Genetic migration supplies the potential for integration of genes or alleles found in one population into another population, genes and alleles that have not been mutationally generated within the second population nor which have been subject to natural selection also within this second population. Whether those genes will be retained in their new population depends on two factors. The second factor is a standard evolutionary consideration, i.e., will selection or drift lead to retention of the alleles within the population? The first factor, though, is a more proximate concern, and that is that a mechanism must exist whereby the transferred gene or genes are retained at all within the actual recipient organism. The latter requires a combination of replication and subsequent segregation of the acquired genetic material into progeny. One means of achieving these two ends is for the gene to be present on, and indeed arrive on, a somewhat autonomous genetic element such as a plasmid. Seemingly more commonly, however, the newly acquired genetic material instead becomes integrated in some manner into the chromosome of the recipient organism. Mechanisms by which such integration can occur can be described as various means of molecular recombination.

Mechanisms of recombination actually can be differentiated into two or three scenarios for the occurrence of genetic recombination. First is genetic recombination that is mediated by organism-encoded mechanisms of gene exchange. Second is genetic recombination that is mediated by mechanisms of gene exchange that are encoded by parasitic or otherwise by semiautonomous genetic elements. A third category would be genetic recombination that occurs somewhat independently of organism-encoded mechanisms of gene exchange, i.e., genetic recombination that is largely accidental. Towards gaining a fuller appreciation of the processes of sex as they occur within microorganisms, such as bacteria, I discuss as follows these various mechanisms.

Genetic Recombination

Genetic recombination means that the genetic material of two otherwise separate and distinct organisms has in some manner come to be present within the same organism, particularly the same cell, and to some degree has become functionally combined. Typically this occurs via either independent assortment (or reassortment), molecular recombination, or some combination of both. The state of karyogamy in fungi – where distinct haploid nuclei sourced from different parents occupy the same cytoplasm – by the above definition also could be considered to exist in a genetically recombined state, though potentially only minimally so. A cell following acquisition of an obligate endosymbiont, though likely to display subsequent molecular recombination at a much lower rate than fungi exhibiting karyogamy, also can be viewed as having achieved some minimal level of genetic recombination (e.g., as is the case for mitochondria or plastids and their host cells). In both of these examples the genetic material associated with two different parents has come to be located within the same cell and particularly with one set of genetic material not overtly parasitizing the other. Mere acquisition of symbiotic organisms, on the other hand, such as Escherichia coli residing in our colons, should not be viewed as a form of genetic recombination since there has not been a combining of genetic material into a single location (cytoplasm, nucleus, virus capsid) and the two organisms (e.g., us and E. coli) otherwise remain completely genetically separate. Indeed, it can reasonably be argued that genetic recombination in fact has not occurred in the case of karyogamy or endosymbiont acquisition, though clearly these examples are closer to achieving genetic recombination than have an E. coli's genes with those of our own bodies. All of these examples nevertheless represent some degree of acquisition of the genes of one organisms by another.

For clarity, I will use the phrase “Pre-genetic recombination state” to distinguish between what clearly is not genetic recombination on the one hand (i.e., we and E. coli clearly do not represent a pre-genetic recombination state) and what clearly is. In addition to karyogamy and endosymbiont acquisition as pre-genetic recombination states, other circumstances that involve gene acquisition but not quite or yet genetic recombination include simply the post-fertilization but pre-meiosis state seen with various eukaryotes as well as the point of acquisition of DNA by transformation or transduction in bacteria. A spectrum of interactions between the genetic material associated with different organisms thus can be described, ranging from no interaction between organisms to interaction that are less intimate to intimate interaction (symbioses) to mutualistic associations of genetic material within the same cell (i.e., a pre-genetic recombination state) to actual genetic recombination. Importantly, and presumably obvious, this pre-genetic recombination state often serves as a prelude to actual genetic recombination.

The independent assortment mechanisms seen during eukaryotic meiotic division unquestionably is considered to be a form of genetic recombination, at least once the meiotic progeny have formed—before that point, in the pre-meiotic diploid cell, the DNA as noted can be viewed as being in a pre-genetic recombination state. The crossing over that also occurs during meiosis is also a form of genetic recombination, particularly molecular recombination. Similarly, viral reassortment, the mixing of genomic segments from different viruses coinfecting the same cell, would be considered to be a form of genetic recombination, at least once viral progeny have formed (i.e., such that the genetic material sourced from different parents has come to be co-encapsidated). The crossing over seen with many additional virus types also is a form genetic recombination. The preceding coinfection that can give rise to these forms of viral genetical recombination thus can be considered to represent a pre-genetic recombination state.

Neither independent assortment nor viral reassortment would be considered to be forms of molecular recombination. Thus, again, a spectrum of recombination can be described, with each step involving a greater degree of mixing together of the genetic material sourced from different parents. These are (1) all of the circumstances in which recombination between the genetic material of different organisms is not seen, though with different degrees of intimacy of organism association nonetheless present, (2) what can be described as a pre-genetic recombination state, (3) genetic recombination itself, and (4) among genetic recombination mechanisms, specifically what can be described as molecular recombination. Large distinctions thus can exist in terms of organism interactions that differ as a function of the extent to which genetic material might become functionally integrated.

Molecular Recombination

Molecular recombination is that form of genetic recombination that is in addition to the independent assortment that is observed during eukaryotic meiotic division. Molecular recombination is what is going on when crossing over between chromosomes takes place during meiosis, whereby a portion of the nucleic acid making up the genome of one individual becomes covalently integrated into the genome of a second individual—or, more generally, two otherwise distinct molecules of poly-nucleic acid become combined chemically into a single molecule of poly-nucleic acid. Many organisms encode one or more proteins that facilitate this stitching together of DNA from different chromosomes, and these proteins are employed either in the course of replication or, alternatively, in the course of nucleic acid repair (or both). Alternatively, these proteins may be encoded by parasitic genetic entities that use these proteins to integrate their genomes into the chromosome of their host organism. There may also exist mechanisms whereby poly-nucleic acid is accidentally recombined in a manner that in some way is independent of enabling proteins (give rise to illegitimate recombination?), though presumably the latter is a relatively rare occurrence.

Before delving into discussion of the different mechanisms of molecular recombination, an additional concept should be considered, and that is the idea of homology. Homology refers to the similarity of nucleic acid sequence, that is, their identity. Two identical sequences may be said to display 100% or full homology, though it is important also to consider the length of that homology, i.e., 10 bases of complete or partial homology is much different from 1,000 bases, both phylogenetically and in terms of the potential for molecular recombination to occur. Alternatively, at the opposite extreme, no homology may be present. That is, lining up sequences it may be impossible to identify more than a few contiguous bases of full or even significant homology. These considerations are important because recombination-enabling proteins typically require some homology between molecules for recombination to take place with reasonable likelihood.

Homologous Recombination

Homologous recombination is that which typically occurs during meiotic processes, during recombination-mediated DNA repair mechanisms in bacteria, and during the replication-associated DNA recombination observed in viruses. Generally homologous recombination is considered to be a highly evolved mechanism that gives rise to an unlinking of loci in the course of outcrossing (i.e., sex among genetically non-identical individuals). This mechanism of unlinking loci is not necessarily the reason that homologous recombination exists but nonetheless unlinking loci appears to be important toward the long-term viability of species. The mechanism explaining why homologous recombination evolved is not agreed upon by researchers, though I must confess that I was trained by researchers who championed the idea that homologous recombination, at least in its most primitive form, evolved as a mechanism of repair of DNA damage (Bernstein et al., 1985).

An important consequence of the existence of mechanisms facilitating homologous recombination is that homologous DNA, which by various mechanisms may be found in the vicinity of an organism's, chromosome has some reasonable likelihood of being swapped into that chromosome. Homologous recombination thus is both a conservative and at least potentially promiscuous process. On the one hand, the overall sequence of nucleotides on the resulting recombinant DNA tends to be more or less preserved including in terms of gene order. Also more or less conserved is existing genetic information. That is, homologous recombination tends to not disrupt existing genetic patterns, at least as seen on larger scales of genetic organization (>> single nucleotides).

Just what DNA may recombine into a chromosome is limited mostly by what DNA is available within a cell along with the sequence (homology) of that DNA, which need not be highly extensive for homologous recombination to occur. Thus, homologous recombination, despite its demand for at least some similarity between donor and recipient genetic material, can be somewhat promiscuous. In less sexual organisms, the result can be a disruption of genetic patterns that is counter the tendency of homologous recombination under other circumstances to preserve genetic patterns. Thus, molecular recombination can unlink alleles while still retaining gene order but on other occasions instead can disrupt gene order. More fundamentally, disruption may be associated with lower levels of homology to begin with, which at an extreme can be seen with what can be described instead as nonhomologous or illegitimate recombination.

Homologous recombination in highly sexual organisms thus tends to preserve the overall genetic patterns observed across a species (e.g., as seen in our own species). In less sexual organisms homologous recombination can still serve to preserve genetic patterns across species, such as the substantial consistency across different strains of bacteria. Less frequently, however, homologous recombination instead can serve to disrupt these genetic patterns, resulting in the insertion of genes in new places, swapping genes between different locations within genomes, or even deleting or otherwise destroying genetic information. The latter, as noted, are rarer consequences of molecular recombination and this rarity may stem at least in part from its being associated with recombination among less homologous nucleotide sequences.

Non-Homologous or Illegitimate Recombination

The potential for DNA to recombine into a chromosome declines with both reduced availability – such as likelihood of finding its way into a given cell, that is, achieving a pre-genetic recombination state – and reduced homology. To a degree these two variables can correlate. That is, the less likely that DNA will find its way into a new organism, so that genetic recombination at least might occur, then so too the less likely that homology will be great enough between that DNA and the recipient organism's for recombination to occur with high likelihood. Thus, for example, with two obligately sexual organisms of the same species there is a high tendency for DNA to come together within the same cell and similarly a high tendency for that DNA to display high homology. For two relatively distantly related individuals, say a horse and a sea urchin, the potential for the genetic material to make its way into the same cell can be somewhat low, and the genetic material itself will be somewhat divergent, in terms of its sequence, that is, the less closely related the two organisms are. These tendencies can be viewed as protective of the recipient organism since non-homologous recombination, recombination that can occur absent significant homology between two organisms, can result in disruption of existing genetic material.

As noted above, homologous recombination usually is a relatively conservative process, giving rise to what in many cases is relatively slight changes in nucleotide sequence. By contrast, the recombinogenic process that can be described as non-homologous or illegitimate recombination – though, semantics aside, may instead be a consequence of otherwise not-appreciated micro homologies – can result in substantial changes in the genetic structure of chromosomes when they do occur. Furthermore, these rare but potentially disruptive recombination events can serve as an important engine of evolutionary innovation. In other words, through non-homologous recombination, genetic material that evolved in one, otherwise distantly related lineage can come to be located within and thereby contribute to the success of a second lineage and do so particularly by supplying new loci to recipients (contrasting the new alleles or combinations of alleles that are the typical products of homologous recombination). In general, both the occurrence of and consequences of illegitimate recombination must be considered to be so rare, so potentially diverse, and so unlikely to be selectively beneficial when they do occur, that specific prediction of their impact on evolutionary innovation is impossible. That is, illegitimate recombination is perhaps the biggest of wild cards with regard to the evolutionary trajectory of lineages that are capable of participating in these processes.

Heterologous Recombination

As noted, when functioning properly the mechanisms of homologous recombination are conservative in the sense that they tend to preserve the original gene order and even much of the original nucleotide sequence of the recombination-encoding organism. Mistakes can happen, however, giving rise to what may be described instead as heterologous recombination. This would include especially relatively long deletions and insertions of genetic material, i.e., more than a few nucleotides and ranging up to even thousands or millions. More generally, heterologous recombination may be viewed as less conservative and therefore may be represented as the products either of errors occurring during homologous recombination or, alternatively, products of non-homologous recombination. As with non-homologous recombination in particular, heterologous recombination can be a source of significant evolutionary innovation though, at the same time, it is an evolutionary mechanism that has a low likelihood of passing the test of subsequent natural selection.

In bacteriophages, novel products of heterologous recombination are described as "morons" which is short for "more DNA", i.e., DNA acquired from diverse sources – such as bacterial chromosomes – that does not so much replace existing genes as add to the genetic repertoire of an organism. The "Moron accretion hypothesis" (Hendrix et al., 2000; 2003) posits that phages evolve, in part, via the accumulation of genes (morons), some of which supply fitness benefits and which therefore are retained. Note an equivalent perspective for bacteria, which have been hypothesized to have been pieced together from plasmid DNA (Holcík and Iyer, 1997), though with the difference that plasmid genes presumably have already stood the test of natural selection within the context of a bacterium prior to incorporation into the chromosome (versus morons, as bacterial genes, which would instead newly find themselves within the chromosome of a virus). By contrast with plasmids potentially contributing to the gradual accretion of bacterial genomes, phages have proteins that predominantly contribute to phage fitness, that is, the fitness of a virus. Phage-acquired bacterial genes, however, are not necessarily inherently able to contribute to phage fitness.

Site-Specific Recombination

Site-specific recombination is encoded by entities such as viruses so that they can insert their genomes into the DNA of host organisms. Certain types of viruses, plus various non-viral genetic parasites such as transposons, can thus assure their long-term retention within host organisms by integrating their genomes into the DNA that is already present and maintained within a host cell. The mechanism of integration is termed site-specific because the location of integration into the chromosome tends to be fairly rigidly defined, at least so long as a proper integration site is available. Alternatively, in a number of instances the site of integration is less well defined, resulting in insertions that are still controlled by the invading organism but which cannot otherwise be strictly described as site specific. These insertion events in either case have the potential of adding genes to the host cell, including genes that provide benefits to the host. Insertions that are not well controlled, i.e., which are not site-specific, and even some which are, however, can lead to inactivation of host genes, often to the host's detriment.

Site-specific recombination is important to bacterial evolution particularly because the temperate phages that are able to carry out this process can also carry genes between bacteria. These genes are either morons or instead alternative aspects of what is described as a specialized transduction. The carried genes give rise to what is known as lysogenic conversion. Among converting genes are those encoding numerous bacterial virulence factors including numerous bacterial exotoxins such as Shiga toxin, cholera toxin, and diphtheria toxin. Integrase genes are responsible for this site-specific recombination and such genes are often found in association with various genetic islands such as pathogenicity islands. The result is a very efficient means of acquisition of new genes that can give rise to fairly dramatic changes in bacterial life styles such as from commensalistic to pathogenic, or from pathogenic to more pathogenic.

Table: Recombination-Associated Terms.

Genetic recombination

Stable co-location of genetic material sources from two different organisms into the same organism, particularly in a manner in which the resulting organism consists of something other than just half of one parent's genetic material and half of the other's (that is, especially just the fusion of gametes is not a process of genetic recombination). Similarly, the combination of all of the genetic material derived from two organism, as one sees during the formation of endosymbiotic interactions, is not strictly an example of genetic recombination. In both these latter examples the genetic material can be said to have been "combined" rather than recombined.
Molecular recombination

Combining of the genetic material from two organisms such that individual DNA molecules come to consist of a combination of so multiply sourced genetic material.
Homologous recombination

Molecular recombination in which the two poly-nucleic acid molecules involved (typically DNA) possess similar sequences at least over the areas initiating recombination.
Non-homologous recombination Recombination that occurs without significant homology and thus that can occur between even distantly related organisms. This type of recombination is likely to result in disruptions of recipient genetic material, but also can result in insertions of new genetic loci.
Illegitimate recombination

Equivalent to non-homologous recombination.
Micro-homologous recombination A form of recombination that either resembles or is identical to non-homologous and therefore illegitimate recombination. Here at least some homology is assumed to be present, to allow for the use of normal recombinational machinery – which typically requires homology in order to operate – to actually carry the recombination mechanisms. The resulting recombination, however, is sufficiently rare as to be difficult to detect and therefore difficult to study, though not so rate that it cannot occur with substantial frequency over large populations and evolutionary time.
Heterologous recombination

Recombination that is unequal among two parents with regard to gains and losses of genetic material. This type recombination can either result from mistakes during homologous recombination or be the consequence of illegitimate recombination. In the latter case, the donor molecule may be inserted into the recipient's chromosome, resulting in an increase in genetic material on the part of the latter.
Site-specific recombination

Molecular machinery that biases recombination to specific locations plus which typically results in an insertion into the recipient. Temperate phages that insert their chromosomes into that of their hosts employ site-specific recombination to do so.