Male mice have unexpectedly been found to switch sex if they lack a gene previously linked with lung development.
If the same is true in humans, it might help to explain why some people who are genetically "male" - they have a Y chromosome - develop female genitals.
"It could be involved in cases of human sex reversal," says David Ornitz, head of the team that made the discovery at Washington University Medical School in St Louis, Missouri.
The researchers suggest screening the DNA of people with sex reversal to see if they have mutations in the gene, or in other genes which activate it.
Sex reversal occurs in one person in every 20,000, but only 10 per cent of cases have been traced to known genetic causes. Could it be, says Ornitz, that defects in a gene called Fgf9 are to blame for some of rest? The gene makes a substance called fibroblast growth factor nine.
Ornitz and his colleagues stumbled upon the sex link during experiments to investigate how lungs develop in mice. They genetically engineered the mice so that Fgf9 no longer worked.
Sure enough, the lungs did not develop and the mice were all stillborn. But the surprises came when he took a look at the prostate gland, which was also suspected to rely on Fgf9 for development. To his surprise, there were no prostate glands, or male genitals in the 10 of the 12 "male" mice with Y chromosomes.
"It looked like complete sex reversal," says Ornitz. "The male sex organs had completely regressed." Instead, the "males" had all developed female sex organs.
Geneticists in Britain established in 1990 that in mice, and probably in humans, the master gene determining maleness isSRY. The gene makes a switch called a transcription factor that activates other genes and sets in motion the formation of male sex organs in most mammals.
Ornitz believes that Fgf9 works further along the production line. He suggests that fibroblast growth factor nine performs at least two crucial functions. First, it triggers multiplication of cells crucial to the development of the testes. Second, it tells them to migrate to the site where the testes develop.
"In the absence of these first two events, the testicular cords fail to develop," says Ornitz. Ultimately, female sex organs develop instead by default.
Robin Lovell-Badge, the geneticist at the National Institute of Medical Research in London who discovered SRY in 1990, says that the finding is important.
The crucial thing is that unlike the transcription factors produced by SRY and another important "downstream" gene called SOX3, the newly discovered gene leads to signalling between cells.
"That seems to be very important," says Lovell-Badge. "The evidence for migration looks very strong."