Mammalian X Chromosome Imprinting: Conflict or Agreement?

Tom Moore
Department of Developmental Genetics; Barbraham Institute

The proposal that the occurrence of sex chromosome heteromorphy has promoted the evolution of various mechanisms of dosage compensation between the sexes is supported by the existence of such mechanisms in groups with low chromosome numbers (e.g. Drosophila, Glossina, Caenorhabditis), but not, apparently, in those with high chromosome numbers (e.g. Aves, Lepidoptera), where small, gene-poor X chromosomes may not encode significant dosage imbalances between the sexes (1).

The predominant system of mammalian dosage compensation, involving the imprinted inactivation of the paternally derived X chromosome (Xp) in females, may be the result of more complex selective forces than those which promoted the evolution of dosage compensation in non-viviparous phyla. Specifically, under polyandry, paternally-derived X-linked loci may increase maternal investment in female offspring at the expense of current male sibs and future maternal reproduction (2,3). Moreover, such loci might ultimately cause maladaptive skewing of the sex ratio, and Suppressors of such "selfish" Xp-linked loci are predicted to evolve. Imprinted Xp inactivation may be such a suppressor.

In order to understand more clearly the evolution of such suppression, we have used a genetic model and computer simulations to analyze the dynamics of suppressor invasion. We find that the model agrees with the -predictions of previous verbal models (2,3) as follows. (1) Systems of selfish Xp-linked loci and their unlinked suppressors may readily evolve under polyandry, but are more restricted under monogamy. (2) "Selfish" Xp-linked loci will go to fixation under polyandry even when the cost to future maternal reproduction is very high. (3) Maternal suppressors of a selfish Xp will readily invade and undergo fixation. (4) Unlinked paternal suppressors of a selfish Xp will not invade if the costs due to Xp are expressed solely in terms of reduced maternal reproduction, but may invade if (as seems likely) the costs are shared by current male sibs.

Once the extant pattern of mammalian X-inactivation had evolved, X-linked conflictor loci might subsequently evolve novel imprinted expression rules, which subvert chromosomal imprinting patterns. The putative maternally imprinted X-linked locus affecting cognitive function in Turner's syndrome (4) may be an example of a "selfish" locus promoting putative "social intelligence-emotional blackmail" types of behavior involving competition in post-natal resource acquisition and interactions among sibs. The existence of such imprinted loci affecting behavior has been predicted because of the significant transfer of maternal resources which occurs post-natally (2). In contrast, it has recently been suggested (5) that such imprinted X-linked loci support a model of "genetic agreement" between parental interests, in promoting mutually beneficial sex-limited growth differences (in mice) and behaviors (in human; see above). However, we suggest that these arguments misrepresent the "genetic conflict" theory and misinterpret the origin of growth differences between XmO and XpO mice.


  1. Moore, T. Abstr. Genet. Res. 61: 146, 1993.

  2. Moore, T. and Haig, D. Trends Genet. 7: 45-49, 1991.

  3. Moore, T. et al., Dev. Genet. 17: 206-211, 1995.

  4. Skuse, D.H. et al., Nature 387: 705-708, 1997.

  5. Iwasa, Y. and Pomiankowski, A.J. Theor. Biol. 197: 487-495, 1999.