Genetic conflicts: the usual suspects and beyond

Abstract

Selfishness is pervasive and manifests at all scales of biology, from societies, to individuals, to genetic elements within a genome. The relentless struggle to seek evolutionary advantages drives perpetual cycles of adaptation and counter-adaptation, commonly referred to as Red Queen interactions. In this review, we explore insights gleaned from molecular and genetic studies of such genetic conflicts, both extrinsic (between genomes) and intrinsic (within genomes or cells). We argue that many different characteristics of selfish genetic elements can be distilled into two types of advantages: an over-replication advantage (e.g. mobile genetic elements in genomes) and a transmission distortion advantage (e.g. meiotic drivers in populations). These two general categories may help classify disparate types of selfish genetic elements.

Keywords: Antagonism; Host–pathogen; Meiosis; Mitochondria; Mutualism; Red Queen; Toxin–antitoxin; Wolbachia.

Fig. 1.

Adaptation to static versus dynamic environments. (A) A change in environmental conditions (for example, a move to a higher temperature, represented as an increase in sun exposure) selects for bird genotypes (green) better adapted to the altered abiotic environment. As a result, these fit genotypes increase in frequency. Even though other genotypes might appear later in the population, selection maintains the high prevalence of the fittest genotype. The change in the allele composition of the bird population has no effect on the environmental conditions. (B) A pathogen infects a population, and selects for a resistant genotype (green) in the host population. As the green genotype increases in frequency, this selects for pathogen genotypes that can infect the most prevalent (now green) genotype. This, in turn, selects for a new resistant host genotype (white) that resists infection by the prevalent pathogen (green) genotype. Thus, cycles of adaptation in the host and pathogen populations drive perpetual cycles of adaptation, typical of Red Queen interactions.

Fig. 2.

Rapid evolution of amino acid sequences identifies interface and timing of host–virus conflict. (A) Where? Host restriction factors block the replication of pathogenic viruses, but this places pressure on the virus, which selects for variants that evade host restriction (red hexagon virus, red circle virus). Similarly, host variation at the interaction interface that restricts evasive viruses is selected for (red surface on host factor), shifting selection back to the virus. As there is always a benefit to adapt to evade (on the virus side) or restrict (on the host side), there is no stable equilibrium and always an advantage to adapt. Often, the difference between a restrictive and ineffective host factor may be determined by a single amino acid change, as described for TRIM5alpha and MxA. (B) When? Positive selection can date the origin of a host–pathogen arms race. Episodic positive selection in a subset of primate species suggests the initiation of the ‘arms race’ began in the common ancestor of these species, as seen in the Cercopithecinae subfamily of Old World monkeys when an ancestral lentivirus gained the ability to antagonize SAMHD1. (C) How? Top: error-prone RNA viruses adapt through generating a diverse pool of offspring, some of which contain adaptive mutations that evade or counteract host restriction. These adaptive mutations increase in frequency in the virus population. Bottom: in the genomes of more slowly evolving DNA viruses like the poxviruses, the copy number of genes frequently expands. If one duplicate gene samples a mutation that increases the fitness of the virus (for example, the white variant evades host restriction), this variant will be selected for and the copy number expansion may contract to contain only the newly adaptive (white) gene variant.

Fig. 3.

Selfish genetic elements that distort transmission in their own favor by eliminating the competition. (A) Toxin–antidote genes inherited on episomes (plasmids) encode two components: a long-lasting toxin and a more unstable antitoxin. Constant protection from the toxin therefore requires constant production of the antitoxin. If a bacterial cell does not inherit the toxin–antidote plasmid, it is ‘killed’ by the action of the persistent toxin once the unstable antidote degrades. Thus, the toxin–antidote gene pair acts as a selfish genetic element, ensuring its own preferential transmission. (B) Restriction–modification (R–M) plasmids were initially widely believed to represent a form of symbiosis, where they act to protect the host genome against invading phage. The modification protein acts to modify the bacterial chromosome, protecting it from the DNA-endonuclease activity of the restriction enzyme. However, on an episome, R–M systems act more as a classic toxin–antidote system (outlined in A), ensuring their own transmission via similar dynamics to the canonical toxin–antidote systems. (C) Toxin–antidote systems can also act following meiosis in eukaryotes. Loci that encode toxin–antidote systems are protected from toxin action by producing a constant supply of the short-lived antidote, whereas homologous sister products of meiosis that lack the locus are eliminated owing to the action of the toxin. Such ‘spore-killers’ are quite abundant in fungi, although only a handful have been completely characterized at the molecular genetic level. (D) In a variation of the toxin–antidote systems, toxin genes can act preferentially on chromosomes bearing a susceptibility locus (e.g. Responder satellite in Drosophila melanogaster) but do not act on themselves owing to the absence of this locus. As a result, such systems can drive themselves to high frequency by eliminating competitor chromosomes bearing the susceptibility locus and so preventing them from efficiently completing gametogenesis.

Fig. 4.

Uniparentally inherited cytoplasmic bacteria distort transmission in their own favor. (A) The two parents equally contribute to the nuclear genomes of their male and female progeny (sex chromosomes are an obvious exception). In contrast, cytoplasmically inherited bacteria such as Wolbachia and Spiroplasma (as well as mitochondria and chloroplasts in plants) are maternally inherited. As a result, males represent an ‘evolutionary dead-end’ for any cytoplasmically inherited bacteria, and they adopt a number of insidious strategies to avoid this fate (B–E). (B) Because male fitness is inconsequential (and may even be harmful) to Wolbachia, it can eliminate male larvae via lethal interaction between the male genotype and the cytoplasmically inherited Wolbachia (Fukui et al., 2015; Hurst and Jiggins, 2000). This maximizes resources available to the female larvae, which can transmit Wolbachia to future generations. (C) If males are a dead-end, it is not in Wolbachia's genetic interests to have any males at all. By converting sexual females into parthenogenetic females (Adachi-Hagimori et al., 2008), it can ensure it is passed on productively to all of the progeny of the parthenogenetic females. (D) One means by which Wolbachia ‘trapped’ in males can skirt their inevitable elimination is to convert genetic males into reproductive females (Kageyama et al., 2002), thereby ensuring they can be passed on to its offspring. (E) Another means by which Wolbachia can quickly spread in the population is by reducing the reproductive fitness of uninfected females relative to infected ones. For instance, under cytoplasmic incompatibility, all progeny resulting from a cross between an infected male and an uninfected female are inviable (Bordenstein and Werren, 1998). In contrast, crosses between infected males and infected females are viable and, thus, successfully increase the proportion of Wolbachia-infected individuals (especially females) in the next generation.