Small molecules mediate bacterial farming by social amoebae

The origin of domestication persists as an important question at the nexus between biology, archaeology, and history (1). As intensive research continues apace on organisms whose domestication accompanied the emergence of modern human societies (2, 3), the study of agricultural evolution has expanded to include nonhuman organisms (bark beetles, termites, ants) whose natural history is interwoven with the dispersal, cultivation, and sustainable harvest of microbial symbionts (4). The most recent addition to the roster of nonhuman farmers is Dictyostelium discoideum, a social amoeba that preys upon soil bacteria (5, 6). The life cycle of D. discoideum makes it a favorable model system for the study of cooperation, altruism, and other aspects of social evolution (7–9). When food is scarce, single-celled amoebae aggregate into a multicellular “slug,” which eventually forms a stalked fruiting body from which spores disperse to new habitats. Up to a third of all wild clones of D. discoideum are “farmer strains” that incorporate bacteria into fruiting bodies and establish new food crops during spore dispersal (5), using, as it were, a “take out” foraging strategy that confers advantages under conditions of food scarcity. However, only half of the bacterial cells (Pseudomonas fluorescens) carried by farmer strains of D. discoideum are eaten, begging a closer investigation of the bacterial crop to determine whether its domestication reflects hidden costs and benefits of its association with social amoebae.

In PNAS, Stallforth et al. (6) use an integrative and comparative approach to address this conundrum. Differential metabolomics and chemical analysis revealed that the inedible strain of P. fluorescens (Pf-A) produces two small aromatic molecules, pyrrolnitrin and a unique chromene, which were shown to benefit host amoebae by enhancing spore production in farmer strains of D. discoideum at natural dosages, whereas spore production by nonfarmer strains was depressed in the presence of these compounds, presumably due to their toxicity. Considering its antibiotic properties (10), pyrrolnitrin may also confer competitive advantages to farmer strains of D. discoideum that carry Pf-A, through allelopathy against soil-borne pathogens. In contrast, the edible strain of P. fluorescens (Pf-B) lacks these secondary compounds, but instead produces two specific enantiomers (4′S, 4″S) of another aromatic compound, enantio-pyochelin. Because the structural genes responsible for the biosynthesis of these compounds had identical

The study by Stallforth et al. highlights the primacy of small molecules as mediators of pattern and process in microbial ecology.

coding sequences in Pf-A and Pf-B strains, the authors reasoned that regulatory mutations were more likely to underlie their different chemical profiles and examined the GacA–GacS two-component system responsible for regulating virulence factors in related Gram-negative bacteria (11). Indeed, a point mutation in the global activator A (gacA) gene of the Pf-B strain inserts a stop codon directly upstream of a sequence encoding a helix-turn-helix motif, resulting in a truncated gene product that cannot bind DNA and activate GacS, the cognate response factor. Like Pf-B, artificial GacA KO mutants of P. fluorescens produce enantio-pyochelin, lack chromene and pyrrolnitrin, and can be eaten by amoebae. Finally, a phylogenetic analysis of gacA sequences, including all known strains of P. fluorescens, showed that the Pf-B strain has the only nonfunctional gacA sequence, suggesting that the mutation is a derived condition.

Regulatory Mutation as a Path to Domestication?

The implications of this study extend beyond the growing body of insights gained through research on social amoebae (12). The identification of a single mutation that is necessary and sufficient to account for the shift from a dangerous, beneficial symbiont (Pf-A as a Mastiff) to a benign, portable food source (Pf-B as a Chihuahua) resonates with two questions central to most studies of domestication: how many steps are required and what kinds of genes are responsible? In a recent review (2), six of the seven genes thought to be responsible for domestication traits (e.g., inflorescence architecture) of maize, wheat, rice, and tomato were identified as transcriptional regulators, whereas 9 of the 19 mutations responsible for varietal differences (e.g., grain color, nutritional content, and timing) among such crops were found in structural genes. This makes sense, as the pleiotropic networks of regulatory genes are far more extensive, by definition, than those of structural genes embedded in specific biosynthetic pathways. However, there are examples of mutant structural genes that fundamentally convert beneficial symbionts into pathogens (13, 14). Although the null gacA mutant responsible for the Pf-B strain of bacteria farmed by social amoebae more closely resembles the genes that alter glume formation and shattering in grain crops, it is unclear whether the shift from Pf-A to Pf-B represents a domestication event or the origin of a new variety, because the relative dependence of each strain on the host amoebae remains unknown. One logical follow-up would be a competition experiment that tracked the conditional reproductive success of Pf-A and Pf-B strains through clonal kin selection, under different starvation conditions for host amoebae.

Comparative Biology and Small Molecules Provide Relevance for Informatics Approaches

Another important lesson from this study is the resurgence of comparative biology as a source of insight in the genomics age. The authors’ analytical strategy was built on a foundation of knowledge gained through applied studies of related Pseudomonas bacteria known primarily as opportunistic pathogens of humans (Pseudomonas aeruginosa) and their domesticated crop plants (Pseudomonas syringae). The two-component GacA–GacS regulatory system found to be nonfunctional in Pf-B was described from P. aeruginosa, and the phylogenetic comparisons of GacA sequences were rooted using P. syringae as an outgroup. The closer relatedness of these species to each other (as opposed to a forced comparison between more disparate model systems) was critical to the authors’ success. Finally, the study by Stallforth et al. highlights the primacy of small molecules as mediators of pattern and process in microbial ecology, and by extension, of microbial communities as drivers of human health and agriculture. A key difference between allelopathic Pf-A and benign Pf-B strains of P. fluorescens is the presence/absence of low-molecular-weight compounds with distinctive structures and functions. As the authors state, “these findings highlight the usefulness of investigating the small molecule chemistry that underlies so much of bacterial-eukaryotic symbiotic associations as a source of both new chemistry—molecules and pathways—and biology—defense, development, cooperation and evolution.” In fact, small molecules with similar properties mediate the most destructive activities of related bacteria. P. aeruginosa produces biofilms in the lungs of human cystic fibrosis patients as a consequence of quorum sensing (15). Dual-component signaling systems characterize quorum sensing in Gram-negative bacteria, in this case using homoserine lactones as the critical signal molecules by which bacteria sense each other’s presence (16). P. syringae infects host plant tissues through coronatine, a small molecular “code breaker” that chemically mimics jasmonic acid and thus short-circuits salicylic acid–mediated plant defenses through cross-talk (17, 18). However, as was demonstrated elegantly for another model organism, the Caenorhabditis elegans flatworm (19), the value of small molecules is not limited to human health and agricultural applications. Rather, they are elements of an ancient chemical language through which the microbial world—which is to say, most of the biosphere—is bound in complex dialogues of competition and cooperation (20).