NEW YORK (AP) - One clue was a region of mouse brains that looked like a cut-and-paste job. Other hints came from methodically trolling through the human genes. And now, scientists hope they've picked up the scent for identifying genes that set up some babies to develop autism.

Nobody can declare a clear victory yet. A proposed autism gene that captured headlines in November, for example, will have to be confirmed by further research before it's widely accepted.

But gene-hunters say the search has been heating up in the past few years, with more funding luring in more investigators, sophisticated tools and techniques being brought to bear, and an encouraging run of tantalizing results.

"It's snowballing at this point," says Dr. Edwin Cook of the University of Chicago, who points to research published in just the past two years.

Once they start finding genes - nobody knows how many might be involved - researchers hope to acquire clues into the roots of the dimly understood disorder.

Autism usually appears by age 3, mostly in boys. Affected children have trouble communicating and interacting with others. They may not respond to their names or even look at other people. In severe cases, they may become aggressive or injure themselves.

Scientists hope that finding genes can answer several key questions:

-What exactly goes wrong in the brain to cause autism? -Are there particular targets for drugs to treat or prevent it, even before birth?

-Do symptoms result from a single event in the brain, or from a process that could be interrupted?

-Could people destined to develop autism be recognized at birth, before symptoms appear, allowing for early intervention?

The hunt for autism genes is complicated. It's not a case of a single flawed gene causing a disease, as in Huntington's or cystic fibrosis. Rather, autism appears to be brought on by unknown environmental influences coupled with a dimly understood combination of genes that makes people vulnerable.

That's the case with some other conditions such as diabetes, heart disease and schizophrenia. And it makes these "susceptibility genes" much harder to find.

When a person inherits a disease-promoting version of such a gene, it only tips the scale toward illness, rather than causing it outright. So the same version of such a gene can show up in Joe and Bill, but only Bill gets sick. Or it may show up in Mary but not Jane, yet both are sick due to different sets of susceptibility genes and environmental influences. So a link between a gene and a disease can be difficult to demonstrate.

Yet autism also offers an advantage for gene-hunters: The genetic effect on susceptibility is strong. The rate of autism in the general population is about two-tenths of a percent or less, but for siblings of an autism patient it jumps to around 3 percent. And for an identical twin of an autism patient, someone who shares all the patient's genes, the rate is 60 percent or more.

That powerful genetic influence lured investigators into the autism gene hunt around the mid 1990s, giving hope for quick success.

"I thought this would be an easy find. I really did," said Susan Santangelo of the Tufts School of Medicine and Harvard School of Public Health.

As she and others looked for the genes, however, they came to believe that this inherited influence was split up among more genes than expected, leaving no gene with a powerful and easily traceable influence. That made it harder to find any of them.

If only three or four genes were responsible, "we could have done it," said neuroscientist Joseph Buxbaum of the Mount Sinai School of Medicine in New York. "We now think our estimates of the number of genes is too low, because we haven't done it."

Nowadays, the "wild optimism" of the mid-90s has given way to a more realistic appreciation of the challenge, says Gerard Schellenberg, a research professor of medicine of the University of Washington. Yet he and some others say it's possible that autism will be the next common psychiatric disorder, after Alzheimer's, to yield susceptibility genes.

Scientists can point to encouraging results from a variety of approaches.

One is to scan all the human chromosomes, the rodlike structures that hold genes, for locations that show evidence of harboring an autism gene. That involves analyzing inheritance patterns for particular bits of DNA along the chromosomes.

A half-dozen of these scans have been done for autism, and while results have not been uniform, researchers say at least two locations show promise. The big job now, of course, is to search those locations to identify any autism genes they hold. The hunt is already on.

Scientists also are following up on clues from some accidents of nature. Researchers at the University of California at Irvine, for example, recently reported finding that a girl with the condition had been born without a chunk of chromosome 15. So genes found in that chunk become candidates for study.

The proposed autism gene that grabbed headlines in November resulted from a different approach. A mutant mouse and a good memory put Patricia Rodier of the University of Rochester School of Medicine and Dentistry onto a gene called HOXA1.

In 1995, she and colleagues examined the brain stem of a woman with autism who had died in the 1970s. The brain stem had developed in a peculiar way, they found - as if a particular band of tissue had been removed and the adjacent areas had simply moved together.

Rodier realized she'd seen this pattern before, in a paper about mice with a disabled version of a gene that helps the brain stem develop.

The next step was to inspect the human version of that gene, called HOXA1, in people. Rodier and colleagues reported in November that a particular version of the gene showed up in autism patients more often than predicted by chance, suggesting it may promote the disease.

Experts call the finding intriguing. But they caution that, like all initial work, it has to be confirmed by further studies. Schellenberg and his collaborators have already begun that follow-up research.

"I wouldn't be trying to replicate it if I didn't think there was a pretty good chance it was true," Schellenberg said.

So now, more than a half-century after American psychiatrist Leo Kanner gave autism its name, researchers may be closer to the genetic roots of the disorder. And a quest that looked so deceptively easy just a few years ago may finally pay off.

Copyright 2001 The Associated Press. All rights reserved.