The genetics of thermosensitive sex determination

Temperature-dependent sex determination, TSD for short, was first discovered in reptiles and has been studied for decades. However, the mechanism underlying TSD has been remarkably hard to define. Recent work has identified a new TSD gene and lays out a powerful approach for finding additional thermosensors. The concept of sex chromosomes and sex determination in mammals is a familiar one: individuals with 2 X chromosomes develop into females, while those with an X and a Y chromosome develop into males. In stark contrast, temperature determines sex in many reptiles. This prompts questions about how and why something as stochastic as climate and temperature influences the decision to become female or male. While we have a solid understanding of the evolutionary significance of temperature-dependent sex determination (TSD), the mechanistic basis of TSD has remained an enigma. Recent work by Schroeder et al. provides novel insight into how temperature determines sex and outlines a new approach for identifying more thermosensitive sex-determining (TSD) genes.¹

Before discussing this study in detail, it is important to understand the basic process of sexual development. Vertebrate embryos are initially bipotential, which means they are capable of developing either a female or a male phenotype. This also applies to the gonadal primordium, which at first develops in an identical fashion in all embryos regardless of genotype or temperature. The bipotential gonad grows in size as somatic and germ cells proliferate and primary sex cords form. Temperature acts on gonads during this critical period to trigger ovarian or testicular development.² Differentiation of testes entails further development of primary sex cords into testis cords (i.e., seminiferous tubules), while differentiation of ovaries involves thickening of the coelomic epithelium (i.e., ovarian cortex) and disintegration of the primary sex cords. The key point is that sex determination, like other developmental decisions, involves a specification stage when extrinsic and intrinsic factors interact to tip the balance toward one developmental fate or another. Schroeder et al. capitalized on a brief temperature sensitive period in snapping turtles to explore the interplay between genes and the environment during specification of gonad fate.¹

Based on reports of variation in TSD pattern within and among snapping turtle populations,^3,4 we reasoned that differences in thermosensitivity have a genetic basis and that polymorphisms could help distinguish TSD genes from downstream effectors that play a conserved role in development of ovaries and testes and thermosensitive genes not involved in sex determination per se. Variants that alter the function of TSD genes, by definition, will be associated with gonadal phenotype. Polymorphisms could alter coding sequences and affect the way a protein responds to temperature. It is also plausible that regulatory variants could change the way temperature influences gene expression. In contrast, polymorphisms in genes not directly involved in sensing temperature and determining sex would not be associated with phenotype.

To test this idea, we focused on a novel candidate identified in an unbiased screen for genes that are differentially expressed at female- versus male-producing temperatures.⁵ In our new study, we confirmed that cold-inducible RNA binding protein (CIRBP) mRNA is differentially expressed during specification of gonad fate and showed that CIRBP protein is expressed in primary sex cords and the coelomic epithelium of bipotential gonads.¹ We then investigated the relationship between CIRBP expression and sex determination by incubating 9 clutches of eggs in a thermal regime that produces mixed sex ratios. Clutches with higher CIRBP expression produced more females, while clutches with lower CIRBP expression produced more males. Sanger sequencing of CIRBP revealed a single nucleotide polymorphism at codon 63 (c63A > C) that was associated with CIRBP expression and sex ratios.

Two additional experiments showed that expression of allele A is induced by brief exposure to a female-producing temperature, while expression of allele C is not affected by temperature As expected based on temperature-dependent expression of these alleles, AA homozygotes were more likely to develop ovaries than AC heterozygotes. Similarly, AC heterozygotes were more likely to develop ovaries than CC homozygotes. Overall, one copy of the A allele increased the odds of developing ovaries 4-fold, while 2 copies increased the odds ratio 16-fold. In addition to associations between CIRBP and sex at the individual and family level, 2 more studies revealed differences in allele frequency between populations with distinct TSD patterns. The A allele is found at higher frequencies in populations that are more susceptible to feminization by high temperatures.

Figure 1.

A female-producing temperature (31°C) increases expression of the CIRBP A allele, but not the CIRBP C allele in bipotential gonads of snapping turtle embryos. In response to brief exposure to 31°C, AA homozygotes have higher CIRBP expression and are 4x more likely to develop ovaries than are AC heterozygotes and 16x more likely to develop ovaries than are CC homozygotes.

These alleles encode the same amino acid. We therefore searched for other polymorphisms within CIRBP and in 6 genes linked to CIRBP. All variants detected were synonymous or in non-coding regions. Moreover, linked polymorphisms were not associated with sex when tested via multiple logistic regression. Together, these studies strongly suggest that temperature-dependent, allele-specific expression of CIRBP plays a role in transducing temperature into a biological signal for ovary vs. testis development in the snapping turtle. Higher CIRBP expression tips the scale toward ovarian fate, while lower expression tips the scale toward testicular fate.

This is particularly intriguing because CIRBP is in the same family as sex-linked genes in mammals (i.e., X-linked RNA binding motif protein and Y-linked RNA binding motif proteins) and Sex-lethal and Transformer, which are sex-determining genes in fruit flies. We hypothesize that CIRBP could influence translation, splicing, or stability of mRNA from other sex-determining genes. Further research will be required to experimentally manipulate CIRBP expression and assess effects on conserved sex-determining genes and gonad fate. For example, will overexpression of CIRBP at male temperatures induce expression of ovarian genes and ovary development? Vice versa, will testes develop if CIRBP expression is knocked down at female temperatures?

Although CIRBP is the first convincing target for mechanistic studies of TSD, the question of how temperature determines sex is turning out to be more complex than anticipated. We estimated that CIRBP accounts for 25% of the genetic variance in thermosensitivity in the snapping turtle. This means there is not a single TSD gene. Instead, TSD appears to be a complex trait in which 2 or more genes act together as an integrated system to sense temperature and determine gonad fate. Our research also demonstrates for the first time that genetic associations can effectively be used to identify TSD genes in reptiles. This will be a powerful approach to use in future studies of thermal biology.