BALTIMORE — Molecular tests may be able to distinguish homosexual from heterosexual men, a small study of twins suggests.
Chemical modifications to DNA that change the activity of genes without changing the genes’ information differ between homosexual and heterosexual men, researchers from UCLA David Geffen School of Medicine have discovered. Results of the unpublished study on the link between these modifications, called epigenetic tags, and sexual orientation were presented October 8 at the annual meeting of the American Society of Human Genetics. Comparing one type of epigenetic tag known as DNA methylation in pairs of twins in which one brother is gay and the other straight revealed patterns that distinguish one group from the other about 67 percent of the time, computational geneticist Tuck Ngun and colleagues say.
The work already has provoked controversy, with some scientists questioning its methodology and others worried about how the research could be used. Some are concerned that the research could be misinterpreted as one step in an effort to “cure” homosexuality. Nothing could be further from the researchers’ intentions, say Ngun and Eric Vilain, the geneticist who heads the research group. “None of us see homosexuality as a disorder or something to be fixed,” Ngun said. “We’re just interested in what makes us tick.”
Very little is known about how human sexual preferences of any type arise, Vilain adds. That’s especially true on a biological level. “Our research is not about homosexuality,” he says. “It’s about understanding sexual attraction, the biology of desire.”
Previous studies have found tentative genetic links to male sexual orientation, but no one has identified a “gay gene” or genes. Still, the development of sexuality seems to have origins early in life, maybe even stemming from cues in the womb. For instance, for each biological older brother a man has, his likelihood of being homosexual rises by 33 percent. That finding has been replicated in several studies and could indicate that some condition in the womb sets epigenetic marks, which later influence preference of sexual partners.
Epigenetic marks have been shown to influence behaviors in rodents such as maternal care and drug addiction (SN: 5/24/08, p. 14). Whether and how these marks are involved in human behavior is still a matter of intense debate, says Peng Jin, a human geneticist at Emory University in Atlanta. Exploring whether they are associated with sexual preference isn’t unreasonable, Jin says, he’s just not sure the researchers have gone about it correctly. He also doubts that a study of less than 100 men has the statistical power to predict sexual orientation.
Ngun and colleagues measured DNA methylation levels in the saliva of 37 pairs of identical twins in which one twin self-identified as homosexual and the other as heterosexual. Another 10 pairs of twins in which both were gay also participated in the study. A computer program dubbed the FuzzyForest algorithm examined data from half of the gay and straight twins to learn how their DNA methylation patterns differed from each other. The initial round of training found 6,134 spots in the genome where the twins differed, but together those sites could correctly identify gay twins in the remaining pairs only 44 percent of the time. Narrowing down the number of sites to nine improved accuracy to 64 percent.
Further analysis involved the set of 10 pairs of gay twins. The researchers asked the computer program to pick out the spots that were different in the mixed orientation twins, but the same in the gay twins. That left five sites that could correctly identify 67 percent of gay twins in the test group.
Some of the regions may be involved in controlling activity of two genes: CIITA, which regulates activity of some immune system genes, and KIF1A, a gene involved in the transport of communication molecules in the brain.
The study raises many issues. Scientists question whether the finding will hold up in larger groups of unrelated people. Also, the computer algorithm hasn’t been tested on other datasets, raising concerns about whether the method is valid.
Vilain agrees that the study has limitations. “We’re looking at the wrong tissue at the wrong time,” he laments. The right tissue would be the brain, and ideally, researchers would be able to track DNA methylation changes over time from fetal stages on. Such research is not ethical in humans, so the team measured DNA methylation patterns in saliva taken from adult men, long after their sexual orientation had been determined. “There were no other choices,” he says.
DNA methylation patterns in saliva may not accurately reflect what is going on in the brain where behavior is controlled, Jin says. Saliva is not good material for epigenetic studies, he adds. The types of cells present in saliva can change dramatically depending on when and what a person has eaten and other factors. Different mixes of cells in the saliva would probably have different DNA methylation patterns that could further confuse the results. Blood would have been a more stable material to examine, although it also doesn’t always match what happens in the brain.
Vilain says his work has “zero clinical application.” This is not molecular gaydar, it’s simply a statistical measure that epigenetic marks differ between men of opposite sexual orientations, he says.
Whether the epigenetic changes are a determining factor in sexual orientation or a result of differing experiences in life, the study can’t determine, he says. But it may offer invaluable insight to the development of human sexuality, says pharmacologist Margaret McCarthy of the University of Maryland in College Park. "This study provides a major step forward in our understanding of how the brain can be affected by factors outside of the genome,” she said in a statement to the Genetic Expert News Service. “Regardless of when, or even how, these epigenetic changes occur, their findings demonstrates a biological basis to partner preference.”