Sexual and asexual oogenesis require the expression of unique and shared sets of genes in the insect Acyrthosiphon pisum

Abstract

Background: Although sexual reproduction is dominant within eukaryotes, asexual reproduction is widespread and has evolved independently as a derived trait in almost all major taxa. How asexuality evolved in sexual organisms is unclear. Aphids, such as Acyrthosiphon pisum, alternate between asexual and sexual reproductive means, as the production of parthenogenetic viviparous females or sexual oviparous females and males varies in response to seasonal photoperiodism. Consequently, sexual and asexual development in aphids can be analyzed simultaneously in genetically identical individuals.

Results: We compared the transcriptomes of aphid embryos in the stages of development during which the trajectory of oogenesis is determined for producing sexual or asexual gametes. This study design aimed at identifying genes involved in the onset of the divergent mechanisms that result in the sexual or asexual phenotype. We detected 33 genes that were differentially transcribed in sexual and asexual embryos. Functional annotation by gene ontology (GO) showed a biological signature of oogenesis, cell cycle regulation, epigenetic regulation and RNA maturation. In situ hybridizations demonstrated that 16 of the differentially-transcribed genes were specifically expressed in germ cells and/or oocytes of asexual and/or sexual ovaries, and therefore may contribute to aphid oogenesis. We categorized these 16 genes by their transcription patterns in the two types of ovaries; they were: i) expressed during sexual and asexual oogenesis; ii) expressed during sexual and asexual oogenesis but with different localizations; or iii) expressed only during sexual or asexual oogenesis.

Conclusions: Our results show that asexual and sexual oogenesis in aphids share common genetic programs but diverge by adapting specificities in their respective gene expression profiles in germ cells and oocytes.

Figure 1

Effect of prenatal application of kinoprene on the type of embryonic ovaries. Acetone or kinoprene dissolved in acetone were topically applied to sexuparae individuals, 24 h after the fourth instar molt. Progeny batches were collected after initiation of larviposition, for five days every 24 h. The reproductive phenotypes of the progeny were determined and the percentage of each morph (asexual females, ambiphasic females that contain a mixture of eggs and embryos, sexual females and males) is indicated. In control conditions (acetone), sexuparae sequentially produced sexual females and males: only sexual females were produced during the first three days after the onset of larviposition; males appeared the fourth day. Sexuparae treated with kinoprene produced exclusively parthenogenetic progeny the first day after the onset of larviposition. On day two, almost all the progeny were parthenogenetic. On day three, the majority of the progeny were still parthenogenetic. On days four and five, sexual morphs represented the majority of the progeny.

Figure 2

Transcripts with identical localizations in both sexual and asexual ovaries. Histone H1, gle1, gld2, bicC, pop2, arl6ip1, ACYPI25088, ACYPI007465 and ACYPI10052 transcripts were detected in germaria (hollow arrowheads) and oocytes (black arrowheads) of both sexual and parthenogenetic ovaries. Sense riboprobes were used as negative controls. For gle1, bicC, ACYPI25088, ACYPI007465 and ACYPI010052, a residual signal was observed in the developing embryo germ band (arrows). Bar scale: 50 μm.

Figure 3

Transcripts with different localizations in sexual and asexual ovaries. Uhrf1 transcripts were detected in germaria (hollow arrowheads) and oocytes (black arrowheads) of asexual ovaries whereas a weak signal was detected in sexual oocytes (black arrowheads). Orb transcripts were detected in the basal parts of germaria in asexual and sexual ovaries (hollow arrowheads) with distinct localizations. In asexual germaria, these cells correspond to oocytes, whereas the cellular structures containing orb transcripts in sexual ovaries are undetermined. Sense riboprobes were used as negative controls. Bar scale: 50 μm.

Figure 4

Detection of orb transcripts in sexual and asexual ovaries by fluorescent WISH. (A) orb riboprobes were detected with an antibody coupled to alexin fluorochrome (red) and DNA was stained with DAPI (white). In sexual ovaries (upper pictures), orb transcripts were detected at the basal part of the germaria (hollow arrowheads). In asexual ovaries (lower pictures), orb transcripts were detected in oocytes (white arrowheads) before ovulation (stage 0), and after ovulation (stage 1). (B) Actin filaments and DNA in sexual ovaries were stained with phalloidin (green) and propidium iodide (red), respectively. Cells at the basal part of the sexual germaria possibly correspond to presumptive sexual oocytes (double arrows). Orb sense riboprobes were used as negative controls. Bar scale: 30 μm.

Figure 5

Transcripts detected specifically in one type of ovary. Lsd1 transcripts were detected specifically in the germaria (hollow arrowheads) and haploid oocytes (black arrowheads) of sexual ovaries. lodestar, cyclin J, ACYPI39770, ACYPI54656 transcripts were detected specifically in asexual ovaries. Lodestar signal was detected specifically in asexual oocytes (black arrowheads). Cyclin J, ACYPI54656 and ACYPI39770 signals were detected in germaria (hollow arrowheads) and in oocytes (black arrowheads) of asexual ovaries. ACYPI54656 and ACYPI39770 showed a residual signal in the germ bands (arrows) of developing embryos. Sense riboprobes were used as negative controls. Bar scale: 50 μm.