Transcriptomic and functional screening of weapon formation genes implies significance of cell adhesion molecules and female-biased genes in broad-horned flour beetle

Abstract

For understanding the evolutionary mechanism of sexually selected exaggerated traits, it is essential to uncover its molecular basis. By using broad-horned flour beetle that has male-specific exaggerated structures (mandibular horn, head horn and gena enlargement), we investigated the transcriptomic and functional characters of sex-biased genes. Comparative transcriptome of male vs. female prepupal heads elucidated 673 sex-biased genes. Counter-intuitively, majority of them were female-biased (584 genes), and GO enrichment analysis showed cell-adhesion molecules were frequently female-biased. This pattern motivated us to hypothesize that female-biased transcripts (i.e. the transcripts diminished in males) may play a role in outgrowth formation. Potentially, female-biased genes may act as suppressors of weapon structure. In order to test the functionality of female-biased genes, we performed RNAi-mediated functional screening for top 20 female-biased genes and 3 genes in the most enriched GO term (cell-cell adhesion, fat1/2/3, fat4 and dachsous). Knockdown of one transcription factor, zinc finger protein 608 (zfp608) resulted in the formation of male-like gena in females, supporting the outgrowth suppression function of this gene. Similarly, knockdown of fat4 induced rudimental, abnormal mandibular horn in female. fat1/2/3RNAi, fat4RNAi and dachsousRNAi males exhibited thick and/or short mandibular horns and legs. These cell adhesion molecules are known to regulate tissue growth direction and known to be involved in the weapon formation in Scarabaeoidea beetles. Functional evidence in phylogenetically distant broad-horned flour beetle suggest that cell adhesion genes are repeatedly deployed in the acquisition of outgrowth. In conclusion, this study clarified the overlooked functions of female-biased genes in weapon development.

Fig 1. Sexual dimorphism of broad-horned flour beetle (Gnatocerus cornutus).

a) male, b) female, top, frontal view; middle, dorsal view; bottom (inset), dorsal view of magnified mandibles. Lateral region of male head (gena, ge) is enlarged, and two head horns (hh) are located between eyes (top and middle panels). Females lack mandibular horn, head horn, and lateral protrusion of gena (b, top to bottom panels).

Fig 2. Comparative transcriptome shows the predominance of female-biased genes.

a) Multidimensional scaling (MDS) plot of 12 samples shows transcriptomic difference of male (blue) and female (red) early prepupal head (n = 6 each). b) MA plot of comparative transcriptome. Among 20565 genes, 89 genes were male-biased and 584 genes were female-biased. Threshold of statistical significance (red dots) is FDR < 0.05 and expression changed >50% (|logFC| > log2(1.5)). Twenty genes in highlighted zone (yellow) were subjected to functional screening by RNAi.

Fig 3. zfp608 ^RNAi female head exhibited male-like gena structure.

a) In normal female (TE), clypeus(cl) and gena (ge) were fused to form a fan-like head structure. Sclerites are only distinguished by striation (white arrow head). b) In zfp608 ^RNAi female, gena is detached from clypeus (white arrowhead) and laterally protruded to form male-like structure (black arrowhead). This phenotype was observed in all zfp608 ^RNAi females (n = 18).

Fig 4. fat1/2/3 ^RNAi males had thicker head horn.

a) Lateral view of normal (TE) and fat1/2/3 ^RNAi male head. Head horn (white arrowheads) became thicker in fat1/2/3 ^RNAi male. b) Morphological measurement shows head horn thickness is increased in in fat1/2/3 ^RNAi males (ANCOVA: treatment, F = 104.9957, p < 0.001). Phenotypic effects on legs were not detected. Male foreleg is shown.

Fig 5. fat4 ^RNAi phenotypes in male and female.

a,b) fat4 ^RNAi caused a severe systemic effect (ventral view). Antenna became short and thick (a, arrowhead). c) all legs were shortened and tarsomeres were fused in fat4 ^RNAi adults. Male foreleg is shown. d) In fat4 ^RNAi male, mandibular horn (mh) was malformed to show thick and short, rectangular column-like morphology. Mandible (md) became thick and short. e) In in fat4 ^RNAi female, a bump was abnormally formed at outer ridge of mandible (arrowhead) where male mandibular horn develops in corresponding region. f) Gena was clearly reduced in male and slightly reduced in female (g). Dashed lines indicate the edge of gena and clypeus (f, g).

Fig 6. dachsous^RNAi males exhibited thick and short mandibular horns.

a) Dorsal view of dachsous^RNAi (ds^RNAi) male. Unlike fat4^RNAi, antenna was unaffected. Gena ridge curled upwards and gena size was reduced (arrowhead). Elytra extension was incomplete. b) ds^RNAi adults showed short and thick legs but tarsomere fusion was not observed. Male foreleg is shown. c) Dorsal view of male right mandible. Mandibular horn (mh) was shortened in ds^RNAi males (e, ANCOVA: treatment, F = 8.0793, p < 0.01). d) Lateral view of male right mandible. ds^RNAi males had thicker mandibular horn (mh) (f, ANCOVA: PW×treatment, F = 4.7988, p = 0.03).