Sex and Meiotic Recombination. Definitions, Origin, and Maintenance

Definitions, Origin, and Maintenance.Definitions of sex vary considerably (Michod and Levin 1988). Here it will be considered to be the bringing together in a single cell of genes from two (or rarely more) genetically different genotypes or individuals (Maynard Smith 1978b). In most but not all organisms, sex is tied to reproduction, i.e., to the production of a new individual that arises from the zygote or its functional equivalent. The most important consequence of sex is the acquisition by the zygote of new genes (mutations in the germ cells of one or both parents) and new gene combinations. The latter arise from reciprocal exchange (crossing-over) of genes between homologous chromosomes, and reciprocal exchange of chromosomes (independent assortment) during meiotic recombination as part of gametogenesis.

Processes that are sexual are general (though maybe not universal; see Sect. 2.4) among the prokaryotes and eukaryotes. Among extant prokaryotes, genetic exchange can occur in three ways (discussed in detail later in the next section): (i) by direct uptake of DNA from the environment under specific conditions (transformation); and as mediated by (ii) phage infection (transduction) or (iii) plasmids (conjugation). Transformation is likely very ancient. Arguably, it can be traced as far back as the emergence of protocells in primordial communities of progenotes that existed ca. 4.0-3.6 billion years ago, prior to the divergence of the Last Common Ancestor of the three lineages Archaea, Bacteria, and Eukaryotes (see Chap. 4). Woese (2002), among others, has postulated an era dominated by the lateral transmission of information with communities as a whole varying in descent and selection operating largely at the level of community optimization.

Meiotic sex, traceable to the last eukaryotic common ancestor (Speijer et al. 2015), is postulated to have arisen from mitosis (Wilkins and Holliday 2009). It may have begun by the interposition of one new step (homolog synapsis) followed by ‘parameiosis’ in the occasional diploid protocells within an otherwise haploid cell population of early protists. It is perhaps vestiges of this innovation that remain today in ‘parasexual’ cell cycles, most notably among the fungi, as discussed later. Over evolutionary time, sex cells, and sexual fusion arose with enhanced inter-genic recombination during the pairing step, as well as related meiotic properties such as synaptonemal complexes. With meiosis thus began the classic alternation of generations, haploid/diploid phases of the eukaryotic life cycle (Chap. 6). Wilkins and Holliday (2009), however, reason that mitosis originated before meiosis because it occurs universally among the eukaryotes, whereas meiosis is both more complex and, though very widely represented as noted, is not universal.

But what was the major selection pressure for the evolution of sex and the origin of meiosis? While increased recombination would imply the creation of new, potentially favorable gene combinations and disruption of unfavorable ones (see below), this feature alone is not generally regarded as conveying sufficiently immediate benefit to constitute a potent evolutionary force. It conveys instead future benefits to the population or lineage rather than to the individual. An alternative, interesting hypothesis (one of many) is that meiosis facilitates repair of DNA damage (Bernstein et al. 1985). If damage occurs to only one DNA strand, it can be repaired by using the other strand as template. However, when both strands are damaged, correction requires proximity of the homologous chromosome for a template and a process akin to recombination. A variation of the Bernstein argument is that sex arose in prokaryotes as a side-effect of processes to promote DNA replication and repair. Indeed, much of the biochemical machinery and the underlying genes are homologous, including the RecA family of so-called recombination enzymes and their eukaryotic homologs (Cox 1999; Marcon and Moens 2005). Notwithstanding the name (‘Rec’ for recombination), DNA replication and repair appear to be their primary function (Redfield 2001). In the broader context of the implications of sex for endogenous and exogenous repair, Stearns (1992, p. 183) says that the evolution of sex can be viewed as the evolution of the mechanisms preventing the ageing of the germ line.”

For eukaryotes, Wilkins and Holliday (2009) have modified the Bernstein repair hypothesis by arguing that the benefit of meiosis was not in restoration of the original wild-type DNA message but prevention of recombination-induced injury. Consequently, meiosis improved recombinational accuracy and confined the process to a localized period in the cell cycle (while also likely increasing the frequency of genetic recombination mainly among the ‘right’ sequences; Wilkins and Holliday 2009). Recombination is error-prone because the ‘wrong’ or ectopic pairing may occur leading to various irregularities in the message including deletions, duplications, or aneuploidy. The invention of homolog synapsis in meiosis would have enforced accurate alignment so that only identical regions were in register, not diverged homologous sequences elsewhere on the chromosome.

The foregoing may explain why sex arose but, having arisen, why is it maintained in some form in virtually all taxa? The fact that sexual reproduction is ubiquitous yet carries significant costs is commonly referred to as the paradox of sex and has been called by Bell (1982, p. 19) “the queen of problems in evolutionary biology.” This controversial matter has been debated in countless papers, review articles, and books (e.g., Williams 1975; Maynard Smith 1978b; Bell 1982; Michod and Levin 1988; Otto and Lenormand 2002; Rice 2002; Otto 2009). At the risk of trivializing a very complex issue, the following general points can be made in passing.

In eukaryotes the costs, relative to an alternative of asexual reproduction, are various but principally of three sorts: (i) in anisogamous species (i.e., those in which the male and female sex cells contribute unequally in terms of gamete characteristics to the production of progeny) the twofold ‘cost of producing males’, since only females produce offspring; (ii) in sexual eukaryotes, the twofold ‘cost of meiosis, or more accurately the cost of genome dilution, since each parent’s genes are diluted by one-half in their progeny; and (iii) the disruption of favorable gene combinations resulting from past selection, analogous to deciding to reshuffle your hand of cards when you already have a good hand in a game of poker (Otto 2009). With respect to (i), a significant general distinction between eukaryotic microorganisms—many species of which are single celled—as opposed to macroorganisms, is that the former typically are isogamous. In such cases there is no ‘cost of producing males’.

Against these handicaps of sexual reproduction is set the traditionally acclaimed advantage, namely the ability to combine beneficial alleles from different individuals, restoring variation that would otherwise become dissipated in asexual reproduction (Otto 2009). Simultaneously, in changing environments, sex also would eliminate genetic associations that may have been favorable in a previous selective environment but are no longer so (Otto 2009). An example of such oscillating conditions as they influence coevolving species is that sex can produce novel genotypes that enable lineages of macroorganisms to survive attack by much shorter lived (hence more rapidly evolving) microbial parasites. Conversely, asexually reproducing lines would be vulnerable both because the parasite quickly evolves virulence to overcome host resistance genes, and because as the size of the host clone increases from generation to generation it presents a progressively larger target. This is one form of the ‘Red Queen Hypothesis’ (Hamilton 1980; Clay and Kover 1996; Lively and Morran 2014).

Through sex, deleterious mutations that would otherwise accumulate in a finite population in the absence of recombination (Muller’s ‘ratchet’ 1964; Bell 1988a) also are purged. These advantages accrue over time (Rice 2002). Crow (1994) has shown conceptually how sexual species can in effect clump harmful mutations and eliminate several at once by a mechanism such as truncation selection (i.e., selection eliminating all individuals beyond a certain phenotypic state or value). In contrast, asexual species can only eliminate them in the original genotype in which they occur. Also, in outbreeding sexual species, genes influencing the mutation rate will become separated from the corresponding mutations, whereas they will not in asexual species (Drake et al. 1998), so evolutionary processes may be quite different in the two situations. A recent test of the longstanding dogma that sex accelerates adaptation, executed by comparing evolutionary events in sexual and asexual populations of Saccharomyces cerevisiae, confirms that sex acts by providing a sorting mechanism to separate the beneficial from the deleterious mutations: advantageous mutations are combined into the same background, whereas deleterious mutations are separated from advantageous backgrounds that would otherwise carry them to fixation (McDonald et al. 2016).

The above arguments do not imply that sex necessarily increases variation or that such variation necessarily increases fitness (Otto and Lenormand 2002; Otto 2009). So while competitively superior genotypes can be produced, sexual recombination overall may not have a net advantage. While an earlier generation of models suggested fairly constrained conditions wherein sexuality would be maintained, recent evolutionary models (Otto 2009) run under more realistic conditions imply the evolution of sexuality when (i) selection varies over time (past genetic associations no longer favorable); (ii) selection varies over space (where migration-driven genetic associations are locally disadvantageous); (iii) rates of sex as opposed to asexual reproduction are relatively high for less fit individuals and relatively low for fit individuals, i.e., when facultatively sexual individuals in poorer condition allocate more resources to sexual reproduction; and (iv) populations are finite. With respect to (iv), most evolution- of-sex models are deterministic and assume infinitely large populations. This is not realistic and can lead to the wrong conclusions because the best genotype can be lost by drift in finite populations. Sexual recombination would allow it to be regenerated relatively quickly whereas asexuality would not. For discussion of this and the other conditions, see Otto (2009).