Somatic Variation, Heritable Variation, and the Concept of the Genet

The idea that the genetic individual or genet is the central entity in which mutational and recombinational events are expressed was introduced in Chap. 1. To recap very briefly, genets are classically viewed as the developing products of zygotes; they arise from a sexual process and therefore represent new genetic entities; in a sense they are independent genetic colonizations of a landscape (Harper 1977). Following our review in this chapter of sexuality/asexuality and the mechanisms of genetic variation, let us now return to the genet concept in a more detailed fashion. This is important, not only because ‘the genetic individual’ is referred to repeatedly in this book, but also because of its major influence in evolutionary biology.

The utility of the genet concept hinges on the premise that although genetic variation can occur in somatic cells, such variation cannot be transmitted to progeny. This view, propounded most forcefully by August Weismann in the late 1800s, can be stated more formally as: (i) the zygote produces somatic cells mitotically and germ cells meiotically; (ii) genetic variation developing during ontogeny cannot be inherited; and hence (iii) heritable variation is expressed only in the zygote or during meiosis in the formation of gametes (Weismann 1892; summarized by Buss 1987, p. 13). If true, this so-called Weismannian doctrine has enormous implications: it restricts evolutionary change to a matter of selection among individuals, as Buss (1987) insightfully develops in his book. So then, to what extent is the genet a valid and useful common denominator in phylogenetic comparisons?

First of all, it should be reemphasized that somatic variation occurs in all organisms and rates are at least as high and typically substantially higher than germline rates (earlier discussion and Lynch 2010). They are presumed to be a significant source of phenotypic and genotypic variation in plant populations (Silander 1985; Klekowski 1988) and in clonal, modular animals (Hughes 1989; more on this in Chap. 5).

Secondly, there is no doubt that somatic variation can affect the life of the organism. Probably the most dramatic evidence of this is that the main changes that lead to various cancers involve somatic mutations (Griffiths et al. 2015). The cells in most forms of cancer have aberrant chromosomes (e.g., deletions, inversions, translocations, aneuploidy) as well as numerous point mutations (ranging from a few to more than 1,000; Vogelstein et al. 2013). Accumulating somatic mutations also form the basis of one of the theories of aging (see Chap. 6). Other sorts of somatic change may be neutral or beneficial, either in allowing the organism to adapt to biotic or abiotic challenges (e.g., generation of antibody diversity, acquisition of acquired immunity), or by concomitant adjustment to specific genotypic alterations. Recently, the history of somatic mutation accumulation within individual neurons of the human brain has been traced by single-cell sequencing (Lodato et al. 2015). Remarkably, each neuron (which lives and remains transcriptionally active for decades) has its own unique genome as a result of as many as ~1,580 single nucleotide variants (SNVs), among other genetic changes. These mutations appear to arise during transcription, unlike the standard case of errors being introduced during DNA replication. Highly expressed genes were enriched for the SNVs.

The extent to which somatic variation can enter the germline, which is the real issue in evolutionary terms, depends on the ontogenetic program of the organism. It is ‘the real issue’ because, as seen above, while somatic genetic changes can be devastating to the individual, possibly even causing death, they are limited to that individual, not the lineage. For dipterans, as illustrated by Drosophila, it would be highly improbable for somatic variants to enter germ cells. The totipotent lineage in Drosophila is restricted to only the first 13 nuclear divisions per generation (Buss 1987, pp. 13-25)—a fleeting opportunity for the origin of a somatic variant. Similarly, in humans, germ cells established in the 56-day-old embryo remain sequestered for up to about three decades (Buss, op. cit. p. 100).

Based on current evidence, both dipterans and humans come as close as any organism does to being a homogeneous genetic entity. That the period of accessibility to the germline is short for vertebrates has been confirmed elegantly in studies where foreign genes are introduced during early embryogenesis (Robertson et al. 1986; Jaenisch 1988). For instance, early embryonic cells can be infected with retroviruses in vitro and reintroduced to the embryo at the blastocyst stage of ontogeny. The infected cells contain integrated provirus that contributes to both the somatic and germ cell lineages, as confirmed biochemically and by the chimeric phenotype of the transgenic animal and its progeny. Infection of pre-implantation stage mouse embryos results in transmission to the germline, whereas infection at the post-implantation stage (between days 8 and 14 of gestation) results in transmission to the somatic but generally not to the germline (Soriano and Jaenisch 1986). Thus dipterans, humans, and mice are examples of a type of ontogeny where all cell lineages are determined early in ontogeny, known as preformistic development (Buss 1987). A correlate of this developmental mode is the absence of ramet production. (Ramets, also discussed in Sect. 1.3 of Chap. 1, are the asexual counterpart to genets and will be taken up further in Chap. 5.)

At the other extreme, an ontogenetic program known as somatic embryogenesis is characterized by absence of a distinct germline and the ability to regenerate a new individual from some tissues at any life stage (Buss 1983, 1987). Somatic variants can be transmitted to progeny either by the mutated cell lineage passing directly to the ‘new’ individual during the process of asexual fragmentation, fission, etc. or by the lineage entering the gametes (Otto and Orive 1995; Orive 2001). For instance, in the fungi, mutations arising in any tissue can be transmitted sexually or asexually. Because asexual reproductive rates are so high, favorable mutants, such as those containing virulence alleles (Clay and Kover 1996), can be rapidly increased through natural selection (Caten 1987). Thus, by clonal growth, fungi can evolve significantly by mutation in absence of recombination. Among some simple animals such as Hydra, the zygote divides to produce an interstitial and a somatic cell lineage. The former remains totipotent and mitotically active. By the time gametes are differentiated it is highly likely that somatic variation will have arisen in the forerunners of those cells. Likewise, corals are totipotent and in Buss’s words (1987, p. 107) ... “a 20,000-year-old reef coral had passed uncounted millions of fruit fly generations.”

Plants are a particularly interesting example of the somatic embryogenesis mode of development. As we discuss in Chap. 5, they grow by virtue of the activity of meristematic cells in their shoot and root apices, and additionally in some cases by lateral meristems encircling the axis. The meristems accumulate mutations over repeated cell divisions as the cell lineage increases. A practical consequence of such mutations is that horticulturists have exploited them for centuries to develop most varieties of fruit trees, potatoes, sugar cane, and bananas, not to mention countless vegetatively propagated ornamental and floricultural plants (Silander 1985).

The spread and impact of a somatic mutation depends on many factors, including the strength of selection at the level of cell lineage; when the mutation occurs in the timing of the mitotic lineage; and the number of cell generations per individual generation (Otto and Orive 1995; Otto and Hastings 1998; Orive 2001). Long-lived and particularly large-statured plant species have higher mutation rates per individual generation than do short-lived species because of the greater number of cell divisions before gamete formation (Klekowski and Godfrey 1989; Schultz and Scofield 2009).

A biological consequence of somatic mutation is that plants can develop as mosaics where one component, say a shoot, is genetically quite different from another. This implies that beneficial somatic mutations (e.g., resistance to parasites or insect grazers) could potentially spread easily, whereas at least some kinds of deleterious mutations would be inconsequential because the affected part could be shed (hence, it is argued, no increase in mutational load would occur) (Whitham and Slobodchikoff 1981; Gill et al. 1995). This variation has been hypothesized as being one way by which long-lived plants could contend with rapidly evolving pests and pathogens. The evidence is mixed (see Chap. 5 and Whitham and Schweitzer 2002; Folse and Roughgarden 2011).

Because of the totipotency of plant meristematic cells and the clonal aspect of development, somatic mutations in precursors of a floral lineage can be transmitted to gametes. As alluded to earlier in this chapter, these events happen occasionally and have been documented in the groundbreaking work with transposable elements of maize pioneered by Barbara McClintock (McClintock 1956; Fedoroff 1983, 1989). For example, if a genetic change occurs during the first embryotic cell division, a plant with genotypically and phenotypically distinct halves is created. Each half will go on to produce different gametes. If a similar change is delayed until ears form, two different sectors with correspondingly distinct kernels will develop. Indeed, the order of genetic events can be surmised from the timing in appearance of the sectors. The important point here, however, is that somatic changes can be reflected in the gametes and ultimately zygotes. Hence, somatic variation can not only alter the fitness of the carrier in which they arise, but they can, at least in some instances, most notably with modular organisms, be passed on to offspring produced sexually (Otto and Orive 1995; Pineda-Krch and Lehtila 2004; see Chap. 5). This mechanism extends the conventional forms of genetic variation discussed previously (Sect. 2.3).

Microorganisms have always posed a challenge for the genet idea. As noted above (The Asexual Lifestyle), although members of a clone are essentially identical initially, they diverge genetically over time due to somatic mutation and other processes. For the genet idea to hold, it is not necessary for daughter cells in a mitotic lineage to be genetically identical. Rather, the complication for microbes is mainly that the nature, occurrence, and transmission of a genetic change is often haphazard and may occur outside the conventional sexual cycle. Formation of the bacterial recombinant is not tied to a particular divisional event, a morphological structure, or a characteristic life cycle stage such as reproduction, dormancy, or dispersal. The recombinant cells are frequently not even evident in a mixed population unless identifiable phenotypic traits such as auxotrophic markers are involved. Some eukaryotic microbes (e.g., the ascomycete and zygomycete fungi) typically are diploid only very briefly, so the nature of the event triggering a new genet is not clear, unlike the usual case among plants and animals (for elaboration of this point see Anderson and Kohn 1995).

Fungi, perhaps uniquely among organisms, tend to fuse upon contact. While such comingling tends to be restricted by vegetative incompatibility systems to close relatives, a single fungal thallus may exist as a genetic mosaic, with genetically different nuclei operating within a common hyphal cytoplasm (Peabody et al. 2000; James et al. 2008; Roper et al. 2013) or spore (Kuhn et al. 2001). As we have seen, by way of heterokaryosis and parasexuality, mutational and recombinational events can be expressed, transferred clonally, and exposed to natural selection independently of fertilization. The evolutionary implications of migration of new genes through an existing genet, followed by change in phenotype and outgrowth of a new genet, present complications to conventional modular theory, a topic taken up in detail in Chap. 5. This fungal situation does not arise with unitary organisms because the germ cells are segregated from the soma, and within other modular life forms this kind of gene migration would be rare, if not unique. Thus, while genets can be visualized clearly for most macroorganisms, the concept must be applied somewhat abstractly for some and perhaps most microorganisms.

The preceding foray into potentially heritable somatic variation is necessary because it documents that genetic variation occurs at many levels, including the cellular, as well as those of the so-called ‘physiological’ and ‘genetic’ individual. This challenges the dogma that the developing product of the zygote (by which is implied a single entity arising from gametic fusion) is the unit of variation. It means, on balance, that the concept of a genet is an ideal that is more or less approximated in various phyla. The unitary organisms, most clearly illustrated by the vertebrates, come closer than do modular life forms (many invertebrates, plants, fungi) in behaving as genetic individuals. As reviewed earlier in this chapter, molecular biology is showing that mobile genetic elements can move among chromosomes of a cell, among cells, and between the somatic and germ lines. One consequence of this fluidity is increased somatic variation and potentially a direct route from soma to gametes. Even in the case of unitary macroorganisms, the concept of the genetic individual must now be revised to reflect more flexibility and fluidity.