Pseudo-B₁₂ Joins the Cofactor Family

Over 50 years ago, a deadly autoimmune disease known as pernicious anemia, which is characterized by neurological and hematological abnormalities, had no known cure. The isolation of the anti-pernicious anemia factor from liver extracts by Folkers and Smith in 1948 represented a breakthrough in the treatment of this disease (21, 25). Named vitamin B₁₂, this compound could be isolated as pure, red crystals from various animal and microbial sources, and the X-ray crystal structure of this complex molecule, one of the largest nonpolymeric natural compounds, was determined in the 1950s by Dorothy Hodgkin et al. (Fig. 1A) (8). Since that time, thousands of researchers have identified and characterized a variety of enzymes that utilize B₁₂ cofactors, determined the complex pathway for B₁₂ absorption in humans, and elucidated the microbial pathways for B₁₂ biosynthesis. Other forms of vitamin B₁₂ have been found in nature, but their functions are not fully understood. In this issue of the Journal of Bacteriology, Anderson et al. provide evidence that a form of B₁₂ known as “pseudo-B₁₂” (Fig. 1B) is a physiologically important form of B₁₂ in Salmonella enterica serovar Typhimurium (1).

FIG. 1.

Corrinoids produced by S. enterica. Cobalamin (A) and pseudo-B₁₂ (B) differ in their α-axial ligands. The β-axial ligand (R) is 5′-deoxyadenosine or -CH₃ in B₁₂ cofactors and -CN in vitamin B₁₂. Collectively, these molecules are abbreviated in the text as B₁₂[DMB] (A) and pseudo-B₁₂ (B).

Unlike all other vitamins, vitamin B₁₂ is not produced by plants; rather, its biosynthesis is restricted to certain prokaryotes (23). Animals obtain B₁₂ from food or, in some cases, directly from resident B₁₂-producing microbes (28). S. enterica has long been used as a model organism for studying B₁₂ biosynthesis (23).

B₁₂ structure and biosynthesis.

B₁₂ and B₁₂-like compounds are collectively termed corrinoids. The structural core of corrinoids is the corrin ring in which the central cobalt atom is coordinated by four pyrrole groups (Fig. 1) (30). Two axial ligands also coordinate the cobalt atom. In the cofactor forms, the upper (β) axial ligand (R) consists of either 5′-deoxyadenosine or a methyl group, which facilitate radical or methyl transfer reactions, respectively (14). The lower (α) axial ligand is attached to the corrin ring via the nucleotide loop which is characterized by an unusual α-glycosidic linkage to C-1 of the ribosyl moiety (30). The B₁₂ cofactors found in animals and many bacteria contain 5,6-dimethylbenzimidazole (DMB) as the α-axial ligand (Fig. 1A; B₁₂ with DMB will be abbreviated hereafter as B₁₂[DMB] to distinguish it from pseudo-B₁₂) (26). B₁₂ cofactors with α-axial ligands other than DMB have also been observed, with the most common example being pseudo-B₁₂ (Fig. 1B), in which N7-linked adenine replaces DMB as the α-axial ligand (9).

The complete biosynthesis of vitamin B₁₂ requires ∼30 gene products. The B₁₂ biosynthetic pathway is characterized by two converging branches: the synthesis of the corrin ring and the synthesis and activation of the α-axial ligand (23, 30). Two pathways for corrin ring synthesis have been discovered and are hypothesized to be derived from a common ancestor (10). The main difference between these pathways is in the participation of oxygen. The “aerobic” pathway requires oxygen and has been characterized in depth in Pseudomonas denitrificans (7, 30). The “anaerobic” pathway, characterized in the aerotolerant anaerobe Propionibacterium shermanii and the facultative anaerobe S. enterica, is sensitive to oxygen, and thus, the expression of B₁₂ biosynthetic enzymes is restricted to anaerobic or microaerobic conditions (22, 30). Nucleotide loop assembly is similar in both pathways (23, 30). Two independent pathways for DMB synthesis also exist (18). Aerobic DMB synthesis occurs by fragmentation of a flavin cofactor and is catalyzed by the BluB enzyme (6, 19, 27). DMB synthesis in some anaerobes appears to branch from the purine biosynthetic pathway, although enzymes involved in this process have not been isolated (11, 18). Interestingly, S. enterica synthesizes the corrin ring anaerobically and derives DMB aerobically from flavin mononucleotide (FMN), although no ortholog of BluB appears to exist in the genome (1, 12).

Probing the origin of the α-axial ligand of B₁₂ in S. enterica.

The observation that S. enterica produces B₁₂[DMB], but that no DMB biosynthetic enzyme has been identified in S. enterica, has puzzled researchers for many years (23). Anderson et al. explored this issue by selecting for mutants of S. enterica that grow under conditions that normally require exogenous DMB (1). One key observation that made this selection possible is that traditional agar contains traces of DMB, and thus, much of the DMB in the B₁₂[DMB] produced by S. enterica can be attributed to its presence in agar. Using purified agar enabled the authors to study only the endogenously produced corrinoids. Another important difference between this and other studies was that growth was monitored on ethanolamine rather than methionine or propanediol. This alteration was critical because (i) the requirement for B₁₂ during growth on ethanolamine is more stringent than on methionine, and (ii) exogenous DMB is required for growth on ethanolamine but not on propanediol (1).

Unexpectedly, rather than uncovering genes involved in synthesizing DMB, the authors identified 25 point mutations that increase the cellular concentration of free adenine, a precursor of pseudo-B₁₂ (1). Thus, all of the characterized mutants produced an elevated level of pseudo-B₁₂ that supported growth on ethanolamine. The authors further show that the attachment of adenine to form pseudo-B₁₂ occurs by the same pathway as the attachment of DMB to form B₁₂[DMB]. This result confirms previous biochemical findings that the nucleotide activation and attachment enzymes accommodate a variety of substrates but prefer DMB (4, 29). The genetic evidence that pseudo-B₁₂ functions as a cofactor for all three B₁₂-dependent enzymes in S. enterica also suggests that the formation of pseudo-B₁₂ is a natural physiological process (29).

Why were no DMB biosynthetic enzymes identified in this selection? One intriguing possibility is that none exist in S. enterica. This possibility was obscured in previous studies by the presence of DMB in agar, coupled with the 15-fold-higher activity of the nucleotide activation enzyme CobT with DMB compared to that with adenine (29). It is curious that S. enterica produces only ∼100 molecules per cell of B₁₂[DMB] when DMB is not provided exogenously (1, 2). It is unlikely that an enzyme-catalyzed reaction would produce such a low level of DMB under conditions in which more is required. Furthermore, DMB is formed from FMN in the presence of S. enterica cell extracts and oxygen, suggesting that S. enterica produces DMB by a method similar to the BluB-catalyzed reaction (12, 27). However, the low level of DMB formed in aerobic S. enterica cultures could be due to the nonenzymatic transformation of FMN to DMB, which has been observed in vitro (16, 20). Alternatively, DMB could be formed as a side reaction of flavin oxidoreductase enzymes (1); this possibility is consistent with the evidence that most flavins in the cell exist in a protein-bound form (15). Also intriguing is the fact that S. enterica produces pseudo-B₁₂ or incomplete corrinoids anaerobically but requires oxygen for DMB biosynthesis (3, 13). Together, these observations suggest that DMB is neither synthesized by a dedicated enzyme nor required for B₁₂ cofactors in S. enterica. If true, this means that pseudo-B₁₂ is the natural corrinoid in S. enterica.

Function of pseudo-B₁₂ revisited.

The identity of pseudo-B₁₂ was known as early as 1952 (5). However, pseudo-B₁₂ was shown not to function in humans and therefore has received less attention than B₁₂ (DMB). The 500-fold-lower binding affinity of pseudo-B₁₂ for the human B₁₂ receptor, intrinsic factor, largely contributes to the inability of humans to use pseudo-B₁₂ (26).

The diversity of lower ligands in corrinoid compounds has been understood for several decades, as numerous benzimidazoles, purines, and phenolic compounds have been isolated as lower ligands of corrinoids (4, 18). Several recent reports have shown that pseudo-B₁₂ is the dominant corrinoid produced by the anaerobes Clostridium cochlearium (9) and Lactobacillus reuteri (24), the cyanobacteria Nostoc commune (32) and Aphanizomenon flos-aquae (17), and the alga Aphanothece sacrum (31). Historically, these corrinoids have been viewed as “alternate” forms. However, these observations and the genetic studies of Anderson et al. (1) have brought to our attention the reality that pseudo-B₁₂ and other “alternate” corrinoids are in fact the natural B₁₂ cofactors in many organisms.