The keys to sex

Of the genes that are responsible for decisions in mammalian development, two on the sex chromosomes stand out as key molecular switches. The Y-linked Sry gene encodes a DNA-binding protein that specifies male gender. The X-linked Xist encodes an untranslated RNA that controls X-inactivation. Genetic analysis of Sry and Xist has revealed phenotypes of mutational loss-of-function and transgenic gain-of-function that established these two genes as both necessary and sufficient for their respective roles.

An idea about the location of the gene that specifies testes, and subsequent male development, was decisive to its isolation. It had to lie in the region where the sequence of the Y chromosome is unique, but close enough to the region shared with the X chromosome to account for the accidental, albeit rare, crossover with the X chromosome to generate XX males or XY females. Peter Goodfellow, Robin Lovell-Badge and colleagues showed that SRY, which is localized within 35 kb of the boundary of the Y-unique sequence on the human Y chromosome, encodes a testis-specific protein that is similar to other DNA-binding HMG-box-containing proteins. Around the same time, Philippe Berta, Goodfellow, Marc Fellous and colleagues, and Gerd Scherer and colleagues, found sex-reversed XY female humans who carried mutations in the SRY gene, thereby demonstrating that it was necessary for maleness. The Lovell-Badge and Goodfellow teams then proved Sry to be sufficient for testis development by showing that an Sry transgene conferred a male phenotype on XX mice. The transgenic mice were infertile, however, as the XX genotype is incompatible with sperm production. This result indicates that the sex chromosomes are important for both sex determination and gamete production, but that the two functions are independently regulated.

Tortoiseshell cats are a common reminder that most female mammals are mosaics containing clusters of cells that express the X chromosome inherited from their father, whereas others are expressing the mother's X chromosome. This phenomenon — termed dosage compensation by X-inactivation — was first described by Mary Lyon in the mouse. It ensures that most genes on the X chromosome are expressed at similar levels in males and females. The first clues to what underlies this process came from Huntington Willard and colleagues, who discovered that the XIST gene is only expressed from the inactive human X chromosome, so it presumably acts in cis on the chromosome that produces the XIST RNA. Neil Brockdorff, Sohaila Rastan and colleagues found the same for Xist in mouse, and went on to identify the 15 kb untranslated nuclear transcript that is conserved in sequence and structure between mice and humans. The targeted knockout of Xist in XX mouse embryonic stem (ES) cells showed conclusively that the gene is essential for X-inactivation.

In eutherian mammals, X-inactivation is random, but in marsupials, imprinting ensures that the paternal X is preferentially inactivated. The same also happens in extra-embryonic tissues of eutherian mammals, and probably in pre-implantation mouse embryos too. A chromosome-counting mechanism seems to allow inactivation of all but one X chromosome during development. By studying transgenic ES cells and embryos harbouring deletions of sequences in the antisense strand of the DNA, Jeannie Lee showed that Xist is regulated by its 40 kb antisense partner, Tsix, which codes for another cis-acting untranslated RNA. It now seems that both the imprinting and counting mechanisms control Xist through Tsix, thereby ensuring that only one X chromosome will remain active in the cells.