Will two walk together, except they have agreed? Amos 3:3


B. Rinkevich, National Institute of Oceanography, Tel Shikmona, PO Box 8030, Haifa 31080, Israel.
Tel.: 972 4 856 5275; E-mail: buki@ocean.org.il

Pineda-Krch & Lehtila (2004) review the costs and benefits associated with intraorganismal genetic heterogeneity (IGH), trying to address the impacts of IGH formation ‘on the fitness of the individual’. They conclude that as ‘cellular self-recognition systems prevent fusion with all nonself cells except close kin’, it seems likely that ‘the majority of genetically heterogeneous individuals will be mosaics rather than chimeras’. To what extent this tenet is generally valid remains to be seen as, when dealing with the phenomenaon of chimerism vis-à-vis mosaicism, the literature reveals a much more complex nature of chimerism and its evolutionary sequels. Three additional aspects are discussed below.

A window in ontogeny

The use of experimental manipulations during the ontogeny of vertebrate immune systems has revealed that immunity of these systems is developed through a series of complex biochemical and molecular processes, cumulatively called ‘immune maturation’. It is interesting, therefore, to find that all types of human and mammalian chimeras (cytomictical chimeras, whole-body chimeras, foetal–maternal chimeras, germ cell chimeras and tumour chimeras; Rinkevich, 2001) are developed only during the maturation of the immune system, at the ontogeneic window of pregnancy. Some chimeras, foetal–maternal chimeras (where foreign blood cells have passed between mother and fetus across the placenta), are so common that probably most adult women remain in a chimeric state decades after pregnancy, as after even a single pregnancy, where the blood cells of the offspring engraft maternal organs, forming what is called ‘microchimerism’.

As regards marine colonial invertebrates, such as hermatypic corals, several reports dating back to more than a century have documented that allogeneic planula larvae which settle close to each other may fuse whereas parent coral colonies never used allogeneic assays but rather responded with a variety of rejection mechanisms. By studying the ontogeny of these allospecific responses, allorecognition in the young colonies of the branching coral Stylophora pistillata, Frank et al. (1997) elucidated three types of responses (stable chimeras, transitory chimeras and histoincompatibility), corresponding to three distinctive stages in the maturation of the developed coral colony ontogeny. Similar windows in ontogeny that allow the creation of chimeric entities, were also revealed in a study of four soft coral species (Barki et al., 2002).

All of the above vertebrate and invertebrate chimeras were found to carry significant ecological costs to the formed entities such as reduced growth rates, morphological resorption and necroses in invertebrates (Frank et al., 1997; Barki et al., 2002), autoimmune diseases, freemartins and other abnormal syndromes in vertebrates (Rinkevich, 2001). It should be stated that adults of these species elaborate very efficient allorecognition systems that prevent the formation of chimeras. Although the formation of chimeras was established in the early stages of development in these, and probably in many other colonial marine invertebrates (but not in all, see the elaborated case of the colonial tunicate Botryllus schlosseri; reviewed in Rinkevich, 2002a), it was considered as a mere error in recognition, arising from the limitation of an ‘imperfect ‘system’ (Feldgarden & Yund, 1992). In the vertebrate system, on the contrary, it was considered (Rinkevich, 2001) to be a byproduct of a novel evolutionary role to ‘educate’ the developing embryo with the armamentarium of effector mechanisms, dedicated to purge the individual from pervasive somatic and germ line variants.

Multipartner chimerism

Several studies (Rinkevich, 1996; Rinkevich & Shapira, 1999; Stoner et al., 1999; Barki et al., 2002; Paz & Rinkevich, 2002) reveal the existence of coral and tunicate chimeras made of more than two partners. Multi-partner chimerisms in corals result with degenerative outcomes (Barki et al., 2002) but within stable and vigorous entities in tunicates (Rinkevich & Shapira, 1999; Paz & Rinkevich, 2002). In the tunicate's multi-partner entities, the different intraspecific conflicts represented by all partners within each chimera may alleviate each other, generating an improved entity where natural selection acts on the ‘group’ level rather than on each individual genome. The state of multi-chimerism incurs, however, some costs (such as reduced potential and maximal chimerism size).

Within multi-chimeras, the question of the potential impact on each individual's fitness has not yet been resolved and the key question for the fate of somatic and germ cell parasitism (Pancer et al., 1995; Stoner et al., 1999) has not yet been tested. As oozooids of botryllid ascidians in nature preferentially form kin aggregates (Grosberg & Quinn, 1986; Grosberg, 1997), the establishment of multichimeras may further increase inclusive fitness of closer relatives. As Botryllus classifies a conspecific as a kin by a shared fusibility allele, it ensures that the multichimeras developed will include the genetic background in different combinations of few parental genotypes. Multichimeras can also develop from larval co-settlement with parental colonies (Rinkevich & Weissman, 1987).

Although intraspecific interactions within multichimeric entities occur, the large botryllid multichimeras successfully control feeding substrates. This should effectively prevent colonization of that surface area by other competitive species and/or increase interspecifc competitive abilities (Rinkevich & Shapira, 1999).

Germ line hitchhiking

The phenomenon of intraspecifc germ cell parasitism in colonial tunicates (Pancer et al., 1995; Stoner et al., 1999) may reveal a theoretical challenge to the concept of Darwinian selection (Rinkevich, 2002b). In this system, cells of the parasitic germ line hitchhike with positively selected genotypes passing through successive generations without being visible to natural selection forces. These cells further fulfil the four criteria cited by Pancer et al. (1995) for germ cell parasitism in Botryllus (transitive/nontransitive hierarchies, disproportionate share of gametic output, frequency increase, somatic embryogenesis type of development).

Hitchhiking onto the soma of positively selected genotypes provides the parasitic forms with the inevitable advantage of establishing new progenies, which may eventually turn into a Pyrric victory (Rinkevich, 2002a,b). By employing three possible defence mechanisms, the inferior partners in this ‘germline war’ can reduce the threat from the development of super-parasitic entities through the: (1) diversification of fusibility allele repertoire, a process that will eventually decrease fusion rates (Yund & Feldgarden, 1992); (2) development of ‘chimeric death’, by activating the process of programmed life span (Rinkevich et al., 1992) and by eradication of all genotypes involved in the germ line war; (3) establishment of multipartner chimeras where firstly, even the most inferior partners in the germ cell hierarchy may have access to reproductive niches, and secondly, where already established multichimeras fail to further fuse with additional fusibility compatible genotypes.

The above three facets of chimerism reveal that this biological system intermingles natural tissue transplantation phenomena with stem cell biology in complex ecological mechanisms that tangle various evolutionary concepts. I posit that this type of intraorganismic selection is of utmost evolutionary significance than when compared with the phenomenon of organismic mosaicism. Scientific questions that seek for the evolution of biological phenomena such as multicellularity, immunology, stem cell and developmental biology may be associated with natural chimerism rather than with an organismic mosaicism.


This study was supported by the NIH (R01-DK54762) and by the Israel Science Foundation (456/01).