Fig 1.
Deletions of Tmy and Nmy hpRNAs reveal distinct and profound male reproductive defects.
(A) Biogenesis of hpRNAs into siRNAs via core RNAi factors Dcr-2/AGO2. We used CRISPR/Cas9 and multiple gRNAs to generate Dsim tmy (B-E) and nmy (F-J) deletion alleles for phenotypic analysis. (B) Tmy genomic region showing RNA-seq and small RNA-seq data from w[XD1] in the vicinity of the Tmy hpRNA. We used multiple gRNAs to delete tmy and replace it with a ubi-GFP marker, which is visible throughout the animal (C). (D) Assays of male reproduction reveal complete sterility of tmy[GFP] mutants at 3 different temperatures. (E) Nmy genomic region showing RNA-seq and small RNA-seq data from w[XD1] in the vicinity of the Nmy hpRNA. We used multiple gRNAs to delete nmy and replace it with a 3xP3-DsRed marker (F). (G) Assays of male reproduction reveal nearly complete loss of male progeny by nmy[DsRed] mutants at 18°C. (H) Summary of nmy (“Winters” suppressor) and tmy (“Durham” suppressor) knockout phenotypes in the w[XD1] background. Statistical tests were Wilcoxon rank-sum test of differences between medians. hpRNA, hairpin RNA; siRNA, small interfering RNA.
Fig 2.
Overviews of the Drosophila testis and male meiotic divisions.
(A) Schematic organization of the Drosophila testis. Spermatogenesis begins at the apical tip and progresses along in a spatiotemporal manner. The apical tip contains the hub, which contains a small number of self-renewing germline stem cells (purple). By asymmetric division, the stem cells give rise to gonial cells, which will in turn undergo 4 synchronous mitotic divisions with incomplete cytokinesis, forming a cyst of 16 interconnected primary spermatocytes. After a period of growth, all spermatocytes undergo meiosis I and II in synchrony (detailed in the inset, B) to form 64 interconnected spermatids. The 64 spermatids differentiate synchronously into mature spermatozoa (only 8 are shown for clarity), during which their nuclei undergo a morphological change from round to needle-shaped. This is achieved by replacing histones (red) with transition proteins (orange), and then by SNBPs (i.e., protamines, green), yielding nuclei with a compact and elongated shape. At the same time, each cell produces a flagellum that extends almost the entire length of the testis. At the end of the histone-to-protamine transition, spermatids are individualized by actin cones (blue) that assemble around the spermatids and move coordinately from head to tail, removing excess cytoplasm. The actin cones end up in a waste bag near the apical part of the testis. Finally, mature spermatozoa coil and are released at the proximal end of the testis into the SV. (B) Drosophila male meiosis. Nuclei are outlined, red shading indicates chromatin. For clarity, only 1 primary spermatocyte, out of the 16 present inside a cyst, is shown. Meiosis produces 4 haploid (1N) daughter cells from a diploid primary spermatocyte (2N). After S and G2 phases, the homologous chromosomes pair up. Three territories are formed, separating the non-homologous chromosomes from each other. Two territories are formed by the 2 autosomes (chromosomes II and III) and the third by the X, Y and fourth chromosomes. The chromosomes condense in prophase I and align on the metaphase plate. Then, homologous chromosomes are separated. At the end of meiosis I, 2 daughter cells are formed, which proceed to meiosis II to separate sister chromatids and then give 2 haploid (1N) daughter cells each. SNBP, sperm nuclear basic protein; SV, seminal vesicle.
Fig 3.
Cytological basis of male reproductive defects in nmy and tmy knockouts.
(A, B) Confocal images of whole mount testes from control flies (w[XD1] as wild-type control) and mutants (nmy, tmy, and ago2) stained with a pan-Histone antibody (red) and phalloidin (F-actin green) to reveal ICs and DAPI to label nuclei (blue). (A) Whole testis images, with the stem cell regions labeled with asterisks and seminal vesicles labeled with SV. (A) Control w[XD1] testis shows the orderly progression from mitotic (1), meiotic (2), and differentiation (3) regions. (B) nmy mutants exhibit grossly normal spermatogenesis with normal ICs (arrows, A’). Spermatogenesis is highly disturbed in tmy and ago2 homozygous mutants, with disorganized cysts and aberrant (tmy) or absent (ago2) IC (A’). The empty SVs. Scale bars: 100 μm. (B) Focus on the SV shows they are filled with sperm in control and nmy mutants, as seen with needle-shaped DAPI staining, but are empty in tmy and ago2 mutants; the large round nuclei at the borders correspond to the cells of the somatic wall. (C, D) Cysts of 64 spermatids in control (w[XD1]) and nmy mutants (grown at 18°C). The stages of the histone-to-protamine transition indicated were determined with pan-Histone staining and nuclear shape. Histones are eliminated in both control (C) and nmy (D) spermatids. However, individualization of control spermatid nuclei yields regular needle shapes that stain uniformly with DAPI, whereas about half of nmy mutant nuclei fail to elongate and exhibit abnormal shape (arrow). In addition, DAPI staining in nmy mutant nuclei is uneven, with both less (triangle) or more (DAPI foci, arrowhead) dense regions, suggesting aberrant chromatin organization. Scale bars: 10 μm. (E, F) Meiosis in control and tmy homozygous testes stained for histones (magenta) and α-tubulin (green). The stages indicated were determined by counting nuclei number in cysts (16 for meiosis I and 32 in meiosis II). In tmy mutants, chromosomes are fragmented and chromatin bridges are observed in anaphase and telophase of both meiosis I and meiosis II. Scale bars: 10 μm. IC, individualization complex; SV, seminal vesicle.
Fig 4.
Specific and overlapping roles for Nmy and Tmy hpRNAs in suppressing Dox family targets and male reproduction.
(A) Gain-of-function assays of hpRNAs and luciferase targets of PDox1 and PDox2 in D. melanogaster S2 cells. Only wild-type Tmy has capacity to repress both PDox sensors. (B) Loss-of-function assays of testis RNA from D. simulans hpRNA knockouts. qPCR assays show that nmy mutants specifically derepress Dox and MDox, while tmy mutants specifically misexpress PDox genes (amplicons could not distinguish these highly similar transcripts). (C, D) Genetic interactions support an endogenous role for Tmy in sex ratio control. (C) tmy/+ heterozygous fathers exhibit even sex ratios in their progeny, similar to w[XD1] fathers. (D) nmy mutant males are temperature sensitive and their progeny sex ratio bias decreases from 90%–95% at 17°C to modest distortion at 25°C. Removal of 1 allele of tmy clearly enhances nmy SR defects at 22°C and 25°C. (E) Cytological analysis demonstrates that tmy is a dose-sensitive enhancer of nmy during spermatogenesis. Cysts of 64 differentiating spermatid nuclei before and during individualization were stained for histones and DAPI. In control w[XD1] and tmy/+, spermatids elongate normally. At restrictive temperature (18°C), about half of nmy spermatids fail to elongate normally, but this is strongly suppressed at permissive temperature (25°C). Loss of 1 tmy allele in nmy mutants at 25°C phenocopies strong nmy defects seen at 18°C. Scale bar: 10 μm. (F) Summary of specific and cross-regulatory repression of de novo X-linked Dox family genes by autosomal hpRNAs Nmy/Tmy. The raw data for the luciferase sensor assays are provided in S1 Data and raw data for qPCR assays in S2 Data. hpRNA, hairpin RNA.
Fig 5.
Evidence that Dox family loci actively drive sex ratio and sterility phenotypes in hpRNA mutants.
(A) Screening of D. simulans wild lines identified X chromosomes with natural variation in Dox family loci, including deletion alleles of Dox, MDox, and PDox2 (but not of PDox1). (B) Amplicon sequencing confirms that the transcribed regions of Dox, MDox, and PDox2 are deleted in these natural alleles. (C–E) The X chromosomes of the indicated wild lines were crossed into nmy or tmy mutants and assayed for modification of these hpRNA knockout phenotypes. Each dot represents sex ratio or fertility quantification of progeny of an individual male. (C) Sex ratio assays in nmy knockout males at 17°C. In control w[XD1], loss of nmy results in near-complete loss of male progeny, but introduction of X chromosomes from either MD15 or NS40 restores equal sex ratio to nmy knockouts. (D) Fertility assays in tmy knockout males at 22°C. Loss of tmy confers complete sterility in both w[XD1] and MD15, but NS40 restores strong fertility to tmy knockout males. (E) Sex ratio assays in tmy knockout males at 22°C. Sex ratio cannot be assessed in sterile w[XD1]; tmy or MD15/Y; tmy males, but NS40/Y; tmy males sire equal numbers of male and female progeny. (F–I) Effect of X suppressor chromosomes on the nmy loss-of-function mutants. Cysts of 64 differentiating spermatid nuclei during individualization were stained with DAPI (cyan) and an anti-pan-histone antibody (red); scale bar: 10 μm. (F) Control cysts in w[XD1] with normal organization. (G) At the restrictive temperature (18°C), nmy cysts show high frequency of misshapen nuclei. (H, I) The MD15 (H) and NS40 (I) X-chromosomes restore homogeneously elongated nuclei to nmy cysts. (J–M) Effect of X suppressor chromosomes on tmy loss-of-function mutants. Whole mount testes stained with phalloidin to reveal individualization actin cones (F-actin, green), an anti-pan-histone antibody (red) and DAPI (blue); scale bar: 100 μm. (J) Control w[XD1] shows normal testis development with mature sperm in the SV. (K) tmy mutant shows scattered ICs and the SV is devoid of sperm; the large round nuclei in this optical section correspond to the somatic sheath. (L) NS40/Y; tmy with low fertility exhibits partial restoration of germline development (n = 2/14). (M) NS40/Y; tmy with normal fertility shows corresponding rescue of normal spermatogenesis and abundant sperm in the SV (n = 12/14). The uncropped genotyping gels are provided in S1 Raw Images. hpRNA, hairpin RNA; IC, individualization complex; SV, seminal vesicle.
Fig 6.
Dox family loci encode protamine-like proteins that disrupt chromosome dynamics.
(A) The HMG Box domains of Dox family are more similar to protamine than other testis HMG Box proteins. (B) However, Dox loci lack the protamine N-terminus that mediates cysteine cross-linking and instead contain putative C-terminal transmembrane domains. (C–G) Localization of Dox peptides in nmy mutants. Whole mount testes stained for V5 (green), and DAPI (white) shows that Flag-V5-GFP-Dox is derepressed in nmy/nmy compared to nmy/+ control. In mutants, tagged Dox is initially detected in primary spermatocytes and persists through meiosis. Tagged Dox was mostly cytoplasmic where it may accumulate on cytoskeletal structures and was not detected in postmeiotic stages. (H–L) Flag-V5-GFP-PDox2 protein is derepressed in tmy/tmy compared to tmy/+ control. Tagged PDox2 is detected in the cytoplasm of meiotic cells and is also enriched at cell membranes. In telophase of meiosis I and II, PDox2 forms nuclear foci and accumulates on chromatin bridges that connect meiotic nuclei. (M–P) Analysis of protamine-RFP shows normal transition in wild type (M) and nmy mutant (N). Spermatogenesis is defective in tmy but protamine is incorporated (O), while the transition to protamine is blocked in ago2 testis (P). SV, seminal vesicle.
Fig 7.
Summary of intrinsic and recurrent sex chromosome conflicts and their suppression by endogenous RNAi.
(Top) Wild-type individuals harboring intragenomic conflict systems typically exist in a cryptic state, in which de novo sex-linked meiotic drive loci are silenced posttranscriptionally by autosomal hpRNA loci that generate endogenous siRNAs. Their existence of meiotic drivers is hidden in this state, since animals have normal reproductive performance, even though there can be numerous X-linked drive loci (e.g., Dox/MDox/PDox1/PDox2 in D. simulans). (Middle) The deleterious activities of meiotic drive loci are unleashed upon mutation of the hpRNA suppressors (e.g., Tmy and Nmy). In D. simulans males, loss of nmy results in near-complete absence of male progeny, whereas loss of tmy causes complete sterility. Genetic interaction tests show that Tmy protects against sex ratio drive, but evidently in the contemporary derepressed state, X-linked drivers likely off-target to the X or autosome yielding absence of female gametes as well. (Bottom) We can infer that Dox family meiotic drives are selfish, rather than contributing to normal reproduction, since removal of all of these loci restores normal fertility and sex balance to hpRNA mutants. In a sense, this returns D. simulans to a D. melanogaster-like state, which lacks the Nmy/Tmy hpRNAs as well as all X-linked Dox family genes. hpRNA, hairpin RNA; siRNA, small interfering RNA.