As scientists seek new ways to stop the human immunodeficiency virus, they are pinning hopes on hiding or blocking a kind of protein doorway that HIV uses to attack human immune cells.

A Richmond company is trying a new idea: Eliminate the lock altogether by actively mutating the genes of HIV patients.

On Saturday, Dale Ando, the chief medical officer of Sangamo Biosciences, reported to a conference of the American Society of Microbiology that cells genetically modified by the firm were fighting off HIV infection in the laboratory.

Sangamo plans in late 2006 on starting human trials involving removal of billions of immune cells from HIV patients, genetically modifying those cells and replacing them as newly armored warriors against HIV.

What remains to be seen is how well such a strategy works, in lab mice and humans, against a virus that repeatedly has proven too nimble to destroy in any single stroke.

Still, infectious-disease specialists find the notion of genetically disrupting HIV's primary gateway into human immune cells full of promise.

As the fight against AIDS enters its third decade, scientists are looking for a new strategybeyond the three-drug cocktail that has managed to slow, but not eliminate HIV infection.

The top target for new drugs is the protein coating on human immune cells that act as receptors. HIV uses those receptors to gain entry to human immune cells and infect them. One receptor used by the most prevalent form of HIV is critical to infection and called CCR5.

It was discovered in the mid-1990s when scientists found a small number of Caucasian intravenous drug users were exposed routinely to HIV but not getting sick or had a much slower progression of the disease. It became clear that they had a mutation in one or both copies of the CCR5 gene and that the mutation apparently wasn't harming their immunity to other diseases.

Evolutionary geneticists have traced the mutation to Northern Europe but disagree about when it emerged and why its prevalence has grown to nearly 1 percent of the Caucasian population, as much as 16 percent in some Eurasian countries.

It is absent in Africans, Native Americans and East Asians. Some scientists believe the CCR5 mutation arose 700 years ago during the two centuries of Black Death, when millions of Europeans died of bubonic plague.

A team of British scientists argue the mutation got a foothold 2,000 years earlier, when viral hemorrhagic fevers swept Europe and perhaps used the same gateway to infect humans as HIV does today.

So far, most drug scientists have targeted the CCR5 receptor with chemicals designed to mask or block the receptor. But each human T-cell is covered in more than 10,000 receptors, and the necessary doses to frustrate HIV infection can risk side effects such as liver toxicity.

Scientists at Sangamo propose instead to do for HIV patients what plague or viruses indirectly may have done long ago for a portion of Europeans but without waiting for evolutionary selection.

"What we're trying to do is give patient cells this mutation and give them back to the patient so they can mount an immune response," Ando said Friday.

The firm's specialty is designing chains of roughly finger-shaped proteins that can lock on to any specific address in human DNA and change its content or function.

In the case of CCR5, scientists have devised proteins to break the double-strand of DNA at the gene and disable it. To do this, HIV patients would undergo something similar to T-cell infusion therapy, with their blood coursing through a machine that removes some of their immune cells.

Scientists then would stimulate those cells with a small electrical charge to open pores and admit the proteins. So far, the company reports successfully altering the gene in 3 to 18 percent of the cells, a high rate of genetic modification. After four weeks to make sure the cells are free of contamination, they would be reinfused in the patient.

Richard Sutton, a professor of molecular virology and microbiology at Baylor College of Medicine, said the approach is "a completely new method."

"I know everyone's very excited about it," he said. "It's certainly an interesting approach, but my sense is the virus is smarter than that."

Once HIV infection begins, the virus can change and use other receptors to enter immune cells besides CCR5. Whether CCR5 is blocked or eliminated, Sutton said, "my feeling is the virus is just going to ignore it and go on to another receptor."

In lab experiments so far, isolated immune cells that have been given the mutation manage to fight off infection by HIV. "The proof of the pudding," Ando said, "is when we get into patients."

Contact Ian Hoffman at ihoffman@angnewspapers.com.