A genetic basis for facultative parthenogenesis in Drosophila

Abstract

Facultative parthenogenesis enables sexually reproducing organisms to switch between sexual and asexual parthenogenetic reproduction. To gain insights into this phenomenon, we sequenced the genomes of sexually reproducing and parthenogenetic strains of Drosophila mercatorum and identified differences in the gene expression in their eggs. We then tested whether manipulating the expression of candidate gene homologs identified in Drosophila mercatorum could lead to facultative parthenogenesis in the non-parthenogenetic species Drosophila melanogaster. This identified a polygenic system whereby increased expression of the mitotic protein kinase polo and decreased expression of a desaturase, Desat2, caused facultative parthenogenesis in the non-parthenogenetic species that was enhanced by increased expression of Myc. The genetically induced parthenogenetic Drosophila melanogaster eggs exhibit de novo centrosome formation, fusion of the meiotic products, and the onset of development to generate predominantly triploid offspring. Thus, we demonstrate a genetic basis for sporadic facultative parthenogenesis in an animal.

Keywords: Drosophila; asexual; development; embryogenesis; facultative parthenogenesis; functional genomics; genome; polyploid; reproduction; transcriptiomics.

Figure 1:. Comparison of sexually reproducing and parthenogenetic D. mercatorum genomes to each other and D. melanogaster.

(A) Reproductive modes of D. mercatorum. (B) Sexually reproducing and parthenogenetic D. mercatorum genome assembly data, quality control, and annotation metrics. (C) Sexually reproducing D. mercatorum genome assembly aligned against the D. melanogaster reference genome (release 6). Alignment is over 75.28Mbp and the gap compressed identity is 75.85%. A raw version of this panel is presented in Figure S2A. Purple dots/lines represent sequences matching against the forward strand and blue the reverse. (D) In situ localization of the indicated genes by HCR onto mitotic chromosomes of third instar larval neuroblasts from sexually reproducing and parthenogenic D. mercatorum. The karyotypes match except for a polymorphism in size for the 4th chromosome arm/Muller element F. Gene probes were selected to represent the indicated contigs. Chromosomes showing localization of the gene probe are outlined with a white line. The indicated chromosome was marked in all karyotypes analyzed (n≥42, N≥3). DAPI/DNA (white). Scale bar, 1μm. (E) Schematic of the arrangement of the chromosome arms in sexually reproducing and parthenogenetic strains of D. mercatorum and in D. melanogaster. Related to Figure S1, S2, and DataS1

Figure 2:. D. mercatorum genome comparison and physical mapping of the largest D. mercatorum contigs on polytene salivary gland chromosomes.

(A) Alignment of the parthenogenetic D. mercatorum genome against the sexually reproducing genome. Red arrows indicate inversions/translocations. Alignment is over 151.14Mbp, including 142.25Mbp of matching base pairs. The gap compressed identity is 98.67%, while the “block” identity is 94.11%. The smallest contigs were removed from the visualization. Purple dots/lines represent sequences matching against the forward strand and blue the reverse strand. (B) Mapping of the largest 14 contigs of the sexually reproducing and parthenogenetic D. mercatorum genomes onto salivary gland polytene chromosomes stained with DAPI. The contig and the chromosome arms are indicated with the arrows pointing to the HCR in situ hybridization band. These gene localizations show that the computationally assigned chromosome arms are appropriately grouped together in the biological samples. In situ hybridizations were carried out using the indicated HCR DNA probes corresponding to indicated genes. DNA/DAPI (white). Scale bars, 10μm. (C) Schematic of mapping of the contigs onto D. mercatorum chromosomes and their corresponding positions on D. melanogaster chromosomes.

Figure 3:. Facultative parthenogenesis and reproductive potential in Drosophila.

(A) Double gene variant combinations resulting in facultative parthenogenesis in D. melanogaster. Summary of Data S6B. Only significantly facultative parthenogenetic combinations are shown. It is important to note that the genotype is heterozygous for the alleles indicated; all but slmb⁻ are homozygous viable (Data S6B). p value was calculated using the Fisher’s exact test. (B) The three gene variant combinations leading to significant development in induced facultative parthenogenetic D. melanogaster and facultative parthenogenetic D. mercatorum were screened for a second generation of parthenogenesis. TM6B is shown in parentheses because the balancer was not always present as the alleles it was balancing were homozygous viable. (C) Parthenogenetically produced offspring from the induced facultative parthenogenetic D. melanogaster lines or produced naturally from the facultative parthenogenetic D. mercatorum were backcrossed to the parental line to determine if they could produce progeny in the F2 generation. GFP-polo⁺ is a transgene on the X, Desat2⁻ is a 5’UTR deletion, Desat1⁻ is a TE insertion, Myc^dp is a gene duplication on the 3rd chromosome, and slmb⁻ is a null allele. Related to Figure S4 and Data S6B.

Figure 4:. The inception of parthenogenesis Drosophila embryos.

(A-C) Schematic showing the categorizations of development in embryos used in this study: formation of polar body (PB) aggregate; first mitosis; and subsequent mitotic divisions shown alongside examples of facultatively parthenogenetic D. mercatorum embryos of each category. DNA/DAPI (white) and acylated-Tubulin (ac-Tub) (magenta). (D) Histogram displaying the proportion of fertilized, sexually reproduced, facultative parthenogenetic, and parthenogenetic D. mercatorum embryos that have completed meiosis and only have a PB aggregate present; initiated the first mitosis; or have undertaken multiple mitotic nuclear divisions. (E) Histogram displaying the proportion of wild-type fertilized and unfertilized D. melanogaster embryos that have completed meiosis and only have a PB aggregate present; initiated the first mitosis; or have undertaken multiple mitotic nuclear divisions. (F) Histogram displaying the proportion of unfertilized D. melanogaster eggs/embryos from GFP-polo⁺, Desat2⁻, Myc^dp Desat2⁻, GFP-polo⁺;Desat2⁻/TM6B, and GFP-polo⁺;Myc^dp Desat2⁻/TM6B mothers that have completed meiosis and only have a PB aggregate present; initiated the first mitosis; or have undertaken multiple mitotic nuclear divisions. (G-H) D. melanogaster embryos from GFP-polo⁺;Desat2⁻/TM6B and GFP-polo⁺;Myc^dp Desat2⁻/TM6B mothers that have undertaken multiple mitotic nuclear divisions. GFP-polo⁺ is a transgene on the X, Desat2⁻ is 5’UTR deletion, and Myc^dp is a locus duplication on the 3rd chromosome. DNA/DAPI (white), and α-Tubulin (magenta). Scale bars, 10μm. Fisher’s exact test used to calculate p values. Related to Figure S5.

Figure 5:. Initiation of parthenogenesis in genetically modified D. melanogaster embryos.

(A) Stacked column chart showing the proportion of embryos that had a central DNA puncta, polar body (PB) aggregate, 1 single metaphase nuclei, 1 single metaphase nuclei and PB aggregate, 2–4 mitotic nuclei, 2–4 mitotic nuclei and PB aggregate, 5 or more mitotic nuclei, and 5 or more mitotic nuclei and PB aggregate from fertilized and unfertilized wild-type D. melanogaster embryos and embryos from GFP-polo⁺, Desat2⁻, Myc^dp Desat2⁻, GFP-polo⁺;Desat2⁻/TM6B, and GFP-polo⁺;Myc^dp Desat2⁻/TM6B mothers. Fisher’s exact test used to calculate p values for all categories that had PB aggregates. (B) PB aggregate in an unfertilized embryo from a GFP-polo⁺ mother. (C) PB aggregate in an unfertilized embryo that initiated mitosis from a GFP-polo⁺ mother. (D) A nucleus in mitosis in an embryo from a GFP-polo⁺; Desat2⁻/TM6B mother. Arrowheads point to ectopic centrosomes. When mitosis is initiated multiple GFP-Polo puncta form, and GFP-Polo becomes associated with the DNA and mitotic spindle in the PB aggregates that initiate mitosis and nuclei that continue mitosis alike. (E) Nuclei in an embryo, from a GFP-polo⁺; Myc^dp Desat2⁻/TM6B mother, that has passed the cellularization stage of development. DNA/DAPI (white), GFP-Polo (cyan), Histone 2A (magenta), and acetylated-Tubulin (ac-Tub) (yellow). (F) Cellularized embryo, from a GFP-polo⁺;Desat2⁻/TM6B mother, at extended germ-band stage. DNA/DAPI (white), centrosomes/cnn (cyan), and α-Tubulin (magenta). (G) Mitotic karyotype chromosomes of third instar larval neuroblasts from sexually reproducing and parthenogenic GFP-polo⁺;Myc^dp Desat2⁻/TM6B D. melanogaster mothers. Neuroblasts in animals derived by sexual reproduction were all diploid (n=12), whereas 50% of the parthenogenetically-derived animals were triploid and 8%, tetraploid (n=12). Individual chromosomes are labelled in white lettering. Scale bars, 10μm in a-f and 1μm in g. GFP-polo⁺ is a transgene on the X, Desat2⁻ is 5’UTR deletion, and Myc^dp is a locus duplication on the 3rd chromosome. Related to Figure S6.

Figure 6:. Proposed model for how Myc, Polo, and Desat2 cause parthenogenesis in D. melanogaster.

(A) In a fertilized egg the paternal pronucleus (pn) from the sperm and the maternal pn fuse and initiate cell divisions that are organized by centrosomes organized around a centriole and procentriole-like body provided by the sperm. (B) Unfertilized eggs having greater maternal Myc expression exhibit increased expression of genes for cell cycle regulatory proteins. The polar bodies (PBs) become positioned away from the egg membrane and can enter mitosis more readily as a result of the Desat2⁻ mutation. Elevated Polo kinase promotes the de novo formation of centrosomes and facilitates entry into the mitotic cycles. As a consequence, the PB nuclei are exposed to all the components necessary to initiate development and the syncytial nuclear division cycles, thus initiating parthenogenesis. Once zygotic transcription is initiated, the elevated expression of Myc promotes cell proliferation favoring diploid or polyploid cells.