Few phenotypes are as familiar—and as interesting—to us as mammalian sexual differentiation; nearly every identification form that we fill out from childhood asks us to check the box marked “M” or “F.” Direct experiments investigating the biological basis for mammalian sex determination have been done for more than two centuries, and speculation about the biological origins of mammalian sex differences are far more ancient than that. We want to point out that we are again using “mammals” here to refer to placental mammals, and are not including marsupials or monotremes in this description.
One of the most important early observations about mammalian sex determination came from cattle. It was observed that a cow sometimes gives birth to a calf that is essentially female (and chromosomally XX once the sex chromosomes were recognized) but has masculinized features and is infertile; these are known as freemartins. Freemartins are always one of a twin, and the other twin is always a male. In the early twentieth century, it was demonstrated that freemartins occur because twins in cattle usually share a placenta and chorions; the testes of the male twin produces a substance, now known to be testosterone, that circulates freely through the shared chorions and placenta to the other twin, resulting in its masculinization, despite being chromosomally XX. Thus, in mammals, testosterone induces male features and inhibits some female features, and testosterone is produced by the testes.
What then induces the sexually indifferent gonad in the embryo to develop into testes? This was called testis-determining factor, or TDF; its biochemical and molecular nature was unknown for decades. Note that, in most mammals, including humans, twins arising from separate fertilizations like this do not have shared chorions, and thus testosterone does not circulate from one twin to the other, as happens in cattle. Human freemartins are found only in fiction and folklore, and not in nature.
It was also recognized more than a century ago that males have a Y chromosome, and it was hypothesized that a gene on the Y chromosome encodes TDF. Many candidate genes were initially proposed and then subsequently rejected as the gene encoding TDF, before the proper gene was found. Among the candidates was a gene in mice known as Sry. Sry was defined by the mutant phenotype—the mice were XX males and were sex-reversed. This sex reversal was due to the sex-determining region, so Sry is also used to refer to both the sex reversal and the region on the Y chromosome responsible for sex reversal. Sry is one of the few functional genes that was mapped to the Y chromosome, although the number of Y chromosome genes was not yet fully known when it was mapped. In addition, in mice, Sry maps very close to the region that crosses over between the X and the Y chromosome. This proved to be very important and requires a bit more explanation.
The pseudoautosomal region and XX males
Recall from Chapter 6 that every pair of chromosomes must cross over to segregate normally in meiosis. Since the X chromosomes in females can pair and synapse, this need for homology is no different from that for an autosome. However, in males, the Y chromosome is not of the same size as the X and does not have the same genes or the same DNA sequence as the X. Nonetheless, the X and the Y do pair in male mammals, but the pairing and synapsis are limited to a relatively small region at one end. Crossing over occurs in this small region, allowing normal meiotic segregation of the X and the Y chromosome, as shown in Figure A.
Figure A: Sry and the pseudoautosomal region. The X and the Y chromosomes in males normally pair and cross over in a region known as the pseudoautosomal region; the two chromosomes do not synapse over their entire length as homologs do. This is shown at the top. The outcome is that sperm with an X chromosome have a small fragment of the Y chromosome, but typically this region does not have any genes in it. The Sry gene is adjacent to this region but is not usually involved in the crossover. In the Sry mutant strains of mice, the pseudoautosomal region is slightly extended and includes the Sry locus, which gives rise to the mutant phenotype. Thus, some sperm have an X chromosome with a fragment of the Y, and this fragment that is translocated includes Sry. When such a sperm fertilizes an ovum, the embryo will be XX but will have the Sry gene on its paternal X chromosome; these embryos are sex-reversed and develop as males, as shown in the lower panel.
Sequences from the X chromosome in this region are crossed over, or translocated, to the Y chromosome, and sequences from the Y chromosome are crossed over to the X chromosome. Since there are few or no genes in this region (for most mammals at least), chiasmata are observed, but recombination is detected only from analysis of the sequences, and there are few or no functional consequences. This region that crosses over between the X and the Y chromosome is called the pseudoautosomal region, or more precisely the pseudoautosomal region 1 (PAR1), since two other regions of the X and Y also appear to cross over.
The precise boundaries of the PAR1 are slightly variable in different laboratory strains of mice, and probably in nature as well. Thus, while no genes are affected by the crossovers most of the time, a gene that lies near the PAR1 boundary on the Y chromosome will occasionally be crossed over or translocated onto the X chromosome in a few sperm cells of some mouse strains. This sperm with a bit of the Y can then produce a mouse that is XX, but in which one of the X chromosomes has a small translocation of part of the Y chromosome, as shown in Figure A. The mouse is, in effect, XXY, but the Y is only a small fragment of the total Y chromosome.
Recall that, in mammals, XXY develops as a male. Thus, it was recognized that, in some strains and some mutant lines, there were XX mice that were sex-reversed because a male-determining fragment has been moved onto the X chromosome; this locus was named Sry. These Sry sex-reversed XX mice all had the same fragment of the Y chromosome, so somewhere in this fragment lay the gene encoding TDF.
Meanwhile, studies in humans were identifying the same small region. Translocations between the X and the Y chromosome with functional consequences are fairly uncommon, but such individuals are males with greatly reduced fertility because of the failure of the gonad to develop. Some of these males visited fertility clinics, and the region of the Y chromosome that had been moved to the X could be mapped. Again, this was identified as a small region near PAR1, the same region being identified in mice by the Sry translocations. Furthermore, a few of the women who came to fertility clinics were identified to be XY, but with a deletion of part of their Y chromosome; they also had a failure of the gonad to develop. The deletion in these XY women must be removing the gene encoding TDF.
Both the translocations in XX males and the deletions in XY females were of varying sizes, many of them encompassing many millions of base pairs of the Y chromosome, so the key was to find the smallest region that was common to both and was also found in Sry XX mice. It also turned out to be important that the alteration affected only this region, and not some other region as well; one false lead had a small change here and a larger and more easily detected deletion elsewhere on the Y chromosome. This region included only two or three protein-coding genes, one of which was Sry, but any of these could have been the gene for TDF.
The proof that the Sry gene did encode TDF came from several lines of evidence. First, the gene is transcribed in the embryonic gonad during the time that testis determination occurs, and it is transcribed in the appropriate cells of the gonad. The other candidate genes did not have this precise pattern of transcription. Second, the Sry gene was conserved among placental mammals and always Y-linked; this also was not true for all other candidate genes. Third, each of the 14 different women who were chromosomally XY with a failure in gonad development had a mutation in this same gene. Fourth, and most compelling, the Sry gene from mice was cloned, and the DNA sequence of only this gene, and no other, was inserted into mouse embryos. Embryos that were XX but had only this additional gene developed testes and male sexual differentiation. This then proved that Sry encodes the TDF.
Sry and testis determination
What then is the molecular function of Sry, and how does the TDF actually trigger testis formation? Sry encodes a transcription factor of the SOX family; the family is named for the mechanism by which it binds to the DNA, as discussed in Chapters 2 and 12, and Sry was the member by which the family was named. (SOX is a contraction of “Sry-like box.”) The Sry protein forms a heterodimer with another transcription factor called SF1 (for steroidigenic factor 1, since it regulates the expression of genes needed for steroid production, including Sry). This Sry/SF1 dimer regulates the expression of several other genes, most notably the gene SOX9, which itself encodes a transcription factor of the SOX type. It is the SOX9 protein—as a dimer in combination with SF1—that directly regulates most of the differentiation of the gonad into the testis. (This includes the DMRT genes described in Box 7.5.) Thus, Sry sets off a cascade of transcription factors, which then activate the expression of genes involved in testis formation, testosterone production, and so on.
Interestingly, the sequence of the Sry protein is not highly conserved among mammals. The main regions of conservation are found in the regions that bind to DNA and regulate the expression of the SOX9 gene; the SOX9 gene is more highly conserved. This lack of conservation suggests that the Sry gene is needed primarily for one function—to turn on the expression of SOX9.
FIND OUT MORE
Kashimada, K. and Koopman, P. (2010) Sry: the master switch in mammalian sex determination. Development 137: 3921–30