Abstract
In this thesis, I examine the mechanisms of evolution at different levels, from evolutionary
conflict between selfish genes within a single individual (Chapter 1), through social evolution
acting within a species (Chapter 2), to genetic divergence and incompatibility between closely
related species (Chapters 4 & 5). The thesis therefore investigates how tiny genetic
differences occurring in individuals accumulate and produce discontinuous groups.
The first chapter explores an interesting form of natural selection, acting independently on
different genomes within the same cell. Natural selection can act at the level of individual
genes: an allele that promotes its own transmission can increase in frequency despite reducing
the fitness of the rest of the genome (Dawkins 1978). This phenomenon, known as intragenomic
conflict (Hurst 1992), has long been hypothesized to drive evolution, forcing
different lineages to adapt to the genes within their own genomes and therefore causing their
genomes to diverge, and potentially, to become incompatible types. Here I test whether intragenomic
conflict drives evolutionary change by evolving yeast populations in the laboratory,
to see if intra-genomic conflicts would lead genomes in independent populations to become
incompatible. After allowing populations to evolve under two treatments of strict vertical
transmission of mitochondria, or mixed horizontal/vertical transmission, I tested the
evolutionary changes in interactions between mitochondrial and nuclear genomes in the
continuum of mutualism and selfishness. As predicted, increasing the independence of
mitochondria from their hosts (by increasing outbreeding) reduced the evolved fitness benefit
that mitochondria provided to their un-evolved hosts. The results presented in this chapter
hint that intra-genomic conflicts can speed up the evolution of cyto-nuclear reproductive
isolation between allopatric populations.
The second chapter also looks at whether conflict, this time between individuals in a
population rather than between genes within an individual, can lead to diversification, not just
in the form of single nucleotide replacements but at the under-examined form of copy number
variation. The sharing of the secreted enzyme invertase (encoded by SUC genes) by yeast
cells is a well-established laboratory model used to test social conflict models. Moreover,
yeast populations vary in SUC gene copy numbers. The observed copy number variation has
been suggested to be the result of natural selection acting at the level of social conflict.
However, genetic variation might instead be explained by adaptation of different populations
to different local availabilities of sucrose, the substrate for the SUC gene product. Here, I
6
provide evidence showing that the variation observed in natural populations is better
explained by the environmental adaptation hypothesis rather than the social conflict
hypothesis (Bozdag & Greig 2014).
The final chapters take a different approach: rather than at bottom-up approach testing how
natural selection (intra-genomic conflict, social conflict and environmental adaptation) may
drive diversification or divergence into different types, I take a top-down approach, testing
which genetic changes are responsible for the discontinuities between already established
types (between two species of yeast, S. cerevisiae and S. paradoxus).
In chapter three, I look at how nucleotide sequence variation can accumulate to such an extent
that it prevents the segregation of diverged chromosomes, causing sexual incompatibilities
between established types (different species). Here, I have genetically manipulated
interspecific hybrids with the aim of inducing crossovers between their diverged
chromosomes. This manipulation increased recombination rates significantly compared to unmanipulated
hybrids. Increased recombination caused a remarkable increase in the fertility of
the yeast hybrids, from 0.5% viable gametes to over 30% viable gametes. I conclude that the
reduced recombination in interspecific hybrids is responsible for at least one third of the
hybrid gamete death.
And finally in chapter four, I determine how individual genetic changes can cause
incompatibility, potentially preventing certain individuals from breeding together and
therefore allowing the accumulation of further genetic changes. Here I assayed a hybrid strain
for two-locus incompatibilities (Bateson-Dobzhansky-Muller genic incompatibilities)
between the two parental yeast species. If such genic incompatibilities exist, the proportion of
viable offspring bearing the hybrid combination for a pair of loci should be significantly
lower than the proportion bearing the non-hybrid (i.e. parental) combination. To check this, I
exploited the improved viability of interspecific hybrids obtained in the chapter three. As a
result, I present seven putative BDMI regions between the two sibling species of yeast.
