Common and unique strategies of male killing evolved in two distinct Drosophila symbionts

Abstract

Male killing is a selfish reproductive manipulation caused by symbiotic bacteria, where male offspring of infected hosts are selectively killed. The underlying mechanisms and the process of their evolution are of great interest not only in terms of fundamental biology, but also their potential applications. The two bacterial Drosophila symbionts, Wolbachia and Spiroplasma, have independently evolved male-killing ability. This raises the question whether the underlying mechanisms share some similarities or are specific to each bacterial species. Here, we analyse pathogenic phenotypes of D. bifasciata infected with its natural male-killing Wolbachia strain and compare them with those of D. melanogaster infected with male-killing Spiroplasma We show that male progeny infected with the Wolbachia strain die during embryogenesis with abnormal apoptosis. Interestingly, male-killing Wolbachia infection induces DNA damage and segregation defects in the dosage-compensated chromosome in male embryos, which are reminiscent of the phenotypes caused by male-killing Spiroplasma in D. melanogaster By contrast, host neural development seems to proceed normally unlike male-killing Spiroplasma infection. Our results demonstrate that the dosage-compensated chromosome is a common target of two distinct male killers, yet Spiroplasma uniquely evolved the ability to damage neural tissue of male embryos.

Keywords: Drosophila; Spiroplasma; Wolbachia; male killing; symbiosis.

Figure 1.

Analysis of apoptosis in infected D. bifasciata embryos. (a–l) Male-killing Wolbachia-infected female (a–f) and male (g–l) embryos from stages 9–16, stained for DNA (red) and apoptotic cells by TUNEL (cyan). Single-channel images of TUNEL staining are also shown. (m) Quantification of TUNEL-positive cells during embryogenesis. Infected embryos from stages 8–15 onwards and uninfected embryos from stages 10–15 onwards were analysed. Asterisks indicate statistically significant differences (*p < 0.0001; n.s., not significant, p > 0.05; Mann–Whitney U test). Box plots indicate the median (bold line), the 25th and 75th percentiles (box edges) and the range (whiskers). Dot plots show all data points individually. Sample sizes are shown at the bottom. (n) Female (i) and male (ii) embryos stained for Sxl. Embryos in (d) and (j) are shown as single-channel images of Sxl staining. Scale bars, 100 µm.

Figure 2.

Accumulation of DNA damage on the dosage-compensated male chromosome. (a–d) Epithelial cells of uninfected D. bifasciata female (a,b; n = 8) and male (c,d; n = 8) embryos at stage 13 stained for Sxl (green), H4K16ac (magenta) and DNA (blue). Yellow-boxed regions in (a) and (c) are magnified and shown as single-channel images in (b) and (d). White-boxed regions in H4K16ac images are highlighted in insets. (e,f) Embryonic epithelial cells of uninfected (e; n = 27) and infected (f; n = 25) male embryos at stage 11, stained for DNA damage (pH2Av; green), H4K16ac (magenta) and DNA (blue). Overlaps between DNA damage and H4K16ac signals are white in the merged image of (f). Scale bars, 20 µm (a,c) and 10 µm (b,d,e,f).

Figure 3.

The quantification and classification of chromatin bridges. The datasets used in figure 2e,f were analysed and the number of chromatin bridges was counted manually. (a) The number of chromatin bridges found in stage 11 embryos uninfected and infected with male-killing Wolbachia. Sample sizes are shown at the bottom. (b) The classification of chromatin bridges (H4K16ac +, +/− and −). For more details, see the text. (c–h) Examples of chromatin bridges observed in infected male embryos stained for DNA (green) and H4K16ac (magenta). (c) A representative image of normal mitotic cells. (d,e) and (f,g) bridges that contain only the H4K16ac-labelled male chromosome (H4K16ac +) and both H4K16ac-labelled and unlabelled chromosomes (H4K16ac +/−). (h) A bridge without H4K16ac signals (H4K16ac −). Yellow and light blue arrows indicate H4K16ac-labelled and non-labelled chromosomes, respectively. Scale bar, 5 µm (c–h).

Figure 4.

Neural development of infected embryos. (a,b) The ventral view of stage 15 female (n = 11) and male (n = 12) embryos infected with male-killing Wolbachia, stained for differentiated neural cells (Elav; red) and axonal projections (22C10; cyan). Single-channel images are also shown. (c–f) The lateral view of stage 15 female (n = 13) and male (n = 10) embryos infected with male-killing Wolbachia, stained as in (a) and (b). Boxed regions in (c) and (e) are highlighted in (d) and (f), respectively, with single-channel images. Scale bars, 50 µm.

Figure 5.

The localization and titre of male-killing Wolbachia. (a–d) The ventral view of stage 14 uninfected (a,b) and infected (c,d) male embryos stained for Wolbachia (HSP60; cyan) and differentiated neural cells (Elav; red). Single-channel images are also indicated. Twenty z-slices (yellow brackets in b and d) were projected to produce stack images in (a) and (c). In (b) and (d), z-section cut along the dotted yellow lines in (a) and (c) were three-dimensionally reconstructed, respectively. (e) Quantification of HSP60 puncta in uninfected and infected embryos. Sample sizes are shown at the bottom. No significant differences (n.s.) were detected between females and males in both cases (p > 0.05; Mann–Whitney U test). Box and dot plots are as in figure 1m. (f) Epithelial cells of a gynandromorphic embryo (n = 1) stained for Wolbachia (HSP60; green) and H4K16ac (magenta). Dotted yellow lines represent the border of male (with strong H4K16ac signals) and female (with weak H4K16ac signals) cell area. Scale bars, 20 µm (a,c) and 10 µm (f).