The role of salivary androgen-binding protein in reproductive isolation between two subspecies of house mouse: Mus musculus musculus and Mus musculus domesticus

Previous behavioural studies using inbred lines have suggested that the gene ( Abpa ) for the alpha subunit of salivary androgen-binding protein (ABP) plays a role in prezygotic isolation between house mouse Mus musculus subspecies. We tested this hypothesis in animals from wild allopatric (121 individuals from four samples) and parapatric (320 animals from 15 samples) populations sampled on the Czech–Bavarian transect across the hybrid zone between M. m. domesticus and M. m. musculus . The study did not reveal a consistent statistically significant trend of homosubspecific preferences in individual allopatric and parapatric populations. Nonetheless, the whole pattern of preference was skewed toward homosubspecific preference mostly on the M. m. musculus side of the hybrid zone. The pattern of homosubspecific preferences was stronger for the time spent sniffing than it was for the first choice of the signal (the ratio of homosubspecific vs. heterosubspecific preferences for both sexes was 6 : 2 in allopatric and 21 : 9 in parapatric populations, while the same rates were 4 : 4 and 16 : 14 for the first choice). To the extent that Y-maze tests reflect preference under wild conditions, we suggest that this slight preference may not in itself be sufficient to impede gene flow between the two subspecies and thus act as a reproductive barrier. ABP most probably participates in a complex system of subspecies-specific recognition in the hybrid zone, but the picture is far too complex at this time to allow a conclusive evaluation of the importance of this role.  © 2005 The Linnean Society of London, Biological Journal of the Linnean Society , 2005, 84 , 349–361.     

Salivary androgen-binding protein variation in Mus and other rodents.

We have searched for genetic variation in the expression of salivary androgen-binding protein (ABP) in a wide variety of mice and other rodents. ABP was present in the salivas of mice of all species and subspecies studied. Genetic studies have identified three common variants of the ABP Alpha subunit (Abpaa, Abpab, and Abpac) in Mus musculus populations with distributions that correspond roughly to those of the subspecies studied (domesticus, musculus, and castaneus, respectively). It appears that the ABP a and b polymorphisms conform to the hybrid zone between the domesticus and musculus subspecies characterized by others. Our studies suggest that the presence of Abpab in inbred strains may be due to a M. m. musculus contribution, perhaps via oriental fancy mice bred to European mice in the early lines leading to the common inbred strains. The relatively common occurrence of the ABP a type in other Mus species leads us to conclude that it is the ancestral type in mice. Further, the observation of what amounts to unique alleles in the three different subspecies indicates that microevolution of the protein has occurred. In a broader survey, ABP was also found in the salivas of Murid and Cricetid rodents generally. These findings suggest that ABP has an important functional role in rodent salivas.     

SALIVARY ANDROGEN-BINDING PROTEIN (ABP) MEDIATES SEXUAL ISOLATION IN MUS MUSCULUS

We wanted to determine whether the microevolution of the mouse salivary androgen-binding protein (ABP) Alpha subunit gene (Abpa) could mediate sexual selection and thereby have a potential role in maintaining gene pool integrity where radiating mouse subspecies make secondary contact. This hypothesis is based upon previous work in this laboratory, which has shown that each subspecies apparently has its own allele and that these alleles have a 25-fold excess of nonsynonymous/synonymous base substitutions compared to an average protein under purifying selection. We provide direct evidence for ABP-assortative mate selection in a laboratory setting: Mus musculus domesticus and M. m. musculus female mice recognize and discriminate between the territories of male mice that essentially differ solely in their Abpa genotype and, when the males are present, the female prefers to mate with the one of her own ABP type. The observation that females could differentiate between the territories of the two males when those mice were absent suggests that the males marked their territories with ABP. In this study, we also detected ABP on the pelts of male mice and in their environment. It is likely that the animals apply the protein to their pelts by licking and that it is then deposited in their surroundings. We suggest that females of the two subspecies are able to discriminate between males of those subspecies on the basis of this protein molecule. Mouse salivary ABP might present a worthwhile system with which to study a prezygotic isolation mechanism in a mammal.     

Signalling components of the house mouse mate recognition system

Subspecies-specific mate recognition may represent significant barrier to gene flow between diverged genomes potentially leading to speciation. In the house mouse, assortative mating involves the coevolution of several signals and receptors. We compared signalling ability of bedding material, faeces, urine, saliva, salivary androgen binding proteins (ABP) and combinations of urine with saliva and urine with ABP in mate choice in two wild-derived inbred strains (one of Mus musculus musculus and one of Mus musculus domesticus origin). We observed high levels of variation in assortative preferences between the two strains and sexes. The strongest preferences were observed in M. m. musculus-derived individuals in tests where urine was present either alone or as part of a composite signal target. M. m. domesticus-derived mice displayed strain-specific preferences for faeces. Saliva was the least preferred stimulus in both strains and sexes. No effect of two-compound cues was detected. We conclude that there is divergence across both the stimulus and preference parts of the recognition system for both house mouse strains. Of the tested stimuli, those that have the capacity to carry a signal for extended periods under natural conditions (such as urine and faeces) seem to be the most important substances in strain-specific recognition.     

Congenic strain analysis reveals genes that are rapidly evolving components of a prezygotic isolation mechanism mediating incipient reinforcement

Two decades ago, we developed a congenic strain of Mus musculus, called b-congenic, by replacing the androgen-binding protein Abpa27(a) allele in the C3H/HeJ genome with the Abpa27(b) allele from DBA/2J. We and other researchers used this b-congenic strain and its C3H counterpart, the a-congenic strain, to test the hypothesis that, given the choice between signals from two strains with different a27 alleles on the same genetic background, test subjects would prefer the homosubspecific one. It was our purpose in undertaking this study to characterize the segment transferred from DBA to the C3H background in producing the b-congenic strain on which a role for ABPA27 in behavior has been predicated. We determined the size of the chromosome 7 segment transferred from DBA and the genes it contains that might influence preference. We found that the "functional" DBA segment is about 1% the size of the mouse haploid genome and contains at least 29 genes expressed in salivary glands, however, only three of these encode proteins identified in the mouse salivary proteome. At least two of the three genes Abpa27, Abpbg26 and Abpbg27 encoding the subunits of androgen-binding protein ABP dimers evolved under positive selection and the third one may have also. In the sense that they are subunits of the same two functional entities, the ABP dimers, we propose that their evolutionary histories might not be independent of each other.     

Characterization of two forms of mouse salivary androgen-binding protein (ABP): implications for evolutionary relationships and ligand-binding function

Mouse salivary androgen-binding protein (ABP) is a member of the secretoglobin family produced in the submaxillary glands of house mice (Mus musculus). We report the cDNA sequences and amino acid sequences of the beta and gamma subunits of ABP from a mouse cDNA library, identifying the two subunits by their pIs and molecular weights. An anomalously high molecular weight of the alpha subunit is likely due to glycosylation at a single site. A phylogenetic comparison of the three subunits of ABP with the chains of other mammalian secretoglobins shows that ABP is most closely related to mouse lachrymal protein and to the major cat allergen Fel dI. An evaluation of the most conserved residues in ABP and the other secretoglobins, in light of structural data reported by others [Callebaut, I., Poupon, A., Bally, R., Demaret, J.-P., Housset, D., Delettre, J., Hossenlopp, P., and Mornon, J.-P. (2000) Ann. N.Y. Acad. Sci. 923, 90-112; Pattabiraman, N., Matthews, J., Ward, K., Mantile-Selvaggi, G., Miele, L., and Mukherjee, A. (2000) Ann. N.Y. Acad. Sci. 923, 113-127], allows us to draw conclusions about the critical residues important in ligand binding by the two different ABP dimers and to assess the importance of ligand binding in the function of the molecule. In addition to the cDNAs, which represent those of the musculus subspecies of Mus musculus, we also report the coding regions of the beta and gamma subunit cDNAs from two other mouse inbred strains which represent the other two subspecies: M. musculus domesticus and M. musculus castaneus. The high nonsynonymous/synonymous substitution rate ratios (K(a)/K(s)) for both the beta and gamma subunits suggest that these two proteins are evolving under strong directional selection, as has been reported for the alpha subunit [Hwang, J., Hofstetter, J., Bonhomme, F., and Karn, R. (1997) J. Hered. 88, 93-97; Karn, R., and Clements, M. (1999) Biochem. Genet. 37, 187-199].     

The complex history of a gene proposed to participate in a sexual isolation mechanism in house mice

Previous behavioral experiments showed that mouse salivary androgen-binding protein (ABP) was involved in interindividual recognition and might play a role in sexual isolation between house mouse (Mus musculus) subspecies. The pattern of evolution of Abpa, the gene for the alpha subunit of ABP, was found to be consistent with this hypothesis. Abpa apparently diverged rapidly between species and subspecies with a large excess of nonsynonymous substitutions, a lack of exon polymorphism within each of the three subspecies, and a lack of intron polymorphism in the one subspecies studied (M. musculus domesticus). Here we characterized the intron and exon sequence variations of this gene in house mouse populations from central Eurasia, a region yet unsampled and thought to be close to the cradle of the radiation of the subspecies. We also determined the intron and exon sequences in seven other species of the genus Mus. We confirmed the general pattern of rapid evolution by essentially nonsynonymous substitutions, both inter- and intraspecifically, supporting the idea that Darwinian selection has driven the evolution of this gene. We also observed a uniform intron sequence in five samples of M. musculus musculus, suggesting that a selective sweep might have occurred for that allele. In contrast to previous results, however, we found extensive intron and exon polymorphism in some house mouse populations from central Eurasia. We also found evidence for secondary admixture of the subspecies-specific alleles in regions of transition between the subspecies in central Eurasia. Furthermore, an abnormal intron phylogeny suggested that interspecific exchanges had occurred between the house mouse subspecies and three other Palearctic species. These observations appear to be at variance with the simple hypothesis that Abpa is involved in reproductive isolation. Although we do not rule out a role in recognition, the situation appears to be more complex than previously thought. Thus the selective mechanism behind the evolution of Abpa remains to be resolved, and we suggest that it may have changed during the recent colonization history of the house mouse.     

Physical linkage and mate preference generate linkage disequilibrium for behavioral isolation in two parapatric crickets

Behavioral isolation is a potent barrier to gene flow and a source of striking diversity in the animal kingdom. However, it remains unclear if the linkage disequilibrium (LD) between sex‐specific traits required for behavioral isolation results mostly from physical linkage between signal and preference loci or from directional mate preferences. Here, we test this in the field crickets Gryllus rubens and G. texensis. These closely related species diverged with gene flow and have strongly differentiated songs and preference functions for the mate calling song rhythm. We map quantitative trait loci for signal and preference traits (pQTL) as well as for gene expression associated with these traits (eQTL). We find strong, positive genetic covariance between song traits and between song and preference. Our results show that this is in part explained by incomplete physical linkage: although both linked pQTL and eQTL couple male and female traits, major effect loci for different traits were never on the same chromosome. We suggest that the finely tuned, highly divergent preference functions are likely an additional source of LD between male and female traits in this system. Furthermore, pleiotropy of gene expression presents an underappreciated mechanism to link sexually dimorphic phenotypes.     

Reinforcement selection acting on the European house mouse hybrid zone

Behavioural isolation may lead to complete speciation when partial postzygotic isolation acts in the presence of divergent‐specific mate‐recognition systems. These conditions exist where Mus musculus musculus and M. m. domesticus come into contact and hybridize. We studied two mate‐recognition signal systems, based on urinary and salivary proteins, across a Central European portion of the mouse hybrid zone. Introgression of the genomic regions responsible for these signals: the major urinary proteins (MUPs) and androgen binding proteins (ABPs), respectively, was compared to introgression at loci assumed to be nearly neutral and those under selection against hybridization. The preference of individuals taken from across the zone regarding these signals was measured in Y mazes, and we develop a model for the analysis of the transition of such traits under reinforcement selection. The strongest assortative preferences were found in males for urine and females for ABP. Clinal analyses confirm nearly neutral introgression of an Abp locus and two loci closely linked to the Abp gene cluster, whereas two markers flanking the Mup gene region reveal unexpected introgression. Geographic change in the preference traits matches our reinforcement selection model significantly better than standard cline models. Our study confirms that behavioural barriers are important components of reproductive isolation between the house mouse subspecies.     

Female guppies agree to differ: phenotypic and genetic variation in mate-choice behavior and the consequences for sexual selection

Variation among females in mate choice may influence evolution by sexual selection. The genetic basis of this variation is of interest because the elaboration of mating preferences requires additive genetic variation in these traits. Here we measure the repeatability and heritability of two components of female choosiness (responsiveness and discrimination) and of female preference functions for the multiple ornaments borne by male guppies (Poecilia reticulata). We show that there is significant repeatable variation in both components of choosiness and in some preference functions but not in others. There appear to be several male ornaments that females find uniformly attractive and others for which females differ in preference. One consequence is that there is no universally attractive male phenotype. Only responsiveness shows significant additive genetic variation. Variation in responsiveness appears to mask variation in discrimination and some preference functions and may be the most biologically relevant source of phenotypic and genetic variation in mate-choice behavior. To test the potential evolutionary importance of the phenotypic variation in mate choice that we report, we estimated the opportunity for and the intensity of sexual selection under models of mate choice that excluded and that incorporated individual female variation. We then compared these estimates with estimates based on measured mating success. Incorporating individual variation in mate choice generally did not predict the outcome of sexual selection any better than models that ignored such variation.     

Phylogeographic origin of Hokkaido house mice (Mus musculus) as indicated by genetic markers with maternal, paternal and biparental inheritance

We examined intraspecies genetic variation in house mice (Mus musculus molossinus) from the northern third of the Japanese Islands, in order to obtain evidence of the history of mouse colonization that might have shaped the current genetic diversity. We extended the previous sampling of mitochondrial cytochrome b sequence and added information from the Y-linked Sry gene and ribosomal RNA gene surveys. We distinguish mitochondrial haplotypes characteristic of the North Asian musculus subspecies group (involving M. m. musculus and M. m. molossinus) as 'MUS', and that of the Southeast Asian castaneus subspecies group as 'CAS' (although the mice resemble MUS morphologically). There was a clear geographic partition of MUS and CAS types into southern and northern Hokkaido, respectively. Conversely, on Tohoku, the MUS and CAS types were interspersed without clear geographic subdivision. In contrast to the mtDNA data, all Hokkaido and Tohoku mice examined were found to possess a unique type for the Y-linked Sry gene, specific to Korea and Japan. Restriction site analysis of nuclear rDNA probe showed a consistent distribution of MUS and CAS types, as major and minor components, respectively, in the Hokkaido and Tohoku mice. These data support the previous notion that the Hokkaido and Tohoku mice experienced genetic hybridization between primary residents of CAS origin and MUS newcomers arriving via a southern route. The invasion of the MUS type could correspond with the evidence for arrival of prehistoric peoples. There are, however, alternative interpretations, including genetic admixture between MUS arriving by a southern route and CAS from a northern route.     

Mosaic genealogy of the Mus musculus genome revealed by 21 nuclear genes from its three subspecies

Patterns of genetic variation provide insight into the evolutionary history of a species. Mouse (Mus musculus) is a good model for this purpose. Here we present the analysis of genealogies of the 21 nuclear loci and one mitochondrial DNA region in M. musculus based on our nucleotide sequences of nine inbred strains from three M. musculus subspecies (musculus, domesticus, and castaneus) and one M. spicilegus strain as an outgroup. The mitochondrial DNA gene genealogy of those strains confirmed the introgression pattern of one musculus strain. When all the nuclear DNA data were concatenated to produce a phylogenetic tree of nine strains, musculus and domesticus strains formed monophyletic clusters with each other, while the two castaneus strains were paraphyletic. When each DNA region was treated independently, the phylogenetic networks revealed an unnegligibly high level of subspecies admixture and the mosaic nature of their genome. Estimation of ancestral and derived population sizes and migration rates suggests the effects of ancestral polymorphism and gene flow on the pattern of genetic variation of the current subspecies. Gene genealogies of Fut4 and Dfy loci also suggested existence of the gene flow between M. musculus and M. spicilegus or other distant species.     

On the logic of Fisherian sexual selection*

In Fisher's model of sexual selection, a female preference for a male trait spreads together with the trait because their genetic bases become correlated. This can be interpreted as a “greenbeard” system: a preference gene, by inducing a female to mate with a trait-bearing male, favors itself because the male is disproportionately likely also to carry the preference gene. Here, we use this logic to argue that Fisherian sexual selection in diploids proceeds via two channels: (i) trait-bearing males are disproportionately the product of matings between preference-bearing mothers and trait-bearing fathers, and thus trait and preference genes are correlated “in trans”; (ii) trait and preference genes come into gametic phase disequilibrium, and thus are correlated “in cis.” Gametic phase disequilibrium is generated by three distinct mechanisms that we identify. The trans channel does not operate when sexual selection is restricted to the haploid phase, and therefore represents a fundamental difference between haploid and diploid models of sexual selection. We show that the cis and trans channels contribute equally to the spread of the preference when recombination between the preference and trait loci is free, but that the trans channel is substantially more important when linkage is tight.     

Mate Choice in Mus musculus Is Relative and Dependent on the Estrous State

Mate choice is a critical behavioral decision process with profound impact on evolution. However, the mechanistic basis of mate choice is poorly understood. In this study we focused on assortative mate choice, which is known to contribute to the reproductive isolation of the two European subspecies of house mouse, Mus musculus musculus and Mus musculus domesticus. To understand the decision process, we developed both full mating and limited-contact paradigms and tested musculus females'' preference for musculus versus domesticus males, mimicking the natural musculus/domesticus contact zone. As hypothesized, when allowed to mate we found that sexually receptive musculus females exhibited a robust preference to mate with musculus males. In contrast, when non-receptive, females did not exhibit a preference and rather alternated between males in response to male mount attempts. Moreover in a no-choice condition, females mated readily with males from both subspecies. Finally, when no physical contact was allowed, and therefore male''s behavior could not influence female''s behavior, female''s preference for its own subspecies was maintained independently of the estrous state. Together, our results suggest that the assortative preference is relative and based on a comparison of the options available rather than on an absolute preference. The results of the limited-contact experiments highlight the interplay between female''s internal state and the nature of the interaction with prospective mates in the full mating conditions. With these experiments we believe we established an assortative mate preference assay that is appropriate for the investigation of its underlying substrates.     

The Genes for Mouse Salivary Androgen-Binding Protein (Abp) Subunits Alpha and Gamma Are Located on Chromosome 7

We demonstrate that the previously described gene Androgen binding protein (Abp; Dlouhy and Karn, 1984) codes for the Alpha subunit of ABP and rename the locus Androgen binding protein alpha (Abpa). A study of recombinant inbred strains demonstrates that Abpa is located on chromosome 7 near Glucose phosphate isomerase-1 (Gpi-1). Biochemical and genetic evidence indicates the existence of another ABP subunit, Gamma, and its locus, Androgen binding protein gamma (Abpg), that is closely linked to Abpa. Although no polymorphism has yet been found for the previously described Beta subunit of ABP (Dlouhy and Karn, 1983; 1984), we suggest that it represents a third locus. Androgen binding protein beta (Abpb). ABP subunits appear to dimerize randomly and a model is presented demonstrating the origin of six ABP dimers in the salivas of AbpaaAbpga/AbpabAbpgb heterozygous mice. The results of cell-free translation studies in which the pre-ABP subunits are identified specifically by immunoprecipitation with anti-ABP antibody supports the idea that independent mRNAs code for the Alpha, Beta and Gamma subunits.     

The genes for SHBG/ABP and the SHBG-like region of vitamin K-dependent protein S have evolved from a common ancestral gene

Sex hormone binding globulin (SHBG) is the most important sex steroid transport protein in human plasma. It is the product of the same single gene as the androgen binding protein (ABP) of testis. Protein S is another protein, which is an important cofactor in the anticoagulation system and, as far as is known today, functionally unrelated to SHBG/ABP. Protein S also has a role in the complement system. A comparison of the human genes for SHBG/ABP and protein S reveals a sequence similarity, which is of a low grade only, between the SHBG/ABP protein and a similar sized COOH-terminal domain of protein S. However, the intron-exon organization exhibits a striking similarity in the two genes, illustrating evolutionary events leading to the appearance of two functionally different proteins from common ancestral genetic elements.     

Novel cytochrome P450, cyp6a17, is required for temperature preference behavior in Drosophila

Perception of temperature is an important brain function for organisms to survive. Evidence suggests that temperature preference behavior (TPB) in Drosophila melanogaster, one of poikilothermal animals, is regulated by cAMP-dependent protein kinase (PKA) signaling in mushroom bodies of the brain. However, downstream targets for the PKA signaling in this behavior have not been identified. From a genome-wide search for the genes regulated by PKA activity in the mushroom bodies, we identified the cyp6a17 Cytochrome P450 gene as a new target for PKA. Our detailed analysis of mutants by genetic, molecular and behavioral assays shows that cyp6a17 is essential for temperature preference behavior. cyp6a17 expression is enriched in the mushroom bodies of the adult brain. Tissue-specific knockdown and rescue experiments demonstrate that cyp6a17 is required in the mushroom bodies for normal temperature preference behavior. This is the first study, to our knowledge, to show PKA-dependent expression of a cytochrome P450 gene in the mushroom bodies and its role as a key factor for temperature preference behavior. Taken together, this study reveals a new PKA-Cytochrome P450 pathway that regulates the temperature preference behavior.     

Asymmetrical reproductive character displacement in the house mouse

Our study addressed reproductive character displacement between two subspecies of the house mouse, Mus musculus musculus and Mus musculus domesticus that hybridize in Europe along a zone where selection against hybridization is known to occur. Based on a multi-population approach, we investigated spatial patterns of divergence of mate preference in the two taxa. Mate preference was significantly higher in the contact zone than in allopatry in both subspecies, suggesting that reproductive character displacement occurs. Moreover, patterns of preference were stronger in M. m. musculus than in M. m. domesticus, indicating an asymmetrical divergence between the two. In the context of selection against hybridization, our results may provide empirical support for the hypothesis of reinforcement in a parapatric hybrid zone. We discuss factors that could explain the asymmetrical pattern of divergence and the possible impact of a unimodal structure on the maintenance of premating divergence between the two subspecies.     

The causes of variation in the presence of genetic covariance between sexual traits and preferences

Mating traits and mate preferences often show patterns of tight correspondence across populations and species. These patterns of apparent coevolution may result from a genetic association between traits and preferences (i.e. trait–preference genetic covariance). We review the literature on trait–preference covariance to determine its prevalence and potential biological relevance. Of the 43 studies we identified, a surprising 63% detected covariance. We test multiple hypotheses for factors that may influence the likelihood of detecting this covariance. The main predictor was the presence of genetic variation in mate preferences, which is one of the three main conditions required for the establishment of covariance. In fact, 89% of the nine studies where heritability of preference was high detected covariance. Variables pertaining to the experimental methods and type of traits involved in different studies did not greatly influence the detection of trait–preference covariance. Trait–preference genetic covariance appears to be widespread and therefore represents an important and currently underappreciated factor in the coevolution of traits and preferences.     

Genotype-by-environment interactions for female preference

Sexual selection is responsible for many of the most spectacular displays in nature, and female preference for certain males is central to much of this. However, female preference is relatively poorly understood, particularly the relative importance of a female's genes, the environment and their interaction on her preference. We investigated preference in a no-choice design using Drosophila melanogaster iso-female lines and find that there are genotype-by-environment interactions for female preference. Whereas the choosiness of some female genotypes differed little across environments, that of others differed greatly, so that the choosiness rank of females in one environment did not necessarily predict their rank in another. Furthermore, the genetic variance underlying preference also varied across environments. These findings have important consequences for the evolution of female preference and the male sexual traits preference targets.