plankton [gene] identification. Also, people have a need for custom spotted microarrays when they want to do a limited number of genes and want a defined gene set.
What is the biggest challenge you and your customers face in working with microarrays?
For cDNA microarrays, sample prep is the largest stumbling block. I am not a fan of cDNA arrays. I am much more a fan of oligo arrays, as cDNA arrays require multiple rounds of purification and have to be highly purified before they are spotted. Precipitates from the purification process can interfere with the printing process. You also have different sized cDNAs on the arrays, and this can be complicated downstream.
So you now only do oligo arrays at TeleChem/arrayit?
We have a FlexChip format that allows people to send any samples they want arrayed, including oligos, cDNAs, or proteins. We work in a class 100 cleanroom and have superaldehyde substrates. We are coming up with proteomic services as well that [are performed] in class 100 cleanrooms.
Do you make your own arraying equipment?
No, we have capital equipment from PerkinElmer, Virtek, and Cartesian. We decided to start making and selling the Spotbot arrayer because we found that there was a niche for a desktop microarrayer. We aren’t in [the arraying equipment business] on a large scale.
Do you still do microarray-based genetic screening of newborns with NuGen sciences, and how is that going?
We are no longer in that collaboration. We have a company to do [microarray-based screening] called arrayit/Next Generation screening.
What future developments would you like to see in the microarray field?
Overall, I would like to see a standardization of the field, especially standardization of normalization protocols and analytical protocols, so data can be cross-validated among experiments and researchers. This way people doing similar experiments will not be talking about two different things.