Sexual conflict through mother's curse and father's curse

Abstract

In contrast with autosomes, lineages of sex chromosomes reside for different amounts of time in males and females, and this transmission asymmetry makes them hotspots for sexual conflict. Similarly, the maternal inheritance of the mitochondrial genome (mtDNA) means that mutations that are beneficial in females can spread in a population even if they are deleterious in males, a form of sexual conflict known as Mother's Curse. While both Mother's Curse and sex chromosome induced sexual conflict have been well studied on their own, the interaction between mitochondrial genes and genes on sex chromosomes is poorly understood. Here, we use analytical models and computer simulations to perform a comprehensive examination of how transmission asymmetries of nuclear, mitochondrial, and sex chromosome-linked genes may both cause and resolve sexual conflicts. For example, the accumulation of male-biased Mother's Curse mtDNA mutations will lead to selection in males for compensatory nuclear modifier loci that alleviate the effect. We show how the Y chromosome, being strictly paternally transmitted provides a particularly safe harbor for such modifiers. This analytical framework also allows us to discover a novel kind of sexual conflict, by which Y chromosome-autosome epistasis may result in the spread of male beneficial but female deleterious mutations in a population. We christen this phenomenon Father's Curse. Extending this analytical framework to ZW sex chromosome systems, where males are the heterogametic sex, we also show how W-autosome epistasis can lead to a novel kind of nuclear Mother's Curse. Overall, this study provides a comprehensive framework to understand how genetic transmission asymmetries may both cause and resolve sexual conflicts.

Keywords: Genomic conflict; Mitochondria; Sex chromosomes; Sexual antagonism.

Figure 1

Graphical representation of the 9 two-locus two-allele models considered. Models 1–5 capture the dynamics of a mitochondrial Mother’s Curse mutation and a nuclear restorer located on an autosome or sex chromosome, respectively. In each model, Locus 1 is the primary sex-asymmetric locus, and Locus 2 harbors alleles that act to restore the fitness of the disadvantaged sex, except Model 5, where it further contributes to the advantaged sex. In Models 6–9, we model how selection acting on a mutation on the uniparentally inherited sex chromosome (Y, Models 6–7; W, Models 8–9) can lead to the spread of a mutation that is beneficial in the heterogametic sex but deleterious in the homogametic sex, Father’s Curse and Nuclear Mother’s Curse respectively.

Figure 2

Fixation time for Mother’s Curse compensators. The rate of fixation is shown for a compensator located on an autosome (a; Model 1), X chromosome (b; Model 2), Y chromosome (c; Model 3), or Z chromosome (d; Model 4) for a given female benefit (s_f = 0.105; simulations were run for 19 different values evenly spaced between 0.02 and 0.4). Colors show the number of generations (up to 1000) required for a nuclear compensator that restores male fitness to fix. The x-axis is the strength of restoration of the compensator (s_a, s_x, s_y, and s_z respectively).

Figure 3

In Father’s Curse (Models 6 and 7) and Nuclear Mother’s Curse (Models 8 and 9), the uniparentally inherited sex chromosome (Y or W) can drive the fixation of an allele that is deleterious in the homogametic sex. Analytical work examines invasion of the sex-specific alleles(Y and W) and the sexually antagonistic alleles (X, Z, or autosomal) from the lower left corner, but higher frequency starting points better illustrate the dynamical behavior. These figures clearly demonstrate that both variants will fix as long the selective cost in the ‘cursed’ sex is smaller than the selective advantage in the benefitting sex. For Models 6–9, both alleles start with an initial frequency of either 0.25 (a-d) or 0.5 (e-h), and the change in allele frequency at every generation is tracked for a total of either 5000 (a,e,c,g) or 10,000 (b,f,d,h) generations. Simulations were run for 100 randomly selected combinations of s_f and s_m for each model. Trajectories where the absolute value of the selection coefficient in females (s_f) is less than that in males (s_m) are shown in red, and the opposite in black.