Conceived and designed the experiments: Received Feb 16; Accepted Sep Copyright Gempe et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. This article has been cited by other articles in PMC.
Abstract Organisms have evolved a bewildering diversity of mechanisms to generate the two sexes. The honeybee Apis mellifera employs an interesting system in which sex is determined by heterozygosity at a single locus the Sex Determination Locus harbouring the complementary sex determiner csd gene. Bees heterozygous at Sex Determination Locus are females, whereas bees homozygous or hemizygous are males.
Little is known, however, about the regulation that links sex determination to sexual differentiation. To investigate the control of sexual development in honeybees, we analyzed the functions and the regulatory interactions of genes involved in the sex determination pathway. We show that heterozygous csd is only required to induce the female pathway, while the feminizer fem gene maintains this decision throughout development. By RNAi induced knockdown we show that the fem gene is essential for entire female development and that the csd gene exclusively processes the heterozygous state.
Fem activity is also required to maintain the female determined pathway throughout development, which we show by mosaic structures in fem-repressed intersexuals. We use expression of Fem protein in males to demonstrate that the female maintenance mechanism is controlled by a positive feedback splicing loop in which Fem proteins mediate their own synthesis by directing female fem mRNA splicing.
The csd gene is only necessary to induce this positive feedback loop in early embryogenesis by directing splicing of fem mRNAs. Finally, fem also controls the splicing of Am-doublesex transcripts encoding conserved male- and female-specific transcription factors involved in sexual differentiation. Our findings reveal how the sex determination process is realized in honeybees differing from Drosophila melanogaster. Author Summary Sexual differentiation is a fundamental process in the animal kingdom, and different species have evolved a bewildering diversity of mechanisms to generate the two sexes in the proper proportions.
Sex determination in honeybees Apis mellifera provides an interesting and unusual system to study, as it is governed by heterozygosity of a single locus harbouring the complementary sex determiner gene csd , in contrast to the well-studied sex chromosome system of Drosophila melanogaster. We show that the female sex determination pathway is exclusively induced by the csd gene in early embryogenesis.
Later on and throughout development this inductive signal is maintained via a positive feedback loop of the feminizer fem gene, in which the Fem protein mediates its own synthesis. The findings reveal how the sex determination process in honeybees is realized by the regulation and function of two genes differing from Drosophila. Introduction In Dzierzon reported that the sex in the honeybee Apis mellifera is determined by the fertilization and non-fertilization of eggs  , and this was more than 50 years before the discovery of sex chromosomes  , .
Dzierzon's key observation was that a virgin queen that has not taken a mating flight queens mate only while in free flight away from nest produces only male progeny. From this result he inferred that unfertilized eggs develop into males, whereas fertilized eggs differentiate into queens and worker bees, which was later confirmed by cytological studies .
The unfertilized eggs have a haploid set of 16 chromosomes when compared with fertilized eggs, in which 32 chromosomes were identified . Despite this, neither the fertilization process nor the haploid or diploid state of the eggs provides the sex determination signal in honeybees.
This is shown by the regular occurrence of males in inbred crosses that are derived from diploid, fertilized eggs  — . This finding led to the hypothesis of complementary sex determination in honeybees, a mechanism that was first provided by genetic studies in another hymenopteran insect, the parasitic wasp Bracon hebetor  , .
Fertilized eggs are either homozygous at the Sex Determination Locus SDL and differentiate into diploid males or are heterozygous and develop into females. The diploid males, however, don't survive in a bee colony as they are eaten by worker bees shortly after hatching from the egg. Fertile males are produced by the queen's unfertilized, haploid eggs that are hemizygous at SDL. The single-locus nature of complementary sex determination in honeybees was confirmed by genetic linkage analysis  ,  , physical mapping  , and the genetic linkage map .
Part of the SDL was characterized by positional cloning and a fine scale mapping approach that led to the identification of the complementary sex determiner csd gene . The gene encodes an SR-type protein and is a potential splicing factor. The csd gene satisfies the criteria of a primary signal of complementary sex determination : The latter has been shown in RNAi-induced knockdown experiments of csd.
Females treated with csd dsRNA develop entire male gonads, whereas the treatment of males had no sex-transforming effect. Disappointingly little, however, is known about the regulatory interactions and mechanisms that link sex determination to sexual differentiation. So far we have no evidence that SDL encodes another gene that, in conjunction with csd, operates to establish the sex determined state by heterozygosity. We have recently isolated the entire genomic region of SDL and identified the feminizer fem gene, which is the ancestral progenitor gene from which csd derived by gene duplication  and lies 12 kb upstream of csd.
The fem gene is required for female development as shown by the sexual transformation of the head of fem-repressed females.
Fem activity is not achieved by heterozygosity. Instead, the fem pre-mRNAs are sexually processed into the productive female mode. In the current study we characterized the sex-transforming function of other candidate genes located at SDL. We then investigated the function of sex-determining genes in controlling all aspects of development. Our previous studies were restricted to the control of basic aspects of soma differentiation  ,  , but the signals that specify the sex of germ cells may differ from those utilized in the soma .
For example, in Drosophila the gene transformer tra , which is the likely ortholog of the fem gene  , is required for the sexual development of the soma but not directly for the sexual fate of germ line cells  — . We repressed sex-determining genes in early embryogenesis and scored the sexual development of subtle soma and germ line characters. Finally, we analyzed the regulatory interactions of the sex-determining genes and addressed the question of how these interactions are utilized to maintain sexual fate throughout development.
In a previous study we proposed that continuously expressed csd is a potential source of information to maintain sexual fate throughout development . In order to study these regulatory interactions, we either repressed or expressed genes and assayed the sexual expression of target genes. Our findings reveal how the regulation and function of the csd and the fem gene realizes the sex determination process throughout development.
The involvement of two SDL genes, fem and csd, in sexual development have thus far been characterized  , . We hypothesized that this region may harbour additional genes that operate in conjunction with csd in the establishment of the primary sex determined state. Three other genes at SDL have been previously predicted, but their involvement in sex determination is unknown.
We explored whether the SDL harbours further genes. Potential exons were identified by exon-finding algorithms and homology searches to EST and gene databases. Exons testing positive in these experiments were combined, but no further transcription units beside the three previous predicted genes were identified. The same sequences were isolated from both males and females implying that the transcripts are not sex-specifically processed.
We obtained 3, bases of the transcript of gene GB, which divides into four exons Figure 1B. This partial transcript encodes a amino acid protein with partial similarity to a domain from a Tribolium castaneum hypothetical protein LOC of unknown function. We isolated bases of the transcript of gene GB, which splits into six exons Figure 1B and which is located on the opposite strand from the other SDL genes.
The partial ORF encodes a protein of amino acids. The protein contains a DUF domain of unknown function that is conserved from worms to humans. We injected dsRNA into male and female syncytial embryos in order to repress transcripts of the new genes and recorded gonad differentiation of 5th instar larva. We used gonad differentiation as an informative indicator of sex determination as it is induced early in development .
We also analyzed the fem gene, for which we have no information on gonad differentiation, and csd, which served as a control for entirely switched gonad development. The syncytial female and haploid male embryos were obtained from single-male drone inseminated queens and virgin queens, respectively.