Structured growth and genetic drift raise relatedness in the social amoeba Dictyostelium discoideum

Abstract

One condition for the evolution of altruism is genetic relatedness between altruist and beneficiary, often achieved through active kin recognition. Here, we investigate the power of a passive process resulting from genetic drift during population growth in the social amoeba Dictyostelium discoideum. We put labelled and unlabelled cells of the same clone in the centre of a plate, and allowed them to proliferate outward. Zones formed by genetic drift owing to the small population of actively growing cells at the colony edge. We also found that single cells could form zones of high relatedness. Relatedness increased at a significantly higher rate when food was in short supply. This study shows that relatedness can be significantly elevated before the social stage without a small founding population size or recognition mechanism.

1. Introduction

The fitness and success of a group of cooperators is under constant threat of collapse by the infiltration and spread of selfish individuals, known as cheaters, that reap the benefits of cooperation without paying the costs [1,2]. Kin selection theory predicts that cooperative behaviours will be selected if relatives preferentially interact with one another by kin recognition or high population viscosity [3–5]. High relatedness is key to kin selection and has been shown in social insects [6,7], birds [8,9], mammals [10,11] and microbes [12–14].

When a microbial colony grows from a single point, zones of highly related individuals can form when genotypes, identical except for their fluorescent labels, are mixed in moderate numbers (approx. 1000) [15]. Models [16] suggest that the mechanism behind this is genetic drift; only a few cells are actively dividing in any region of the growing colony edge and there are no phenotypic differences between cells (figure 1a). Selection could also produce zones, but it is not required. Simulations suggest that the segregation rate is dependent upon the availability of nutrients; diffusible nutrients cannot penetrate far into the colony, lowering the effective population size and promoting segregation by drift. This outward growth results in the formation of growing zones of clonemates (figure 1a).

Figure 1.

Zones of high relatedness form by genetic drift. (a) A thin layer of red and green cells, identical except for their colour actively grow at the colony edge, simple sampling error creates small groups of the same genotype. Cells growing in the direction of the arrows form zones of coloured cells. (b) Relatedness increases within fruiting bodies with distance from centre of origin (O). Each aggregation territory (blue–grey lines) is of a similar size, so the probability a territory is contained within a single zone increases with distance from O.

Further simulations show that when genotypes have clear altruist or cheater phenotypes, altruists obtain the group benefit of their behaviour because of increased relatedness through structured growth [16]. However, the effect structured growth has on relatedness has not been quantitatively assessed for any social microbe, and so to address this issue, we use the social amoeba Dictyostelium discoideum, now a model organism for social evolution studies [17–19]. When food is scarce, starving Dictyostelium cells aggregate to form a fruiting body where around one-fifth die to form a stalk, lifting the remaining cells aloft as viable spores [20]. Such a costly act of altruism must have evolved in groups of highly related individuals [3,4,21]. However, during the aggregation stage, several genotypes mix forming chimaeras [22]. Cheater genotypes avoid contributing their fair share of cells to the stalk and are over-represented in the sporehead [18,22].

We investigate whether genetic drift during the non-social growth stage can act as a mechanism for the formation of highly related groups prior to the social stage. Whether it will form clear zones such as bacteria is not obvious given the high mobility of Dictyostelium amoebae. Natural formation of zones of high relatedness during growth should increase the opportunity for altruism to evolve and be maintained, even in initially mixed populations (figure 1).

Relatedness within fruiting bodies should increase with distance from the origin because the zones widen with distance while the area of an aggregation territory for a given density of amoebae remains constant, increasing the likelihood that cells of only one genotype will fall within an aggregation territory [23] (figure 1b).

Here, we demonstrate that even with large initial populations and mobile cells, founder effects and genetic drift at the growing edge act as a mechanism to increase and maintain relatedness in D. discoideum and that this effect is greater in low food conditions. We also show that clear zones can form from a single cell.

2. Material and methods

Wild D. discoideum clones were grown in the presence of Klebsiella aerogenes (Ka) bacteria on SM agar (10 g peptone, 1 g yeast extract, 10 g glucose, 1.9 g KH₂PO₄, 1.3 g K₂HPO₄, 0.49 g MgSO₄ (anhydrous), 17 g agar per litre) plates and incubated 22°C. We obtained amoebae by spreading spores evenly over SM plates in the presence of Ka. Log-phase cells were harvested and washed by repeated centrifugation in KK2 buffer (16.1 mM KH₂PO₄, 3.7 mM K₂HPO₄).

Clones were transformed with actin15-red fluorescent protein (RFP) following the procedure in Pang et al. [24]. Clones expressing the markers were then further selected on G418-SM agar plates (30 µg G418 ml⁻¹). Mixes were performed with two clones that were isogenic except for the RFP marker, to eliminate any effects of kin recognition.

(a). Genetic drift assay

Log-phase NC28.1 and NC28.1-RFP cells were mixed in three proportions (wt : RFP; 50 : 50, 95 : 5 and 99 : 1) and suspended at a density of 4 × 10⁸ cells ml⁻¹ in KK2 with 1 × 10⁷ total cells pipetted per plate in the centre of a 150 mm wide SM plate spread with Ka. The low-nutrient treatment diluted nutrients 50-fold. Images of whole plates were taken using an SLR digital camera with a 660 nm filter and a 514 nm LED as a light source [25].

Ten fruiting bodies were taken at each of four distances (at less than 1, 3, 5 and 7 cm) from the centre and suspended in spore buffer (20 mM EDTA and 0.1% NP-40). The proportion of the two genotypes in each fruiting body was then determined by direct counting of spores using a fluorescent microscope. The relatedness (r) within individual sporeheads was calculated as:

where p and q are the proportion of the two genotypes in the sporehead and are the proportions of p and q for the whole population [26].

3. Results

For all three mixtures of NC28.1 and NC28.1-RFP, clear zones formed readily (figure 2a), and as a consequence of structured growth, relatedness within fruiting bodies increased with distance from the origin, where relatedness was zero (R² = 0.5944; p = 0.0033; figure 2b). Clones did mix; there was low relatedness in fruiting bodies at the ‘borders’ between zones.

Figure 2.

Zones of high relatedness form readily without recognition. (a) NC28.1 mixed in three different proportions with NC28.1-RFP. The following mixes were (wt : RFP): (i) 50 : 50, (ii) 95 : 5 and (iii) 99 : 1. In each case, distinct zones of 28.1-RFP formed. (b) Relatedness (r) within sporeheads increases from the point of origin in a 50 : 50 mix of NC28.1 and NC28.1-RFP. When food was plentiful (dashed line) r increased significantly with distance from the origin (O) (R² = 0.5944, p = 0.003). When food was reduced (solid line), r also increased (R² = 0.7576, p < 0.001) and to a significantly higher level than when grown on plentiful food (two-way ANOVA: nutrient level F_1,24 = 12.7, p = 0.003; distance form origin: F_3,24 = 18.1, p < 0.001). (c) A non-fruiting mutant forms a zone from a population of originally clonal cells of NC34.1. The mutation occurred at some point between O and the apex of the zone (dashed line). The wide angles suggest the mutant had an accelerated growth rate.

A fortuitous mutation showed that a clonal zone can form from a single mutant cell. On a clonal plate of the genotype NC34.1, a spontaneous mutation occurred during growth producing a zone of cells that could not form normal fruiting bodies (figure 2c).

When the availability of food was decreased, relatedness almost doubled, reaching 0.68 at the plate edge, compared with 0.36 with the regular SM plate (R² = 0.7576, p < 0.001; two-way ANOVA: nutrient level, F_1,24 = 12.7, p = 0.003; distance from origin, F_3,24 = 18.1, p < 0.001; figure 2b).

For data, see the electronic supplemental material.

4. Discussion

In any given population, the fitness of an altruistic gene is dependent on the structure of the population [3,4,27,28]. Our results show that high relatedness can be obtained by structured growth and genetic drift without a small initial group size (figure 2a). This result is important for two reasons. First, high relatedness means there will be distinct local groups of cooperators and cheaters, so that the cooperators are protected from exploitation. Second, these results provide one explanation for the disparity between ready chimaera formation in vitro [22,29], and high relatedness in wild fruiting bodies [19].

This, however, provides another problem; if local dispersal produces areas of high relatedness to neighbours, then it also makes neighbours closest competitors, cancelling out the relatedness effects [5]. However, combining structured growth with regular spore dispersal can maintain altruism, as spore dispersal relaxes the effects of competing with highly related neighbours [5].

During the evolution of multicellularity, a degree of unrelatedness—and therefore conflict—had to be overcome, perhaps by recognition or via some other mechanism [30–33]. We show that a sufficient degree of structured growth may be sufficient to achieve this, despite the mobility of amoebae that tends to mix cells. Relatedness also increased at a higher rate when there were fewer nutrients, creating more distinct zones (figure 2b). This suggests that in the wild where there are likely to be even fewer nutrients, high relatedness may be generated even more readily. Thus, combined with the added friction of a rough natural substrate this effect upon relatedness may be further enhanced. Other mechanisms such as recognition [30] may have evolved to inflate relatedness further, an important adaptation if motile vegetative cells mix slowing down the rate of increase.

Our study shows that structured growth is a potentially important cause of high relatedness. Once altruism evolves, groups of altruists can form while excluding cheaters, allowing more complex mechanisms to evolve.