Life is hierarchically organized (1): Genes cooperate to form genomes, discrete genomes to produce eukaryotic cells, and millions of cells to produce a human or an oak tree. In each case, lower units formed a cooperative group and lost their ability to reproduce independently—each underwent a major evolutionary transition (1, 2). This raises a new question: Cooperative groups are vulnerable to the “selfish” interests of the individuals that comprise them, so how is cooperation maintained? (3) In a population of cells comprising a multicellular organism, selection might favor cells that neglect producing soma to increase investment in reproductive germline (4–6) or that selfishly proliferate as cancers (see Box 1). While the potential consequences of within-organism selection have been discussed widely (e.g., refs. 7–12), the technologies required to test these ideas have only recently become available and developmental biology has therefore recently begun to incorporate them. Testing these ideas, however, requires a cross-disciplinary approach built on a common foundation. Here, we provide this common foundation: an introduction to current thinking in evolutionary biology on multicellularity for developmental biologists and a guide to relevant aspects of developmental biology for evolutionary biologists. We assess whether our understanding of cellular, developmental, and reproductive biology across the Metazoa confirms or challenges expectations from evolutionary biology and highlight avenues of future research. While we concentrate on animals, this discussion applies to all multicellular groups (see refs. 12 and 13 for previous treatments of intraorganismal conflict). These discussions apply most strongly to “unitary” organisms rather than “modular” organisms built from physically connected units that are at least partially self-sustaining and able to reproduce, as in tree branches or coral polyps. The complete interdependence of parts in a unitary organism means they survive and reproduce as a whole (9) (for discussions of modular organisms see refs. 12 and 13). We provide a background of the relevant theory, which generates clear expectations of the conditions required for multicellularity to evolve and persist. We then explore the biology of some of the many animals that defy these expectations. Here, we apply a major evolutionary transitions framework to understand the evolution of multicellularity (see refs. 14–17 for reviews with a more mechanistic perspective). We use the analogous transition to sociality in insects, arguably the best-studied major transition, to highlight where our understanding of multicellular evolution is lacking. As is common in evolutionary literature, we use “intentional” language like “selfish” (SI Appendix, Glossary).
Multicellularity has arisen independently at least 25 times (18). The majority of multicellular groups, however, are simple facultative aggregations: Each cell can survive and reproduce independently and can establish another multicellular group (18). Obligate multicellularity is rarer, seen in animals, plants, fungi, red algae, brown algae, and potentially some ciliates and cyanobacteria (19, 20). Under obligate multicellularity, cells cannot survive or reproduce independently of the multicellular group, and groups contain multiple sterile cell types (19). While cells in obligately multicellular organisms still divide, only a fraction can establish a new group, so this within-organism division is not considered reproduction.
Two conditions are often described as necessary for the evolution of obligate multicellularity: development from a single cell (1, 19, 21) and an early and strict separation of sterile somatic cells from reproductive germ cell lineages (7, 22). These mechanisms align all cells’ fitness interests and are predicted to minimize conflict within multicellular groups, by creating a clonal group with limited access to future generations.
Conclusion： Research into the evolution of multicellularity has focused on facultatively multicellular organisms, such as Dictyostelium slime molds, as they enable manipulation of the relative costs and benefits to cells of independent or group living. Yet, those same characteristics that make them good experimental systems may prevent them from evolving obligate multicellularity. Obligately multicellular organisms, like us, can be considered boringly predictable. However, we highlight that our conflict-free expectations are colored by our narrow set of model organisms, that within-organism selection among cell lineages may be more prevalent than often assumed and that, by exploring more broadly across the Metazoa, a broader understanding of multicellularity could emerge.
