Abstract
The majority of arthropod species carry facultative heritable microbes, bacteria that are passed from mother to offspring, and which may contribute to host function. These symbionts are coinherited down the maternal line with mitochondria, and selection favouring either new symbionts, or new symbiont variants, is known to drive loss of mitochondrial diversity as a correlated response. More recently, evidence has accumulated of episodic directional selection on mitochondria. We therefore examined the reciprocal interaction and model the impact of selection on mitochondrial DNA (mtDNA) on symbiont frequency. We performed this for three generic scenarios: a fixed benefit to the host carrying the symbiont, a benefit that decreased with symbiont frequency, and a benefit that increased with symbiont frequency. We find that direct selection on mtDNA can drive symbionts out of the population under some circumstances. Symbiont extinction occurs where the positively selected mtDNA mutation occurs initially in an individual that is uninfected with the symbiont, and the symbiont is initially at low frequency. When, in contrast, the positively selected mtDNA mutation occurs in a symbiont infected individual, the mutation becomes fixed and in doing so removes symbiont variation from the population. Given low frequency symbiont infections are common in natural populations, and selection on mtDNA is also considered to occur frequently, we conclude that mtDNA driven loss of symbionts represents a novel mechanism driving loss of facultative heritable microbes. We conclude further that the molecular evolution of symbionts and mitochondria, which has previously been viewed from a perspective of selection on symbionts driving the evolution of a neutral mtDNA marker, should be reappraised in the light of positive selection on mtDNA. Where low mtDNA and symbiont genetic diversity are observed, it should not be assumed to be a consequences of selection acting on the symbiont.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
↵Ethics: This work is in silico, with no ethics requirements.
Data availability: no data was generated in this project. Code for simulations can be found at: https://doi.org/10.6084/m9.figshare.c.5354999.v1