Understanding microbial interactions is a fundamental objective in microbiology and ecology. The synthetic community system described here can set into motion a range of research to investigate how the diversity of a microbiome and interactions among its members impact its function, where function can be measured as exometabolites. The system allows for community exometabolite profiling to be coupled with genome mining, transcript analysis, and measurements of member productivity and population size. It can also facilitate discovery of natural products that are only produced within microbial consortia. Thus, this synthetic community system has utility to address fundamental questions about a diversity of possible microbial interactions that occur in both natural and engineered ecosystems.

Nongrowth states are common for bacteria that live in environments that are densely populated and predominantly nutrient exhausted, and yet these states remain largely uncharacterized in cellular metabolism and metabolite output. Here, we investigated and compared stationary-phase exometabolites and RNA transcripts for each of three environmental bacterial strains.

15Microbial exponential growth is expected to occur infrequently outside of the 16 laboratory, in the environment. Instead, resource-limited conditions impose non-growth 17 states for microbes. However, non-growth states are uncharacterized for the majority of 18 environmental bacteria, especially in regard to exometabolite production. To investigate 19 exometabolite production in response to nutrient limitation, we compared 20 exometabolites produced over time in stationary phase across three environmental 21 bacteria: Burkholderia thailandensis E264 (ATCC 700388), Chromobacterium 22 violaceum ATCC 31532, and Pseudomonas syringae pathovar tomato DC3000 (ATCC 23 BAA-871). We grew each strain in monoculture and investigated exometabolite 24 dynamics over time from mid-exponential to stationary phase. We focused on 25 phase is not at all "stationary" for these bacteria, and sets the stage for understanding 49 how individual metabolisms support interspecies interactions in resource-limited 50 environments. 51 52 Keywords 53 Burkholderia thailandensis, Chromobacterium violaceum, Pseudomonas syringae, 54 secondary metabolism, RNAseq, mass spectrometry, metabolomics, persistence, non-55 growth state 56 57 Introduction 58Much of microbiology research in the laboratory is conducted with bacterial or 59 archaeal cells that are growing exponentially. However, it is estimated that 60% of 60 microbial biomass in the environment is in a non-growing state (1, 2). Various abiotic 61 and biotic stressors are known to induce non-growth states, but perhaps the most 62 common is resource limitation. Labile resources can be low in an environment, as 63 characteristic of the oligotrophic open ocean, or they can be available but inaccessible, 64 as typical in heterogeneous soil matrices. Thus, unlike most cultivated laboratory 65 strains, environmental microbes experience short periods of accessible resources 66 punctuated by long periods of famine (3, 4). 67 While Gram-positive bacteria can survive resource limitation through sporulation 68 (5), Gram-negative bacteria can persist in stationary phase without entry into a 69 specialized dormant cell structure (6). Instead, Gram-negative bacteria survive in 70 stationary phase by employing various stress response adaptations (7). Stress 71 4 response adaptations include changes to cell morphology, transcription, translation, and 72metabolism. Furthermore, in stationary phase, microbes can re-route metabolic 73 pathways to maintain essential components of the cell and the proton motive force (8). 74While these adaptations are thought to serve as survival mechanisms, the levels and 75 types of metabolic activities in stationary phase are not well understood for most 76 environmental microbes. 77It is known, however, that microbes can exhibit appreciable metabolic activity in 78 stationary phase (9). For example, entry into stationary phase resulted in prolonged 79 protein production in Escherichia coli despite that overall protein levels decreased (10). 80Metabolomic studies of E. coli in...

During prolonged resource limitation, bacterial cells can persist in metabolically active states of non-growth. These maintenance periods, such as those experienced by cells in stationary phase cultures, can, perhaps counterintuitively, include upregulation of cellular secondary metabolism and release of exometabolites into the local environment, at the cost of an energetic commitment to growth. As resource limitation is a characteristic feature of many habitats that harbor environmental microbial communities, we hypothesized that neighboring bacterial populations employ exometabolites to compete or cooperate during maintenance, and that these exometabolite-facilitated interactions can drive community outcomes. Here, we evaluated the consequences of exometabolite interactions over stationary phase among three well-known environmental bacterial strains: Burkholderia thailandensis E264 (ATCC 700388), Chromobacterium violaceum ATCC 31532, and Pseudomonas syringae pv.tomato DC3000 (ATCC BAA-871). We assembled these stains into laboratory-scale synthetic communities that only permitted chemical interactions among them. We compared the responses (transcripts) and behaviors (exometabolites) of each member with and without neighbors. We found that transcriptional dynamics were altered in the presence of different neighbors, and that these changes could be attributed to the production of or response to bioactive exometabolites employed for competition during maintenance. B. thailandensis was especially influential and competitive within its communities, as it consistently upregulated additional biosynthetic gene clusters involved in the production of bioactive exometabolites for both exploitative and interference competition. Additionally, some of these bioactive exometabolites were upregulated and produced in a non-additive manner in the 3-member community. These results demonstrate that the active investment in competition during maintenance can contribute to both bacterial population fitness and community-level outcomes. It also suggests that the traditional concept of defining competitiveness by growth outcomes may be too narrow, and that maintenance competition could be an alternative measure

One interference mechanism of bacterial competition is the production of antibiotics. Bacteria exposed to antibiotics can resist antibiotic inhibition through intrinsic or acquired mechanisms. Here, we performed a coevolution experiment to understand the long-term consequences of antibiotic production and antibiotic susceptibility for two environmental bacterial strains. We grew five independent lines of the antibiotic-producing environmental strain, Burkholderia thailandensis E264, and the antibiotic-inhibited environmental strain, Flavobacterium johnsoniae UW101, together and separately on agar plates for 7.5 months (1.5 month incubations), transferring each line five times to new agar plates. We observed that the F. johnsoniae ancestor could tolerate the B. thailandensis -produced antibiotic through efflux mechanisms, but that the coevolved lines had reduced susceptibility. We then sequenced genomes from the coevolved and monoculture F. johnsoniae lines, and uncovered mutational ramifications for the long-term antibiotic exposure. The coevolved genomes from F. johnsoniae revealed four potential mutational signatures of reduced antibiotic susceptibility that were not observed in the evolved monoculture lines. Two mutations were found in tolC: one corresponding to a 33 bp deletion and the other corresponding to a nonsynonymous mutation. A third mutation was observed as a 1 bp insertion coding for a RagB/SusD nutrient uptake protein. The last mutation was a G83R nonsynonymous mutation in acetyl-coA carboxylayse carboxyltransferase subunit alpha (AccA). Deleting the 33 bp from tolC in the F. johnsoniae ancestor reduced antibiotic susceptibility, but not to the degree observed in coevolved lines. Furthermore, the accA mutation matched a previously described mutation conferring resistance to B. thailandensis -produced thailandamide. Analysis of B. thailandensis transposon mutants for thailandamide production revealed that thailandamide was bioactive against F. johnsoniae, but also suggested that additional B. thailandensis -produced antibiotics were involved in the inhibition of F. johnsoniae . This study reveals how multi-generational interspecies interactions, mediated through chemical exchange, can result in novel interaction-specific mutations, some of which may contribute to reductions in antibiotic susceptibility.