Annu. Rev. Ecol. Syst. 1998. 29:233-261
John H. Werren
Genetic Conflict Over Sex Determination
Genetic conflict will occur when the various components of the sex-determining
system are selected to push zygotic sex determination or parental sex ratios in
different directions. Given the divergent selective pressures acting on genes
with different inheritance patterns (cytoplasmic, autosomal, and sex
chromosomal) and different sites of expression (maternal, paternal, and
zygotic), genetic conflict is an inherent feature of sex-determining systems.
Here we list the general arenas of conflict over sex determination and sex
ratios.
CYTO-NUCLEAR CONFLICT
Conflict between cytoplasmic and nuclear genes over sex determination and sex
ratios is obvious and appears to be common and widespread. Many cytoplasmic
sex-ratio distorters are microorganisms that are transmitted through the egg
cytoplasm but not through sperm (reviewed in 86).
In plants, cyto-nuclear conflict has been documented between maternally
inherited organelles inducing CMS and autosomal suppressors of cytoplasmic male
sterility (CMS) (reviewed in 39,
148).
In the absence of suppression or other counterbalancing forces, cytoplasmic
sex-ratio distorters can spread near or to fixation, potentially driving the
population (and species) to extinction (81,
160).
Cyto-nuclear conflict is discussed in more detail below.
SEX-CHROMOSOME DRIVE AND B-CHROMOSOME DRIVE CONFLICT
Sex-chromosome drive is just one manifestation of selection favoring meiotic
drive loci, which also occur on autosomes (reviewed in 114).
However, the sex-ratio distortion resulting from it can create intense
sex-ratio selection. There is considerable evidence that X-chromosome drive
selects for repressors on the Y chromosome and autosomes (see below). In
species with recombination on the sex chromosomes, selection on linked genes
can favor either enhancement of drive or suppression of drive, depending on
how tightly linked the gene is and whether linkage disequilibria are maintained
(180). However, the possibility that sex-ratio distortion
induced by X-drive favors compensatory shifts in zygotic sex determination (or
maternal-effect sex determiners) has not been extensively explored.
Sex-chromosome drive can also potentially cause population extinction (73,
112,
113).
Many B chromosomes are parasitic genetic elements that have an increased transmission in gametes (transmission drive), by which the chromosomes are maintained within populations despite the fitness costs they impose on the host (127, 129). In many cases, transmission of Bs through males and females (or male and female function in hermaphrodites) is asymmetric. Under this circumstance, selection is expected to lead to the accumulation of sex ratio and sex-determining genes that bias sex toward the transmitting sex. However, detailed studies in a few coccid species with biased transmission of B chromosomes have failed to show an effect of B on sex determination (U Nur, personal communication). One striking example of a sex-ratio distorting B chromosome is the psr chromosome in N. vitripennis described previously (131).
PARENT-OFFSPRING CONFLICT
Trivers (163) originally formulated the idea that parents and offspring
can have divergent genetic interests due to the fact that they are genetically
related but not genetically identical. Studies of parent-offspring conflict
usually concern conflict over the amount of resources allocated to offspring.
However, Trivers & Hare (164)
proposed that conflict should exist between a queen social insect and her
worker progeny over sex ratios in social insects. Empirical studies provide
strong support that such conflict exists (159).
Given the growing evidence for maternal-effect sex-determining genes, the possibility of conflict over sex determination needs to be considered more thoroughly. There are two situations in which such conflict is likely: (a) when fitness costs to a parent of a son and daughter differs, and (b) under partial inbreeding or local mate competition. When one sex is more costly to the parent to produce than the other, natural selection will favor the parent overproducing the less-costly sex (57). However, selection acting on the zygote will generally favor a more-balanced sex ratio. This is particularly true when the cost to the mother is in terms of future survival and reproduction. For example, in red deer (Cervus elaphus), producing a male is more reproductively costly to the mother than producing a daughter, and the mother often fails to reproduce in the year following a male birth (38). The dynamics of this interaction have not been explored theoretically. Depending on the mating system, paternal-effect sex determiners will have genetic interests more concordant with either zygotic or maternal genes.
Under partial inbreeding or local mate competition, maternal-effect genes will be selected to produce a more female-biased sex ratio. Zygotic-effect sex determiners will also be selected to produce a female bias, but the equilibrium ratio should be less biased because of asymmetries in genetic relatedness. The result will be conflicting selective pressures. A possible outcome would be the accumulation of maternal modifiers and zygotic modifiers pushing in opposite directions. Again, the interacting system has not been explored theoretically. Conflict also clearly occurs between parental sex-chromosome drivers and zygotic sex-determining genes. In principle, the sex-ratio distortion resulting from driving sex chromosomes should lead to compensatory shifts in sex determination to the underrepresented sex (113).
MATERNAL-PATERNAL CONFLICT
Interest has focused primarily on intragenomic conflict between maternally
derived and paternally derived genes over resource allocation to developing
zygotes and on intergenomic male-female conflict over female reproductive
effort (70). Nevertheless, there are some interesting applications to
sex determination evolution. Brown (13)
and Bull (15, 17)
have shown that maternal gene/paternal gene conflict can lead to the evolution
of paternal genome loss and haplodiploid sex determination. Basically, there
is a selective advantage to maternal genes that "eliminate" the
paternal genome. This advantage (termed the automatic frequency response by
Brown) results from a higher maternal genome transmission in the next
generation in haploid males relative to diploid males (i.e. no reduction due to
meiosis). The advantage accrues as long as haploid males have a fitness greater
than one half that of diploid males.
In addition, intergenomic maternal-paternal conflict clearly occurs in species with haplodiploid and paternal genome–loss sex determination (71). In haplodiploids, males are under selection to increase the proportion of fertilized eggs (proportion of females) produced by their mates. However, it is unclear what opportunities are available to males for affecting female sex ratios. In paternal genome–loss systems [e.g. coccids (130)], paternal genes will be selected to escape or suppress paternal genome loss. Some supernumerary chromosomes have evolved escape mechanisms from paternal genome loss, such as in the mealy bug (127) and the flatworm Polycelis nigra (9).
Alternative Models for Sex-Determination
Evolution
Genetic conflict is an inherent feature of sex-determining systems. However, a
number of models have been proposed for the evolution of sex determination
besides that of genetic conflict. We briefly review some models currently in
the literature, focusing on factors that destabilize sex-determining systems
and cause evolutionary transitions in the sex-determining mechanism.
TRANSIENT COVARIANCE OF FITNESS AND SEX (HITCHIKING)
Bull (17) has proposed that transient linkage disequilibrium between
sex-determining alleles and genes under strong positive selection could
destabilize sex determination by causing distorted population sex ratios. These
distorted sex ratios would create counter-selection for sex-determining loci
producing the opposite sex. Such an affect may explain the diversity of sex
determination found in M. domestica, in which some sex-determination
variants appear to be linked to pesticide-resistance alleles (106,
124,
147). In the platyfish, several body-color genes are tightly linked
to sex-determining loci (104).
ACCUMULATION-ATTRITION
Graves (67) proposed an "addition-attrition" model to explain the
evolution of mammalian sex determination. According to the model, mammalian sex
determination evolves by a series of autosomal additions (translocations) to
the Y chromosome followed by degeneration of these pseudo-autosomal regions.
Only genes that evolve functions in male sex determination escape
mutational degradation that results when crossing over is suppressed between X
and Y chromosomes. A series of translocation events could result in turnover of
sex-determining genes on the Y. The model is consistent with the view that
sexually antagonistic genes can accumulate on the sex chromosomes (e.g.
Y-linked genes that enhance male fitness and diminish female fitness) (141,
142)
and the idea that male growth enhancers will accumulate on the Y (87,
88).
POPULATION STRUCTURE AND INBREEDING
Hamilton (73) pointed out that subdivided populations with local mating
(and inbreeding) select for parents that have female-biased sex ratios. There
is considerable empirical evidence that local mate competition does lead to
female-biased sex ratios (reviewed in 3,
74).
However, there has been little consideration of how inbreeding and local mate
competition shape the zygotic sex-determining mechanism in species without
parental sex-ratio control.
Two other population-structure effects relevant to sex-determination evolution are local resource competition (33) and local resource enhancement (152). Whenever fitness returns differ through males and females (or male and female function for hermaphrodites) as a function of amount of investment in that sex (e.g. because of differential dispersal), biased sex ratios will be selected (58, 59). However, most models of these effects implicitly assume parental sex-ratio control. The same selective force should also select for biases in the zygotic sex-determining genes, although less strongly than for parental sex ratio and parental-effect sex-determining genes. Such effects have not been investigated theoretically.
VARIABLE FITNESS OF MALES AND FEMALES
Facultative adjustments in sex ratio and sex determination are expected when
male and female fitness are differentially affected by some environmental
factors. For example, Trivers & Willard (165)
pointed out that when maternal condition varies, and this variation translates
into a greater fitness effect on sons versus daughters, then selection will
favor mothers in good condition overproducing sons and mothers in bad condition
overproducing daughters. Variable fitness affects are also invoked to
explain age-specific sex change in sequential hermaphrodites and host-size
effects on sex in parasitic wasps (31).
Variable fitness effects almost certainly are important in the evolution of environmental sex determination (16, 32). Environmental sex determination is observed in some marine worms and molluscs, in parasitic nematodes such as mermithids, in some plants (136), in a few fish, and in some lizards, turtles, and crocodillians (reviewed in 17, 97). In invertebrates, crowding or poor nutrition is typically associated with increased male determination. Sex determination is temperature sensitive in a variety of reptiles, although the selective factors favoring such sex determination are still unclear.