Echolocation call frequency variation in horseshoe bats: molecular basis revealed by comparative transcriptomics

Sex Differences in Cochlear Transcriptomes in Horseshoe Bats

Sexual dimorphism of calls is common in animals, whereas studies on the molecular basis underlying this phenotypic variation are still scarce. In this study, we used comparative transcriptomics of cochlea to investigate the sex-related difference in gene expression and alternative splicing in four Rhinolophus taxa. Based on 31 cochlear transcriptomes, we performed differential gene expression (DGE) and alternative splicing (AS) analyses between the sexes in each taxon. Consistent with the degree of difference in the echolocation pulse frequency between the sexes across the four taxa, we identified the largest number of differentially expressed genes (DEGs) and alternatively spliced genes (ASGs) in R. sinicus. However, we also detected multiple DEGs and ASGs in taxa without sexual differences in echolocation pulse frequency, suggesting that these genes might be related to other parameters of echolocation pulse rather than the frequency component. Some DEGs and ASGs are related to hearing loss or deafness genes in human or mice and they can be considered to be candidates associated with the sexual differences of echolocation pulse in bats. We also detected more than the expected overlap of DEGs and ASGs in two taxa. Overall, our current study supports the important roles of both DGE and AS in generating or maintaining sexual differences in animals.

Characterization of microRNA and gene expression in the cochlea of an echolocating bat ( Rhinolophus affinis )

MicroRNAs (miRNAs) are important post‐transcriptional regulators of gene expression and play key roles in many biological processes, such as development and response to multiple stresses. However, little is known about their roles in generating novel phenotypes and phenotypic variation during the course of animal evolution. Here, we, for the first time, characterized the miRNAs of the cochlea in an echolocating bat (Rhinolophus affinis). We sampled eight individuals from two R. affinis subspecies with significant echolocation call frequency differences. We identified 365 miRNAs and 121 of them were novel. By searching sequences of these miRNAs precursors in multiple high‐quality mammal genomes, we found one specific miRNA shared by all echolocating bats but not present in all other nonecholocating mammals. The targeted genes of this miRNA included several known hearing genes (e.g., KCNQ4 and GJB6). Together with the matched mRNA‐seq data, we identified 1766 differentially expressed genes (DEGs) between the two subspecies and 555 of them were negatively regulated by differentially expressed miRNAs (DEMs). We found that almost half of known hearing genes in the list of all DEGs were regulated negatively by DEMs, suggesting an important role of miRNAs in call frequency variation of the two subspecies. These targeted DEGs included several important hearing genes (e.g., Piezo1, Piezo2, and CDH23) that have been shown to be important in ultrasonic hearing of echolocating mammals.

Comparative transcriptome analysis provides insights into the molecular mechanisms of high-frequency hearing differences between the sexes of Odorrana tormota

Background

Acoustic communication is important for the survival and reproduction of anurans and masking background noise is a critical factor for their effective acoustic communication. Males of the concave-eared frog (Odorrana tormota) have evolved an ultrasonic communication capacity to avoid masking by the widespread background noise of local fast-flowing streams, whereas females exhibit no ultrasonic sensitivity. However, the molecular mechanisms underlying the high-frequency hearing differences between the sexes of O. tormota are still poorly understood.

Results

In this study, we sequenced the brain transcriptomes of male and female O. tormota, and compared their differential gene expression. A total of 4,605 differentially expressed genes (DEGs) between the sexes of O. tormota were identified and eleven of them were related to auditory based on the annotation and enrichment analysis. Most of these DEGs in males showed a higher expression trend than females in both quantity and expression quantity. The highly expressed genes in males were relatively concentrated in neurogenesis, signal transduction, ion transport and energy metabolism, whereas the up-expressed genes in females were mainly related to the growth and development regulation of specific auditory cells.

Conclusions

The transcriptome of male and female O. tormota has been sequenced and de novo assembled, which will provide gene reference for further genomic studies. In addition, this is the first research to reveal the molecular mechanisms of sex differences in ultrasonic hearing between the sexes of O. tormota and will provide new insights into the genetic basis of the auditory adaptation in amphibians during their transition from water to land.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-022-08536-2.

Geographic variation in the skulls of the horseshoe bats, Rhinolophus simulator and R. cf. simulator: Determining the relative contributions of adaptation and drift using geometric morphometrics

The relative contributions of adaptation and genetic drift to morphological diversification of the skulls of echolocating mammals were investigated using two horseshoe bat species, Rhinolophus simulator and R. cf. simulator, as test cases. We used 3D geometric morphometrics to compare the shapes of skulls of the two lineages collected at various localities in southern Africa. Size and shape variation was predominantly attributed to selective forces; the between-population variance (B) was not proportional to the within-population variance (W). Modularity was evident in the crania of R. simulator but absent in the crania of R. cf. simulator and the mandibles of both species. The skulls of the two lineages thus appeared to be under different selection pressures, despite the overlap in their distributions. Difference in the crania of R. cf. simulator was centered largely on the nasal dome region of R. cf. simulator but on the cranium and mandibles of R. simulator. It is likely that the size and shape of the nasal dome, which acts as a frequency-dependent acoustic horn, is more crucial in R. cf. simulator than in R. simulator because of the higher echolocation frequencies used by R. cf. simulator. A larger nasal dome in R. cf. simulator would allow the emission of higher intensity pulses, resulting in comparable detection distances to that of R. simulator. In contrast, selection pressure is probably more pronounced on the mandibles and cranium of R. simulator to compensate for the loss in bite force because of its elongated rostrum. The predominance of selection probably reflects the stringent association between environment and the optimal functioning of phenotypic characters associated with echolocation and feeding in bats.

Mitonuclear mismatch alters nuclear gene expression in naturally introgressed Rhinolophus bats

Background

Mitochondrial function involves the interplay between mitochondrial and nuclear genomes. Such mitonuclear interactions can be disrupted by the introgression of mitochondrial DNA between taxa or divergent populations. Previous studies of several model systems (e.g. Drosophila) indicate that the disruption of mitonuclear interactions, termed mitonuclear mismatch, can alter nuclear gene expression, yet few studies have focused on natural populations.

Results

Here we study a naturally introgressed population in the secondary contact zone of two subspecies of the intermediate horseshoe bat (Rhinolophus affinis), in which individuals possess either mitonuclear matched or mismatched genotypes. We generated transcriptome data for six tissue types from five mitonuclear matched and five mismatched individuals. Our results revealed strong tissue-specific effects of mitonuclear mismatch on nuclear gene expression with the largest effect seen in pectoral muscle. Moreover, consistent with the hypothesis that genes associated with the response to oxidative stress may be upregulated in mitonuclear mismatched individuals, we identified several such gene candidates, including DNASE1L3, GPx3 and HSPB6 in muscle, and ISG15 and IFI6 in heart.

Conclusion

Our study reveals how mitonuclear mismatch arising from introgression in natural populations is likely to have fitness consequences. Underlying the processes that maintain mitonuclear discordance is a step forward to understand the role of mitonuclear interactions in population divergence and speciation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12983-021-00424-x.

Echolocation call frequency variation in horseshoe bats: molecular basis revealed by comparative transcriptomics

Recently diverged taxa with contrasting phenotypes offer opportunities for unravelling the genetic basis of phenotypic variation in nature. Horseshoe bats are a speciose group that exhibit a derived form of high-duty cycle echolocation in which the inner ear is finely tuned to echoes of the narrowband call frequency. Here, by focusing on three recently diverged subspecies of the intermediate horseshoe bat (Rhinolophus affinis) that display divergent echolocation call frequencies, we aim to identify candidate loci putatively involved in hearing frequency variation. We used de novo transcriptome sequencing of two mainland taxa (himalayanus and macrurus) and one island taxon (hainanus) to compare expression profiles of thousands of genes. By comparing taxa with divergent call frequencies (around 15 kHz difference), we identified 252 differentially expressed genes, of which six have been shown to be involved in hearing or deafness in human/mouse. To obtain further validation of these results, we applied quantitative reverse transcription-PCR to the candidate gene FBXL15 and found a broad association between the level of expression and call frequency across taxa. The genes identified here represent strong candidate loci associated with hearing frequency variation in bats.