Abstract
Many multicellular organisms produce two cell lineages: germ cells, whose descendants form the next generation, and somatic cells which support, protect, and disperse the germ cells. This distinction has evolved independently in dozens of multicellular taxa but is absent in unicellular species. We propose that unicellular, soma-producing populations are intrinsically susceptible to invasion by non-differentiating mutants which ultimately eradicate the differentiating lineage. We argue that multicellularity can prevent the victory of such mutants. To test this hypothesis, we engineer strains of the budding yeast Saccharomyces cerevisiae that differ only in the presence or absence of multicellularity and somatic differentiation, permitting direct comparisons between organisms with different lifestyles. We find that non-differentiating mutants overtake unicellular populations but are outcompeted by multicellular differentiating strains, suggesting that multicellularity confers evolutionary stability to somatic differentiation.
One Sentence Summary Using a synthetic biological approach, we show that multicellularity protects species that produce somatic cells from exploitation by common mutants.
Main Text
Somatic differentiation, a permanent change in gene expression inherited by all of a cell’s descendants, produces somatic cells from a totipotent germ line. Though somatic cells may divide indefinitely, they cannot beget the complete organism and are thus considered non-reproductive. Generation of such sterile cells has clear fitness costs that must be offset by somatic functions which improve the viability or fecundity of germ cells. The absence of a soma in unicellular species (1), as well as the persistence of undifferentiated multicellular groups among the volvocine algae (2) and cyanobacteria (3), has fueled speculation that multicellularity must arise before somatic differentiation can evolve (4-7). It has been argued that somatic differentiation is not observed in unicellular species because the potential fitness benefits are insufficient (6-8): while soma can contribute motility and protective structures to multicellular organisms, somatic cells in a unicellular species could only benefit the germ line through secretion of useful products into a shared extracellular milieu. However, nutrient exchange between members of microbial consortia (9, 10) demonstrates the potential for productive interactions between cell types in the absence of physical adhesion. Benefits associated with somatic differentiation, including reproductive division of labor (11) and suppression of germ line mutations through lineage sequestration (12), are thus likely accessible to unicellular species.
We propose the alternative hypothesis that unicellular somatic differentiation could offer fitness benefits in a population of genetically identical cells, but remains rare because it is not an evolutionarily stable strategy (13). Commonly-occurring mutants that do not differentiate (“cheats”) could take advantage of somatic cell products in the shared media without paying the reproductive costs of differentiation, thus increasing in frequency until their genotype prevails. We further posit that if multicellularity results from cells of a single lineage failing to disperse (rather than cells aggregating from different lineages), differentiating populations may outcompete cheats: although cheats initially arise through mutation in a group with somatic cells, their descendants will eventually be confined to their own multicellular groups composed entirely of cheats and thus cannot benefit from the local accumulation of somatic cell products (14).
To test this hypothesis, we designed strains of the budding yeast Saccharomyces cerevisiae that produce soma, are multicellular, or both: one strain is a multicellular, differentiating organism and the other two represent both possible intermediates in its evolution from a non-differentiating, unicellular ancestor (Fig. 1A). Employing synthetic strains which differ from one another at only a few, well-defined loci ensures that no undesired variables confound the direct comparison of fitness and evolutionary stability, and permits the study of a unicellular, soma-producing lifestyle not found in nature. Performing experimental tests with living organisms also avoids the potential pitfall of biologically-unrealistic parameter regimes in purely analytical models.
We mimicked somatic differentiation by engineering fast-growing “germ” cells which can give rise to slower dividing, differentiated “somatic” cells that secrete invertase (Suc2), an enzyme that digests sucrose (which this yeast strain cannot take up directly) into the monosaccharides glucose and fructose, which any cell in the shared medium can then import (15, 16) (Fig. 1B). These somatic cells thus perform a digestive function, ensuring the availability of monosaccharides which serve as the sole carbon source during growth in sucrose minimal media. In nature, differentiation is often achieved through multiply-redundant gene regulatory networks that stabilize cell fate (17): to simplify our system, we instead made differentiation permanent and heritable by forcing the expression of somatic cell-specific genes to depend on a site-specific recombinase that excises the genes needed for rapid cell proliferation (Fig. 1C).
Germ and somatic cells must be present at a suitable ratio for fast growth on sucrose: germ cells have the higher maximum growth rate, but monosaccharides become limiting when somatic cells are rare. We predicted that the ratio between cell types would reach a steady-state value reflecting the balance between unidirectional conversion of germ cells into somatic cells and the restricted division of somatic cells. We designed tunable differentiation and division rates to allow us to regulate the ratio between cell types and thus control the growth rate of the culture as a whole.
Both features depend on a single, genetically-engineered locus (Fig. 1C). In germ cells, this locus expresses the fluorescent protein mCherry and a gene that accelerates cell division, the cycloheximide resistant (cyh2r) allele of the ribosomal protein L28 (ref. 18); in somatic cells, the locus expresses a different fluorescent protein (mCitrine) and the invertase Suc2. The germ line form is converted to the somatic form by a version of Cre recombinase engineered by Lindstrom et al. (19) to be active only in the presence of β-estradiol. Adding β-estradiol to a growing culture induced conversion of germ to somatic cells, apparent as the onset of mCitrine expression and slow loss of mCherry fluorescence by dilution (Figs. 1D and S1; Movie S1). Conversion rates ranged from undetectable levels (< 10-3 conversions per cell per generation) to approximately 0.3 conversions per cell per generation as the β-estradiol concentration increased (Figs. 1E and S2A). Expression of the codominant, cycloheximide-sensitive wild-type allele of CYH2 from its native locus permitted continued growth following cyh2r excision, but at a reduced rate which depended on cycloheximide concentration. The growth rate deficit of somatic cells ranged from undetectable (< 1%) to nearly 30% as the cycloheximide concentration increased (Figs. 1F and S2B).
The combination of irreversible differentiation and restricted somatic cell division caused cultures to approach a steady-state ratio between the two cell types over time (Fig. 2A). The steady-state fraction of somatic cells increased with the conversion rate and decreased with the somatic cells’ growth disadvantage, as expected (Fig. 2B). The steady-state ratio between cell types could be tuned over four orders of magnitude by altering the cycloheximide and β-estradiol concentrations (Fig. 2B).
To investigate the ability of invertase secretion from somatic cells to support germ cell proliferation, we determined how the culture’s growth rate on sucrose depended on the ratio between cell types. In the presence of cycloheximide, cultures containing both cell types at an intermediate ratio grew more quickly in sucrose media than cultures of either cell type alone (Fig. 2C), confirming that somatic cells benefit their germ line through invertase secretion.
Clonal multicellularity arises through the maintenance of contact between daughter cells following cytokinesis (20). In budding yeast, clonal multicellularity can be produced, either through engineering (21) or evolution (22, 23), by mutations that prevent degradation of the septum, the specialized part of the cell wall that connects mother and daughter cells after their cytoplasms have been separated by cytokinesis (24). Deletion of CTS1, a chitinase gene required for septum degradation, causes formation of “clumps” (groups of daughter cells attached through persistent septa) that typically contain 4-30 cells during growth in well-mixed liquid medium (ref. 25, Fig. 3A). In the presence of β-estradiol, differentiating strains that lack CTS1 (Δcts1) produced clumps that frequently contained both germ and somatic cells, as evaluated by fluorescence microscopy (Fig. 3A) and flow cytometry (Figs. 3B and 3C). Combining our gene excision-based differentiation system with CTS1 deletion thus allowed us to produce strains exhibiting all life strategies needed to compare the evolutionary stability of unicellular and multicellular differentiation.
Somatic cells in unicellular strains can only benefit germ cells by secreting useful products into a shared medium; non-differentiating cheats (e.g., Cre- germ cells) and germ cells in well-mixed media have equal access to somatic cell products, but cheats do not pay the reproductive toll of differentiation. We therefore predicted that cheats would enjoy a fitness advantage over germ cells, allowing them to invade unicellular, differentiating populations (Fig. 4A, top). In multicellular species, however, significant local accumulation of somatic cell products (in our experiment, monosaccharides) within multicellular groups can give differentiating lineages an advantage over cheats as long as the benefit of better nutrition overcomes the cost of producing slower-replicating, somatic cells (21). In clonally multicellular species, such as our Δcts1 strain, novel cheats arising by mutation will eventually be segregated into cheat-only groups by cell division and group fragmentation. We hypothesized that cheats would then experience reduced access to somatic cell products, potentially negating their growth advantage over germ cells (Fig. 4A, bottom).
To test this prediction, we introduced cheats into unicellular or multicellular differentiating cultures and investigated their fate. We mixed differentiating cultures expressing a third fluorescent protein, Cerulean, with cheats that lack this third color and cannot differentiate because they lack the recombinase whose action gives rise to somatic cells (Cre-). We monitored the relative frequency of the two strains over a series of growth and dilution cycles. In sucrose media, cheats invaded unicellular, differentiating populations but were outcompeted in multicellular, differentiating populations (Fig. 4B). The growth advantage of multicellular, differentiating strains was nullified in monosaccharide-containing media, where somatic cells should confer no fitness advantage to clumps (Fig. 4B). Moreover, the growth advantage of differentiating strains depended strongly on the conversion rate (β-estradiol concentration): higher conversion rates were advantageous only in the multicellular differentiating case (Fig. 4B), where they increased the fraction of somatic cells overall as well as the fraction of clumps containing at least one somatic cell (Fig. 3C). In unicellular cultures grown on sucrose, or in cultures grown on glucose (unicellular or multicellular), increasing the conversion rate increased the growth advantage of cheats by reducing the growth rate of the germ cell population (Fig. 4B).
Our results show that unicellular, soma-producing strains are evolutionarily unstable to invasion by non-differentiating cheats. This finding is unlikely to depend on the specific molecular mechanisms that effect differentiation or allow somatic cells to assist germ cells: any form of differentiation in a single-celled species would require that the two cell types exchange resources through a shared medium. From the spontaneous mutation rate in budding yeast [≈ 3 × 10-10 per base pair per cell division (26)] and the size of the recombinase gene, we estimate mutations that inactivate our engineered recombination system would occur in about one out of every 107 cell divisions; this is likely an underestimate of the frequency of inactivating mutations in natural differentiation, which typically requires more loci and thus presents a larger target for mutation (17). Because this frequency is high relative to typical microbial population sizes, cheats would thus likely arise and sweep to fixation shortly after the appearance of unicellular somatic differentiation, explaining the absence of extant species with this life strategy. The short persistence time of unicellular differentiating species makes them an unlikely intermediate in the evolution of development relative to undifferentiated multicellular species, which have persisted for hundreds of millions to billions of years in some clades (1-3). We cannot, however, rule out the possibility that the transient existence of a unicellular differentiating species might suffice for the secondary evolution of multicellularity, which has been observed experimentally in small populations on the timescale of weeks (22, 27). In any case, the differentiating phenotype cannot be stably maintained against cheats until clonal multicellularity evolves.
Our study demonstrates that synthetic biology can directly test hypotheses about evolutionary transitions, complementing retrospective inference through comparison of existing species, experimental evolution, mathematical modeling, and simulation. We note that other major evolutionary transitions, including the appearance of body plans (spatially-ordered arrangements of cell types) and life cycles (temporal sequences of growth and dispersion), could be studied through experimental evolution or further engineering of the strains described above. Furthermore, our differentiating strain provides an experimentally-tractable version of a common simplifying assumption in population genetics: independent deleterious mutations are often modeled as having the same fitness cost (28). In our strains, cyh2r excision is a form of irreversible mutation that always produces the same fitness disadvantage, but the magnitude of the fitness cost can be experimentally varied by altering the cycloheximide concentration. Thus engineered organisms, including those we have developed, can permit robust experimental testing of a wide variety of outstanding hypotheses in evolutionary biology.
Acknowledgments
We thank Derek Lindstrom and Dan Gottschling for providing the estradiol-inducible Cre construct PSCW11-cre-EBD78; Alex Schier, Michael Desai, Michael Laub, Cassandra Extavour, David Haig, and members of the Murray and Nelson labs for helpful discussions during manuscript preparation; and Beverly Neugeboren, Linda Kefalas, and Sara Amaral for research support. This work was supported in part by National Science Foundation and Department of Defense National Defense Science and Engineering Graduate Research Fellowships and by NIH/NIGMS grant GM068763.