The heterogametic sex differs from the homogametic sex in that the pair sex chromosome (X or Z) is contained in its cells in only a single copy. While, for autosomes, the F1-hybrids of both sexes contain the complete chromosomal set from each species; for the X-chromosome this is true only for females. In the first half of the 20th century, Dobzhansky and Muller (Dobzhansky 1936; Muller 1940) pointed out that, in cases where some of the genes on the autosomes derived from a species from which an allosome is derived, i.e. an unpaired sex chromosome (Y or W), are not sufficiently compatible with the corresponding genes on the X-chromosome (Z-chromosome) derived from the second species, the hybrid members of the heterogametic sex will probably have reduced fitness (Fig. XXI.9). It is apparent that a second necessary precondition for the reduced fitness of these hybrids lies in the highly recessive nature of the interactions between the participating alleles. The term dominance hypothesis is derived from this; the designation began to be used for the Dobzhansky - Muller hypothesis after H.A. Orr demonstrated, on the basis of analysis of a mathematical model of this phenomenon, that key importance in the formation of phenomena responsible for the Haldane rule lies in the average degree of dominance of phenotype manifestations of the relevant interactions (Orr 1993). If the functions of the given genes were ensured by the products of mutually interacting genes from the chromosome set of the species from which the X-chromosome was derived, the incompatibility of the autosomes of one species with the X-chromosome of the other species should not be manifested in the phenotype. If the average degree of dominance of the negative effects of the products of interacting genes were greater than 0.5, the homogametic sex would be affected more. As females have two X-chromosomes, they should have twice as many incompatible genes. The requirement of low dominance of negative manifestations of common products of incompatible genes will apparently not be very restrictive. The commonest type of disorder will tend to be loss of functioning of a certain molecular complex, i.e. an effect that is, in its nature, usually recessive.
For a long time, the most serious obstacle in accepting the dominance hypothesis came from the results of genetic experiments in which female drosophila with an “unbalanced” genome were prepared by an ingenious procedure, i.e. with a genome containing one complete set of autosomes from each parent species, and simultaneously both X-chromosomes derived from the same species (Coyne 1985a). On the basis of the dominance hypothesis, we would expect that these females would have viability and fertility reduced to the same degree as hybrid males. However, it was found that this assumption is valid only for their viability; they have the same fertility as normal hybrid females, i.e. substantially greater than hybrid males. Several possible explanations of these experiments have been proposed to date (Turelli & Orr 1995; Orr & Turelli 1996; Gorshkov & Makarieva 1999); a frequently accepted explanation consists in the faster male hypothesis (XXI.4.3.2) and another very probable explanation will be discussed in the section concerned with the somatic mutation hypothesis (XXI.4.3.5).
The fact that X-chromosomes are important for existence of the Haldane rule for sterility is confirmed by comparison of species in which the X-chromosomes are large with species in which the X-chromosomes are small and thus contain a smaller number of all genes. Such a comparison can be performed for the Drosophila genus, where there are groups of species whose X-chromosomes bear approximately 20% of the genes and groups of species whose large X-chromosomes bear approximately 40% of all the genes present in the genome of the particular species. A study encompassing 81 interspecific hybridizations of Drosophila with smaller X-chromosomes and 44 interspecific hybridizations of Drosophila with large X-chromosomes indicated that the F1-males of species with large X-chromosomes have relatively more reduced fitness than the F1-males of species with smaller X-chromosomes (Turelli & Begun 1997).
The major effect of the X-chromosome on the fertility of hybrids is sometimes explained not only in that they are present in only one copy in the genome of the males, but also its anticipated greater content of genes responsible for interspecific incompatibility. This effect, in itself, has become the subject of a great many discussions and independent studies. Usually, the possible greater content of incompatible genes is explained as a result of faster fixation of recessive adaptive mutations on this chromosome and thus greater interspecific divergence of the X-chromosomes compared with the divergence of the autosomes (Charlesworth, Coyne, & Barton 1987). While the presence of recessive mutations is not manifested on the autosomes, as its effect is masked by the standard allele on the other chromosome, the presence of a recessive mutation on the X-chromosome of males is manifested in males and can immediately be an object of selection. Negative mutations are understandably eliminated faster on X-chromosomes than on autosomes; however, these mutations are rarely fixed and thus mostly do not contribute to interspecific genetic divergence.
However, it should be pointed out that the very existence of this phenomenon, i.e. of a large content of genes responsible for the sterility of hybrids on the X-chromosome, is occasionally thrown into doubt. For example, experiments have been performed in which the methods of classical genetics were employed to introduce, into the genome of an individual of one species, in place of its own chromosome sections, the corresponding sections from a foreign species. In some cases, it was demonstrated, and in others not demonstrated (Hollocher & Wu 1996; True, Weir, & Laurie 1996; Jiggins et al. 2001) that there is a substantial difference between situations when a DNA section was inserted into the X-chromosome of a male and when an identically long homological section of foreign DNA was inserted into both its autosomes.
The dominance hypothesis is not relevant only for the interactions of autosomes and X-chromosomes, but can also be applied to other types of interactions leading to manifestation of the Haldane rule. It has been found that interactions between X- and Y-chromosomes, Y-chromosomes and autosomes and between chromosomes, especially an X-chromosome and cytoplasm, can be very important here (Turelli & Orr 2000). Interactions involving the Y-chromosome increase manifestations of the Haldane rule on the sterility of members of the relevant heterogamete sex in all species. However, from the viewpoint of viability, because of the low number of functional genes on the usual, i.e. differentiated type of Y-chromosome, they are apparently of very little importance. Interactions with participation by the cytoplasm, include both interactions between genetic elements in the cytoplasm derived from one and from the other species, i.e., for example between proteins synthesized in the zygote and in the parent cell. Interactions between the cytoplasm and genes on the sex chromosomes weaken the manifestations of the Haldane rule in species with heterogametic males. The cytoplasm and the X-heterochromosome of hybrids of male sex are derived from the same species here. In species with heterogametic females, to the contrary, these interactions increase the manifestations of the Haldane rule, as here the cytoplasm and Y (in fact W) chromosome are derived from different species.