Uniparental inheritance of cytoplasmic genomes: its evolution and consequences

Abstract

Many eons ago, a proto-eukaryote engulfed a prokaryote, giving rise to the most enduring symbiotic partnership in the history of life. That bacterium evolved into the mitochondrion and with it evolved an array of innovations. Mitochondria play a crucial role in the energy production of the cell, a function that has released energy constraints on the eukaryotic cell and enabled the evolution of complex life. Mitochondria have retained their own genome, although its size has been greatly reduced in many organisms, particularly in animals. As a result, eukaryotic cells contain both a nuclear and a mitochondrial genome. In addition to mitochondria, eukaryotes can carry other cytoplasmic genomes: chloroplasts and bacterial endosymbionts. Together, these cytoplasmic genomes share a number of common features: all exist within their host's cytoplasm and are generally inherited via a single parent. The evolutionary reasons behind uniparental inheritance are not well understood. We do know that many organisms go to great pains to actively avoid the biparental transmission of cytoplasmic genomes. Clearly, uniparental inheritance is important, but why? The most widely accepted explanation for the evolution of uniparental inheritance is conflict between the nuclear and cytoplasmic genomes. According to this hypothesis, uniparental inheritance evolved to protect hosts against "selfish" cytoplasmic genomes - those that invest in their own replication to the detriment of the host. In the first part of this thesis, I challenge this hypothesis, arguing that it requires unrealistic biological conditions. Instead, I propose two alternative hypotheses for the evolution of uniparental inheritance: (1) avoidance of costly mixing of different cytoplasmic genomes within hosts; and (2) selection for the accumulation of beneficial cytoplasmic mutations within hosts. I conclude that the need to avoid costs associated with the mixing of cytoplasmic genomes has the strongest support of any existing hypothesis. Irrespective of the evolutionary reasons behind uniparental inheritance, this mode of inheritance has implications for the spread and evolution of cytoplasmic genomes, which is the focus of the remainder of my thesis. Cytoplasmic genomes are asexual and generally lack recombination. Both theoretical and empirical work have shown that the absence of sexual reproduction and recombination should impair adaptive evolution. In fact, asexual genomes should suffer from irreparable mutational meltdown in a process known as Muller's ratchet. Increasingly, empirical evidence suggests that the mitochondrial genome, particularly that of animals, shows pervasive signatures of adaptive evolution despite lacking sex and recombination. In the second part of this thesis, I show that uniparental inheritance dramatically alters the evolutionary dynamics of cytoplasmic genomes, explaining why these genomes have higher levels of adaptive evolution than predicted by existing theory. I then move on to investigate the consequences of uniparental inheritance on the bacterial endosymbionts of arthropods. Many bacterial endosymbionts manipulate the reproduction of their host to promote their own spread. I show that uniparental inheritance of cytoplasm protects arthropods from invasion by harmful bacteria. This places evolutionary pressure on endosymbionts to evolve mechanisms to manipulate their host's reproduction, explaining the pervasiveness of reproductive manipulation by the endosymbionts of arthropods.