Cooperation is widespread across life, but its existence can be threatened by exploitation. The rise of obligate social cheaters that are incapable of contributing to a necessary cooperative function can lead to the loss of that function.  In the social amoeba Dictyostelium discoideum, obligate social cheaters cannot form dead stalk cells and in chimeras instead form living spore cells. This gives them a competitive advantage within chimeras. However, obligate cheaters of this kind have thusfar not been found in nature, probably because they are often enough in clonal populations that they need to retain the ability to produce stalks.  In this study we discovered an additional cost to obligate cheaters. Even when there are wild-type cells to parasitize, the chimeric fruiting bodies that result have shorter stalks and these are disadvantaged in spore dispersal.  The inability of obligate cheaters to form fruiting bodies when they are on their own combined with the lower functionality of fruiting bodies when they are not represent limits on obligate social cheating as a strategy.

Strains and culture conditions 

To prepare food bacteria for D. discoideum clones to prey upon, we first spread non-pathogenic K. pneumoniae KpGe (Dicty Stock Center, dictybase.org) from stocks frozen in 80% KK2 [2.25 g KH₂PO₄ (Sigma-Aldrich) and 0.67 g K₂HPO₄ (Fisher Scientific) per liter] and 20% glycerol on an SM/5 agar media [2 g glucose (Fisher Scientific), 2 g yeast extract (Oxoid), 0.2 g MgCl₂ (Fisher Scientific), 1.9 g KHPO₄ (Sigma-Aldrich), 1 g K₂HPO₅ (Fisher Scientific), and 15 g agar (Fisher Scientific) per liter] and allowed the bacteria to grow at room temperature until colonies appeared. We picked a single colony with a sterile loop, spread it on a new SM/5 plate, and allowed the bacteria to proliferate. We collected these bacteria into KK2 with a sterile loop and diluted them to 1.5 OD₆₀₀ in KK2 (~5 x 10⁸ cells, measured with an Eppendorf BioPhotometer). We used these bacteria as food for amoebas in our experiment and repeated this process anew for each of the three replicate experiments.

To grow NC28.1, the wild-type ancestor D. discoideum clone, from freezer stocks for use in our experiments we added spores frozen in 80% KK2 and 20% glycerol to 200ml of 1.5 OD₆₀₀ K. pneumoniae suspension. We spread the mix of spores and bacteria on SM/5 plates with a sterile glass spreader, then incubated the plates at room temperature for 7 days under constant overhead light until the social cycle was complete and fruiting bodies had formed. We repeated this process for each of the three replicate experiments.

EC2 (also called EC28.2 in (Inglis, Ryu et al. 2017)) was selected as an obligate cheater.  It is the result of an experimental evolution experiment that used unstructured growth and dispersal to evolve a cooperative wild strain called NC28.1 into non-fruiting cheaters.  Past work measuring the degree of cheating (the overrepresentation of spores within chimeric fruiting bodies) suggests that EC2 is a particularly effective cheater, and its recent divergence from a wild isolate should render its behavior more natural than cheaters derived from extensively lab-adapted strains.  To grow EC2, the RFP-labelled obligate social cheater, from freezer stocks for use in our experiments, we added amoebas frozen in HL5 (5 g proteose peptone, 5 g thiotone E peptone, 10 g glucose, 5 g yeast extract, 0.35 g Na2HPO4 * 7H2O, 0.35 g KH2PO4 per liter) to 10% DMSO to 200ml of 1.5 OD₆₀₀ K. pneumoniae suspension. We spread the mix of amoebas and bacteria on an SM/5 plate with a sterile glass spreader, then incubated the plate at room temperature for 24-48 hours until starving EC2 amoebas began aggregating. We then used a sterile loop to transfer a sample to a new plate containing fresh K. pneumoniae for them to prey upon. These were allowed to grow for 24-48 hours until a vegetative front of amoebas had formed. We collected these amoebas with a sterile loop into ice-cold KK2 (see “Experimental procedures”) and ensured that the amoebas we used were clonal by plating 10 SM/5 plates with about 10 amoebas each, then picking a single clonal plaque originating from a single amoeba. We repeated this process for each of the three replicate experiments.

Experimental procedures

In order to obtain cells of both D. discoideum clones for experimental mixing, we plated amoebas (EC2) or spores (NC28.1) previously grown from freezer stocks as described above on separate SM/5 agar plates with 200ml of 1.5 OD₆₀₀ K. pneumoniae suspension.

We collected amoebas to make the mixtures by pouring ice-cold KK2 onto the plates, mixing them into suspension with a gloved fingertip, then collecting and centrifuging the mixture at 10°C for 3 minutes at 1300 rpm in order to pellet the amoebas and leave K. pneumoniae in solution. We decanted the pellets, resuspended them in KK2, and measured their density with a hemacytometer before making the mixtures. For each treatment, we mixed 200ml of fresh K. pneumoniae suspension with a total of 2x10⁵ amoebas then spread the solution evenly with an ethanol-sterilized glass spreader on an SM/5 agar plate. We made mixtures of EC2 and NC28.1 with various initial frequencies of EC2 (0.0, 0.1, 0.3, 0.5, 0.7, 0.9, and 1.0). We repeated this experiment three times, each on a separate day.

We collected fruiting bodies after one week at room temperature under constant overhead light to allow fruiting bodies to fully develop. On each plate, we selected three fruiting bodies at random to represent three independent data points. To do this, we placed a plate of fruiting bodies over a grid of 1cm by 1cm squares. We selected three squares at random using a random number generator and marked each plate at the centers of each of the squares. We then individually collected the closest intact (not collapsed) fruiting body to each mark with fine tweezers. For each, we pressed the sorus, which contains the spores, against the side of a tube containing 100ml of KK2 to dislodge the spores, then laid the stalk on a glass microscope slide. After three fruiting bodies were collected from a single plate, the stalks were covered with a cover slip and sealed with nail polish for later imaging. Stalk length was individually recorded by imaging picked stalks under a Leica S8AP0 dissecting microscope with Leica application suite software v4.1 using the “draw line” tool.

 Analysis

We excluded several data points from the analysis for which we could not accurately measure stalk height due to damage incurred during collection.

In order to test whether increasing cheater frequency yields shorter fruiting bodies, we used a linear mixed-effects model with the function lme in the nlme package in R version 4.2.1 (Team 2013) with stalk height as the response variable, the initial cheater frequency as a fixed effect, the total number of spores per sorus as a fixed effect, and the day of the experiment as a random effect (stalk height ~ initial cheater frequency + total spores + 1|day).  In this model, we included total number of spores as a fixed effect in case fruiting body height could vary due to random variation in the size of aggregates that form across the plate. 

We then compared this model with one lacking the random effect of day (stalk height ~ initial cheater frequency + total spores) using the anova function in base R. The two models were not significantly different (p = 0.27, with day: AIC = 105.67, without day: AIC = 104.89), so we proceeded with the simpler model without the effect of day. We then further simplified the model by removing the effect of total number of spores because it did not significantly affect stalk height (p = 0.20).