Report

A maleness gene in the malaria mosquito Anopheles gambiae

See allHide authors and affiliations

Science  01 Jul 2016:
Vol. 353, Issue 6294, pp. 67-69
DOI: 10.1126/science.aaf5605

Female-fatal product of the Yob gene

Mechanisms of sex determination are highly variable and complex in invertebrates. For vector-borne disease control strategies, it is important to understand the reproductive biology of the organisms involved. Krzywinska et al. have studied sex determination in species of Anopheles mosquitoes, whose females suck blood and can transmit malaria parasites (see the Perspective by Sinkins). One gene called Yob is found only on the Y chromosome of males. Yob is expressed in male embryos within 2 hours of eggs being laid, and transcription continues lifelong. If Yob is injected into embryos younger than 2 hours, only male mosquitoes emerge, because Yob's presence is lethal to female embryos.

Science, this issue p. 67; see also p. 33

Abstract

The molecular pathways controlling gender are highly variable and have been identified in only a few nonmammalian model species. In many insects, maleness is conferred by a Y chromosome–linked M factor of unknown nature. We have isolated and characterized a gene, Yob, for the M factor in the malaria mosquito Anopheles gambiae. Yob, activated at the beginning of zygotic transcription and expressed throughout a male’s life, controls male-specific splicing of the doublesex gene. Silencing embryonic Yob expression is male-lethal, whereas ectopic embryonic delivery of Yob transcripts yields male-only broods. This female-killing property may be an invaluable tool for creation of conditional male-only transgenic Anopheles strains for malaria control programs.

Animals have evolved a variety of mechanisms for sex determination. In insects, a common pattern requires a primary sex-determining signal that in early embryos activates a short cascade of sex-specifically spliced genes. The products of these genes modulate differentiation of an individual into either male or female, depending on the sex determination system (1). In the absence of the primary signal, a default pattern of gene splicing leads to the development of the opposite sex. Only the final element of the cascade, the doublesex gene, is well conserved, whereas upstream genes are highly evolutionarily labile (2) and are difficult to identify with comparative genomics tools.

Drosophila melanogaster has an XX/XY chromosomal sex determination system, in which a double dose of X chromosome–linked signal elements (XSE) represents the primary signal that initiates female-specific pre-mRNA splicing in a cascade of subordinate genes sex lethal (Sxl), transformer (tra), and doublesex (dsx) (3, 4). The resulting transcripts yield functional female-specific proteins, of which SXL regulates splicing of tra, and TRA, along with its cofactor TRA2, promotes splicing of dsx. In males, a single dose of XSE is insufficient to initiate this female-specific cascade, and the three genes are spliced by a default male-specific mode, with only dsx encoding a functional protein. The male and female forms of DSX regulate sex-specific morphologies by modulating the expression of genes involved in downstream sexual differentiation processes.

The dipterans Ceratitis capitata, Lucilia cuprina, and Musca domestica, and the beetle Tribolium castaneum, also possess an XX/XY chromosomal system. However, in these insects an unidentified Y chromosome–linked M factor constitutes a primary signal that determines maleness (58). In none of these species is Sxl involved in sex determination. Instead, maternal tra mRNA is sufficient to direct default female development in the absence of the Y-linked factor, whereas in males the M may be preventing establishment of tra function by an unknown mechanism (912). The XX/XY system incorporating a dominant Y-linked M factor is common in many insect groups (1315). However, with the exception of Nix (a homolog of tra2) from a mosquito Aedes aegypti (16), genes encoding M factor have remained enigmatic.

In anopheline mosquitoes, the only described element of the (morphological) sex determination cascade is dsx (17). To elucidate the initial steps of sex determination in the African malaria mosquito, Anopheles gambiae, we analyzed transcriptomes of male and female embryos (18), whose sexual identity was determined by polymerase chain reaction (PCR) (fig. S1). Separate pools of mRNA were sequenced, yielding ~500,000 Roche 454 reads from each sex. From the male pool, 21 reads uniquely mapped to the previously characterized scaffold AAAB01008227 derived from the A. gambiae Y chromosome (19).

The corresponding gene (hereafter called Yob) consists of three exons and produces transcripts differing in size because of intron retention (fig. S2). Yob likely represents the same gene as gYG2 (20), which partially overlaps Yob, but its annotation lacks over half of the Yob gene sequence. Among the A. gambiae Y-linked genes, only the Yob/gYG2 locus is Y-linked across the examined species from the A. gambiae complex (21), indicative of an important sexual identity function. Consistent with the chromosomal location, transcription of Yob is limited to males. Transcription begins in embryos between 2 and 2.5 hours after oviposition, before blastoderm formation (22) and expression of other Y-chromosome genes (20, 21), and continues throughout the life of the mosquito (Fig. 1A and fig. S3A). Onset of the Yob transcription is nearly contemporaneous with that of orthologs of even skipped (AGAP010279) and hunchback (AGAP010777) (fig. S3B), whose transcripts mark the beginning of zygotic expression in Drosophila (23, 24). We determined that Yob activation precedes establishment of sex-specific splicing of dsx by up to 6 hours. The female isoform of dsx is maternally deposited, but largely degraded in male embryos within 4 hours of oviposition, and only after complete degradation of the female isoform in males is a persistent pattern of sex-specific dsx splicing established (Fig. 1B). These findings are consistent with Yob acting upstream of dsx in the sex-determining hierarchy. If true, introduction of Yob transcripts into female cells should cause an alteration of dsx splicing pattern.

Fig. 1 Transcription of Yob and dsx in A. gambiae early embryos.

(A) Onset of zygotic Yob transcription. (B) Establishment of the sex-specific expression pattern of dsx. Time of RNA sample collection (in hours after oviposition) are shown above gel images; (-) negative control; ribosomal protein S7 (rpS7) transcript levels were used as a gel loading control.

We investigated the effect of in vitro–synthesized mRNA corresponding to the shortest, presumably mature A. gambiae Yob transcript isoform on dsx splicing in the A. gambiae Sua5.1 cell line, which lacks the Y chromosome and Yob expression (fig. S4). We observed a significant loss of the female and gain of the male dsx transcript isoforms in cells transfected with Yob mRNA, as compared with control nontransfected cells (Fig. 2, A, B, and D), or cells transfected with nonproductive forms of Yob (see below; Fig. 2D). Transfection experiments in larvae suggest that Yob exerts the same effect on dsx in vivo (fig. S5). This confirms that Yob is involved in the sex determination pathway as a direct or an indirect upstream regulator of dsx splicing.

Fig. 2 Yob is a protein-coding sex determination gene controlling dsx splicing.

(A) Reverse transcription (RT)–PCR analysis of dsx splicing pattern in the A. gambiae Sua5.1 cells transfected with in vitro–synthesized Yob mRNA, as compared with control nontransfected cells. M and F, A. gambiae male and female pupae; (-) negative control. Similar results were obtained in three independent experiments. (B) Digital representation of the relative amounts of dsx splice forms shown in (A). (C) RT-PCR analysis of ribosomal protein S7 transcript levels used as a sample loading control. (D) Splicing pattern of dsx in Sua5.1 cells transfected with native Yob transcripts; nonproductive forms of Yob, either lacking the methionine start codon, or containing a premature stop codon; and in nontransfected control cells. The data shown are mean + SD. *P < 0.001, one-sided Mann-Whitney test.

In the vast majority of protein-coding genes, different evolutionary constraints on individual codon positions result in synonymous substitutions observed with much higher frequency than the nonsynonymous ones (25). Based on this principle, we evaluated the protein-coding potential of Yob by comparing its sequence to PCR-isolated orthologous sequences from Anopheles arabiensis and Anopheles quadriannulatus, two members of the Anopheles gambiae complex. The analyzed region encompasses two open reading frames (ORFs) longer than 50 codons, of which only the shorter ORF bears a substitution pattern indicative of purifying selection (fig. S6). Guided by these results, we experimentally tested for the protein-coding potential of the shorter ORF. We transfected the Sua5.1 cells with two modified Yob transcripts containing putative nonsynonymous point mutations r.1A>C and [r.5U>A; 6U>G] and investigated the effects of the deduced amino acid changes Met1→Leu and Phe2→Stop, respectively, on dsx splicing. No shift in dsx splicing pattern was observed in cells transfected with transcripts either lacking an initiation codon or containing a premature stop codon, unlike the positive control cells transfected in parallel with the wild-type Yob transcripts (Fig. 2D and fig. S7). These results indicate that Yob encodes a short, 56–amino acid protein that may contain a helix-loop-helix motif (fig. S8). A protein with the same length and similar secondary structure, but nonconserved primary sequence, was proposed to be encoded by an early zygotically expressed Y-linked gene, GUY1, potentially involved in sex determination in Anopheles stephensi (26).

We injected Yob mRNA into nonsexed preblastoderm embryos of A. gambiae and its sibling species A. arabiensis to assess whether ectopic Yob transcripts affect mosquito sex ratios. To control for successful injection, we coinjected a plasmid with a green fluorescent protein (GFP) expression cassette (embryos that receive sufficient nucleic acids develop into larvae transiently expressing GFP; fig. S9). Surviving individuals were sorted at the larval stage into a GFP-positive and a GFP-negative group (Fig. 3A), and at the pupal stage mosquitoes were sexed. In both species, all the GFP-positive individuals developed as phenotypic males, whereas in the GFP-negative group and in the control group of A. gambiae embryos injected with GFP plasmid only, the sex ratio was unbiased (Fig. 3B). The GFP-positive males had the XY karyotype, as indicated by PCR; moreover, they had normally developed reproductive organs, produced motile sperm, and were fertile (table S1). Thus, ectopic delivery of Yob mRNA is lethal to genetic female embryos, but has no discernible effect on the sexual development of genetic males.

Fig. 3 Misregulation of Yob expression causes sex-specific lethality in embryos.

(A) Larvae hatched from embryos injected with an insufficient amount of (left), or successfully delivered (right), nucleic acids, as indicated by lack or presence of GFP expression in the midgut cells. Scale bars, 1 mm. (B) A summary of two and four independent Yob mRNA microinjection experiments in A. arabiensis and A. gambiae, respectively. (C) A summary of four independent Yob double-stranded RNA microinjection experiments in A. gambiae. *P < 0.0001; Fisher’s exact test.

In D. melanogaster, female-specific embryonic lethality is caused by loss-of-function mutations in genes located at the top of the sex determination cascade and invariably results from a misregulation of dosage compensation (3, 2729). The dosage compensation mechanism in that species equalizes X-linked gene products in XX and XY individuals by up-regulation of transcription from the male X chromosome (30). The process is tightly linked to the sex determination through female-specific repression by SXL (31, 32). Mutations in Sxl, or in genes involved in Sxl regulation, cause overexpression of X-linked genes and female death during embryogenesis (3, 2729). In females of T. castaneum, in which Sxl has no sex-determining role, TRA has been proposed to inhibit dosage compensation (12). In a lepidopteran, Bombyx mori, in which females are heterogametic (ZW) and males homogametic (ZZ), male embryo lethality, observed after a knockdown of a sex determination gene, Masculinizer (responsible for male-specific splicing of dsx and repression of transcription from Z chromosomes), has been similarly attributed to misregulation of dosage compensation (33).

Dosage compensation in A. gambiae also relies on hyperactivation of the X-chromosome genes in males (34). Therefore, in the presence of YOB, dosage compensation may be directly or indirectly induced in the XX embryos, leading to their death as a result of abnormal overtranscription from both X chromosomes (fig. S10). Conversely, in the absence of YOB, lack of dosage compensation and concomitant insufficient transcription from the X chromosome should be male-lethal, rather than leading to feminization of the XY individuals. Indeed, we observed highly significant male deficiency in mosquitoes surviving transient knockdown of Yob in nonsexed embryos (Fig. 3C). All tested female survivors had the XX karyotype.

Involvement of the Y chromosome factor in sex determination in Anopheles was first supported by the finding of a single, triploid Anopheles culicifacies male with the XXY sex chromosomes (15). Such a karyotype, apparently extremely rare and not reported elsewhere in numerous mutational and cytogenetic studies in Anopheles, seems to counter overtranscription of the X-chromosome genes as a cause of female embryo lethality observed in our study. However, the genetic background of this male must have been severely compromised through the mutagenic effects of irradiation on his parents (15). It thus seems likely that the XXY male may have carried multiple mutations, including those causing loss of function of dosage compensation machinery that allowed his survival to adulthood despite possessing two X chromosomes.

A. gambiae and A. arabiensis are the most important African vectors of human malaria. Control of the disease depends heavily on the use of insecticides, but emergence of resistance in mosquito populations severely threatens the effectiveness of these approaches (35). The sterile insect technique and other genetic control methods have been proposed to complement current efforts to suppress mosquito populations (36, 37). Such programs must incorporate male-only releases, because released females would contribute to pathogen transmission. However, no effective methods to sex the large number of Anopheles needed for releases currently exist. The fitness and mating competitiveness of adults is highly dependent on larval density; therefore, removing females from the release generation during the embryonic stage would drastically decrease the costs of rearing of high-quality males (38). Yob represents an excellent tool to be used in transgenic technology to conditionally eliminate female embryos and efficiently produce male-only generations of both malaria-transmitting Anopheles species (37).

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/353/6294/67/suppl/DC1

Materials and Methods

Figs. S1 to S10

Tables S1 to S3

References (3948)

REFERENCES AND NOTES

  1. Materials and methods are available on Science Online.
  2. Acknowledgments: We thank H. Ranson, A. Wilson, and anonymous reviewers for comments and L. Revuelta for technical assistance. This study was supported by funds from the Liverpool School of Tropical Medicine (LSTM), an MRC-LSTM Research Studentship, Wellcome Trust grant 089045/Z/09/Z (J.K. and G.J.L.), and funds from the Pirbright Institute (J.K.). Transgenic mosquitoes used for sexing were developed with Biotechnology and Biological Sciences Research Council grant BB/F021933/1 (G.J.L.). J.K. was partly supported by a subcontract CRSC120816JKUND of the NIH contract HHSN272200900039C. The RNA-sequencing data have been deposited in the Sequence Read Archive repository under the accession numbers SRR953486.1 and SRR953487.1.
View Abstract

Navigate This Article