The Scientific Do-Over

What would you do if you could apply modern technologies to yesterday's scientific questions?

Aug 1, 2006
Jeffrey M. Perkel

I knew a number of gifted graduate students when I worked at the bench, but I never met one like Arul Chinnaiyan. His was an exceptionally productive graduate career: Between 1995 and 1998, Chinnaiyan and his advisor at the University of Michigan coauthored 16 papers, including two that describe the identification of apoptotic signaling pathway members, FADD and FLICE - work that has since been cited 1,415 and 2,026 times, respectively.

To identify FADD,1 Chinnaiyan, now the S.P. Hicks Endowed Professor of Pathology at Michigan, performed a yeast two-hybrid (Y2H) screen using the Fas cytoplasmic domain as the bait. "That was really the only reasonable discovery technology at the time to look for novel protein interactions," he says. The expedition paid off for Chinnaiyan, but as is so often the case when one goes fishing with Y2H, he reeled in a pile of red herrings, too. In all, he and a technician spent two to three years running those screens and culling the list of candidate interactors.

Could that time have been better spent? In retrospect, yes. Automated DNA sequencers, microarrays, RNA interference, and GFP fusions all simplify and streamline processes that were grindingly slow only a few years ago. In Chinnaiyan's case, had he today's tools at the time, he says he would have substituted affinity pull-down and mass spectrometry for Y2H to find interaction partners (as he subsequently did with FLICE), a change that he estimates would have compressed his timeline down to a couple of months.

That savings would probably have given him the time to expand his work to include microarrays and RNAi - tools that would have enabled him to survey the protein's roles throughout the cell, rather than limiting himself to apoptotic pathways. "At that time we didn't really have a global way of monitoring many different pathways, or at least surrogates of many different pathways, in a reasonable fashion," he says.

Doubtless Chinnaiyan would have succeeded, just as he did a decade ago. Equally likely, those who couldn't cut it back then would be just as stymied today: You either have the skills or you don't, and no amount of technological development can change that.

As technologies change, so too do expectations, and the bar for scientific excellence keeps rising. If you can accomplish in months what once took years, it clearly no longer merits a PhD. Eric Hoffman, director of the Research Center for Genetic Medicine at Children's National Medical Center in Washington, DC, was a postdoc in 1987 when he cloned dystrophin, the gene that is mutated in Duchenne muscular dystrophy.2 A graduate student in the same lab, Tony Monaco, spent three years' worth of 80-hour weeks mapping that gene to its location on the short arm of the X chromosome. Today, using Affymetrix SNP arrays and five or so patient samples the same task could be accomplished overnight, says Hoffman.

On the other hand, Hoffman's work - cloning and sequencing the 14-kb cDNA itself - would not be so easily recapitulated today. "It was a bitch," he recalls. "What are you going to do? 5' RACE off a 14-kb RNA? I don't think so."

Of course, dystrophin is a genetic anomaly; most genes are far shorter, and modern techniques would suffice to clone them relatively quickly. Bullish as I am on technology, I think that's kind of sad, actually. "Wow! You've cloned a gene!" has morphed into, "So, you cloned a gene."

Marc Vidal, who earned his PhD in the 1980s, told me he wished he could have gotten his thesis in the 1970s. "I always thought that it must have been so fantastic to be around when the first genes were cloned, when restriction enzymes were figured out, when genetic engineering was developed, and that sort of thing," he says. "Those were the real innovative days."

In a decade or two from now, can you imagine what students will see when they look back at us?


1. A.M. Chinnaiyan et al., "FADD, a novel death domain-containing protein, interacts with the death domain of Fas and initiates apoptosis," Cell, 81:505-12, 1995.2. E.P. Hoffman et al., "Dystrophin, the protein product of the Duchenne muscular dystrophy locus," Cell, 51:919-28, 1987.