New research suggests that the AIDS virus, which once appeared to be a manageable single entity, is a complex family of rapidly mutating viruses that like a clever enemy can constantly change its weaponry, its camouflage, its defenses and even its targets in the body.

As a result of the mutations in AIDS viruses there are presumed to be thousands of slightly different forms, some possibly having acquired new specialized abilities to be transmitted, to infect different tissues, to evade the immune system or to resist drug treatments.

According to new findings at the Los Alamos National Laboratory, the AIDS virus is mutating its genetic code as much as five times faster than the influenza virus, thought until now to be the fastest in mutating. The genes of the AIDS virus are mutating between one million and 10 million times faster than the genes of human beings.

The flu virus has taken 50 years to evolve as much as the AIDS virus has in the last 10 years, the Los Alamos study shows. New flu vaccines must be developed every few years to keep up with the changes.

Because genetic mutations lead to modifications of the virus' molecular structure, it is possible that if the right changes happen to occur, the virus' behavior would change.

Some AIDS researchers suspect that some of the variations in the AIDS virus that are already known are the result of mutations in the recent past. There is even evidence that within the lifetime of any one AIDS patient, the original strain of virus that began the infection can give rise to several new strains, all of which continue to proliferate.

The Los Alamos finding "casts bewildering shadows" across the prospects for reliable diagnosis, broadly effective treatment and a vaccine that will block all forms of the virus, Gerald Myers, a molecular geneticist who measured the rate of change at the New Mexico laboratory, said last week.

Los Alamos, better known for its research on nuclear weapons, operates a computerized AIDS virus data base. It contains the specific genetic codes, or sequences, from about 30 different AIDS viruses isolated from 1976 to 1986. This is believed to be most of the sequences that have been determined so far in what is a technically difficult process.

Researchers have known for some time that the AIDS virus can mutate, spawning new, slightly differing lineages. And they have been aware of major differences that distinguish two families of AIDS viruses, called Human Immunodeficiency Virus 1 (HIV-1) and Human Immunodeficiency Virus 2 (HIV-2). HIV-1 includes the vast majority of AIDS cases around the world but a few cases caused by HIV-2 have been found in West Africa and France. There are preliminary reports of a third and possibly a fourth family.

Although AIDS virus classification has few agreed-upon rules, differences within a family are usually very small while those between families are relatively large. The Los Alamos findings reveal that the rate of change is considerably faster than previously assumed.

"The AIDS viruses now manifest themselves as a complex family tree, sprouting new genetic branches -- and apparently very quickly at that," Myers said.

Myers' study of the rate has led him to estimate that the best known major families, HIV-1 and HIV-2, split about 40 years ago, long before the disease was recognized. Although there are no known viruses from that time, Myers was able to calculate the rate at which the viruses have been accumulating changes in the past decade and to calculate that it would require about four decades to accumulate the number of differences between the two families.

Myers' interpretations also challenge the widespread suspicion that the two HIVs evolved from a similar virus that infected African monkeys, called Simian Immunodeficiency Virus or SIV. Myers' analyses show that SIV is more closely related to HIV-2 than to HIV-1. This indicates that SIV arose as a branch of the HIV-2 family after the two human AIDS virus families diverged.

The most disparate HIV-1 viruses studied so far differ by about 30 percent in their genetic sequences, an amount that would have taken about 20 years to accumulate at the virus's mutation rate of 1 to 2 percent a year. By comparison, the influenza A virus of today has changed by less than 19 percent from the virus of 50 years ago, when the earliest samples were preserved.

Myers said his evidence supports the view that the AIDS virus has only recently come into existence, having arisen from some unknown ancestor, which he said could still turn out to have been a primate virus that remains undiscovered. When the virus evolved the ability to infect humans, it discovered a vacant ecological niche, an environment to which it was ideally suited and in which it had no competition, and began an explosive period of evolution.

"I think probably the virus has made an evolutionary breakthrough," Myers said. "It's entered a new niche, and we'll have to expect to see a lot of new variation continuing to arise for the foreseeable future."

Myers' findings emerged from a computerized comparison of the genetic codes of various AIDS viruses isolated at different times. Like all genes, those of an AIDS virus consist of a long chain of smaller molecules called bases. Each base is the equivalent of a letter in a sentence except that the genetic alphabet has only four letters (four kinds of bases) and all its "words" are three letters long.

Each genetic sentence, consisting of hundreds or thousands of words in an exact sequence, dictates the sequence of building blocks, called amino acids, that a cell must string together to make a protein molecule. (Viruses cannot make proteins. They must invade living cells and hijack the cell's protein-making machinery to make new viruses.)

A mutation, in an AIDS virus as in a cell, consists of a change from one base to another at any position in the sequence.

Depending on the position and what the new base is, the message may still make genetic sense and the resulting altered protein may confer new abilities on the viral progeny.

It is the exact structure of the proteins that envelop a virus that determine such things as which cells it can infect and whether the immune system will be able to attack it.

If a person has antibodies specifically shaped to attack one kind of enveloped protein and the virus evolves a different one, the antibodies may be useless.

Myers suggested that if large enough changes in the AIDS virus arise, some strains could be different enough that a vaccine against one fails to protect against another. This is one of the reasons that vaccine workers are becoming pessimistic about achieving their goals soon.

For the same reasons, the AIDS antibody test could fail to detect the presence of an infection. The test looks for a specific kind of antibody and if the person's immune system has manufactured a different one -- appropriate to a mutated AIDS virus -- the test could fail to detect an infection.

Myers said he is organizing an international effort, which he calls molecular epidemiology, to collect more virus sequences from around the world to produce a more detailed viral family tree. Conceivably, a pattern would emerge to shed new light on a dark but scientifically intriguing phenomenon.