April 9, 2001

Biotech CEO Says Map Missed Much of Genome

by Raja Mishra
Globe Staff

Somewhere within the vast microprocessors that hum in the Maryland offices of biotech firm Human Genome Sciences Inc. are the sequences of 60,000 never-before-seen human genes.

In these genes lie the key to hundreds of diseases and explanations, perhaps, of human nature. And the massive effort to map the human genome, which riveted the world's attention over the last year, missed them all.

Or so says Dr. William Haseltine, the CEO of Human Genome Sciences. He wants a recount. And he is not alone.

The genome map makers concluded in February, to much surprise, that there are only about 30,000 human genes, perhaps 35,000, meaning humans are only 50 percent more genetically complex than the lowly roundworm. But in the weeks since, researchers, including those in the Human Genome Project itself, have said the count is anything but etched in stone. It probably will move upward by thousands of genes a year, they expect. Exactly how high is subject to intense, often acrimonious, debate. Haseltine, one of the most respected - and wealthiest - geneticists of his generation, has a guess: 120,000. He said the world's best genetic minds missed tens of thousands of genes in their haste to produce the completed map.

''We believe they have missed as many as two-thirds of the genes that exist,'' said Haseltine in an interview. ''I think they are guilty of sloppy science and sloppy conclusions.''

Scientists at MIT's Whitehead Institute, part of the public genome effort, scoffed at his boast and offered a challenge: ''Anybody who wants to claim a larger gene number, we'll randomly pick 3 percent of the genome and count,'' said Eric Lander, head of the genome sequencing center at the Whitehead Institute. ''Let's see his data.''

Haseltine said he is preparing to make his gene count public, though he refused to name a forum.

''That there are tons more seems unlikely,'' said Lander. ''But by no means is gene enumeration complete.''

While the exact number of human genes has no intrinsic value, knowing that all of them are accounted for is invaluable to those in the disease-fighting trenches.

''The most important issue here is identifying all the sequences, so researchers can have some handle in studying them,'' said Dr. Linday Farrer, chief of genetics at Boston University, who does work on Alzheimer's disease and hypertension.

The human genome is a vast string of chemical instructions, symbolized by the letters A, T, C, and G. Two microscopic copies of the entire genome reside in every human cell.

Most of the human genome serves no apparent function. But within these long tracts of so-called junk DNA are genes, small strings of A's, T's, C's, and G's that are templates for making proteins.

It is proteins that do the work of making and regulating life. The C.elgans roundworm has about 19,000 genes, scientists have found; the fruit fly has 13,000. The human has between 30,000 and 35,000, reported researchers at the publicly funded Human Genome Project and the private company Celera Genomics, which mapped the genome in a parallel effort.

But humans are billions of times more cellularly complex than C.elgans. Scientists explained that our complexity probably lies at the protein level and that most of our genes, contrary to popular belief, can make more than one protein.

Misdesigned proteins, caused by faulty genes, result in hundreds of diseases. But often, disease involves the interaction of numerous genes. That's why disease researchers, hunting for these guilty genes, need to know that all the possible genetic suspects are accounted for.

Haseltine, for one, says researchers are being deprived of thousands of targets.

''What's wrong is that the data is not complete. It is not fully assembled,'' he said. ''There are gaps, big gaps.''

Haseltine became extremely rich, worth more than $300 million in stocks and options, after founding Human Genome Sciences with geneticist J. Craig Venter in 1992. Venter left in 1997 to start Celera.

He is known for his outspoken style and glamorous, jet-setting life. But his resume is hefty. Haseltine was a student of biologist and Nobel laureate Walter Gilbert at Harvard and went on to join the Harvard Medical School faculty. There he became a pioneer in studying the genetic underpinnings of HIV, the virus that causes AIDS.

His company, and numerous other biotech start-ups, have been advertising access to 60,000 or more genes for years. For instance, Oakland-based Double Twist said it could sell researchers the codes for 65,000 to 105,000 genes. But when the Human Genome Project and Celera announced their findings, it seemed as if Double Twist, Haseltine, and others had egg on their faces.

Nick Tsinoremas, Double Twist's vice president of genomics, said his company's prediction was based on incomplete evidence. He said he still believes the number is higher than 35,000, and he will submit a paper to prove it within a month.

''Most of the approach [the Human Genome Project and Celera] used was a very conservative approach,'' he said, adding that his guess, which he would not reveal, is not as high as Haseltine's.

The approach involved computer-aided estimations. Over three decades, researchers have painstakingly sequenced thousands of genes, a small part of the total, almost hand-counting each genetic letter. To decode the rest of the massive genome, scientists programmed high-powered computers to search for sequences that were similar to those they already knew of. ''The only genes you have a hope of finding are ones that have already been isolated or is closely related to it,'' said Haseltine, explaining his criticism of the method.

Haseltine, and dozens of other researchers, say unique genes, unlike any others previously decoded, still await discovery. The question is: How many? Terry Gaasterland, who heads the lab of computational genomics at Rockefeller University in New York, recently published a paper showing that a similar estimation technique used to map the fruit fly genome probably resulted in an undercount. The fly was thought to have about 13,000 genes. Gaasterland's team found 19,410 ''plausible'' ones.

''They look like genes, they smell like genes,'' she said.

She noted that both the Human Genome Project and Celera left the door open to a gene count of 57,000, though it was not well publicized.

''That's a bottom line. It could be up to 57,000 and there's potential for it to be more,'' she said.

Marc Vidal, a geneticist at the Dana Farber Cancer Institute, also has developed a technique to make more accurate gene counts. He said he believes the human count is higher than 35,000.

''It's such a vast, huge discovery, you don't expect it to be perfect,'' he said of the Human Genome Project.

Though many academic researchers say the gene count is higher, they have sniped that business people like Haseltine are arguing for higher gene counts because more genes means more profit potentential from patenting and licensing genes. Haseltine firmly denied the accusation.

''It's not a matter of commerce. It's a matter of medical science,'' he said. ''If people take the 30,000 figure seriously, we go into the battle against disesase without a full complement of weapons.''

In fact, many in the scientific establishement seem certain that thousands more genes will emerge, and they are willing to put their money where their mouths are. A pool run by the Sanger Centre, the British partner in the Human Genome Project, is taking bets on the topic, at $5 per bet this year and $20 in 2002. The final tally will occur at a scientific meeting in Cold Spring Harbor, N.Y., in 2003.

The mean bet, as of last week, was 61,710 genes.

Copyright 2001 Boston Globe. All rights reserved. International copyright secured.
Filel Date: 4.12.01