AIDS researchers have discovered a gene mutation widespread in the population which appears to confer resistance to HIV.

It is the second such gene to emerge in the past two years, and the scientists who found it think the two may act together to give even greater immunity.

But an expert warns today that further research is needed before people who carry the genetic quirk can consider themselves HIV-proof.

Both mutations affect the formation of a molecule called CCR5, which the virus uses to enter its victim's cells.

Individuals resistant to HIV-1 infection were found to have two identical copies of the first gene - delta32 - with the effect that the normal gene for the CCR5 receptor was absent.

However this did not explain all resistance to HIV-1.

The latest discovery, which may provide the answer, has been called the m303 mutation and was identified by Dr Caroline Quillent and colleagues at the Aids research unit based at the St Joseph Hospital, in Paris.

Reporting in the Lancet today, they say that m303 also prevents the formation of the CCR5 molecule. They carried out genetic tests on 18 men who had remained uninfected with HIV-1 in spite of frequent unprotected sex with HIV-1 positive partners.

One of these men, identified as ExU2, was found to carry both a delta32 and an m303 mutation, and concluded that these two distinct alterations to the CCR5 gene conferred resistance to infection.

Further tests on the man's father and sister showed that the sister carried the same mutation, indicating that the m303 variant was inherited as a single genetic trait.

The researchers then screened 209 healthy blood donors and found that three of them had the m303 mutation.

''This indicates the m303 variant is not a sporadic point mutation in an isolated family, but is also present in the general population.''

The abnormal CCR5 genotype of subject ExU2 probably reflected his remarkable resistance to HIV-1, particularly since his target cells were unreceptive to the particular strain isolated from his partner, with whom he had frequent and unprotected sexual intercourse.

Dr Quillent said they were unable to be precise about the frequency of m303 in the population because they did not take account of the ethnic origin of the 209 healthy unexposed subjects who were screened. But she said: ''The presence of m303 in three of these individuals suggests that the frequency in Western European countries is appreciable.''

However, the difficulties in carrying these discoveries are spelled out in an accompanying commentary by Danish immunologist Peter Garred.

So far no criteria exist for selecting people to take part in international studies on host genetics, and setting up an international database may be impossible in such a competitive field, he points out. Without further research, requiring international co-operation, the gene mutations should not be used to assess prognosis in individual cases.