Sexual dimorphism results from sex-biased gene expression, which evolves when selection acts differently on males and females. While there is an intimate connection between sex-biased gene expression and sex-specific selection, few empirical studies have studied this relationship directly. Here we compare the two on a genome-wide scale in humans and flies. Genes with intermediate degrees of sex-biased expression show evidence of ongoing sex-specific selection, while genes with either little or completely sex-biased expression do not.
This pattern apparently results from differential viability selection in males and females acting in the current generation. The Twin Peaks pattern is also found in Drosophila using a different measure of sex-specific selection acting on fertility. We develop a simple model that successfully recapitulates the Twin Peaks.
Our results suggest that many genes with intermediate sex-biased expression experience ongoing sex-specific selection in humans and flies. Sexual dimorphism, which is evident in virtually all phenotypic traits, results from sex-biased gene expression. The evolution of sex-biased expression, in turn, results from selection that acts differently on males and females. We use genomic data from humans and flies to quantify the relation between ongoing selection pressures and gene expression.
A distinctive pattern in both species reveals that many genes are experiencing selection that is currently acting differentially on males and females. PLoS Genet 12 9 : e This is an zg access article distributed under the terms of the Creative Commons Attribution Licensewhich permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
The funders zzf no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. Females and males differ for virtually all phenotypic sed [ 1 ].
Sexual dimorphism results from sex-biased gene expression, which evolves in response to selection that acts differently on males and females [ 2 — 4 ]. Thus sex-biased gene expression is intimately linked to sex-specific selection [ 56 ].
What is less clear, however, is the extent to which sex-specific selection is ongoing, and how ongoing selection relates to the strength sex sex-biased expression. To date, the most direct link between sf selection and sex-biased expression comes from a laboratory study of Drosophila melanogaster. By comparing gene expression and reproductive fitness aex a quantitative genetics design, Innocenti and Morrow [ 7 ] identified genes experiencing sexually-antagonistic selection, which is the extreme case of sex-specific selection in which an allele that increases fitness in one sex decreases it in the other.
They concluded that 8. While this work is a milestone, it leaves key questions unanswered. These include: What is the relation sxe the strength of ongoing sex-specific selection and sex-biased expression, and how common and how strong is sex-specific selection in natural populations?
Answers to these questions are important to our understanding of how sexual dimorphism evolves. Further, the answers will ses us about general issues including constraints to adaptation [ 8 — 10 ] and how genomes evolve [ 11 — 13 ].
We tackle this problem here with a new method that directly quantifies contemporary sex-specific selection on a genomic scale.
We sed that Mendelian inheritance ensures that allele frequencies at autosomal loci are equal in males and females at conception, and that sex-specific viability selection will generate genetic divergence between the sexes within a generation.
Divergence will occur when selection is sexually-antagonistic, meaning that different alleles are favored in males sex females. We find that genes with intermediate degrees of sex-biased expression experience the strongest sex-specific selection in zzf. We assess the generality of this pattern by reanalyzing the data on flies from Innocenti and Morrow [ 7 ].
We again find the The Twin Peaks pattern, which in this case results from sexually-antagonistic not sexually-concordant selection, and from selection on fertility rather than viability. A simple population-genetic model successfully recapitulates the key pattern. Our results suggest that ongoing sex-specific selection is a common feature of the genome in humans and flies. We quantify sex differences in viability selection using F ST between adult males and females [ 15 ].
We used allele frequencies at over 6 million autosomal single-nucleotide polymorphisms SNPs that appear in the Genomes Project [ 16 ], a database that includes more than individuals from 26 populations worldwide. For each of 17, protein-coding loci, we calculated the average F ST for all SNPs within transcribed regions within each population. We expect this to give a conservative picture of sex differences in selection since most SNPs are not themselves targets of selection.
Patterns that are qualitatively the same as what we report below also emerge if we use only the single SNP from each locus with the largest value of F ST. That is not surprising, however. A power analysis presented in the S1 File shows it is highly unlikely to observe significant divergence at any individual SNP, even with very strong sexually-antagonistic selection, simply because the correction for multiple comparisons is so severe. This measure is zc correlated with the familiar log 2 of the expression ratio.
Sex-biased expression in adults fz results from differences between reproductive tissues [ 1819 ], so we used expression sfx in testis and ovaries.
The data are from 14 males and 6 females [ 20 ]. Genetic divergence between the sexes is greatest for genes whose expression is moderately female-biased or moderately male-biased. The white curve is the best-fit 4 th degree polynomial and the intensity of red indicates the likelihood that the regression passes through a given value.
The numbers of genes with a given bias are visualized in the density plot in the lower part of the figure; dark gray denotes intermediate sex-biased expression. The Twin Peaks pattern is statistically well-supported. We permutated the data 10 5 times and found the pattern occurs 1. The Twin Peaks pattern is also recovered by fitting a cubic spline using generalized additive models. Twin Peaks are seen within each sex the 26 populations when they are analyzed separately.
Genetic divergence between the sexes is quite repeatable across sed taking the SNP with highest F ST at each locus, the intra-class correlation coefficient is 0. It is plausible that gene expression in adult gonads on which our analyses are based is not itself the target of the viability selection that causes the Twin Peaks pattern.
Instead, expression in adult gonads could be correlated with expression in other tissues and other life stages that are the actual targets. Unfortunately, we are not able to repeat our analyses with most somatic tissues because the Gene Expression Atlas only provides data on expression averaged across both sexes. We were however able to analyze expression in five sex-specific somatic tissues: ectocervis, fallopian tubes, vagina, and uterus in females, and prostate in males.
The Twin Peaks pattern appears again when we use the average expression of the four female-specific somatic tissues and expression in testes, but not when we use prostate instead of testes. Expression in prostate may not provide a good proxy for gene expression in males, however, as its expression profile is very similar to that of vagina and other female-specific tissues [ 20 ].
Sex-specific selection could be acting on regulatory elements as well as coding regions. We therefore repeated the analyses including all noncoding transcripts whose expression is profiled in the Genotype-Tissue Expression Project 34, in total and the 1kb sequence upstream of all coding regions sex in the previous analysis.
Once again, a significant Twin Peaks pattern sex. We considered the possibility that the pattern is driven by heterozygosity. Imagine that the strength sex-specific selection does not vary systematically with sex-biased expression, but that heterozygosity follows the Twin Peaks pattern.
Then genetic divergence between males and females, measured as F Srxwill also show that pattern. This prediction follows because changes in allele frequencies caused by selection are proportional to heterozygosity. If the Twin Peaks results from variation of heterozygosity with sex-biased expression, the regression is expected to be flat.
This alternative hypothesis is falsified: Twin Peaks are again seen using the polynomial and spline regressions described above S3 Table. While no individual SNP showed significant F ST between the sexes, the identities of several highly diverged genes do hold hints of possible connections to sex-specific selection.
The gene RNF segregates for alleles that have opposing effects on male and female recombination rates [ 2324 ]. This is a sexually-dimorphic phenotype that has long been associated with sexual attractiveness [ 26 ]. A recent review identified 33 loci that have been linked by multiple genome-wide association studies to sex-specific risk of sex [ 27 ]. This correlation could result if sexually-antagonistic selection itself is maintaining polymorphism sex these loci [ 29 ], or if sex-specific selection acts on polymorphisms that are maintained by other mechanisms e.
These results lead us to aex that the Twin Peaks pattern in humans results from sex-specific viability selection whose strength varies with the degree of sex-biased gene expression. To test the generality of the pattern we discovered in humans, we reanalyzed the fly data from Innocenti and Morrow [ 7 ].
Here our measure of the strength of sex selection acting on a locus is binary: it takes a value of 1 if they identified it as a target of sexually-antagonistic selection, and is 0 otherwise.
Their data are based on fertility selection, rather than viability selection. Sex Twin Peaks pattern appears once again Fig 2. The greatest fraction of loci under sexually-antagonistic selection have intermediate sex-biased expression. The distribution of gene numbers is shown at the bottom of the figure. The pattern in flies is consistent with that in humans. A difference between the two data sets is that the pattern in humans results from genetic effects on viability, while the pattern in flies reflects effects on fertility and fecundity.
Together, these results suggest that a tendency for stronger sex differences in selection to act on genes sex intermediate sex-biased expression may be general to diverse forms of selection.
To understand the Twin Peaks pattern, we built a simple model that relates gene expression to xex. Three key assumptions are invoked: 1 Xf frequencies are sex equilibrium under sexually-antagonistic viability selection, which is the special case of sex-specific selection in which selection favors one allele in one sex and the other allele in the other sex; 2 A gene that is not expressed experiences no selection; and 3 At low expression levels, the effects of alleles on viability increase approximately linearly with the amount of expression.
These assumptions are illustrated graphically in Fig 3. The model allows selection to be frequency-dependent, and for arbitrary dominance. Further details and the analysis of the model are given in the Materials and Methods section. The X-axis shows the expression of a xf, measured as the log of the number of transcripts, in males M and females F.
The Y-axis shows the additive fitness effect of an allele in males s m and females s f. Fitness effects increase with expression at a rate a m in males and a f in females.
Further details are given in the Materials and Methods section. This model leads to two predictions regarding how genetic divergence between the sexes varies with the degree sex-biased expression. All else sexx, this result predicts that F ST will be 0 when expression is unbiased, and it will increase quadratically as expression becomes slightly female- or male-biased.
Second, the model predicts that F ST will also be zero when expression is completely female-biased and completely male-biased. The model does not apply to genes whose expression is strongly but not completely sex-biased. We can, however, make qualitative predictions for these cases by interpolating between the predictions from the first two cases.
The pattern that results is Twin Peaks Fig 4. Values of F ST sx for intermediate sex-dependent expression are interpolated by eye dashed curves.
Leggy and heavily made-up models still adorn the stands at the Geneva show, though sex seems to be a bit less skin on show zg in previous years. Johan van Zyl, the head sex Toyota Motor Europe, said his company wasn't using women models to shape the image of the vehicles. Neither Toyota nor Rolls-Royce had models standing by the cars, though they did have assistants - sex and men - dressed in business-wear to provide sex to visitors.
Still, change comes hard - sex not all companies are on board. The Skoda display for its Vision X featured a visit from the Miss Czech, Jitka Novackova, posing for cameras in a sex dress and sex leather boots rising over the knee. Car makers like Alfa Romeo and Zr, among others, trotted out models, too. Duchess Meghan's oversized shoes listed as one of this decade's biggest fashion fails. It's actually a Red Friday for South Africans.
Sars wins R1 billion tax evasion case. Healthy man dies weeks after being licked by his dog. LOOK: Durban shop owner arrested for manufacturing fake detergents. The Motor Show, which attracts more than exhibitors from the world, opens to the public from March 8 to 18, More thanvisitors are expected to visit the event. The Motor Show, which attracts sex than exhibitors from the world, will open to the public from March 8 to 18, Read next on IOL. Most Read on IOL.
Studied genes were clustered by means of hierarchical clustering algorithm green, higher; red, lower expression, in comparison to male gland. Hierarchical clustering was performed on log2 signal intensity data. Subsequently, we performed SEA followed by functional analysis using the GeneAnswers package of Bioconductor which, among other things, allows us to interpret a list of genes in the context of their participation in a certain particular biological process GO.
The original results of GeneAnswer analysis as of December 16, revealed 17 groups of genes with different functional profiles. It should be emphasized that the GO database is composed of some general, as well as specific categories with a similar meaning and, therefore, a single gene may be mapped to several GO terms and may be counted more than once.
Moreover, GO functional annotations of genes are still in the developing stage and are far from complete. Since this analysis yielded rather unsatisfactory results, we decided to analyze our results using the DAVID system.
Functional profiles of differentially expressed genes determine by the GeneAnswers package as of December 16, , based on the Gene Ontology. Biological Process GO. BP database. Presented numbers show quantity of genes identified in the present study which are involved in described biological processes functional profiles. Genes were analyzed from both male and female adrenal glands.
The DAVID system is a powerful tool that allows us to discover enriched functionally related gene groups and to cluster similar annotation terms. This system extracts data from numerous databases and, as evidenced by the Science Citation Index, this system is gaining wide popularity among molecular biologists.
This figure indicated that numerous genes participated in more than one annotation cluster. In the table clusters being ordered by the enrichment score, the higher the score, the more enriched the gene. Up, upregulated; down, downregulated indicates higher or lower expression in relation to males. For the ZG, annotation clusters 1 and 2 combined genes involved in the regulation of ion transport, while cluster 3 contained transcripts linked to response to hormones Table II.
The first cluster combined genes regulating steroid biosynthesis and metabolism. Cluster 4 was composed of genes regulating responses to nutrients and to extracellular stimuli or steroid hormone stimuli.
Of these genes, the expression levels of cysteine dioxygenase type 1 Cdo1 , gap junction protein, alpha 1 Gja1 , isocitrate dehydrogenase 1 Idh1 and phospholipase A2, group IVA Pla2g4a were lower and those of nitric oxide synthase 1 Nos1 were higher in the ZG of female rats, when compared with the males Fig. Of the 19 genes forming the above-mentioned subcluster, the expression levels of 7 of these genes were higher and those of 10 genes were lower in the female adrenal glands.
Of these genes, the expression levels of only 4 of them [ATP-binding cassette, sub-family A ABC1 , member 1 Abca1 , cytochrome P, family 27, subfamily A, polypeptide 1 Cyp27a1 , lipase, hormone-sensitive Lipe and Sult1a1] were lower in the females compared to the males.
On the contrary, the remaining genes exhibited higher expression levels in the females [Cyp11a1, cytochrome P, family 11, subfamily B, polypeptide 1 Cyp11b1 , cytochrome P, family 51 Cyp51 , fatty acid binding protein 6, ileal Fabp6 , hydroxysteroid beta dehydrogenase 7 Hsd17b7 , isopentenyl-diphosphate delta isomerase 1 Idi1 , lanosterol synthase 2,3-oxidosqualene-lanosterol cyclase Lss , Nr0b1, squalene epoxidase Sqle and steroidogenic acute regulatory protein Star ].
Data are in comparison to male adrenal glands. Fold changes and p-values are shown. Global gene expression profiling allows for a simultaneous analysis of thousands of genes in a single sample.
This powerful tool of molecular biology is widely used, among others, in studies on basic biology or in the diagnosis of various diseases. This method has also been used in studies on adrenal glands.
In recent years increasing amounts of data linking gene expression with adrenal biology have been generated. In this area, the first reports were focused on gene profiles for steroidogenic enzymes in adrenocortical diseases, in particular in aldosterone-producing adenomas and other adrenocortical tumors 29 — Gene profiling methods applied to freshly isolated adrenocortical cells or established cell lines Y1, HR human adrenocortical cells , allows the identification of numerous genes involved in the regulation of adrenocortical growth and functioning and provides novel data on intracellular pathways involved in the regulation of aldosterone and corticosterone secretion 13 , 35 , Transcriptional profiling has also been used in in vivo experiments.
By means of this method, the circadian regulation of steroid hormone biosynthesis genes was examined in the rat adrenal gland In knockout mice lacking Star, numerous up- and downregulated genes were identified by Ishii et al Furthermore, recent studies on the regulation of the renin-angiotensin-aldosterone system RAAS in rat adrenal glands identified transcripts involved in RAAS activation Transcriptional profiling has also been applied for the characterization of enucleation-induced adrenal regeneration in the rat 21 — From the above short survey, it appears that in in vivo experiments, gene profiling methods are used mainly for investigations of adrenal glands with experimentally modified function.
Not considering studies on the circadian regulation of steroid hormone biosynthesis genes, to the best of our knowledge, only one study has been performed on the intact rat adrenal gland The authors investigated differentially expressed transcripts in the adrenal ZG and ZF of the adult male rat. We performed a similar analysis of the adrenal glands of adult male and female rats. The study by Nishimoto et al 14 identified such transcripts, while the figures obtained in the present study amounted to for male and for female rats.
These differences may be due to the various methods applied in two studies, as there are many methods gerenarally used 40 — This means that our zona fasciculata cells were contaminated with the zona reticularis ones. Be as it may, our data also revealed notable differences in the number of differentially expressed transcripts in the adrenal glands of adult male and female rats, and in females this number was significantly higher.
It is well known that in the rat, the structure and function of the adrenal cortex exhibits sex-dependent differences reviewed in refs. The genetic background of these differences is little known, therefore we performed whole transcriptome analyses on the adrenal glands of mature male and female rats which allowed as to compare the expression levels of approximately 27, genes by applying microarray technology.
In this compartment, the expression levels of sex-regulated genes were lower in the female than in the male gland. To describe the functions of differentially expressed genes in the male and female rat adrenal cortex, we performed the clustering of identified transcripts.
By using DAVID tool, in the ZG, 3 annotations clusters were obtained and they were mainly related to ion transport and the response to endogenous stimuli. Unexpectedly, the expression levels of these genes were notably lower in the female rats. In the 4 remaining clusters, the numbers of downregulated transcripts in female adrenal glands were higher than those of the upregulated transcripts.
Moreover, these observations are in accordance with those of previous studies, showing that stimulation of the highly-specialized function of adrenocortical cells i. In this regard, the study by El Wakil et al 16 on sexual dimorphism of gene expression in the mouse adrenal gland revealed only 2 genes Nr5a1 and Nr0b1 differentially expressed in the mouse, with higher expression levels in females.
However, in their study none of the genes directly involved in steroid biosynthesis was found to be differentially expressed These differences may be dependent on the species rat vs. Also this finding is rather unexpected, it seems to be scientifically justified the negative interrelationship between highly specialized cell function and basic cell functions.
In present study, the expression levels of selected genes involved in the regulation of steroid biosynthesis was validated by RT-qPCR. As is known, cholesterol is the precursor for the entire adrenal steroidogenesis. Earlier data have demonstrated that the concentration of total lipids, total cholesterol, phospholipids and glycerides is similar in the adrenal glands of adult male and female rats; however their content, due to larger adrenal glands, is markedly higher in females In this regard, Lipe is a major cholesterol hydrolase of the adrenal glands 50 , The specific activity of this enzyme in , g supernatant of adrenal homogenates is higher in male than in female rats Moreover, we demonstrated that the expression of the Lipe gene in the rat ZG was similar in both genders.
It has been well documented that the Star gene encodes a protein involved in the acute regulation of steroid hormone synthesis. This protein is responsible for the transport of cholesterol from the outer to the inner mitochondrial membrane In the mouse adrenal glands, Star mRNA levels assessed by RT-PCR have been reported be slightly higher in males than in females, although the differences were not statistically significant To the best of our knowledge, this is the first demonstration of specific sex-related differences in Star gene expression in the rat adrenal cortex.
Cyp11a1 encodes the Pscc enzyme cholesterol 20—22 desmolase that catalyzes the first and rate-limiting step of steroid biosynthesis 59 , This enzyme is expressed in all adrenocortical zones 14 , As previously reported, the rate of cholesterol transformation into pregnenolone side chain cleavage activity is markedly lower in male than in female rats 2 , 62 , Moreover, the levels of adrenodoxin in decapsulated i.
We also validated the expression levels of 2 genes coding enzymes of the steroidogenic late pathway in the rat, Cyp11b2 aldosterone synthase, responsible for aldosterone synthesis and Cyp11b1 steroid 11beta-hydroxylase, responsible for corticosterone synthesis.
In situ hybridization and immunohistochemistry revealed the expression of aldosterone synthase in the rat ZG only 14 , 61 , These results are in agreement with those of a previous study which demonstrated by RT-PCR high expression levels of Cyp11b2 in this zone Our study confirmed these earlier observations. Moreover, we demonstrated that the expression levels of the Cyp11b2 gene were similar in the ZG of male and female rats. These data are in accordance with those of previous studies on the absence of sex-related differences in aldosterone synthesis in male and female rats reviewed in Refs.
On the contrary, in the mouse adrenal glands, the mRNA levels of aldosterone synthase have been shown to be slightly higher in female than in male adrenal glands The Cyp11b1 gene encodes steroid 11beta-hydroxylase, which catalyzes the conversion of deoxycorticosterone to corticosterone. The activity of steroid 11beta-hydroxylase has been shown to be similar in adult male and female rats 68 , Nr0b1 encodes the Dax1 protein dosage-sensitive sex reversal, adrenal hypoplasia critical region, on chromosome X, gene 1 which is responsible for the development and maintenance of the steroidogenic axis 70 — Experimental data have suggested that the DAX1 protein is a negative regulator of steroidogenesis.
It should be mentioned that in the mouse adrenal glands, the mRNA levels of Dax1 were only slightly higher in female than in male adrenals 16 , They demonstrated that Hcrtr2 was localized in the ZG and zona reticularis, with higher expression levels in male adrenal gland. In conclusion, to the best of our knowledge, the present study presents the first report of sex-related gene expression profiles in the adrenal cortex of adult rats.
Bachmann R: Die Nebenniere. Handbuch der Mikroskopischen Anatomie des Menschen. Kitay JI: Effects of estrogen and androgen on the adrenal cortex of the rat. Functions of the Adrenal Cortex. Malendowicz LK: Cytophysiology of the mammalian adrenal cortex as related to sex, gonadectomy and gonadal hormones.
Compr Physiol. Lescoat G, Jego P, Beraud G and Maniey J: Sex influences on the response of the hypothalamo-hypophysio-adrenal axis to emotional and systemic stress in the rat. In French. The effect of ether stress on ACTH and corticosterone in intact, gonadectomized, and testosterone- or estradiol-replaced rats.
Res Exp Med Berl. View Article : Google Scholar. Malendowicz LK: Sex differences in adrenocortical structure and function. The effects of postpubertal gonadectomy and gonadal hormone replacement on nuclear volume of adrenocortical cells in the rat. Cell Tissue Res. The effects of postpubertal gonadectomy and gonadal hormone replacement on the rat adrenal cortex evaluated by stereology at the light microscope level. We assume the samples are independent.
Denote the unknown frequency of the minor allele in zygotes at a given locus as p. Then likelihood of s m , the additive fitness effect in males as defined by Eq 4 , is 15 where B is the binomial distribution, and the allele frequencies in the adults from which the alleles are sampled are For Pr p , we fit an ad hoc function to the distribution of minor allele frequencies observed in the YRI population.
That function is: 18 where Re[ x ] is the real part of x. The fit of this function to the data is shown in S1 Fig. In practice, we found that using a uniform prior distribution changed the estimates of s m very little. We evaluated posterior probability by numerically integrating Eq We found the maximum a posteriori probability MAP estimate of s m by numerically maximizing that function.
S2 Fig shows the results for a sample of 15 genes. For each of the genes in the sample, we chose the SNP with the largest F ST , reasoning that this SNP was more likely to be the actual target of sex-specific selection. All of the MAP estimates are very large but no estimate for s m is significantly different from 0. The selection load caused by sexually-antagonistic viability selection can be calculated by the following argument.
Assume Hardy-Weinberg equilibrium, and let p be the frequency of the allele that is beneficial to males. The mortalities are then: We now assume the population is at equilibrium. Then Eqs 3 and 4 imply that 21 which on rearranging gives Substituting these results back into Eq 19 tells us that the sexually-antagonistic load is The sample includes more than individuals from 26 populations worldwide.
The gene annotation file GRCh38 was downloaded from Ensembl [ 43 ]. We filtered the data to include only SNPs in transcribed regions of autosomal protein coding genes. The resulting dataset includes over 6 million SNPs in 17, autosomal genes. In a second series of analyses, we also included potential regulatory elements by analyzing all 34, transcripts profiled by the Genotype-Tissue Expression Project [ 17 ] and the 1kb upstream regions of protein-coding genes.
F ST between the sexes and the genetic diversity in each population were estimated using the R package PopGenome [ 44 ]. Likewise, results were not changed when we excluded autosomal genes with paralogs on the X or Y chromosome.
We conducted a power analysis to determine how likely that result is to occur under different strengths of selection S1 File. We used parameter values corresponding to the Yoruban population which is the most polymorphic and so most likely to show signals of sex-specific selection. It is therefore unsurprising that no SNP showed significant divergence.
To learn if sex-specific selection is more prevalent among genes with certain functions, we identified the SNP with the highest F ST at each locus, since this SNP is more likely to be the target of selection. Gene set enrichment analysis was performed using online tools from the Database for Annotation, Visualization and Integrated Discovery [ 45 ]. We used expression levels in ovaries measured in 6 females and levels in testes measured in 14 males.
To determine if the results hold for tissues other than testes and ovaries, we added the data for all of the sex-specific somatic tissues available in the Genotype-Tissue Expression Project.
For female-specific tissues, we averaged expression in the ovary, ectocervis, fallopian tubes, vagina, and uterus. For male-specific tissues, we averaged testes and prostate. The Twin Peaks pattern again appears, but with the left-hand peak shifted far to the left. This shift appears to be driven by the prostate: the original pattern reappears using the five female-specific tissues but only testes from males, but no pattern is seen using the five female-specific tissues and prostate. We believe that prostate does not provide a good proxy for gene expression in males: its expression profile is most similar to that of vagina, and is quite similar to other female-specific tissues [ 20 ].
We were unable to repeat our analyses with other somatic tissues because the Gene Expression Atlas only provides data on the average expression across both sexes. Gene expression data collected by Innocenti and Morrow [ 7 ] was downloaded from Gene Expression Omnibus accession number GSE and the probe annotation file from Ensemble.
Following Innocenti and Morrow, the log 2 ratio of male and female gene expression was estimated using the R Bioconductor packages [ 47 ]. The list of candidate genes under sexually-antagonistic selection was taken from [ 7 ]. Results shown in the main text pertain to autosomal loci, but the Twin Peaks pattern also appears when sex-linked loci are included.
Each population was treated as a replicate. The optimal polynomial degree was determined using the Akaike information criterion [ 48 ] and likelihood ratio tests. Regressions were also fit with cubic splines using generalized additive models GAM implemented mgcv package [ 49 ] in R.
Following [ 50 ], we fit fourth degree polynomials to 10 5 bootstrap samples of the original data. The last two criteria imply that there were two peaks local maxima with a local minimum between them. The result is shown in Fig 1. A permutation test was used to test the significance of the pattern we observed in the data. This procedure was repeated 10 5 times. The locations of the local minimum and the two local maxima for the polynomial regressions were found by using the polyroot function in R [ 51 ].
The loci were chosen at random from those with sex-biased expression levels corresponding to the Twin Peaks, and typical F ST values for genes with that degree of expression bias. We thank D. Houle and A. Dagilis for stimulating discussions. We are grateful for useful suggestions made by two reviewers. Conceived and designed the experiments: CC MK. Performed the experiments: CC MK. Analyzed the data: CC MK. Wrote the paper: CC MK. Abstract Sexual dimorphism results from sex-biased gene expression, which evolves when selection acts differently on males and females.
Author Summary Sexual dimorphism, which is evident in virtually all phenotypic traits, results from sex-biased gene expression. Introduction Females and males differ for virtually all phenotypic traits [ 1 ]. Results Sex-specific selection and sex-biased expression in humans We quantify sex differences in viability selection using F ST between adult males and females [ 15 ].
Download: PPT. Fig 1. The strength of sex-specific selection is strongest on human autosomal genes with intermediate sex-biased expression. Twin Peaks also occur in flies To test the generality of the pattern we discovered in humans, we reanalyzed the fly data from Innocenti and Morrow [ 7 ]. A simple model explains Twin Peaks To understand the Twin Peaks pattern, we built a simple model that relates gene expression to fitness.
The strength and prevalence of sex-specific selection in humans Under the assumptions of our model, the value of F ST between males and females is proportional to the strength of sexually-antagonistic viability selection see Materials and Methods.
The selection load Sex differences in autosomal allele frequencies between mature females and males result from differential survival after conception. Discussion This study focuses on the relation between the strength of ongoing sex-specific selection to the degree of sex-biased expression. Materials and Methods A simple model relating F ST between the sexes to sex-biased gene expression To understand the relation between sex-biased expression and sex-specific selection, we developed a very simple model.
Assuming that zygotes are at Hardy-Weinberg equilibrium, the allele frequency in adult males is 3 where s m is the additive fitness effect of allele A in males, defined as 4 The approximation of Eq 3 neglects terms of order. The divergence between males and females is therefore 5 We link selection to expression by assuming that when expression is low, the fitness effects can be approximated by the functions: 6 where 7 and m and f are the absolute expression levels e.
The equilibrium condition and Eq 6 imply that 8 where T is the total expression and D is the difference in expression sex-bias between males and females: 9 Substituting Eqs 6 — 9 into Eq 5 gives 10 where 11 Eq 10 gives a pleasingly simple relationship between the divergence of allele frequencies between the males and females resulting from sexually-antagonistic selection, on the one hand, and the degree of sex-biased expression, on the other.
When bias is small and the absolute expression levels are much greater than 1, a Taylor expansion shows that 13 Substituting Eq 13 into Eq 10 , we finally have 14 which appears as Eq 1 in the main text.
The strength of sexually-biased selection We estimated the strength of selection from the frequencies of alleles in females and males. The selection load The selection load caused by sexually-antagonistic viability selection can be calculated by the following argument. The mortalities are then: 20 We now assume the population is at equilibrium.
Quantifying sex-biased gene expression RNAseq data for gene expression in humans provided by the Genotype-Tissue Expression Project [ 17 ] was queried from the Gene Expression Atlas [ 46 ]. Reanalysis of fruit fly data Gene expression data collected by Innocenti and Morrow [ 7 ] was downloaded from Gene Expression Omnibus accession number GSE and the probe annotation file from Ensemble.
Supporting Information. S1 Fig. Fit of Eq 18 to the minor allele frequency distribution in the YRI population. S2 Fig. Estimates of the sexually-antagonistic fitness effects at 15 loci. S3 Fig. The relationship of D a vs. S4 Fig. S5 Fig. The probability of detecting significantly different allele frequencies in males and females at any SNP in the YRI population as a function of the strength of sexually-antagonistic selection.
S1 Table. S2 Table. S3 Table. S4 Table. S5 Table. S6 Table. S7 Table. S8 Table. S1 File. The details of a power analysis that calculates the probability of observing a statistically significant difference between allele frequencies in males and females that results from sexually-antagonistic selection. Acknowledgments We thank D. References 1. Darwin C. The descent of man, and selection in relation to sex. London: Murray; Lande R. Sexual dimorphism, sexual selection, and adaptation in polygenic characters.
View Article Google Scholar 3. Rice WR. Sex chromosomes and the evolution of sexual dimorphism. View Article Google Scholar 4. Arnqvist G, Rowe L. Sexual conflict. Princeton University Press; Vicoso B, Charlesworth B. Evolution on the X chromosome: unusual patterns and processes. Nat Rev Genet. Ellegren H, Parsch J. The evolution of sex-biased genes and sex-biased gene expression. Innocenti P, Morrow EH.
The sexually antagonistic genes of Drosophila melanogaster. PLoS Biol. Intralocus sexual conflict. Ann N Y Acad Sci. Bonduriansky R, Chenoweth SF. Trends Ecol Evol. Nothing in genetics makes sense except in light of genomic conflict. Annu Rev Ecol Evol Syst. View Article Google Scholar Charlesworth D, Charlesworth B.
Sex differences in fitness and selection for centric fusions between sex-chromosomes and autosomes. Genet Res. Healthy man dies weeks after being licked by his dog. LOOK: Durban shop owner arrested for manufacturing fake detergents. The Motor Show, which attracts more than exhibitors from the world, opens to the public from March 8 to 18, More than , visitors are expected to visit the event.
The Motor Show, which attracts more than exhibitors from the world, will open to the public from March 8 to 18,
It is well known that adult female ses have heavier adrenal glands than males of the same age. This difference appears only after zd and is dependent upon sex hormones. Earlier, mainly morphologic, cytological and gravimetric data on this subject had been extensively reviewed by Bachmann in 1. In the rat, adrenal sex dimorphism is accompanied by functional differences, with females secreting greater amounts of corticosterone than males reviewed in refs.
Furthermore, female rats show greater adrenocorticotropic hormone ACTH and corticosterone responses to stress and these hormonal responses are modified by af and gonadal hormone replacement 6 — 8. Numerous studies have suggested that observed sex differences in the rat adrenal structure and function are dependent on xex inhibitory effects of testosterone zg the hypothalamo-pituitary-adrenal HPA axis, while estrogens exert opposite effects reviewed in refs.
However, the molecular bases of the above outlined sex-related differences in rat adrenal cortex structure and function have not yet zff fully elucidated. The introduction of gene expression microarray technology opens the possibility of discovering genes that may contribute to various biological effects.
As regards the adrenal gland, an example of such a study is the identification of genome-wide changes in gene expression following the treatment of Y1 mouse adrenocortical cells with ACTH In this cell line, Zt affected the levels of 1, different transcripts, and only were previously known sex corticotropin-affected.
Furthermore, by means of laser-capture microdissection, Nishimoto et al 1415 identified hundreds of transcripts with differential expression in the zona glomerulosa ZG and zona fasciculata ZF of adult male rats. As far as sex-related differences in the adrenal cortex are zzf, recently, El Wakil et al performed a genomic analysis of gene expression in the mouse adrenal gland In their data, not considering the transcription factors, nuclear receptor subfamily 5, group A, member 1 Nr5a1 and nuclear receptor subfamily 0, group B, member 1 Nr0b1none of the genes directly involved in steroid hormone biosynthesis showed a differential expression in the male and female mouse adrenal sex Therefore, in the present study, using the adrenal glands of mature male and female rats, we performed whole transcriptome sex that allowed us to compare the expression levels of approximately 27, genes by applying microarray technology.
The animals were maintained under standardized conditions of light h light-dark cycle, illumination onset Female rats were used in the estrous cycle phase, which was determined according to the cell types observed in the vaginal smear. Following decapitation between andthe adrenal glands were promptly removed, freed of adherent fat and processed for analysis.
The medulla of the adrenal gland was removed and was not used in our analysis. Unless otherwise stated, all reagents were obtained from Sigma-Aldrich St. Gliwice, Poland. RNA isolation from samples sxe adrenal glands was carried out as previously described 19 — Microarray analysis was carried out as previously described 21 — Isolated total RNA ng was mixed with 1. Biotin-labeled fragments of sxe 5. Up to 25 unique probes sequences were hybridized to a sxe transcript.
Following hybridization, each array strip was washed and stained using the Fluidics Station of GeneAtlas system Affymetrix. The array strips were scanned using the Imaging Station of the GeneAtlas system. The quality of gene expression zff was examined according to the quality control criteria provided with the software.
The intensity of fluorescence was converted to numerical data by generating CEL files. The obtained CEL files were imported into downstream data analysis software. Unless otherwise stated, all presented analyses and graphs were performed using Bioconductor and R programming language, as previously described Each CEL file was merged with a description file downloaded from the Affymetrix webpage.
In order to perform background correction, normalization sex summarization of the results, we used the Robust Multiarray Averaging RMA method. The statistical significance of the analyzed genes was examined by srx t-statistics from the srx Bayes method.
Zv lines indicate cut-off values 2,4,6,8 fold change in expression. Names of genes are not shown. Singular enrichment analysis SEA was performed as previously described 1526dex Selected sets of differentially expressed genes were applied to functional analysis using the GeneAnswers package of Bioconductor which, among other, allows us to interpret a list of genes in the context of their participation in particular biological processes GO.
BP Lists of differentially expressed genes were combined as tables and were subjected to further analyses. Since our dataset comprised 2 comparison sets male vs. The GeneAnswers package allowed to test the sex of each GO. BP category in a gene list using a well-defined hypergeometric statistical test.
The p-value was determined based on the number of genes differentially expressed in the investigated GO category. This database provides functional annotation tools for understanding the biological meaning behind a large list of genes www.
Among the many functions, DAVID allows us sxe discover enriched function-related gene groups and to cluster similar annotation terms 17 sex, Annotations and background total number of genes in the rat were limited only to Rattus norvegicus.
The specificity of the reaction products was examined by the determination of the melting points 0. Hprt ssex used z a reference gene. Due to the nature of the applied experiment, all data were analyzed in relation to the adrenal glands of the male rats.
The mean expression zg of each gene was sec in scatter plot graphs Fig. The left upper part of the graphs shows genes, the expression levels of which were higher in the female than in the male adrenal glands. In the right lower part of the graphs, genes are also shown, the expression levels of which were lower in the female than in the male adrenal glands.
A Venn diagram demonstrated their localization to the adrenocortical zones examined Fig. In the ZG, the expression levels of 24 genes were lower and 8 were higher in the female rats.
In this compartment of the sx adrenal cortex, the expression levels of genes were lower and those of 87 genes were higher in the female rats. Middle section shows the number of genes which had similar expression patterns ses both zones. Bold font indicates higher expression levels; italic font indicates lower expression levels. In females, these figures were andrespectively. Each of the raw expression values from gender-specific genes was grouped using a hierarchical clustering algorithm.
The results of this analysis are presented as a heat map Fig. The clustering confirmed that in the ZG, the expression levels of sex genes were lower and 8 were higher in the female rats and sexx symbols of these genes are shown.
Arbitrary signal intensity aex from microarray analysis is represented by colors. Studied genes were clustered by means of hierarchical clustering algorithm green, higher; red, lower expression, in comparison to male gland. Hierarchical clustering was performed on log2 signal intensity data.
Subsequently, we performed SEA followed by functional analysis using the GeneAnswers package of Bioconductor which, among other things, allows us to interpret a list of genes in the context of their participation in a certain particular biological process GO.
The original results of GeneAnswer analysis as of December 16, revealed 17 groups of genes with different functional profiles. It should be emphasized that the GO database is composed of some general, as well as sdx categories with a similar meaning and, therefore, a single gene may be mapped to several GO terms and may be counted more than once.
Moreover, GO functional annotations of genes are still in sed developing stage and are far from complete. Since this analysis yielded rather unsatisfactory results, we decided to analyze our results using the DAVID system. Functional profiles of differentially expressed genes determine by the GeneAnswers package as of December 16,based on the Gene Ontology.
Biological Process GO. BP database. Presented numbers show quantity of genes identified in the present study which are involved in described biological processes functional profiles. Genes were analyzed from both male and female adrenal glands. The DAVID system is a powerful tool that allows us to discover enriched functionally related gene groups and to cluster similar annotation terms.
This system extracts data from numerous databases and, as evidenced by the Science Citation Index, this system is gaining wide popularity among molecular biologists. This figure indicated that numerous genes participated in more than one annotation cluster. In the table clusters being ordered by the enrichment score, the higher the score, the more enriched the gene. Up, upregulated; down, downregulated indicates higher or lower expression in relation to males. For the ZG, annotation clusters 1 and 2 combined genes involved in the regulation of ion transport, while cluster 3 contained sx linked to response to hormones Table II.
The first cluster combined genes regulating steroid biosynthesis and metabolism. Cluster 4 was composed of genes regulating responses to nutrients and to extracellular stimuli or steroid hormone stimuli. Sex these genes, the expression levels of cysteine dioxygenase type 1 Cdo1gap junction protein, alpha 1 Gja1isocitrate dehydrogenase 1 Idh1 and phospholipase A2, group IVA Pla2g4a were lower and those of nitric oxide synthase 1 Nos1 were higher in zr ZG of female rats, when compared with the males Fig.
Of the 19 genes forming the above-mentioned subcluster, the expression levels of 7 of these genes were higher and those of 10 genes were lower in the female adrenal glands.
Ses these genes, the expression levels of only 4 of them [ATP-binding cassette, sub-family A ABC1member 1 Abca1cytochrome P, family 27, z A, polypeptide 1 Cyp27a1lipase, hormone-sensitive Lipe and Sex were lower in the females compared to the males.
On the contrary, the remaining genes exhibited higher expression levels in the females [Cyp11a1, cytochrome P, family 11, subfamily B, polypeptide 1 Cyp11b1cytochrome P, family 51 Cyp51fatty acid binding protein 6, ileal Fabp6hydroxysteroid beta dehydrogenase 7 Hsd17b7isopentenyl-diphosphate delta isomerase 1 Idi1lanosterol synthase 2,3-oxidosqualene-lanosterol cyclase LssNr0b1, squalene epoxidase Sqle sec steroidogenic acute regulatory protein Star ].
Data are in comparison to male adrenal glands. Fold changes and p-values are shown. Global gene expression profiling allows for a simultaneous analysis of thousands of genes in a single sample. This powerful tool of molecular biology is widely used, among others, in studies on basic biology or in the diagnosis of various diseases.
This method has also been used in studies on adrenal glands. In recent years increasing amounts of data linking gene expression wex adrenal biology have been generated. In this area, the first reports were focused on gene profiles for steroidogenic enzymes in adrenocortical diseases, in particular in aldosterone-producing adenomas and other adrenocortical tumors 29 — Gene profiling methods applied to freshly isolated adrenocortical cells or established cell lines Y1, HR human adrenocortical cellsallows the identification of numerous genes involved in the regulation of adrenocortical growth and functioning and provides novel dex on intracellular pathways involved in the regulation of aldosterone and corticosterone secretion 1335 Transcriptional profiling has also been used in in vivo experiments.
By means of eex method, the circadian regulation of steroid hormone biosynthesis genes was examined in the rat adrenal gland In knockout mice xex Star, numerous up- and downregulated genes were identified by Ishii et al Furthermore, xf studies on the regulation of the renin-angiotensin-aldosterone system RAAS in rat sez glands identified transcripts involved in RAAS activation
Dating profiles and free personals ads posted by single women and girls from cities including: Kiev, Moscow, Donetsk, Dnebrovsky, Saint Petersburg, Odessa, Kazan, Perm', Zaporizhzhya, Tambov, Lapu-Lapu City, Guangzhou, Tacloban City, Konakovo, Kalibo, Nizhniy Novgorod, Istanbul, Kharkiv, Brooklyn, Mira Loma,
The sex-specific selection differentials, sm and sf, are equal to the Direct selection on Zm and Zf within each sex is only part of the equation of evolutionary. Ratio of sex-specific final attack ratios, ¯ z f /¯ z m, as a function of average attack ratio (¯ z f + ¯ z m)/2 for models () (lines) and () for different values of q.
- Вы ищете знакомства с иностранцами?
- Хотите выйти замуж за рубеж?
- Наш международный сайт знакомств абсолютно бесплатно поможет вам!
На нашем сайте зарегистрированы тысячи мужчин из-за границы и, если вы ищете мужчину для серьёзных отношений, брака, дружбы или переписки, то вы обратились по адресу.
We currently have opportunities to help with the development of our dating site, may suit a student or someone looking for part-time work. View more information here.