CMS

Saur Jacobs M & Wade MJ 2003 A synthetic review of the theory of gynodioecy. Am Nat 161:837-851.

• Charlesworth and Ganders (1979) showed that a cytoplasmic polymorphism could be maintained, in the absence of nuclear restorers, if hermaphrodites are self-compatible and there is pollen limitation
• pollen limitation disproportionately affects females relative to hermaphrodites
• the female-producing cytotypes increase in frequency until their fitness is reduced by pollen limitation
• a stable equilibrium is reached when the number of seeds produced by the hermaphrodites (which are freed from pollen limitation by virtue of selfing) equals that produced by females

• under the simplest models of gynodioecy, gender polymorphism is conditional on whether the action of the nuclear restorer, R, is additive or dominant (Lloyd 1974)
• when the restorer is dominant and has no fitness cost, the gender polymorphism is transient
• the cytoplasmic sterility gene, CMS, and the nuclear restorer, R, both spread to fixation with the equilibrium population consisting entirely of hermaphrodites
• pollen limitation in outcrossing populations can slow the approach to fixation but cannot prevent it (Charlesworth and Gander 1979)

• stable nuclear polymorphism is possible when restorers act additively (Lloyd 1974; Couvet et al. 1986; McCauley and Taylor 1997)
• there is no joint cytonuclear polymorphism
• the CMS allele spreads to fixation
• the restorer allele, R, reaches an intermediate equilibrium frequency if females produce more than twice the amount of ovules as hermaphrodites

• Charlesworth and Ganders (1979, p. 216) hypothesized pleiotropy of the CMS allele
• the CMS genes increase ovule fitness in females but decrease ovule or viability fitness in restored hermaphrodites
• this "pleiotropic" model is the first model that demonstrates stable gynodioecy in the presence of restorer genes
• it relies on selfing and does not maintain joint cytonuclear variation
• the marginal gametic fitness of a nuclear restorer is reduced on CMS background
• this is equivalent to a "cost of restoration," as per Delannay et al. (1981)

• Delannay et al. (1981) also rely on selfing to maintain polymorphism
• further studies of their model reveal that stable gynodioecy is established without selfing under certain fitness parameters (Charlesworth 1981)

• the increased allocation to ovules that results from the presence of females creates an effective sex ratio bias in which male gametes are the minority sex
• the Fisherian advantage favors nuclear restorer alleles because they are associated with pollen (see eq. [10]), the minority gamete, but does not affect the relative fitness of female-inherited CMS genes in the ovules

• when s < 1, the gametic fitness effect of the CMS cytoplasm is negative for a nuclear genome
• when s > 1, the gametic fitness of nuclear alleles is greater than in hermaphrodites because more resources are allocated to gametes in females than in hermaphrodites
• although CMS alleles may lower the transmission of nuclear alleles, genomic conflict does not describe the evolutionary forces acting on alleles that are associated with gender
• specifically, restorers do not evolve as a response to genomic conflict
• the restorers spread because of their association with the minority sex
• the evolution of restorers is not consistent with the idea of cytonuclear conflict as previously argued (Frank and Barr 2001)
• the restorer does not experience the reduction in fitness of the nuclear genome in the presence of the CMS alleles

• the "conflict of interest" also breaks down when discussing selective pressures operating on the nuclear and cytoplasmic alleles through a population genetics perspective
• consistent with a scenario of genomic conflict, a restorer allele, R, lowers the ability of a CMS allele to spread if restored hermaphrodites suffer an ovular fitness cost
• however, the same restorer allele enjoys a transmission bias only in the presence of a CMS allele
• we find it difficult to call this "genomic conflict" because the fitness advantage to the nuclear restorer is dependent on having the cytoplasmic allele, CMS, in the population
• furthermore, as the restorer spreads, it lowers its own transmission advantage by reducing the degree of gametic sex ratio bias in the population
• a restorer has negative frequency-dependent fitness effects on both itself and the CMS allele

• the concept of genomic conflict depends in large part on the ability to isolate the fitnesses of cytoplasmic alleles from those of nuclear alleles and to show that an increment in the fitness of one results in a decrement in the fitness of the other
• whenever gene interactions―that is, epistasis―affect fitness, the interaction effect of the two genes cannot legitimately be viewed as a "property" of either one independent of the other
• gender expression is epistatically determined with gynodioecy
• the marginal fitness of a nuclear gene depends on the frequency with which it experiences the CMS gene
• several kinds of selection affect total fitness and determine the evolutionary equilibrium
• it is not clear whether isolating one component of selection, as in genomic conflict, is sufficient to make predictions about the evolutionary trajectory of a population