Genetic Recombination: Transposition, Retroviruses, and the Evolution of Genomic Diversity

Recombination is a process leading to the rearrangement of nucleotides by the breaking and rejoining of DNA molecules. There are three general categories (Alberts et al. 2015): homologous; conservative site-specific; and transpositional. The term ‘homologous’ in this context pertains to regions of sequence similarity (homology) between the DNA molecules involved. As noted earlier, modifications induced by processes such as transposition, or in aberrant homologous recombination, are also considered to be forms of mutation.

General or homologous recombination involves genetic exchange between members of a pair of homologous DNA sequences (i.e., those of extensive sequence similarity). Its most widespread role is in the accurate repair of double-strand breaks but it also is involved in genetic exchange between two paired chromosomes in eukaryotes (or DNA strands in the case of prokaryotes; Lengeler et al. 1999). Recombination or crossing-over does not occur with every chromosome at every meiosis.

Where recombination is part of the formal meiotic process in eukaryotes, it is frequently referred to as meiotic recombination and includes (i) independent or Mendelian assortment of entire maternal and paternal homologs during metaphase I and anaphase I of meiosis; and (ii) reciprocal recombination of chromosome segments (crossing-over) that occurs between the nonsister chromatids of paired homologs during prophase I of meiosis (Alberts et al. 2015). In this role, recombination is both a mechanical process to insure the equivalent segregation of chromosomes to the two daughter germ cells, as well as a mixing process that reassorts genes on those individual chromosomes.

Although genes may be reassorted, gene order on the chromosomes involved usually remains the same, since the recombining sequences are quite similar (cf. transposition, below). As alluded to earlier, an analogous phenomenon also occurs in bacterial transformation and conjugation, where exogenous chromosome fragments are integrated into the genome of the recipient cell by homologous recombination (see Sex, and Adaptive Evolution in Prokaryotes in Sect. 2.3). Thus, features of homologous recombination are common to all organisms. For eukaryotes, assortment is quantitatively more significant than crossing-over in all organisms with a haploid chromosome number exceeding two (Crow 1988). Chromosomes in all organisms can break spontaneously. An additional key role (nonmeiotic) of homologous recombination is to accurately repair such single- or double-strand breaks. (Double-strand breaks can also be corrected by a cruder mechanism known as nonhomologous end-joining, which reseals the DNA but results in a mutation where the strands broke; Alberts et al. 2015.)

Transposition and conservative site-specific recombination differ from general recombination in that diverse, specialized segments of DNA are moved about the genome (Alberts et al. 2015). The pieces moved, which vary considerably from a few hundreds to many thousands of nucleotide base pairs, go by various colloquial (‘jumping genes’; ‘selfish DNA’), as well as specific names. For example, in bacteria, there are two general types of transposable elements: (i) insertion-sequence or IS elements that can move themselves but do not carry genes other than those related to the movement; and (ii) transposons—carry the movement genes as well as others (Griffiths et al. 2015). Transposition and conservative site-specific recombination differ in the reaction mechanisms involved and because the conservative process requires that there be specialized DNA sequences on both donor and recipient DNA, whereas transposition generally requires these specialized sequences only on the transposon.

Transposition is further divided into three classes involving (i) DNA-only transposons; (ii) retroviral-like retrotransposons; and (iii) nonretroviral retrotransposons. Transposons are typically ‘cut’ from one place and ‘pasted’ into another place in the genome without duplication, whereas the retrotransposons (retroposons) are duplicated because they are transcribed into RNA, reverse-transcribed into DNA, and then reintegrated. Different kinds of TEs predominate in different kinds of organisms: bacterial transposons tend to be the DNA type, whereas it has been estimated that about half of mammalian genomes originate from TEs primarily of the retroelement group (Van de Lagemaat et al. 2003). Similarly, 64% of the genome of the powdery mildew pathogen Blumeria consists of TEs (Spanu et al. 2010). At least four identical transposon families occur in invertebrates and vertebrates, where they have moved horizontally from the former to the latter evidently by parasite-host interactions (Gilbert et al. 2010).

TEs have broad evolutionary implications because they can mutate existing genes, create new genes, and affect gene regulation at both the transcriptional and post-transcriptional levels when they insert nearby (see Epigenetics below; also Slotkin and Martienssen 2007; Freschotte 2008; Elbarbary et al. 2016). The resulting variation can be adaptive and, for example, IS have been shown to promote the evolution of specialists (as opposed to generalists; see Chap. 3) in controlled growth experiments with bacteria (Zhong et al. 2004, 2009). Frequently, they have deleterious consequences because the insertion and rearrangement can lead to disease (e.g., the Alu elements in humans; Kazazian 2004). For this reason and because the elements replicate themselves independently of host chromosomes, they have been called ‘intra-genomic parasites’ or ‘selfish DNA’ (Charlesworth 1985; also see Sect. 2.3), though such terms are simplistic and potentially misleading (Kidwell and Lisch 2001).

Moreover, repetitive DNA may well have a functional role in the physical ordering of the genome (Shapiro and von Sternberg 2005). McClintock (1956) discovered TEs in the late 1940s and 1950s in maize, and called them ‘controlling elements’ because, although distinct from genes, they could modify gene expression. They have since been well documented in many other taxa including bacteria (phage Mu; insertion sequences; transposons conferring antibiotic/metal resistance or surface antigen variation); yeasts (Ty and mating type elements of Saccharomyces); and animals (Drosophila transposable elements and hybrid dysgenesis determinants; vertebrate and invertebrate retroviruses). At least in maize, and presumably in other macroorganisms, transposition occurs at predictable times and frequencies in the ontogeny of the individual. In maize, a controlling element can have a similar effect on genes governing different biochemical pathways and at different places in the genome (McClintock 1956; Fedoroff 1983, 1989). Moreover, a single element can control more than one gene concurrently.

Both transposition and site-specific recombination are complex processes in detail and a molecular biology text such as Alberts et al. (2015) should be consulted for specifics. The key conceptual point is that these phenomena occur broadly if not universally and add considerable genetic versatility or plasticity to organisms beyond the conventional mechanism of recombination normally associated with sexuality. For additional comments, see Genomic Plasticity and Epigenetics sections, below.

An analogous process pertains to integration of some plasmids (small, ancillary, self- replicating extrachromosomal elements) into the bacterial chromosome and ‘promiscuous’ (organellar) DNA into the nuclear chromosomes. To date, plasmids (see Sect. 2.3) are known to occur ubiquitously in bacteria. Though uncommon in eukaryotes, they are found in many fungi and in some higher eukaryotes, often in association with mitochondria (Funnell and Phillips 2004).

Promiscuous DNA has been detected in most eukaryote species examined, including plants, filamentous fungi, yeasts, and invertebrates (Timmis et al. 2004). The term originated with Ellis (1982) for DNA that appeared to move from chloroplasts to mitochondria. Subsequently, evidence has accrued for a broader process, including the insertion of mitochondrial and chloroplast DNA sequences into nuclear DNA (Herrmann et al. 2003; Matsuo et al. 2005; Bock and Timmis 2008).

The extent to which such transpositions produce functional transcripts remains unclear; for example, most of the plastid DNA engulfed by the nucleus may be eliminated by genome shuffling (Matsuo et al. 2005). If the genes are expressed, there are potentially significant evolutionary implications because of the different modes of inheritance of a nuclear as opposed to an organellar gene. Presumably such transpositions also can interrupt nuclear gene function, depending on where they insert.

To what extent is mobile DNA favored by natural selection? At the level of the ‘selfish gene’ (Dawkins 1989) selection is presumably for these mobile elements, especially in eukaryotes with excess DNA, or in bacteria where they add unique (useful but generally described as nonessential) features as plasmids, discussed later under prokaryote recombination. However, once essential gene functions are disrupted, selection at the level of the gene will be offset by counter-selection at the level of the physiological individual. So the tendency should be toward some balance in opposing forces. Note, however, that deleterious genes can still spread in a population by over-replication or if they alter reproductive mechanisms to favor themselves (Campbell 1981; Chap. 6 in Bell 1982). Certain TEs (retroviruses, below) provide an independent mechanism for moving genetic material horizontally.

Perhaps the most intriguing subcategory of site-specific recombination involves the retroviruses. They are unique in having an RNA genome that replicates by reverse transcription through a DNA intermediate, which can then integrate as provirus into host chromosomal DNA. Retroviruses are considered with transposons because of similar structure and functional properties. They do not transpose in the same way that bacterial transposons do, but are analogous in that they can be viewed as intermediates in the transposition of viral genes from proviral integration sites in the host chromosomes (Varmus 1983; Varmus and Brown 1989). Retroviruses and viral-like elements have been described from diverse genomes, including those of mammals, the slime mold Dictyostelium, yeast, fish, reptiles, birds, and plants (McDonald et al. 1988). The most information is on mammalian retroviruses and because of the interesting evolutionary implications, this is summarized briefly below (see also Doolittle et al. 1989).

There are two retroviral categories (Benveniste 1985; Varmus 1988): Infectious or exogenous retroviruses occur as a few copies of proviral DNA per cell, only in the genome of infected cells; they are infectious and often pathogenic (as in HIV); and they are transmitted horizontally, i.e., among individuals rather than from mother to daughter. Endogenous retroviruses occur as multigene families in the host DNA of somatic cells (and occasionally germline cells, in which case they are transmitted vertically) of all animals of the species of origin (Benveniste 1985; Jern and Coffin 2008). They have been viewed as fossil representatives of retroviruses extant in the geological era when they entered the germline (Jern and Coffin 2008). About 7-8% of the human genome is of retroviral origin (Jern and Coffin 2008). Endogenous retroviruses are usually not infectious to cells of the species of origin, but are often so to those of other species. In fact, it is this property of being able to replicate in heterologous cells that sets them apart from conventional cellular genes.

Both types of retroviruses can cause host genes to mutate, or can carry host genes with them. This is significant because of the spreading of somatic variation through the soma, and the possibility of introducing variation directly to the germline (Sect. 2.5). From an evolutionary standpoint, the endogenous group is particularly intriguing because some members have been transmitted horizontally and, once established, may subsequently have been incorporated into the germ line and transmitted vertically (i.e., from mother cell- to-daughter cell and from parent to offspring). Benveniste (1985, p. 362) reviews evidence that retrovirus transfers have included those “from ancestors of primates to ancestors of carnivores, from rodents to carnivores, from rodents to primates, from rodents to artiodactyls, from primates to primates, and from primates to birds.” A specific example is the baboon type C viruses, which are transmitted vertically in primates and which were transferred millions of years ago to ancestors of the domestic cat where they were incorporated into germ cells and inherited thereafter in conventional Mendelian fashion (Benveniste and Todaro 1974, 1975). Benveniste (1985) proposes that retroviruses may promote genetic interaction above the species level, much as do plasmids in bacteria (discussed later).

Are retroviruses a major force in evolution? They do seem to play a major role by influencing gene regulation (McDonald 1990) and by their phjylogenetic implications noted above. The analogue that comes closest to this is the transfer and incorporation of bacterial DNA into plant chromosomes. Crown gall and hairy root diseases of plants involve transfer of a plasmid (tumor-inducing or Ti plasmid) from a pathogen, Agrobacterium tumefaciens, to the plant, where a fragment (T-DNA) is covalently integrated into the host nuclear genome. In essence, the agrobacteria use genetic engineering methods to force the infected plant to synthesize nutrients (opines), which the bacteria utilize (discussed in Chap. 3; Zambryski 1989; Platt et al. 2014). (Although the process of Agrobacterium oncogenesis is often considered with processes involving transposable elements, strictly speaking the T-DNA is not a transposon because it does not jump about the chromosome. The fascinating Agrobacterium/ plant tumor story is summarized in Chap. 3.)