The fruitless gene affects female receptivity and species isolation

Abstract

Female mate rejection acts as a major selective force within species, and can serve as a reproductive barrier between species. In spite of its critical role in fitness and reproduction, surprisingly little is known about the genetic or neural basis of variation in female mate choice. Here, we identify fruitless as a gene affecting female receptivity within Drosophila melanogaster, as well as female Drosophila simulans rejection of male D. melanogaster. Of the multiple transcripts this gene produces, by far the most widely studied is the sex-specifically spliced transcript involved in the sex determination pathway. However, we find that female rejection behaviour is affected by a non-sex-specifically spliced fruitless transcript. This is the first implication of fruitless in female behaviour, and the first behavioural role identified for a fruitless non-sex-specifically spliced transcript. We found that this locus does not influence preferences via a single sensory modality, examining courtship song, antennal pheromone perception, or perception of substrate vibrations, and we conclude that fruitless influences mate choice via the integration of multiple signals or through another sensory modality.

Keywords: Drosophila melanogaster; Drosophila simulans; behavioural isolation; fruitless; prezygotic isolation; speciation.

Figure 1.

Genetic mapping identifies the fruitless gene as influencing female rejection of heterospecific males. (a) Schematic of the 3rd chromosomes (blue and red lines) in deficiency complementation mapping. D. melanogaster (mel, top left) females bearing a deficiency or disruption (broken blue line) are crossed to D. simulans (sim, bottom left; red lines) produce hybrids inheriting intact chromosomes from both species (bottom right; sim/mel^Bal) and hybrids inheriting a disruption in the D. melanogaster chromosome (top right; sim/mel^Dis). Gene(s) within this disrupted region are only expressed from the sim homologue's alleles. (b) Genetic constructs used to assay female rejection of heterospecific males. Rectangular bars represent deficiencies; blurred ends represent imprecisely known breakpoints; scale is approximate. The three deficiencies at top (marked with asterisk) are from [19]. Orange is statistically significant (p < 0.05); grey is not statistically significant. Arrowed box represents location and direction of fru gene compared with the deficiencies; note that image is shown in orientation relative to fru. At bottom, fru is expanded and shown 5′–3′ to represent the fru P1–P5 first exons, common exons C1–C5, and 3′ exons A–D; boxes are exons, black boxes are coding. The relative locations of transposable element insertions are represented by numbered inverted triangles: 1 = fru^GAL4, 2 = fru^MI05459, 3 = fru^NP0021, 4 = fru^MI01850, 5 = fru^KG00116; dashed lines represent targeted fru^ΔP1 and fru^ΔP2 deletion locations. Image not to scale. (c) Representation of fru transcripts. Adapted from [27]. (d) Proportion mated when paired with D. melanogaster males for control pure species females (1 h assay) without a disruption (mel/mel^Bal), with a disruption (mel/mel^Dis), and hybrid females (24 h assay) without a disruption (sim/mel^Bal), when compared with hybrid females with a disruption (sim/mel^Dis). Comparisons where sim/mel^Dis females have a significant reduction in mating have p-values shown in bold. (e) Transcript presence in flies homozygous for the Minos insertion fru^MI05459 compared with flies that have had the Minos element excised. The housekeeping gene RpL32 is used as a control. Each sample was done with a biological replicate. (f) Female receptivity is rescued when the Minos element within fru^MI05459 is removed (n = 32). (Online version in colour.)

Figure 2.

Proportion of females mating when they contain fru^ΔP1 (independent deletions #1 and #2) or fru^ΔP2 targeted deletions of their melanogaster allele. Note that, for all hybrid females, the simulans fru allele is intact. (a) fru^ΔP1 (#1; n = 39) or fru^ΔP2 (n = 50) hybrid females, compared with hybrid Canton-S/simulans controls in a 24 h assay. (b) Hybrid females with a disruption (sim/mel^Dis; light purple) compared with hybrid females without a disruption (sim/mel^Bal; dark purple), pure species females with (mel/mel^Dis; light blue) and without (mel/mel^Bal; dark blue) a heterozygous disruption. Pure-species females assayed for 1 h; hybrid females assayed for 24 h. p-Values are for the interaction term between species (sim/mel versus mel/mel) and genotype (Dis versus Bal). (c) Pure species D. melanogaster females paired with D. melanogaster males for 1 h (n = 30); females are wild-type (Canton-S) or have a homozygous deletion (fru^ΔP1 or fru^ΔP2). (d) RT-PCR product from tissues taken from a female body (thorax and abdomen), head without the brain, brain, and whole fly for fru P2, fru P5, and the control gene RpL32. (Online version in colour.)

Figure 3.

The removal of sensory modalities does not increase female receptivity towards D. melanogaster males. (a) Proportion of hybrid disruption fru^MI05459 and control sim/mel^Bal females that mate when paired with males with wings removed (wing−) versus those with wings intact (wing+). The assay was repeated for this disruption plus two additional disruptions (fru^4-40, fru^GAL4). (b) Proportion of hybrid disruption fru^MI05459 and control sim/mel^Bal females that mate when their last two antennal segments have been removed (ant−), which removes the primary sensory organs for both olfactory and auditory signals, compared with females with intact antennae (ant+). The assay was repeated for this disruption and fru^GAL4. (c) Proportion of hybrid disruption fru^MI05459 or fru^GAL4 and control sim/mel^Bal females that mated when placed on a vibrationless (vib−) versus control (vib+) substrate.