Evan Eichler took his genomic hotspots to patient populations. The result: translational research that could have implications in the clinic. Evan Eichler, long known for his exploration of evolutionary hotspots in the human genome, recently was senior author on a Nature Genetics paper that goes beyond basic research and into the clinical realm. He and his team scanned hundreds of patients and found a microdeletion event that corresponded with a previously uncharacterized syndrome involving mental retardation, seizures, and slight physical abnormalities. GT's Meredith Salisbury caught up with him to ask about the clinical implications of his work, as well as tips for basic scientists who find themselves getting closer to translational research.
Genome Technology: What kind of screening did you use?
Evan Eichler: What we used in a first pass was our homemade BAC array that targets these hotspot regions of the genome. At this point we've screened about 800 individuals. What this microarray does is it uses array comparative genomic hybridization technology where you have a standard reference DNA sample, which is in this case a normal individual, and you're hybridizing that with one fluorochrome and the diseased individual with the other fluorochrome. You cohybridize and look for signal intensity differences in one channel or the other as indicative of gain or loss of sequence.
When we find events that we haven't seen in normals, we then go back and we screen the parents of those individuals to see if it's in fact a de novo event. In that particular case one of the individuals in that family turned out that when we went back and looked at the mother, the mother in fact had the deletion event. So we kind of put that to the side and said, 'Oh, probably not pathogenic.' We screened another 300 or 400 cases and found two more deletions. We were a little bit suspicious now because we had seen it three times and we had never seen it in our control group. We went back to those parents of those patients — both of them were de novo. At that point we went back to the clinicians and we said the [first] mother was unaffected and the father didn't have the event — can you go back and check out that mother? And the clinicians went back and it turns out that she was mildly mentally retarded and had epilepsy. She had three [kids] from another marriage, and one of them was affected — that child also had the deletion event. We said, 'Well, we got it.'
We went back and did fine-scale analysis using oligonucloetide arrays from NimbleGen. Then we test patients as well as the same reference, but now instead of having four BACs we have 11,000 oligos across the region. It helps us refine exactly what is deleted. In some cases we can get some refinement on the breakpoints. In all of these cases the breakpoints were mapping to clusters of segmental duplications.
We then screened another 1,000 individuals as a replicate study and we used the old standard, quantitative PCR.
GT: Today the lines are blurring between basic and clinical research. You come from a basic research background, but here you're looking at patient cohorts and coming up with results that could have a clinical impact. Did crossing that barrier affect how you did your work?
EE: If you go back far enough into my ancient history, I worked on Fragile X syndrome as part of my PhD. I had started to get an idea of what it means to work with families and patients. There's a lot of sensitivities involved with that — things that are frustrating from a basic research perspective because all you want is the answer, but the most important thing is you have to respect the families. Confidentiality, their decision not to submit additional material for testing. We've encountered that in the studies we've done in the past two years.
I think that's the thing that distinguishes human genetics from other types of phenotype-driven genetic systems. You really have this human factor — that in many ways can be an impediment to research — but at the same time it's those subtleties that we find when we look at phenotypes that really help us link the genetic things that we see at the basic research level back to the clinical side. We couldn't do this work without the participation of families, and the willingness to send DNAs from unaffecteds. There must be a lot of things that go through people's minds — or a mother who has four children, and now for a clinician to come back and see that mother and say, 'You're actually borderline.' I never asked how those discussions went. There are people who are well trained in that area, and I'm not one of them. So I know where I don't belong.
The technology is really blurring things. One of the things that we realized about four or five years ago, and the reason that we decided to go back into human phenotype and disease, is that we find all these things — structural changes, copy number variations, segmental duplications — within the genome. You can compare them against another genome, say a chimp genome. But at the end of the day, unless you can link it back to phenotype, you really haven't completed the work of a geneticist. What we really wanted to do is get back to phenotype.
GT: Do you have advice for someone who is coming from basic research and is interested in reaching across to work with clinicians?
EE: The most important one is you have to respect the work that clinicians put into obtaining the DNA and the phenotype associated with it. That's something that's often lost on basic researchers who think of it as DNA in a tube, and they don't realize the hours that went into diagnose this patient or gather the clinical workup on it. Basically if you look at our clinical papers you'll see a lot of clinicians. The one thing that I think people often blow it on is they don't recognize the contributions that clinicians make, and would like to have a four-author paper. Even if they didn't write a word in the paper, they contributed significantly to the development of that research project — for me that's the definition of authorship, intellectual contribution.
The other piece of advice is that you've just got to be patient. You're dealing with families. You have divorces, separations of families, you have people who get upset or want to retract their information. You're dealing not just with human DNA — you're dealing with people. Not respecting those two things, the clinicians and the families, is probably the biggest faux pas that basic researchers do. GT: How have you followed up on this work since the paper was published? EE: What we're looking at specifically right now is epilepsy samples. The idea here is that even though we began by looking at children with mental retardation, what we found was in fact many of the kids also have seizure syndrome or epileptic-like events. The real litmus test is to screen epilepsy patients without mental retardation and see if we can find structural variants which might delimit the epilepsy phenotype and separate it from the mental retardation phenotype.
I'm not committed to one disease, but I'm more interested in the mutational mechanism. We're now applying [this approach] in other disease populations: congenital heart disease, renal disease, diabetes, schizophrenia, and dyslexia, for example.