Cytoplasmic genetics under inbreeding and outbreeding

Abstract

Selfish cytoplasmic factors that gain overrepresentation within an individual at a cost to individual fitness can invade a population only if they are biparentally transmitted. It has been hypothesized that the spread of these selfish genomes provides the conditions for the spread of nuclear genes that enforce uniparental inheritance of cytoplasmic genes. However, not all cell fusions result in uniparental inheritance. By using the same logic as the above hypothesis it is shown that, depending on the rate of in- and outbreeding, a variety of transmission patterns of cytoplasmic factors may be expected. Two selective pressures are considered separately. First, selection on nuclear enforcement of uniparental inheritance. If this enforcement is costly then it is not expected where, in a population with biparental inheritance, either selfish cytoplasmic genes cannot spread, or the cost of those that do spread is always below the cost of nuclear suppression. These conditions are met when the rate of inbreeding is adequately high. Biparental inheritance may thus be expected in fusions between closely related cells. This expectation is consistent with data from sexual fusions in inbred fungi and with somatic cell fusions in diverse taxa. Secondly, in an outbred species with enforced uniparental inheritance, a cytoplasmic factor that resists attempts to prevent its transmission may spread. In contrast, in the inbred condition the optimal strategy is for a cytoplasmic factor to allow its relative to be transmitted free of cost. If such resistance is possible then, in contrast to the above expectations, biparental (or paternal) inheritance may be associated with outbreeding, whereas homothallic or selfing species could have relatively strict uniparental transmission. This expectation is consistent with data on cytoplasmic inheritance in higher plants and chlorophyte algae. Thus all four possible states are found: both uni- and biparental inheritance are associated both with in- and outbreeding. We may conclude that those general models that cannot predict this much variation in cytoplasmic genetics are falsified by the data. The broad diversity of resolutions is consistent with the hypothesis that there exists a conflict between nuclear and cytoplasmic genes over the transmission of the latter. However, as the conflict hypothesis is compatible with any outcome, no set of comparative data can directly falsify the model in the absence of extra information. It would hence be useful to be able to predict the conditions under which any of the four states is to be found. Possible relevant parameters, both biological and historical, are discussed. The most robust prediction is that somatic fusion between close relatives, as guaranteed by com patibility at highly polymorphic loci, is expected to provide the conditions for biparental inheritance.