Alan Hinnebusch earned his scientific chops in the early 1980s, when yeast genetics, he says, was entering its golden age. At the time, Hinnebusch was a postdoc in the Cornell laboratory of Gerald Fink. "DNA transformation had just been developed in Gerry's lab," he says. That technique essentially allowed researchers to "isolate any gene, manipulate it, and put it back into the organism," making yeast "a eukaryote that you could study like Escherichia coli in terms of genetics."
That technologic potential piqued the interest of Hinnebusch, who even before his stint in Fink's lab was eager to study eukaryotic gene regulation. "It was fascinating to me that you could use genetics - isolating mutants and looking at how they behaved - to deduce what was happening at a molecular level," he says. "I just instantly decided that that approach was very cool."
THE BEST MINER
This pioneer of translational control has humble origins. In the early 1970s, Hinnebusch did his undergraduate work at the University of Dayton, where his genetics professor, Ken McDougall, prodded him to apply to some top-name graduate schools. "He told me to take a chance, ?see how high you can go.' That was very important," says Hinnebusch.
The strategy paid off: Hinnebusch was accepted to Harvard University, where he joined the laboratory of Lynn Klotz and started analyzing repeated elements in the genome of a dinoflagellate. He believed, as some scientists did at the time, that such repetitive sequences might represent elements involved in regulating gene expression. "Of course, that turned out not to be the case," says Hinnebusch, who was nevertheless pleased with the project. "I learned how to do good experiments, what controls you need, and how to stick with something when it's not working," he says. "So in terms of learning how to do science, it was perfectly fine."
The environment at Harvard, however, he found intense. "Coming from a small school in the Midwest, I guess I was a little insecure," says Hinnebusch. "Everybody else was from Princeton or Berkeley or someplace like that, and it seemed like they had all done very impressive undergraduate research. I felt like the only way I was going to be successful was if I worked much harder than the next guy."
That attitude has likely played a large part in Hinnebusch's success. "There are prospectors and there are miners," says Fink, now at the Whitehead Institute. Some scientists identify interesting problems and then move on to the next thing; others settle in and start to dig. Hinnebusch has been blasting away at the same question for more than 20 years. "I've had many fine scientists in my lab over the years," says Fink. "But as a miner, Alan's the best. He is dogged and he just does not leave any stone unturned."
MAKE THAT A DOUBLE MUTANT
The clues Hinnebusch started investigating while in Fink's lab (first at Cornell and then at the Whitehead) related to yeast's ability to selectively switch on the genes that control amino acid biosynthesis when the organism is starving. Before Hinnebusch arrived, Fink had isolated a handful of mutants that were unable to regulate this starvation response. Some expressed the biosynthetic genes all the time, "like people who empty the refrigerator after they've had a big meal," says Fink. Others were unable to activate the genes at all, "like anorexics who don't know they're starving to death," says Fink. "The genius was figuring out what these mutants could tell you."
Step one, says Hinnebusch, was to simply make double mutants. "By looking at the phenotype of double mutants, you can figure out which factors are more proximal to the target genes and which are further upstream," he says. That analysis allowed Hinnebusch to develop a flow chart showing which proteins are involved in the process and how the regulatory circuit is arranged. "Of course that's the stuff I love. That's what attracted me to genetics in the first place." He also showed that a protein they dubbed GCN4 was the key regulator of the pathway.
Then he moved to Bethesda. "I liked the idea that, at NIH, I wouldn't have to fight for grant money or worry about teaching," says Hinnebusch, who is currently chief of the Laboratory of Gene Regulation and Development at NIH's National Institute of Childhood Health and Human Development. "I could really just dive in and be my own best postdoc and accomplish a huge amount by spending all day at the bench." That's exactly what he did. "Alan developed a lot of projects and made many of the key discoveries on his own," says former postdoc Ronald Wek of the Indiana University School of Medicine. "A number of his early papers have one author, or very few authors, because Alan was doing a lot of the work himself."
DELIVERED BY EIF2
Hinnebusch's initial goal on arriving at NIH in 1984 was to prove that GCN4 was a DNA-binding transcription factor that controlled directly the activity of the amino acid biosynthetic genes. "But I very quickly got sidetracked into a different aspect of GCN4 biology." In the process of sequencing the GCN4 gene, Hinnebusch saw that "there was all this extra sequence at the 5' end that didn't seem to be encoding anything," he says. What's more, this upstream region contained "four alternative AUG codons, which could be potential translation initiation sites."
Hinnebusch had a decision to make. "I had a lab of one person, so I could either go after these upstream AUGs and figure out if there's some interesting translational control," some way these decoy codons might allow GCN4 to be translated when amino acids are scarce. "Or I could keep on my path, which was to produce GCN4 protein in bacteria [and] see if it would bind in vitro to the regulatory sequences we and others had identified." He chose to chase the unusual - another trait that's pure Hinnebusch. "That's the way he is," says Jon Lorsch of the Johns Hopkins Medical Institutions, a long-term collaborator. "If everything makes sense except this one thing, Alan will say, ?we have to figure out what that one thing is telling us.'" In this case, Hinnebusch says, the AUG stutter meant that "we had to alter the rules about how translation initiation occurs."
Previous research had indicated that in yeast, ribosomes always initiate translation at the first AUG codon they encounter. "Our idea was, ok, the ribosome sees the first AUG and it looks like a perfectly good start codon, so it has to initiate there," says Hinnebusch. "It translates that first little open-reading frame, but then it's able to bypass the other AUGs under the appropriate conditions and use the fifth AUG in the message to translate GCN4 protein." It was an interesting hypothesis, but "this made no sense to people," says Hinnebusch. When he presented the model at a meeting on translation, people asked Hinnebusch how this could be happening at the molecular level. "I said frankly, I was hoping that you guys could give me some clues."
One important clue came from the laboratory of Marilyn Kozak at the University of Medicine and Dentistry of New Jersey, who discovered that ribosomes can actually reinitiate translation at a downstream AUG - but only if the second AUG is far enough away from the first. Translating that finding to his own system, Hinnebusch found that the AUG that serves as the true start codon for GCN4 lies at exactly the right distance from that first alternative AUG in the message. "And if we made the distance between that first AUG and the last decoy too big, none of the ribosomes could slip through to GCN4, and the gene was shut off," he says. "So we started to think that reinitiation is a process that takes time - that there's some assembly that has to occur to make the ribosome ready or competent to initiate again."
That assembly, he discovered, involves eIF2, which is the protein that delivers the charged initiator tRNA to the ribosome. It turns out that one of the genes that switches on GCN4 encodes a kinase that phosphorylates eIF2. Biochemists working on mammalian cells discovered that the modification reduces the factor's ability to ferry tRNA. "This is what we were looking for," says Hinnebusch: "a way of slowing down the assembly of the initiation complexes" so that the ribosome, scanning down the message after translating the first upstream open-reading frame, would not be ready to reinitiate until it hit the fifth and final AUG, the one that serves as the authentic GCN4 start codon.
As to how starvation regulates the process, Hinnebusch says, "we have pretty decent data, but [we] haven't proven 100 percent that the regulatory domain of the kinase can bind the uncharged tRNAs that accumulate in amino acid-starved cells." That interaction boosts the kinase's activity, thus encouraging phosphorylation of eIF2.
No doubt Hinnebusch and his crew will plug away until his idea is proven to his satisfaction. "He's a very careful, persevering, rigorous scientist," says Harvard's Fred Winston, a fellow Fink postdoc. "He'll do something as many times as he needs to, to be absolutely sure he's right."
Those compelling arguments, supported by a surfeit of data, are trademark features of every Hinnebusch publication. "Look at one of his papers and you'll find 20 figures, and one of the most thorough analyses you can imagine," says Wek. Winston agrees. "When you read his papers, you need to set aside a fair amount of time," he says. "At the end you're exhausted, but you're convinced."
Now Hinnebusch is using "what he's learned about translational control to dissect the basic mechanisms of translation initiation," says Lorsch. Wek agrees. "He's actually made some pretty seminal contributions to the understanding of general protein translation," he says. "This is the stuff of textbooks."
The continuing accumulation of high-resolution images of ribosomes and of other bits of the translational machinery will help make things even clearer. "When you open a textbook, there's still just a bunch of blobs: eif-this and eif-that," says Hinnebusch. "Now we've got real structures to look at, so it's no longer just blob-ology. It's a very exciting time."