Genetic Links between Recombination and Speciation

Meiosis distinguishes cells in the germ line from their somatic counterparts. Part of meiosis involves a bizarre series of events in which cells deliberately damage their DNA and then repair it. Of the many double-strand breaks that are generated, a minority are resolved as reciprocal exchanges between homologous chromosomes [1]. In a wide variety of species, recombination diversifies genomes and is required for the formation of viable gametes. Although the functional role of crossing over should impose strong selective constraints, the rate of recombination varies among individuals. Recombination rate shows resemblance among relatives [2], differs among inbred strains raised in a common environment [3], and responds to artificial selection [4], demonstrating a genetic component to individual rate differences that remains mostly unexplained. A handful of genes responsible for variation in recombination rate have been identified, including modifiers of the total number of crossovers [5–8] and loci that determine the genomic placement of recombination events [9]. 
 
As a fundamental part of gametogenesis, meiosis also plays a role in speciation, the means by which one species becomes two. A common observation is that otherwise viable hybrids between genetically differentiated lineages suffer from reduced fertility up to the point of complete sterility—an important reproductive barrier that keeps species distinct [10]. One stage during which problems appear is meiosis, when hybrid gametogenesis sometimes arrests. According to an influential model, hybrid sterility arises from disrupted interactions between different genes that evolved in separate populations [11,12]. The search for incompatibility genes that cause hybrid sterility has so far yielded a short list [13–15]. 
 
The fact that only a small number of genes are known to control variation in recombination rate or to contribute to hybrid sterility makes the discovery that one gene mediates both traits especially noteworthy. Prdm9 encodes a histone methyltransferase that modifies chromatin and is essential for meiosis in mice [16]. Decades of persistent effort from the group of Jiri Forejt demonstrated that incompatibility between Prdm9 and other loci causes sterility in F1 hybrid males formed by crossing two subspecies of house mice [17]. Subsequently, genetic mapping in crosses, population genetic analysis, bioinformatic prediction of DNA binding sites, and association studies jointly identified Prdm9 as a primary determinant of recombination hotspot location (short stretches of sequence within which most crossovers occur) in mice and humans [9,18–20]. The dual roles of Prdm9 reveal a tantalizing link between recombination and speciation at the genetic level and motivate the search for other genes that affect both processes. 
 
Now, progress from Jiri Forejt’s lab provides fresh evidence for a genetic connection between recombination and hybrid sterility. Balcova et al. (2016) [21] profile the genome-wide recombination rate by visualizing the immunolocalization pattern of the MLH1 mismatch repair protein that resolves double-strand breaks into crossovers (Fig 1). This powerful approach enables the total number of recombination events to be counted in individual spermatocytes. The authors first use the MLH1 technique to confirm that the genome of an inbred strain representing the house mouse subspecies Mus musculus musculus experiences an average of 4.7 more crossovers than the genome of an inbred strain descended primarily from M. m. domesticus (a 19% increase). Next, Balcova et al. (2016) measure recombination rates in a panel of inbred strains, each of which harbors single chromosomes from the M. m. musculus strain on the genomic background of the M. m. domesticus strain. By comparing rates in chromosome substitution strains with the appropriate parental strain, the authors identify three chromosomes that affect the genome-wide crossover number. The locus with the largest phenotypic effect is located on the X chromosome. Rate differences among substitution strains carrying pieces of the chromosome further localize the position of this modifier of the global recombination rate to a 4.7 Mb interval (nicknamed Meir1 by the authors). This locus lies within a broader part of the X chromosome previously shown to modulate recombination in crosses involving other subspecies of mice [22,23]. 
 
 
 
Fig 1 
 
Image illustrating immunofluorescent cytology approach to measuring the genome-wide recombination rate (from Balcova et al., 2016). 
 
 
 
Several lines of evidence suggest that this region of the X chromosome genetically links recombination and speciation. First, the same interval is associated with multiple correlates of hybrid sterility. Small testes, low sperm count, and morphological abnormalities in sperm observed in F1 males from crosses between M. m. musculus mothers and M. m. domesticus fathers all map to this genomic location [24,25]. Second, hybrid male sterility due to one of the loci in this X-linked region (Hstx2) involves an incompatibility with Prdm9 [25]. Finally, the effects of this piece of the X chromosome on recombination and sterility show similar properties, including male specificity. 
 
The findings of Balcova et al. (2016), along with previous discoveries about Prdm9, raise the intriguing possibility that recombination and speciation are mechanistically coupled. Remarkably, both the fine-scale placement of crossovers (mediated by Prdm9) and the global recombination rate (controlled by Meir1) seem to be associated with hybrid sterility. The authors propose that the incompatibility between Prdm9 and Hstx2 could inhibit the repair of double-strand breaks, which could in turn lead to the disrupted synapsis of chromosomes and meiotic arrest observed in sterile F1 males [26,27]. Regardless of the mechanism, the larger implication is that evolutionary divergence in recombination genes can generate a frequently observed form of reproductive isolation. 
 
Once the causative gene(s) and mutation(s) corresponding to Meir1 and Hstx2 are identified in this system, important remaining questions can be answered. Do the joint effects on recombination and hybrid sterility reflect the pleiotropic activities of a single gene or multiple linked genes? What are the molecular, cellular, and developmental bases of these phenotypes? What is the evolutionary history of the causative mutations? Did natural selection drive phenotypic divergence, as seems to be the case for the rapidly evolving Prdm9 [28]? 
 
The results from Balcova et al. (2016) should also motivate similar studies in other organisms. The observations that recombination rate evolves and that hybrid sterility is a key reproductive barrier are phylogenetically widespread, leaving no obvious reason that the emerging connection between recombination and speciation should be restricted to mice.

Meiosis distinguishes cells in the germ line from their somatic counterparts. Part of meiosis involves a bizarre series of events in which cells deliberately damage their DNA and then repair it. Of the many double-strand breaks that are generated, a minority are resolved as reciprocal exchanges between homologous chromosomes [1]. In a wide variety of species, recombination diversifies genomes and is required for the formation of viable gametes. Although the functional role of crossing over should impose strong selective constraints, the rate of recombination varies among individuals. Recombination rate shows resemblance among relatives [2], differs among inbred strains raised in a common environment [3], and responds to artificial selection [4], demonstrating a genetic component to individual rate differences that remains mostly unexplained. A handful of genes responsible for variation in recombination rate have been identified, including modifiers of the total number of crossovers [5][6][7][8] and loci that determine the genomic placement of recombination events [9].
As a fundamental part of gametogenesis, meiosis also plays a role in speciation, the means by which one species becomes two. A common observation is that otherwise viable hybrids between genetically differentiated lineages suffer from reduced fertility up to the point of complete sterility-an important reproductive barrier that keeps species distinct [10]. One stage during which problems appear is meiosis, when hybrid gametogenesis sometimes arrests. According to an influential model, hybrid sterility arises from disrupted interactions between different genes that evolved in separate populations [11,12]. The search for incompatibility genes that cause hybrid sterility has so far yielded a short list [13][14][15].
The fact that only a small number of genes are known to control variation in recombination rate or to contribute to hybrid sterility makes the discovery that one gene mediates both traits especially noteworthy. Prdm9 encodes a histone methyltransferase that modifies chromatin and is essential for meiosis in mice [16]. Decades of persistent effort from the group of Jiri Forejt demonstrated that incompatibility between Prdm9 and other loci causes sterility in F 1 hybrid males formed by crossing two subspecies of house mice [17]. Subsequently, genetic mapping in crosses, population genetic analysis, bioinformatic prediction of DNA binding sites, and association studies jointly identified Prdm9 as a primary determinant of recombination hotspot location (short stretches of sequence within which most crossovers occur) in mice and humans [9,[18][19][20]. The dual roles of Prdm9 reveal a tantalizing link between recombination and speciation at the genetic level and motivate the search for other genes that affect both processes. Now, progress from Jiri Forejt's lab provides fresh evidence for a genetic connection between recombination and hybrid sterility. Balcova et al. (2016) [21] profile the genome-wide recombination rate by visualizing the immunolocalization pattern of the MLH1 mismatch repair protein that resolves double-strand breaks into crossovers (Fig 1). This powerful approach enables the total number of recombination events to be counted in individual spermatocytes. The authors first use the MLH1 technique to confirm that the genome of an inbred strain representing the house mouse subspecies Mus musculus musculus experiences an average of 4.7 more crossovers than the genome of an inbred strain descended primarily from M. m. domesticus (a 19% increase). Next, Balcova et al. (2016) measure recombination rates in a panel of inbred strains, each of which harbors single chromosomes from the M. m. musculus strain on the genomic background of the M. m. domesticus strain. By comparing rates in chromosome substitution strains with the appropriate parental strain, the authors identify three chromosomes that affect the genome-wide crossover number. The locus with the largest phenotypic effect is located on the X chromosome. Rate differences among substitution strains carrying pieces of the chromosome further localize the position of this modifier of the global recombination rate to a 4.7 Mb interval (nicknamed Meir1 by the authors). This locus lies within a broader part of the X chromosome previously shown to modulate recombination in crosses involving other subspecies of mice [22,23].
Several lines of evidence suggest that this region of the X chromosome genetically links recombination and speciation. First, the same interval is associated with multiple correlates of hybrid sterility. Small testes, low sperm count, and morphological abnormalities in sperm observed in F 1 males from crosses between M. m. musculus mothers and M. m. domesticus fathers all map to this genomic location [24,25]. Second, hybrid male sterility due to one of the loci in this X-linked region (Hstx2) involves an incompatibility with Prdm9 [25]. Finally, the effects of this piece of the X chromosome on recombination and sterility show similar properties, including male specificity.
The findings of Balcova et al. (2016), along with previous discoveries about Prdm9, raise the intriguing possibility that recombination and speciation are mechanistically coupled. Remarkably, both the fine-scale placement of crossovers (mediated by Prdm9) and the global recombination rate (controlled by Meir1) seem to be associated with hybrid sterility. The authors propose that the incompatibility between Prdm9 and Hstx2 could inhibit the repair of doublestrand breaks, which could in turn lead to the disrupted synapsis of chromosomes and meiotic arrest observed in sterile F 1 males [26,27]. Regardless of the mechanism, the larger implication is that evolutionary divergence in recombination genes can generate a frequently observed form of reproductive isolation.
Once the causative gene(s) and mutation(s) corresponding to Meir1 and Hstx2 are identified in this system, important remaining questions can be answered. Do the joint effects on recombination and hybrid sterility reflect the pleiotropic activities of a single gene or multiple linked genes? What are the molecular, cellular, and developmental bases of these phenotypes? What is the evolutionary history of the causative mutations? Did natural selection drive phenotypic divergence, as seems to be the case for the rapidly evolving Prdm9 [28]?
The results from Balcova et al. (2016) should also motivate similar studies in other organisms. The observations that recombination rate evolves and that hybrid sterility is a key reproductive barrier are phylogenetically widespread, leaving no obvious reason that the emerging connection between recombination and speciation should be restricted to mice.