Evolution of dominance in a Mendelian trait is linked to the evolution of environmental plasticity

Abstract

Allelic dominance and phenotypic plasticity both influence how genetic variation is expressed in phenotypes, shaping evolutionary responses to selection. In both cases, changes in genotype or environment can cause sharp, nonlinear phenotypic shifts, hinting at shared underlying features of development that may link dominance and plasticity. Here, we investigate these links using a Mendelian, female-limited color dimorphism found in many species of the Drosophila montium lineage. In most species, the Dark allele is dominant, but two species-D. jambulina and D. cf. bocqueti-have been reported to have dominant Light alleles. We show that in both Dark-dominant and Light-dominant species, the color dimorphism is linked to the same locus: the POU domain motif 3 (pdm3) transcription factor. We then demonstrate that the interspecific differences in dominance relationships between pdm3 alleles are due to changes in phenotypic plasticity. In the Dark-dominant species D. rufa and D. burlai, the Dark allele is dominant across all developmental temperatures. In contrast, in both Light-dominant species, dominance is temperature-dependent, with the Light allele becoming increasingly dominant at higher temperatures. These results suggest a mechanistic connection between the evolution of dominance and phenotypic plasticity. We propose this connection may emerge from threshold-like properties of developmental systems.

Keywords: Drosophila pigmentation; Evolution of dominance; developmental thresholds; evolution of phenotypic plasticity; genotype-phenotype mapping; sex-limited polymorphism.

Figure 1

Abdominal pigmentation in four species of the Drosophila montium species subgroup (melanogaster species group) used in this study. In each species, females show a Mendelian color dimorphism in the A6 segment (Dark / fully pigmented vs Light / not pigmented) while males are monomorphic with dark A6. The phylogeny is based on Conner et al (2021), and branch lengths are not to scale.

Figure 2

Female A6 pigmentation in D. rufa as a function of rearing temperature. (A) The homozygous parental strains Dark 10_7Y (left) and Light 59BY (right). A slight dependence on temperature is observed in each parent (one-way ANOVA, df = 2; F = 4.992, p = 0.0089 for Dark homozygotes; F = 60.37, p < 10⁻¹⁰ for Light homozygotes). (B) A slight dependence on temperature is observed in the F1 females from the cross between Light females and Dark males (right, F = 22.68, p < 10⁻⁸), but not in the reciprocal cross (left, F = 0.184). Black bars and black dots indicate mean and median values, respectively; the numbers indicate mean values. Different letters indicate significant differences (p < 0.05, Tukey’s HSD test).

Figure 3

Female A6 pigmentation in D. burlai as a function of rearing temperature. Labeling as in Figure 2. (A) The homozygous parental strains Dark 781 (left) and Light 11x (right). A slight dependence on temperature is observed in the Light homozygotes (one-way ANOVA, df = 3; F = 120.7, p < 10⁻¹⁰), but not in the Dark homozygotes (F = 0.985). (B) No temperature plasticity is observed in the F1 females from reciprocal crosses between Dark and Light parents (F = 1.038 for the left cross, and F = 1.29 for the right cross). Black bars and black dots indicate mean and median values, respectively; the numbers indicate mean values. Different letters indicate significant differences (p < 0.05, Tukey’s HSD test).

Figure 4

Female A6 pigmentation in D. cf. bocqueti as a function of rearing temperature. Labeling as in Figure 2. (A) The homozygous parental strains Dark l1 (left), Dark y4 (middle), and Light 85L (right). A slight dependence on temperature (one-way ANOVA, df = 3) is observed in the Light 85L (F = 3.374, p = 0.021) and Dark y4 homozygotes (F = 2.968, p = 0.034), but not in the Dark l1 homozygotes (F = 2.401). (B) F1 females from both reciprocal crosses between the Dark l1 and Light 85L parents show a strong temperature dependence (F = 300.6 for the left cross and F = 636.7 for the right, p < 10⁻¹⁰ for both). (C) F1 females from both reciprocal crosses between the Dark y4 and Light 85L parents show a strong temperature dependence (F = 517.0 for the left cross and F = 748.0 for the right, p < 10⁻¹⁰ for both). Note that the plasticity curves differ between the two heterozygous genotypes, depending on the Dark parent, but no differences are observed between reciprocal crosses (Supplement Table S6). (D) The abdominal pigmentation of an F1 female from the cross between Light 85L females and Dark l1 males, reared at 23 °C. (E) The abdominal pigmentation of an F1 female from the cross between Light 85L females and Dark y4 males, reared at 23 °C.

Figure 5

Female A6 pigmentation in D. jambulina as a function of rearing temperature. Labeling as in Figure 2. (A) The homozygous parental strains Dark 51+2 (left), Dark NHO (middle), and Light 78L (right). No significant differences between temperatures are observed in any of the three strains (one-way ANOVA, df = 2, F = 0.22 for the left cross, F = 0.66 for the cross in the middle, and F = 1.66 for the right cross). (B) F1 females from both reciprocal crosses between the Dark 51+2 and Light 78L parents show a strong temperature dependence (F = 84.38 for the left cross and F = 78.76 for the right, p < 10⁻¹⁰ for both). (C) F1 females from both reciprocal crosses between the Dark NHO and Light 78L parents show a strong temperature dependence (F = 13.66, p < 10⁻⁵ for the left cross and F = 47.38, p < 10⁻¹⁰ for the right). Note that the plasticity curves differ between the two heterozygous genotypes, depending on the Dark parent, but no differences are observed between reciprocal crosses (Supplement Table S7). (D) The abdominal pigmentation of an F1 female from the cross between Light 78L females and Dark 51+2 males, reared at 17 °C. (E) The abdominal pigmentation of an F1 female from the cross between Light 78L females and Dark NHO males, reared at 17 °C.

Figure 6

Female A6 pigmentation in the interspecific crosses between D. chauvacae, D. burlai, and D. cf. bocqueti, as a function of rearing temperature. Labeling as in Figure 2. (A) Female and male abdominal pigmentation in the D. chauvacae Dark l2 strain. (B) D. chauvacae Dark l2 homozygotes show only slight temperature plasticity (one-way ANOVA, df = 2; F = 4.973, p = 0.0083). (C) F1 females from the cross between D. chauvacae Dark l2 females and D. burlai Light 11x males show somewhat greater temperature plasticity than either the D. chauvacae parent (Fig 6B) or the D. burlai heterozygotes (Fig 3B) (one-way ANOVA, df = 2; F = 19.03, p < 10⁻⁷). (D) F1 females from the cross between D. chauvacae Dark l2 females and D. cf. bocqueti Light 85L males show greater temperature plasticity than the D. chauvacae parent, and comparable to D. cf. bocqueti heterozygotes (Fig 4B, C) (one-way ANOVA, df = 3; F = 513.2, p < 10⁻¹⁰). (E) The abdominal pigmentation of an F1 female from the cross between D. chauvacae Dark l2 females and D. cf. bocqueti Light 85L males, reared at 23 °C.

Figure 7

Dominance relationships between the Light and Dark pdm3 alleles vary across species, segments, and sexes, with multiple examples of temperature-dependent dominance reversal. Abdominal segments 5–7 are shown for females, and abdominal segment 5 for males. All other segments have invariant pigmentation. Segments where the pigmentation of heterozygotes is similar to Light homozygotes across all temperatures are designated as “Light dominant” and highlighted in yellow. Segments where the pigmentation of heterozygotes is similar to Dark homozygotes across all temperatures are designated as “Dark dominant” and highlighted in blue. Segments where the dominance relationship between the Light and Dark alleles varies across temperatures are designated as “Temperature-dependent dominance”. Of these, cases where the Light allele is fully dominant at high temperature and the Dark allele is fully dominant at low temperature are highlighted in green. Cases where the Light allele is fully dominant at high temperature and intermediate dominance is observed at low temperature are highlighted in yellow. A case where the Dark allele is fully dominant at low temperature and incomplete dominance of the Dark allele is observed at high temperature is highlighted in cyan. Segments where the pigmentation of heterozygotes is intermediate between Dark and Light homozygotes across all temperatures are designated as “Intermediate” and highlighted in light gray. Segments where no difference is observed between Light and Dark homozygotes are left blank.

Figure 8

A putative model of the relationship between environmental plasticity and allelic dominance. (A) Hypothesized relationship between temperature, pdm3 expression (grey curve), and pigmentation (black curve). pdm3 is a repressor of dark pigmentation (Yassin et al. 2016). We propose that pdm3 expression increases at higher temperatures, leading to lighter pigmentation in a threshold-dependent manner. When pdm3 expression is above the threshold (horizontal dashed line), the result is a Light phenotype (downward arrow); when pdm3 is expressed below the threshold, the result is a Dark phenotype (upward arrow); intermediate temperatures may lead to intermediate pdm3 expression and intermediate pigmentation. (B) Hypothesized relationship between temperature and pdm3 expression in species where the dark pdm3 allele is fully dominant. Light pdm3 alleles cause higher pdm3 expression than Dark alleles; all pdm3 alleles respond to temperature in a similar manner; and the threshold is such that pdm3 expression is below the threshold in DD homozygotes and DL heterozygotes, and above the threshold in LL homozygotes, at all temperatures. (C) Hypothesized relationship between temperature and pdm3 expression in species with temperature-dependent dominance. In this model, the threshold is the same, but the Light pdm3 allele has a steeper temperature response than the Dark allele; as a result, the DL heterozygotes cross the pigmentation threshold (dashed horizontal line) at intermediate temperatures. (D) An alternative hypothesis for the relationship between plasticity and dominance. In this model, all pdm3 alleles respond to temperature in a similar manner, but the pigmentation threshold is lower than in species with full dominance (downward arrow). As a result, DL heterozygotes cross the pigmentation threshold at intermediate temperatures. Models (C) and (D) are not mutually exclusive; different mechanisms may operate in different species, or both mechanisms may operate in the same species.