Corrinoid salvaging and cobamide remodeling in bacteria and archaea

ABSTRACT

Cobamides (Cbas) are cobalt-containing cyclic tetrapyrroles used by cells from all domains of life as co-catalyst of diverse reactions. There are several structural features that distinguish Cbas from one another. The most relevant of those features discussed in this review is the lower ligand, which is the nucleobase of a ribotide located in the lower face of the cyclic tetrapyrrole ring. The above-mentioned ribotide is known as the nucleotide loop, which is attached to the ring by a short linker. In Cbas, the nucleobase of the ribotide can be benzimidazole or derivatives of it, purine or derivatives of it, or phenolic compounds. Given the importance of Cbas in prokaryotic metabolism, it is not surprising that prokaryotes have evolved enzymes that cleave part or the entire nucleotide loop. This function is advantageous when Cbas contain nucleobases that somehow interfere with the function of Cba-dependent enzymes in the organism. After cleavage, Cbas are rebuilt via the nucleotide loop assembly (NLA) pathway, which includes enzymes that activate the nucleobase and the ring intermediate, followed by condensation of activated intermediates and a final dephosphorylation reaction. This exchange of nucleobases is known as Cba remodeling. The NLA pathway is used to salvage Cba precursors from the environment.

INTRODUCTION

The chemical structure of cobamides (Cbas) consists of a cyclic tetrapyrrole containing a central cobalt ion equatorially coordinated to the nitrogen atom of the pyrrolic moieties (Fig. 1). Any compound that features the ring structure of a Cba is referred to as a corrinoid (1). In a biologically active Cba, the cobalt has upper face (Coβ) and lower face (Coα) axial ligands. In AdoCbl, the upper ligand is a 5′-deoxyadenosine (Ado) group derived from ATP and the lower ligand is 5,6-dimethylbenzimidazole (DMB) (2). Alternative bases may be substituted at the Coα position, and base preference varies among microbial organisms and Cba-requiring enzymes (3–6). Here, we review the approaches used by prokaryotes to salvage incomplete corrinoids and nucleobase precursors and remodel Cbas. Because the nomenclature in the B₁₂ field is confusing, in Table 1, we provide the names, abbreviations, and definitions of compounds frequently used in this review. Figure 1 should help the reader visualize the compounds mentioned in Table 1.

Fig 1.

Chemical structure of AdoCba and cleavage sites of remodeling enzymes. “Base” denotes the diversity of nucleobases found in Cbas (3). VcCobS is the AdoCbl 5′-phosphate synthase from Vibrio cholerae that has been proposed to remove the α-adenosine moiety of the nucleotide loop of psCbl (light blue arrow), replacing it with α-ribazole-5′-phosphate containing DMB, yielding 5′ deoxadenosylcobalamin (AdoCbl) (7, 8). Other colored arrows show the cleavage sites for the amidohydrolase activities of CbiZ and CbiS (orange arrow). CbiZ has been found in bacteria and archaea, but CbiS has not been reported in bacteria (9–11). The CbiR protein has phosphoribosyl hydrolase activity and is broadly distributed among bacteria (magenta arrow) (12).

TABLE 1.

Name, abbreviations, and definitions of compounds relevant to B₁₂ biosynthesis

Corrinoid	A general term that refers to cyclic tetrapyrroles in which two of the imidazole rings are directly linked together. This structure is known as the corrin ring. In corrinoids, a cobalt ion is held through interactions with each of the pyrrolic nitrogen atoms.
Cobamide	A corrinoid with a nucleotide attached to a substituent of the corrin ring. The nucleotide is located on the alpha face [(Co(α)] of the corrin ring, features an alpha N-glycosidic bond and its nucleobase may or may not form a coordination bond with the Co ion of the corrin ring. Cbas are also referred to as complete corrinoids.
Incomplete corrinoid	A corrinoid that lacks parts of or the entire α-nucleotide.
Cobalamin (Cbl)	Cba whose nucleobase is 5,6-dimethylbenzimidazole.
Cyanocobalamin (CNCbl)	Cbl with a cyano group as its upper Co(β) ligand; CNCbl is a synthetic derivative Cbl; cells do not synthesize CNCbl.
Vitamin B12	A synonym of Cbl.
Adenosylcobalamin (AdoCbl)	Cbl with a 5-deoxyadenosine (Ado) upper Co(β) ligand.
Adenosylcobalamin-5′ phosphate	AdoCbl containing a phosphoryl group on the 5′ carbon of the ribosyl moiety of the nucleotide.
Coenzyme B12	A synonym of AdoCbl.
Adenosylcobamide	A Cba with a 5-deoxyadenosine (Ado) upper Co(β) ligand, and any nucleobase other than DMB or adenine.
Adenosylpseudocobalamin	A Cba with a 5-deoxyadenosine (Ado) upper Co(β) ligand, and adenine as its nucleobase.
Adenosylcobyric acid	An incomplete corrinoid that lacks the entire nucleotide and the 3-aminopropanol moiety to which it is attached to the corrin ring.
Adenosylcobinamide (AdoCbi)	An incomplete corrinoid that lacks the nucleotide but retains the 3-aminopropanol moiety that links the ring to the loop.
Adenosylcobinamide-phosphate (AdoCbi-P)	AdoCbi with a phosphoryl group attached to the hydroxyl group of the 3-aminopropanol moiety of the structure.
Adenosylcobinamide-GDP (AdoCbi-GDP)	AdoCbi-P to which a GMP moiety has been attached via a phosphodiester bond.
3-Aminopropanol (AP)	The moiety that tethers the nucleotide to the corrin ring.
3-Aminopropanol-phosphate (AP-P)	The phosphoryl group of AP-P is covalently linked to the 3′ OH group of the ribosyl moiety of the α-nucleotide forming a phosphodiester bond.
5,6-Dimethylbenzimidazole	The nucleobase of cobalamin.
α-Ribazole	The α-nucleoside of DMB.
α-Ribazole-5′-phosphate	The α-nucleotide of DMB.

WHICH ORGANISMS SYNTHESIZE COBAMIDES?

Because of their structural complexity, de novo assembly of Cbas requires about 30 enzymes (13), although this number may vary among organisms. Briefly, the corrin ring is assembled first, then the cobalt ion is inserted into the ring followed by the attachment of a 5′-deoxyadenosyl group to the cobalt ion on the upper or beta face of the ring (14–17). The late steps of Cba biosynthesis involve the activation and attachment of the base on the lower face of the ring structure, the branch known as the nucleotide loop assembly (NLA) pathway (18, 19). Genetic evidence suggests that corrinoid intermediates of the NLA pathway must be adenosylated prior to entering the NLA pathway (14).

While most bacteria require Cbas, many cannot synthesize them de novo. Shelton et al. (20) found genes encoding Cba-dependent enzymes in the genomes of 86% of bacteria; however, only 37% of these were predicted to have a functional de novo synthesis pathway. Different Cbas are present in the human gut, but Cba-dependent enzymes may require a specific Cba to be functional, while others are less specific and can use diverse Cbas as co-catalysts (4). Thus, Cba availability and exchange may regulate interactions within microbial communities (21, 22). Despite this widespread requirement, bioinformatics analyses suggest that only some prokaryotes synthesize Cbas de novo, while others rely on salvaging incomplete precursors or on remodeling complete Cbas (23). Here, we review the approaches used by prokaryotes to salvage incomplete corrinoid and nucleobase precursors and remodel Cbas. We also address the physiological importance of these processes for cell growth and survival.

ROLE OF COBAMIDES IN NATURE

AdoCbl and other Cbas are used by diverse enzymes in eukaryotes and prokaryotes to perform metabolic reactions, including methyl transfers, carbon skeleton rearrangements, and elimination reactions (13). These compounds are also used by some bacteria as photoreceptors involved in carotenogenesis or DNA photolyase (24–26).

Cbas-dependent enzymes catalyze diverse reactions in prokaryotes. For example, methanogenic archaea utilize Cba-dependent methyltransferases during methanogenesis (27–31), while some bacteria use Cbas for the catabolism of ethanolamine, glycerol, and 1,2-propanediol (32–38). Other Cba-dependent enzymes are involved in tRNA modifications, organohalide respiration, and one-carbon metabolism (39–43). Humans, other animals, and some prokaryotes utilize Cba-dependent methylmalonyl-CoA mutases to interconvert propionate to succinate (44). In humans, defects in or absence of this enzyme lead to severe neurological effects that are often fatal (45, 46).

CORRINOID TRANSPORT

Due to the complexity, size, and positive charge of the corrin ring at neutral pH, the import of complete and incomplete corrinoids into the cell is an active process requiring energy input. Gram-negative organisms transport incomplete and complete corrinoids across the outer membrane using the calcium-dependent transporter BtuB, whose activity depends on energy made available through interactions with the cell membrane protein TonB (47–50). Transport through the inner membrane occurs through the ABC transporter BtuFCD with BtuF binding corrinoids in the periplasm and delivering them to the BtuCD complex in the inner membrane (51–54). Notably, Bacteroides thetaiotamicron utilizes three homologous BtuF proteins with different corrinoid specificity. Although seemingly redundant, multiple btuF alleles impart a colonization advantage to B. thetaiotamicron in a mouse model (55). B. thetaiotamicron also utilizes the surface lipoprotein BtuG2 to bind Cbas with high affinity and interact with BtuB to facilitate transport into the periplasm (56). A recent report of a novel class of high-affinity, vitamin B₁₂-binding proteins dedicated to the acquisition of Cbas in gut commensal Bacteroidetes illustrates the extent to which organisms that occupy complex environments go to acquire enough of this important cofactor (57). Orthologs of the BtuCDF proteins have been identified in the archaeon Halobacterium sp. NRC1, but detailed knowledge of their functions is limited (58).

SALVAGING OF INCOMPLETE CORRINOIDS AND COBAMIDE REMODELING

Here, we use the term “incomplete corrinoids” to refer to corrinoids that lack the nucleotide loop (e.g., Cbi and Cby). Some prokaryotes can synthesize Cbas de novo but can also salvage Cbas (i.e., complete corrinoids) and incomplete corrinoids.

Cba-requiring organisms that cannot synthesize these molecules de novo rely on salvaging incomplete precursors and have the enzymes to remodel Cbas so they contain a specific nucleobase. To provide a framework for the discussion of corrinoid salvaging and Cba remodeling used by prokaryotes, we describe below the late steps of Cba biosynthesis.

Organisms that lack most or all of the corrin ring biosynthetic enzymes require a complete corrin ring that can be salvaged via the NLA pathway (Fig. 2, black ovals) (20, 59).

Fig 2.

The NLA pathway used by prokaryotes for the conversion of incomplete corrinoids to Cbas and for Cba remodeling. The steps of the NLA pathway of an aerobic bacterium are catalyzed by eight enzymes (i.e., CobA, PduX, CobD, CbiB, CobU, BluB, CobT, CobS, and CobC), which are shown in black ovals. CobB is not considered a Cba biosynthetic enzyme, but it is listed because it compensates for the absence of CobT (60). Archaeal NLA enzymes (CobY and CobZ) are shown in yellow ovals. Bacterial remodeling enzymes CbiR (12) and VcCobS (7) are shown in magenta and light blue ovals, respectively; a detailed discussion of CbiR and VcCobS is presented below. A remodeling amidohydrolase (CbiZ) discovered in archaea and bacteria (9, 61) is shown in orange ovals. CbiS is the only enzyme discovered in hyperthermophilic archaea reported thus far that participates in NLA biosynthesis and remodeling (gray ovals). Finally, some Firmicutes and Bacillota synthesize an enzyme (CblS) that can activate exogenous α-R into α-RP (light purple oval). The product of the pathway (AdoCbl) is shown within a blue square. HOCbi, hydroxycobinamide; L-Thr-P, L-threonine-phosphate; AP, 3-aminopropanol; AdoCby, adenosylcobyric acid; AdoCbi-P, adenosylcobinamide phosphate; AdoCbi-GDP, adenosylcobinamide-GDP; AdoCbl-P, adenosylcobalamin 5′-phosphate; α-R, alpha-ribazole; α-RP, alpha-ribazole phosphate; α-AMP-3-O-AP, alpha-adenosine monophosphate-3′-O-aminopropanol; FMNH₂, dihydroflavin mononucleotide; AdopsCbl, adenosylpseudocobalamin; DMB, 5,6-dimethylbenzimidazole; and AdoCbl, adenosylcobalamin.

The NLA pathway of Salmonella Typhimurium is used for incomplete corrinoid salvaging (14, 62). Shown in Fig. 2 (black ovals) are eight enzymes (CobA, PduX, CobD, CbiB, CobU, CobT, CobS, and CobC) used by the facultative anaerobe S. Typhimurium to generate adenosylcobalamin (AdoCbl) from adenosylcobyric acid (AdoCby). At present, it is not known how S. Typhimurium synthesizes DMB. S. Typhimurium lacks a BluB homolog for the synthesis of DMB. The BluB enzyme shown in Fig. 2 was first identified in Sinorhizobium meliloti and Rhodospirillum rubrum (63, 64).

Below is the list of enzymes that catalyze the late steps (a.k.a., NLA pathway) of Cba biosynthesis. We list the biochemical activity, gene product name, and EC numbers for each of them:

1. ATP:L-threonine O-3-phosphotransferase (PduX, EC 2.7.1.177) (65).

2. L-threonine-O-3-phosphate decarboxylase (CobD, EC 4.1.1.81) (66–68).

3. Adenosylcobinamide (AdoCbi) synthase (CbiB, EC 6.3.1.10) (69).

4. AdoCbi:NTP kinase/AdoCbi:GTP guanylyltransferase (CobU, EC 2.7.1.156, EC 2.7.7.62, respectively) (70–72).

5. DMB synthase (BluB, EC 1.13.11.79) (63, 64).

6. Nicotinate mononucleotide (NaMN): DMB phosphoribosyl transferase (CobT, EC 2.4.2.21) (73). Notably, CobT is a very non-specific enzyme and is responsible for the structural diversity of cobamides (3, 32, 73–77).

7. Adenosylcobamide 5′-phosphate (AdoCba-5′-P) synthase (CobS, EC 2.7.8.26) (19, 32, 78).

8. The final step of the pathway is catalyzed by the AdoCbl-5′-P phosphatase (CobC, EC 3.1.3.73). It is interesting that CbiB and CobS are integral inner membrane proteins that may serve as anchors for interactions with other enzymes (65, 79, 80) to facilitate the flux of intermediates (69, 81).

BIOSYNTHESIS OF DMB IN THE ABSENCE OF OXYGEN AND ITS ACTIVATION TO α-RIBAZOLE-5′-PHOSPHATE

As stated above, the nucleotide loop of AdoCbl contains DMB as its nucleobase. Multiple DMB biosynthesis pathways have been identified. In organisms capable of synthesizing AdoCbl in the presence of oxygen, the C1 carbon of the ribityl moiety of FMNH₂ is converted into the C2 carbon of DMB by the enzyme BluB (5,6-dimethylbenzimidazole synthase, EC 1.13.11.79) (Fig. 2) (63, 64, 82–85). However, under anoxic conditions, a multiple-step pathway that assembles DMB from a purine biosynthesis intermediate and S-adenosylmethionine (SAM) was described in Eubacterium limosum (86, 87) (Fig. 3). Elegant work in Moorella thermoacetica showed that the product of the pathway is not DMB but its α-riboside known as α-ribazole (α-R) (87). In this pathway, DMB biosynthesis starts with the conversion of 5-aminoimidazole ribotide into 5-hydroxybenzimidazole (5-OHBza), a reaction catalyzed by SAM:5-OHBza synthase BzaAB/BzaF (EC 4.1.99.23) (88). Three additional methyltransferases (i.e., BzaC, BzaD, and BzaE) complete the synthesis of DMB. Notably, the BzaC methyltransferase requires the function of the CobT phosphoribosyl transferase (EC 2.4.2.21) to activate 5-OHBza into its α-ribotide, which is presumably dephosphorylated by the CobC phosphatase (EC 3.1.3.73) before BzaC can methylate the hydroxyl group of the benzyne ring, yielding 5-OHMeBza-R (87).

Fig 3.

Pathway for the synthesis of α-R under anoxic conditions. The pathway for the synthesis of DMB was first discovered in Eubacterium limosum (86) and subsequently studied in Moorella thermoacetica (87). In this scheme, the steps catalyzed by BzaD and BzaE are speculative and remain under investigation (89).

The activation of DMB can happen by two mechanisms. As shown in Fig. 2, the CobT enzyme transfers the phosphoribosyl group of nicotinate mononucleotide onto DMB with inversion of configuration yielding what we know as “activated DMB” or α-ribazole-5′-phosphate (α-RP). It has been reported that the CobB sirtuin deacetylase has CobT-like activity and compensates for the absence of CobT in vivo and in vitro (60, 74). However, the mechanism used by CobB to activate DMB to α-RP has not been reported. Finally, it is known that the S. Typhimurium CobT enzyme can use NAD⁺ in lieu of NaMN to generate α-5,6-dimethylbenzimidazole adenine dinucleotide (α-DAD), which could be cleaved to form α-RP (90) (Fig. 4). To date, the enzymatic conversion of α-DAD into AMP and α-RP has not been reported.

Fig 4.

An alternative route to α-RP. When CobT uses NAD⁺ plus DMB as substrates, the product of the reaction is alpha-5,6-dimethylbenzimidazole adenine dinucleotide (90), which could be cleaved by an as-yet-unidentified nudix enzyme into α-RP and AMP.

SALVAGING α-NUCLEOTIDE PRECURSORS

Some Firmicutes and Bacillota (e.g., Listeria and Geobacter, respectively) can salvage the riboside α-ribazole from their environments using two proteins, namely, CblT and CblS (91). CblT is an ECF-type ABC transporter through which α-R is translocated into the cytoplasm, and CblS is an α-R kinase that generates the corresponding α-nucleotide α-RP at the expense of ATP (Fig. 2). A periplasmic protein that binds α-R and delivers it to CblT has not been identified.

COBINAMIDE SALVAGING AND REMODELING IN ARCHAEA

Archaeal enzymes involved in cobinamide (Cbi) salvaging are highlighted by yellow ovals in Fig. 2. CbiZ homologs that have also been found in bacteria are identified in Fig. 2 as orange ovals. Unlike the S. Typhimurium CobU enzyme, archaeal genomes encode a GTP:AdoCbi-phosphate guanylyl transferase enzyme known as CobY (EC 2.7.7.62), which is a non-orthologous replacement for CobU that has guanylyl transferase activity but lacks the kinase activity of CobU (71, 92, 93). Because CobY lacks kinase activity, Cbi cannot be directly converted into AdoCbi-P. Instead, archaea salvage Cbi through the hydrolysis of the 3-aminopropanol group that tethers the corrin ring to α-RP. This cleavage is catalyzed by the CbiZ enzyme (EC 3.5.1.90) to form Cby (Fig. 2). AdoCby and AP-P (generated by CobD) are then condensed by CbiB to yield AdoCbi-P, which is the substrate for CobY (9, 10, 69, 92, 94, 95). Following the guanylylation of AdoCbi-P by CobY, CobS condenses AdoCbi-GDP with α-RP to yield AdoCbl-5′-P, which is dephosphorylated by CobZ (EC 3.1.3.73), the archaeal CobC homolog (96) to yield AdoCbl, the final product of the pathway. Thus, archaea require the amidohydrolase activity of CbiZ and the guanylyl transferase activity of CobY to salvage the incomplete precursor Cbi and to remodel complete Cbas (9, 10).

Some archaea synthesize a protein known as CbiS in which the amidohydrolase CbiZ is fused to CobZ, the archaeal AdoCbl-5′-P phosphatase that catalyzes the final enzyme in Cba biosynthesis. We note that CbiS was first identified in the hyperthermophile Methanopyrus kandleri (11) and that the CbiZ amidohydrolase portion of CbiS shows remarkably little Cba specificity and can cleave nucleotide loops containing various nucleobases, presumably providing flexibility in the assembly of usable Cbas for M. kandleri (11).

The physical linkage in CbiS of the first (CbiZ) and last (CobZ) enzymes involved in Cba remodeling in archaea is intriguing and supports the hypothesis that Cba synthesis and salvaging enzymes co-localize to the cell membrane to facilitate the efficient flow of intermediates (69, 79–81). Since a CbiS-type enzyme has only been identified in hyperextremophiles, it is possible that the fusion of these Cbi and CobZ may contribute to thermal stability or protection from other environmental stressors (11).

APPROACHES USED BY BACTERIA TO REMODEL COBAMIDES

To date, three approaches to Cba remodeling have been described and are discussed below.

Removal of the nucleotide loop

This approach is identified in Fig. 2 by broken orange arrows. Although this approach to Cba remodeling was initially associated with the assimilation of the precursor Cbi in archaea (9, 11), studies of homologs of the cbiZ gene encoded by the genome of in bacteria led to the conclusion that the CbiZ amidohydrolase was also involved in Cba remodeling (97). Furthermore, Cba remodeling is likely critical in complex communities like the human gut microbiome, where Cbas other than Cbl have been detected at much higher abundance than Cbl (4).

CbiZ cleaves the amide bond that attaches AP to the propionamide substituent of pyrolle D of the ring, releasing the nucleotide loop and yielding AdoCby. The latter can then be converted to a functional AdoCba by NLA pathway enzymes (Fig. 2). Initial phylogenetic analyses suggested that less than 10% of bacterial genome sequences available from databases contain a cbiZ gene; however, cbiZ genes are widely distributed and do not seem to diverge from a single common ancestor (97).

Rhodobacter sphaeroides uses this approach to Cba remodeling. R. sphaeroides is a metabolically versatile α-proteobacterium that can obtain carbon and conserve energy via fermentation, photosynthesis, and respiration (oxically or anoxically) (98). This bacterium utilizes Cba-dependent enzymes for methionine synthesis (99) and for the assimilation of acetate (100, 101). However, pseudocobalamin (psCbl), a Cba that contains adenine as the nucleobase of the nucleotide loop, cannot be used by R. sphaeroides; hence, this bacterium uses CbiZ to remodel psCbl into Cbl by converting the resulting Cby precursor to AdoCbl using the NLA pathway (Fig. 2, orange broken arrows) (61).

Interestingly, the genome of R. sphaeroides also encodes a homolog of the bifunctional kinase, guanylyltransferase (CobU) enzyme whose kinase activity can salvage exogenous Cbi, which is also a substrate for CbiZ (97). The presence of CbiZ and CobU likely provides flexibility to R. sphaeroides for the assimilation of Cbas using the CbiZ function or for the assimilation of Cbi using either CbiZ or CobU (see Fig. 2).

Dehalococcoides mccarthyi is a bacterium of particular interest for its role in the detoxification of groundwater contaminants, including chlorinated ethenes via organohalide respiration (102, 103). Reductive dechlorination reactions are catalyzed by Cba-dependent reductive dehalogenases (39). The D. mccarthyi genome lacks genes encoding corrin ring biosynthetic enzymes, so this organism is restricted to salvaging precursors, such as Cbi, or remodeling alternative Cbas. Interestingly, the genome of D. mccarthyi encodes multiple homologs of cbiZ (97, 104), and only Cbl, 5-MeBza, or [5-MeOBza]-Cba support the activity of the reductive dehalogenases of this bacterium (104, 105). However, purinyl- and phenolyl-Cbas can be remodeled to benzimidazolyl-Cbas if the preferred benzimidazole nucleobase is provided (104, 105). Notably, cbiZ alleles of D. mccarthyi share more sequence similarity with archaeal homologs in Methanosarcina mazei than to cbiZ alleles found in bacteria such as R. sphaeroides (97).

At present, it is unclear whether the putative CbiZ proteins encoded by genomes of D. mccarthyi strains have amidohydrolase activity, and if they do, what the physiologic reasons for multiple copies of cbiZ and diversity may be. Future work in D. mccarthyi physiology is needed to reveal the function of CbiZ homologs in this bacterium.

Cleavage of the phosphodiester bond of cobamides

The mucin-degrading organism Akkermansia muciniphila is suggested to be a beneficial gut microbe that may strengthen the host gut barrier by stimulating the production of mucus (106–108). Cbas and Cba precursors are a valuable resource in the gut, as most gut bacteria utilize Cba-dependent enzymes (20). A novel phosphodiesterase CbiR was identified in A. muciniphila capable of remodeling pseudoCbl (psCbl) into Cbl (Fig. 2, blue broken arrows). CbiR cleaves AdopsCbl yielding AdoCbi-P, the substrate for CobU, which converts AdoCbiP into AdoCbi-GDP, and after three additional reactions, this bacterium synthesizes AdoCbl. CbiR homologs were identified in the genomes of organisms across at least 22 phyla, suggesting a widespread alternative mechanism for Cba salvaging (12).

Direct Cba remodeling

Recently, a putative new activity of the Cba-5′-P synthase CobS was reported in Vibrio cholerae (7). Heterologous expression of V. cholerae CobS in Escherichia coli resulted in the conversion of psCbl to Cbl, suggesting that in V. cholerae, CobS (VcCobS) has phosphodiesterase activity similar to that of CbiR. AdoCb-P resulting from the reaction catalyzed by VcCobS is proposed to be used by the enzyme as a co-substrate with α-RP yielding Cbl (Fig. 2, blue broken arrows) (7, 8). Notably, unlike in A. muciniphila, the conversion of AdoCbi-P to Cbl by VcCobS is thought to occur in the absence of AdoCbi-GDP (8). We note that the experiments that led to the proposed “direct remodeling” function of VcCobS were performed with crude extracts and not with highly purified VcCobS protein, leaving open the possibility that the remodeling activity putatively associated with VcCobS may be due to an as-yet-unidentified protein. Should VcCobS indeed have the proposed remodeling function, it would be the first report of such an activity in any CobS enzyme, expanding the number of strategies of Cba remodeling among prokaryotes and our understanding of CobS function.

Concluding remarks and open questions

To date, we have gained valuable insights into the different approaches used by prokaryotes to procure themselves with Cbas needed to perform specific metabolic reactions. The current literature shows that prokaryotes have evolved every possible strategy to remodel unusable Cbas into Cbas that contain nucleobases that allow cells to use them as co-catalysts. Cells have also evolved different strategies to assimilate the valuable incomplete corrinoids such as Cbi. Interesting questions remain, however. As mentioned above, the genome of D. mccarthyi encodes several cbiZ homologs, and to date, no evidence has been reported supporting the idea that these putative proteins have amidohydrolase activity. If all the cbiZ homologs do have amidohydrolase activity, the enzymes must be analyzed in detail to learn more about their specificities for corrinoids so we can broaden our understanding of the physiologic conditions that demand their function. Such efforts would lead researchers to address questions regarding the regulation of the many cbiZ genes in this bacterium, accelerating the identification of signal(s) that control cbiZ gene expression and determining whether the signal(s) are endogenous or exogenous. Open questions raised by the finding of the putative activity assigned to VcCobS AdoCbl-5′-P synthase will likely be addressed in the future. Unanswered questions regarding the proposed bifunctionality of VcCobS (i.e., phosphodiesterase/synthase) should address what drives the formation of the phosphodiester bond between AdoCbi-P and α-RP, given that the phosphodiester bond is formed between the 3′ hydroxyl group of α-RP and AdoCbi-P. If VcCobS has such bifunctionality, the enzyme must be analyzed from a mutational/functional point of view to determine whether the enzyme has distinct catalytic sites or whether a single site can catalyze phosphodiesterase and synthase reactions, which proceed via distinct mechanisms. Mechanistic studies that investigate the source of energy driving such an exchange would provide valuable insights into the functions of VcCobS.

The role of Cba remodeling in the dynamics of complex microbial communities is also of great interest. Questions regarding the fate of excised nucleobases must be addressed since it has been shown that some nucleobases inhibit the growth of some prokaryotes (109, 110). Do nucleobases play a role in population composition? In conclusion, the field of Cba remodeling is rich in mechanistic and physiologic questions of relevance to community dynamics. Further work performed in these areas of Cba biosynthesis and physiology will advance our understanding of the role of this magnificent molecule in nature.

Biographies

Elizabeth A. Villa received her B.S. degree in Biological Sciences from Virginia Polytechnic Institute and State University and her M.Sc. in Biology from Appalachian State University. She received her Ph.D. degree in Microbiology from the University of Georgia. Elizabeth completed her doctoral studies in Dr. Jorge C. Escalante-Semerena’s laboratory, which centered around understanding the biosynthesis and use of cobamides by various prokaryotes. She has worked in medical communications since completing her doctoral degree.

Jorge C. Escalante-Semerena received his B.S. degree in Biochemistry from the School of Chemistry at The Universidad Nacional Autónoma de México, his M.Sc. and Ph.D. degrees in Microbiology from The University of Illinois-Urbana, Champaign (UIUC), and postdoctoral training in Bacterial Genetics at UIUC and The University of Utah. In 1988, he joined the faculty of The Department of Bacteriology of The University of Wisconsin-Madison where he remained for 24 years. At UW-Madison he was the Ira L. Baldwin Professor of Bacteriology. In 2012, he joined The Department of Microbiology of The University of Georgia-Athens where he holds the title of UGA Foundation Distinguished Professor in Microbiology. Dr. Escalante-Semerena’s has worked on prokaryotic metabolism and physiology for the last 36 years trying to understand how the cell works by focusing on the functions of hypothetical genes. He believes that only then will biologists get a minimalistic picture of a cell’s capabilities.