Alleged X chromosome activator may be a dud. (Not a Turn-On).

Female mammals have two X chromosomes, but males only have one. Early in development, the female embryo must shut down one X chromosome in every cell, or an overdose of the genes on those chromosomes will kill her.

Scientists haven't worked out all the details of X inactivation, and a new study prolongs the puzzlement. A gene that scientists have found to regulate X chromosome activity in mice doesn't work in people, report researchers from the Johns Hopkins Medical Institutions in Baltimore.

In a female embryo, the choice of which X chromosome to inactivate--either that from the mother or that from the father--is random in each cell. Thus, women have patches of cells expressing X chromosome genes derived from Mom and others expressing genes derived from Dad. In various tissues of mice and some other mammals, the maternal X chromosome is always on and the paternal X chromosome remains silent. This specification is called imprinting.

Scientists understand the off--but not the on--half of a molecular switch that controls X chromosomes' fates in mice and people (SN: 8/05/00, p. 92). The gene Xist (pronounced "exist") is continually transcribed into RNA from all X chromosomes in both sexes. Unchecked, Xist RNA sets off a cascade of molecular events that compacts the X chromosome into inert material. Scientists have been searching for the gene that halts Xist expression on the active X chromosomes.

In 1999, scientists from Harvard University discovered a mouse gene that could stop Xist activity. They aptly named this gene Tsix--Xist spelled backward--because Tsix RNA is complementary to, and so binds, Xist RNA and knocks it out of commission.

The Harvard researchers showed that Tsix controls imprinting in mouse placental cells. They speculated that Tsix might also control the random X inactivation of women.

Not so, say Barbara R. Migeon and her colleagues in the August American Journal of Human Genetics.

These scientists had suspected that Tsix might function differently in people than in mice because human Tsix is missing elements found in mouse Tsix. Using color-coded RNA tags, they studied expression of Tsix RNA and Xist RNA in human fetal cells. They found that, unlike in mouse placental cells, Tsix is expressed only on the inactive X chromosome and is thus unlikely to be stopping Xist.

Migeon speculates that Tsix controls imprinted, but not random, X inactivation. "Human placenta is not imprinted, so the fact that we don't have an effective Tsix molecule would make sense," she says.

Tsix in people is probably just an evolutionary vestige, Migeon concludes. A gene from a nonsex chromosome might better control random inactivation, she notes.

Not everyone is willing to dismiss Tsix so easily, however. The study looked at human cells that were far past the period of X inactivation, says Harvard researcher Jeannie T. Lee, who led the original Tsix discovery in mice. Legal restrictions make it difficult to get human cells that are young enough, Lee says.

Other scientists found the study more convincing. Migeon's finding may help explain why women don't have imprinted X inactivation, says Andrew Feinberg, a biologist from Johns Hopkins who wasn't involved in the study. The Tsix gene in people is missing an element that is essential to imprinting, he says.

This paper reminds us that it's important to study people, not just mice, Feinberg concludes.