Microbes display broad diversity in cobamide preferences

Abstract

Cobamides, the vitamin B₁₂ (cobalamin) family of cofactors, are used by most organisms but produced by only a fraction of prokaryotes, and are thus considered key shared nutrients among microbes. Cobamides are structurally diverse, with multiple different cobamides found in most microbial communities. The ability to use different cobamides has been tested for several bacteria and microalgae, and nearly all show preferences for certain cobamides. This approach is limited by the commercial unavailability of cobamides other than cobalamin. Here, we have extracted and purified seven commercially unavailable cobamides to characterize bacterial cobamide preferences based on growth in specific cobamide-dependent conditions. The tested bacteria include engineered strains of Escherichia coli, Sinorhizobium meliloti, and Bacillus subtilis expressing native or heterologous cobamide-dependent enzymes, cultured under conditions that functionally isolate specific cobamide-dependent processes such as methionine synthesis. Comparison of these results to those of previous studies of diverse bacteria and microalgae revealed that a broad diversity of cobamide preferences exists not only across different organisms but also between different cobamide-dependent metabolic pathways within the same organism. The microbes differed in the cobamides that support growth most efficiently, cobamides that do not support growth, and the minimum cobamide concentrations required for growth. The latter differ by up to four orders of magnitude across organisms from different environments and by up to 20-fold between cobamide-dependent enzymes within the same organism. Given that cobamides are shared, required for use of specific growth substrates, and essential for central metabolism in certain organisms, cobamide preferences likely impact community structure and function.IMPORTANCENearly all bacteria are found in microbial communities with tens to thousands of other species. Molecular interactions such as metabolic cooperation and competition are key factors underlying community assembly and structure. Cobamides, the vitamin B₁₂ family of enzyme cofactors, are one such class of nutrients, produced by only a minority of prokaryotes but required by most microbes. A unique aspect of cobamides is their broad diversity, with nearly 20 structural forms identified in nature. Importantly, this structural diversity impacts growth as most bacteria that have been tested show preferences for specific cobamide forms. We measured cobamide-dependent growth in several model bacteria and compared the results to those of previous analyses of cobamide preference. We found that cobamide preferences vary widely across bacteria, showing the importance of characterizing these aspects of cobamide biology to understand the impact of cobamides on microbial communities.

Keywords: cobalamin; cobamide; cobamide preference; corrinoid; vitamin B12.

Fig 1

Cobamide structures and cobamide-dependent growth. (A) Structure of B₁₂ (cobalamin), which contains the lower ligand 5,6-dimethylbenzimidazole, is shown in the base-on conformation in which the ring nitrogen is coordinated to the cobalt ion (dashed line) (left). Alternative lower ligands of cobamides in this study (right) are shown, with the names of the cobamides given below each structure. (B–J) Cobamide dose-dependent growth assays showing OD₆₀₀ measured at the indicated times for (B) MetH-dependent growth of S. meliloti, (C) NrdJ-dependent growth of S. meliloti, (D) EAL-dependent growth of E. coli, (E) MetH-dependent growth of E. coli, (F) MetH-dependent growth of E. coli expressing E. coli metH on a plasmid, (G) MetH-dependent growth of E. coli expressing V. cholerae metH on a plasmid, (H) MetH-dependent growth of B. subtilis expressing P. megaterium metH, (I) MetH-dependent growth of B. thetaiotaomicron, and (J) MetH-dependent growth of R. gnavus. The EC₅₀ values calculated from the curves in panels B–J and genotypes of the engineered strains in B–H are shown in Table S1. OD₆₀₀ values for [Bza]Cba-supplemented cultures of B. subtilis were recorded after 72 hours because growth was not observed until after 45 hours. Points represent the means of 3–6 biological replicates; error bars represent standard deviation.

Fig 2

Comparison of EC₅₀ values for cobamide-dependent growth. Organisms examined in the current study (bold) are compared with those from previous studies (8, 13, 19–21). S. meliloti Rm1021 ΔnrdJ cobD::gus Gm^R metH::Tn5 pMSO3-nrdAB(E. coli) was used for S. meliloti MCM-dependent growth (8). EC₅₀ values for S. meliloti are higher than for the other tested microbes likely because wild-type S. meliloti synthesizes Cbl de novo and lacks a high-affinity cobamide uptake system. MetH-dependent growth of C. reinhardtii was tested in a metE mutant (13). Symbols show the mean EC₅₀ values. Capped bars represent 95% confidence intervals, except with organisms labeled with *, which indicates error as standard deviation. Errors of C. reinhardtii and P. lutheri EC₅₀ values were not reported (13). Bars are uncapped on the left or right when lower or upper bounds for 95% confidence intervals could not be determined, respectively. The lower bound for Blastococcus sp. B12_004 grown with [5-OHBza]Cba is 10⁻⁷ nM (21). The base of the leftward and rightward arrows represents maximal and minimal concentrations for EC₅₀ from dose–response assays in which lack of growth or saturating growth was not reached, respectively. For O. tauri and A. carterae, EC₅₀ values could not be calculated, but the capped bars for [Ade]Cba show the upper and lower bounds (13). Symbols in the shaded region on the right represent cobamides that were unable to support growth at any concentration tested.