Cell-based analysis reveals that sex-determining gene signals in Ostrinia are pivotally changed by male-killing Wolbachia

Abstract

Wolbachia, a maternally transmitted bacterium, shows male-killing, an adaptive phenotype for cytoplasmic elements, in various arthropod species during the early developmental stages. In lepidopteran insects, lethality of males is accounted for by improper dosage compensation in sex-linked genes owing to Wolbachia-induced feminization. Herein, we established Ostrinia scapulalis cell lines that retained sex specificity per the splicing pattern of the sex-determining gene doublesex (Osdsx). We found that Wolbachia transinfection in male cell lines enhanced the female-specific splice variant of Osdsx (Osdsx^F ) while suppressing the male-specific variant (Osdsx^M ), indicating that Wolbachia affects sex-determining gene signals even in vitro. Comparative transcriptome analysis isolated only two genes that behave differently upon Wolbachia infection. The two genes were respectively homologous to Masculinizer (BmMasc) and zinc finger-2 (Bmznf-2), male-specifically expressed sex-determining genes of the silkworm Bombyx mori that encode CCCH-type zinc finger motif proteins. By using cultured cells and organismal samples, OsMasc and Osznf-2 were found to be sex-determining genes of O. scapulalis that are subjected to sex-specific alternative splicing depending upon the chromosomal sex, developmental stage, and infection status. Overall, our findings expound the cellular autonomy in insect sex determination and the mechanism through which sex is manipulated by intracellular selfish microbes.

Keywords: Ostrinia scapulalis; Wolbachia; cell culture; host sex determination; sex-specific gene expression.

Fig. 1.

Cell-based analyses of the effects of Wolbachia infection on gene expression. (A) Splicing pattern of Osdsx in cell lines based on RT-PCR. (B) Heatmap2 analysis showing sample-to-sample distances on the matrix of variance stabilized data for overall gene expression using principal component analysis (PCA). Darker colors indicate similar expression, and color key values indicate arbitrary units. (C) MA-plot showing log fold-changes between uninfected and wSca-infected male cell lines. Each dot indicates a single contig. Red dots surrounded by circles are contigs that showed significant changes upon wSca infection. (D) mRNA structure of Osdsx^M, Osdsx^F, OsMasc^M, OsMasc^F, Osznf-2^F, and Osznf-2 ^M. Yellow: coding region. Gray: untranslated region. Arrows: position of qPCR primers that specifically amplify sex-specific isoforms of Osdsx, OsMasc, and Osznf-2.

Fig. 2.

Relative titers of the sex-specific isoforms of Osdsx (A), OsMasc (B), and Osznf-2 (C) in the male-derived O. scapulalis cell line M4. Black: nontreated. Yellow: 3 weeks after transinfection with wSca. Orange: 5 weeks after transinfection with wSca. Red: 10 weeks after transinfection with wSca. Green: tetracycline treatment, started at the third week after transinfection with wSca and continued for 7 weeks thereafter.

Fig. 3.

Relative titers of sex-specific isoforms of Osdsx, OsMasc, and Osznf-2 in O. scapulalis adults. A, D, and G: titers of sex-specific isoforms of each sample of normal ZW female (red), normal ZZ male (blue), Wolbachia-infected ZW female (purple), and cured ZZ male (green). B, C, E, F, H, and I: titers of each isoform compared between normal ZW female (red), normal ZZ male (blue), Wolbachia-infected ZW female (purple), and cured ZZ male (green) individuals.

Fig. 4.

Relative titers of sex-specific isoforms of Osdsx, OsMasc, and Osznf-2 at the first-instar larval stage. A, D, and G: titers of sex-specific isoforms of each sample. B, C, E, F, H, and I: titers of each isoform compared between ZW female (red) and ZZ male (blue) of an uninfected line, ZW (light purple), and ZZ (dark purple) of a Wolbachia-infected matriline, and ZW (light green) and ZZ (dark green) produced by a Wolbachia-eliminated mother. Because some of the data contains zeros, 10⁻⁷ was added to each figure to plot on the log scale. Red dots: ZW individuals (females) of uninfected line. The same letters above the bars indicate no significant difference using generalized linear models with Bonferroni corrections.

Fig. 5.

Chronological changes in titers of sex-specific isoforms of Osdsx, OsMasc, and Osznf-2 at 0, 6, 9, 12, 24, and 48 hpo of embryos laid by uninfected females (gray), Wolbachia-infected females (purple), and Wolbachia-eliminated females (green). For OsMasc and Osznf-2, different scales were used for the titers of male-specific and female-specific variants (Y-axis). Note that both ZW and ZZ embryos are indiscriminately included in every cohort.