Blue=human chromosome, Green/yellow=centromeres, Red=human artificial chromosomePhoto: Katie Rudd / Duke University

It was Willard's search to understand the X chromosome that led him to the centromere. He and his colleagues were searching the centromeric region of the X chromosome for DNA sequences that might somehow control X inactivation. He hypothesized that the repetitive elements specific to the X chromosome might act as tags that would key the silencing mechanism. "We did find repetitive sequences specific to the X chromosome centromere," he recalls, "but they had nothing to do with X inactivation whatsoever. So, the experiment didn't work the way we thought it would, but rather than throw it away, it caught my fascination." Willard devoted his efforts to searching for evidence that the repetitive DNA found in the centromeric region was functional rather than just a pile of garbage.

This may mark the point that Willard assumed an X-man persona. "I think that was probably the low point of my reputation in the scientific community," he says, "because people really thought I was crazy. They asked, 'Why are you working on this? It's total junk. You're wasting your time. There can't be a code in there, because it's the same little sequence over and over again. So, by definition, it can't be telling us anything.' I don't know whether I like a challenge or whether I'm pig-headed, but we kept at it."

So, the "pig-headed" Willard and his colleagues invented analytical genetic techniques that enabled them to decipher DNA sequences in the hall-of-mirrors realm of the centromere. One thought was that maybe this repetitive DNA, called alpha satellite DNA, "was just a camouflage that was hiding some other magic sequence that would be buried in there, and we'd have to develop tools that would allow us to get to that magic sequence."

Willard's search for a "magic sequence" was still causing head shaking among colleagues, he says, when his laboratory dropped a scientific bombshell: It announced the creation of a functioning artificial human chromosome. Willard reasoned that if the stuttering DNA was central to the centromere's function in dividing cells, then an artificial chromosome with an artificial centromere containing only "junk" DNA should waltz right along with its natural brethren during the dance of cell division. Sure enough, when Willard and his fellow centromerists synthesized the chromosome and added it to a human cell, they found it worked beautifully. The human cells harboring the artificial chromosomes divided happily ever after, reproducing the artificial chromosomes along with the natural ones.

"Our cells have forty-six chromosomes, and we stuck in a tiny forty-seventh chromosome, and it worked just the way our natural ones do. That was the key breakthrough in that field," he says. "It showed for the first time, definitively, that these alpha satellite sequences confer centromere function."

Willard insists that nature must have evolved these stuttering sequences to mean something important to the cell, since successful cell division is so critical to all life. Still, the nature of that "something" remains unknown. "I'm the first to admit that the fact that our artificial chromosomes work doesn't tell us the code the centromere uses," says Willard. "That just tells us where the code is."

The scientific community reacted to his attempts to build an artificial human chromosome with "a lot of eye-rolling," Willard says. "It was sort of, 'Here he goes again, not giving up on this idea.'" However, he recalls, when he presented his results formally at a 1997 conference in Madrid, "it put to rest the notion of a magic sequence buried in the DNA and established that it was the alpha satellite DNA itself that was being read by the cell." Willard and his colleagues are now tinkering and testing versions of the artificial chromosome, to search for the key characteristics that make the centromere work. The search will be an arduous one, given that the centromeric region comprises some 3-million DNA units on each chromosome.

As a veteran of the scientific centromere wars, Willard deeply appreciates the fact that the science of genomics is full of unknowns--X's--yet to be determined. He has no fear of confronting the unknown. "We now understand only 2 percent of the genome in terms of how it encodes information," he says. "That leaves 98 percent to go."

http://genomics.duke.edu/