Social Interactions and Distribution of Bacillus subtilis Pherotypes at Microscale

Bacillus subtilis strains communicate through the comQXPA quorum sensing (QS) system, which regulates genes expressed during early stationary phase. A high polymorphism of comQXP′ loci was found in closely related strains isolated from desert soil samples separated by distances ranging from meters to kilometers. The observed polymorphism comprised four communication groups (pherotypes), such that strains belonging to the same pherotype exchanged information efficiently but strains from different pherotypes failed to communicate. To determine whether the same level of polymorphism in the comQXP′ QS system could be detected at microscale, B. subtilis isolates were obtained from two separate 1-cm³ soil samples, which were progressively divided into smaller sections. Cross-activation studies using pherotype-responsive reporter strains indicated the same number of communication pherotypes at microscale as previously determined at macroscale. Sequencing of the housekeeping gene gyrA and the QS comQ gene confirmed different evolutionary rates of these genes. Furthermore, an asymmetric communication response was detected inside the two pherotype clusters, suggesting continuous evolution of the QS system and possible development of new languages. To our knowledge, this is the first microscale study demonstrating the presence of different QS languages among isolates of one species, and the implications of this microscale diversity for microbial interactions are discussed.Quorum sensing (QS), a widespread phenomenon in the bacterial world, controls a wide range of cell density-dependent behaviors. Bacillus subtilis uses QS to control production of antimicrobial peptides, bacteriocins, and antibiotics (20) but also to alternate between two cell types during stationary phase: competent cells, able to take in DNA from the environment, and dormant spores, able to survive harsh environmental conditions (9, 12, 24). Development of genetic competence in B. subtilis is controlled by a QS system encoded by the comQXPA operon (2, 53, 54). This involves the ComX pheromone that accumulates during exponential growth (25, 46, 47) and is initially synthesized as a 55-residue protein that is processed, modified, and released into the extracellular medium as a 5- to 10-amino-acid peptide. The isoprenoidal modification on the tryptophan residue of this peptide is catalyzed by the ComQ protein (2, 25, 34, 35, 42, 52). Upon reaching the threshold concentration, processed and modified ComX binds to the membrane-associated, histidine protein kinase ComP and triggers the QS response, linking autophosphorylation of ComP and transfer of phosphate to the response regulator ComA (59). The level of phosphorylated ComA is also controlled by dephosphorylation, which is dependent on a separate QS system involving competence sporulation factor (CSF) and the RapC phosphatase (3, 59). Phosphorylated ComA directly controls expression of various genes (6, 33), including the srfAB operon that contains the comS gene (15, 41), required for development of competence (55).Previous studies of environmental B. subtilis strains indicate a high polymorphism (approximately 56% identity at the nucleotide level) in the QS locus, which is restricted to comQ, comX, and the N-terminal region of the comP gene. Sequences surrounding this locus, downstream gene comA, a C-terminal region of comP, and the upstream degQ gene, are highly conserved (2, 53, 54). Sequence analysis of the comQXP loci of 13 strains indicated clustering into four distinct similarity groups (2). These groups were congruent for comQ, comX, and the N-terminal region of comP, indicating coevolution of the three genes. In addition, the similarity groups correlated with four pherotypes, able to communicate efficiently within but not between groups. Similar variation has been reported for the agr QS system in staphylococci (19, 56) and in the competence QS system of Streptococcus pneumoniae (17, 19, 37, 38, 60).B. subtilis is often referred to as a soil-dwelling organism, its spores persisting in soil until encountering conditions suitable for germination and growth (10). The basic structural unit of soil ecosystems is the soil aggregate, in which biogeochemical processes occur at scales relevant to microorganisms. Approximately 50% of the volume of a soil aggregate represents open pores, while the remainder consists of mineral particles (sand, silt, and clay) held together by organic material (48), with which B. subtilis may be preferentially associated (16, 43). Soil aggregates can be classified as macroaggregates (diameter, >250 μm) and microaggregates (diameter, 2 to 250 μm) (39), but little is known about the distribution of bacteria within aggregates. Structural organization of the soil creates a mosaic of microenvironments, within which water movement and diffusion of nutrients and other molecules play key roles in functioning of the soil microbiota (7, 13, 39). These roles may vary with the scale at which they operate. Tisdall and Oades (51) suggest that scales at which microorganisms are important in the soil aggregation process range between 2 and 2,000 μm, depending on the specific system being investigated (13). Although the microscale distribution of microorganisms and their associated functions have rarely been studied, it is becoming recognized that greater knowledge of spatial organization at the scale of a soil aggregate (microscale) is essential for a better understanding of soil ecosystem function and of the mechanisms that generate and maintain diversity, including speciation, extinction, dispersal, and interactions within and between species (7, 13, 26).The aim of this study was to assess the potential role of QS in generating and maintaining microscale diversity within the soil. This was achieved by determining the genomic and functional diversification of the B. subtilis QS system with regard to geographical distance and ecological characteristics. Isolates were obtained from two 1-cm³ sandy, riverbank soil samples separated by approximately 5 m, allowing assessment of macroscale diversity. In addition, each riverbank soil sample was treated as a separate macroaggregate that was progressively sectioned to obtain subsamples of different sizes, allowing assessment of microscale diversity. The riverbank soil B. subtilis isolates were compared with Bacillus isolates previously obtained from desert soil samples separated by distances of meters to kilometers (2, 40), representing macroscale distribution. The Bacillus isolates were used to (i) correlate geographical distance (microscale/macroscale) with genomic distance of the QS comQ gene and the housekeeping gyrA gene, (ii) investigate and compare the specificity of the QS response of microscale and macroscale isolates, and (iii) explore dominance of pherotypes inside soil aggregates. To our knowledge, this is the first investigation of a QS system that addresses the genomic and functional diversification of bacterial populations at microscale.