Genetic conflicts and intercellular heterogeneity

Abstract

Adaptation is what elevates biology above mere chemistry and physics, and the salient feature of adaptation is that it is generally associated with individual organisms. The edge of an organism is a crucial boundary, within which we generally find adaptive cooperation and outside of which we find competition and conflict. Therefore, for understanding adaptation, it is important to have a clear conception of the forces that created and maintain the individual organism and to understand when organisms occur in forms that challenge the identity of the individual as an adapted unit. For multicellular organisms, genetic homogeneity is thought to be the most essential element; each cell will be selected to act for the good of the whole if the other cells are genetically identical. Therefore, multicellular individuals that break this rule present a challenge to our conception of the organism. As a preliminary, we would like to clarify the subject of this discussion. The cases of intraorganismal genetic heterogenity discussed by Pineda-Krch & Lehtilä (2004) would be better described as intercellular heterogeneity – different cells of the same individual having different genes. Heterogeneity can also lead to conflicts between genes within the same cells, including meiotic drive, transposition, genomic imprinting, and conflict arising because of atypical transmission of B chromosomes or sex chromosomes. These are also interesting, and for some of the same reasons, but here our focus is on the narrower phenomenon of differences between cells of the same organism. When does intercellular heterogeneity challenge the identity of an individual as an adapted unit? This may depend on the type of heterogeneity. A mosaic arises when there is a mutation in a part of an individual that does not affect the entire individual. It can be in either somatic or germ-line tissues. A chimera is an organism that forms when two genetically separate and different individuals merge. These three forms of intraindividual heterogeneity do not equally challenge our view of what an individual is. Somatic mutations must arise in all multicellular organisms and any individual with a large number of cells and a large number of mutable genes is going to be a somatic mosaic to some degree. Such mutations can be fatal to the individual, as is often the case with cancer, but they are not fatal to our concept of the individual. Such mutations are not inherited. A germ-line mosaic is potentially a greater problem because the variant can compete for access to the next generation. But the competition is usually limited by passage through a single-cell bottleneck (Maynard Smith, 1989). This means that each mutation, however effective a cheater, gets only one opportunity to cheat. This keeps it from being too much of a threat to individuality. Chimeras are more challenging because within-individual variation does not go away, but is recreated each generation although mergers. A single exploitative mutation can therefore spread through the entire population, which is impossible in mosaics. This complicates our view of the individual as an adapted unit because selection will operate on the chimeric individual, but may also favour exploitation by one of the chimeric partners. This in turn can favour genes that resist fusion, or resist exploitation after fusion. Rarity of chimeras is expected as it involves cooperation among unrelated lines. If chimerism is much more common than is generally believed, then it could challenge our view of the importance of clonality in maintaining the individual as an adapted, selected whole. The small number of references given in Pineda-Krch and Lehtilä do not yet make a strong case for high levels of chimerism. On the other hand, there are several strong arguments that clonality and absence of conflict are not essential to organismal function. First, there are a few examples of chimeras where conflict has been demonstrated, but it does not destroy organismal function. Chimerism among unrelated individuals occurs in the social amoeba Dictyostelium discoideum, when individual, predatory amoebae starve. They follow cAMP gradients, forming a slug with up to hundreds of thousands of cells that may be made up of several genetically distinct clones. The slug moves up through the soil and leaf litter, towards light and away from ammonia, then forms a fruiting body consisting of 20% dead stalk cells that support the remaining cells differentiated as spores (Kessin, 2001). The stalk provides a dispersal advantage to the remaining cells at the ultimate cost to themselves. These fruiting bodies may be clonal or chimeric. Even cells in chimeric fruiting bodies have thousands of clone mates present, making overall relatedness high, averaging about 0.5 in small soil samples that might be the arena over which cells aggregate (Fortunato et al., 2003). There is conflict over contribution to stalk cells in chimeras, with some clones successfully contributing proportionally more to spore than to stalk (Strassmann et al., 2000). Another cost of chimerism is that chimeric slugs move less far than slugs of pure clones of the same size, perhaps because of competition within the slug (Foster et al., 2002). However, the study also points to the advantage of chimerism. Joining non-clonemates in an aggregation may be the only way to achieve large size in D. discoideum where amoebae aggregate across very small areas. Large size is advantageous because larger slugs migrate farther, making them more likely to reach up to the forest floor where the spores can be dispersed (Foster et al., 2002). The advantages of cooperation in migration apparently outweigh the costs of potentially being exploited by another clone in the chimera. Attaining larger size is a common advantage of chimerism. A second example comes from the social insects. If a social insect colony is a superorganism, it could be viewed as a chimeric one. This chimera does not usually arise through fusion, but through the production of worker daughters by the queen. These parts may appear to be individuals in their own right, but they are not independent. The workers are a necessary part of the superorganism; without them, it could not reproduce. Queens are related to their own daughters by 1/2, and those daughters, at least in the Hymenoptera (ants, bees and wasps), are related to each other by 3/4, if their mother mated only once. So even in the simplest, most highly related colony type, parts of the superorganism differ in their genetic make-up. A large fraction of genes are shared, but not nearly so large a fraction as would be found in even the most ancient mosaic. In social insect superorganisms genetic variability results in some within-organism conflict. For example, in colonies with a single once-mated queen, the queen should prefer equal investment in males and future queens, while the workers should prefer a 3 : 1 ratio of investment biased towards future queens. In the ant Formica truncorum, workers kill male larvae in colonies where the queen mated only once, but not in colonies where the queen mated multiple times, in accord with worker sex ratio interests (Sundström et al., 1996). Workers in many species can produce the males themselves, something that is facilitated by males coming from unfertilized eggs. On relatedness grounds, the workers would be selected to defer to their mother queen in male production only if the queen has mated more than twice (Ratnieks, 1988). This explains why queens in multiple mated honeybees produce all the males, while in singly mated stingless bees, workers often produce males (Peters et al., 1999; Tóth et al., 2002). Conflict in social insect superorganisms does not depend solely on relatedness. It also depends on power (Alexander, 1974). For example, conflicts are predicted over who gets to be queen, but in many species, including honeybees, such conflict is minimized because workers limit the number of developing queens through their control of food provisions. The cost of not exerting such power is illustrated by stingless bees in the genus Melipona, where all cells are provisioned equally, leaving caste determination in the hands of the individual developing female. In this genus, up to 20% of females develop as queens, although only a few are needed, and the rest are swiftly executed by the workers. The workers still exercise ultimate power, but their failure to do so earlier leads to considerable waste of resources on raising surplus queens. These conflicts must have a price in terms of reproductive success, but is seems that the price is more than compensated by the advantages to cooperation among the members of the superorganism. Clearly, there is a great deal of cooperation within a social insect colony. Foraging, caring for young, and defence, are all highly cooperative (Wilson, 1971). Social insects are ecologically and evolutionarily very successful, making up a large share of the tropical animal biomass. Finally, there is another common form of cooperation, even more striking than chimeras, that involves unrelated partners: mutualisms. Some mutualisms could be viewed as chimeras of zero relatedness. Some are highly physically integrated, for example Rhizobium-legume mutualisms. Rhizobium provides nitrogen while legumes provide carbon, and the two evolve to trade these two necessary substances (Denison, 2000). Such chimeras are often stable, widespread and successful. These examples demonstrate that clonality is not essential to the cooperative enterprise, and that it may be worth looking for more examples of chimeras. On the other hand, social insects and mutualisms differ from standard chimeras in important respects. Unlike cells, social insects do not generally have the option of building their colonies out of clonal units. Perhaps if they did have that choice, chimeric superorganismal colonies would be no more common than standard chimeras. As for mutualisms, an important difference is that mutualistic partners bring different pre-evolved functions to the chimeric individual, while those of the same species do not (Queller, 1997, 2000). Two legumes would not be able to profitably exchange carbon and nitrogen; nor would two Rhizobium cells. We therefore suggest that chimeras are rare for several reasons. Genetic conflicts in chimeras may sometimes be destructive. However, the rarity of chimeras may also stem from the fact that there is rather little that conspecific cells can initially do for each other. Finally, when there are things they can do for each other, this can often be accomplished with clonemates.