Coenzymology: the biochemistry of vitamin biogenesis and cofactor-containing enzymes

Anaerobic synthesis of vitamin B12: characterization of the early steps in the pathway

S. Frank, A.A. Brindley, E. Deery, P. Heathcote, A.D. Lawrence, H.K. Leech, R.W. Pickersgill, M.J. Warren


The anaerobic biosynthesis of vitamin B12 is slowly being unravelled. Recent work has shown that the first committed step along the anaerobic route involves the sirohydrochlorin (chelation of cobalt into factor II). The following enzyme in the pathway, CbiL, methylates cobalt-factor II to give cobalt-factor III. Recent progress on the molecular characterization of this enzyme has given a greater insight into its mode of action and specificity. Structural studies are being used to provide insights into how aspects of this highly complex biosynthetic pathway may have evolved. Between cobalt-factor III and cobyrinic acid, only one further intermediate has been identified. A combination of molecular genetics, recombinant DNA technology and bioorganic chemistry has led to some recent advances in assigning functions to the enzymes of the anaerobic pathway.

  • chelation
  • cobalamin
  • pathway
  • precorrin
  • tetrapyrrole
  • vitamin B12


The structural complexity of vitamin B12 is reflected in an intricate biochemical pathway, which involves more than 30 enzyme-mediated steps for its complete de novo synthesis [1,2]. Consequently, many organisms have given up on cobalamin as a coenzyme and neither make nor require the vitamin in their metabolism. Indeed, one of the most surprising facets of cobalamin biosynthesis is that it is only made by certain members of the eubacteria and archaea, with the pathway never appearing to have made the transition into eukaryotes [3]. Within those prokaryotes that make cobalamin de novo, there are two related, though genetically distinct, pathways in operation [4]. These are referred to as the aerobic and anaerobic routes. The aerobic pathway is so-called because of its requirement for molecular oxygen but is also characterized by the comparatively late chelation of cobalt into the corrin ring [1,57]. This pathway is described in the accompanying paper [7a]. The anaerobic route has no requirement for molecular oxygen [1,2,4], and cobalt is inserted at an early stage of the synthesis [4,8,9]. This paper will concentrate on progress that has been made in understanding the early steps in the anaerobic pathway for cobalamin biosynthesis.

As with all modified tetrapyrroles, the structure of the corrin ring component of vitamin B12 is derived from the uro'gen III (uroporphyrinogen III) template [10]. This first macrocyclic intermediate of tetrapyrrole synthesis is methylated at positions C-2 and C-7 by an enzyme called SUMT (S-adenosyl-L-methionine uro'gen III methyltransferase; encoded by cobA) (Figure 1) [11], whose structure has recently been solved by X-ray crystallography [7]. The enzyme adds two SAM (S-adenosyl-L-methionine)-derived methyl groups in specific order, generating initially precorrin-1 (where the methyl group is added to the C-2 position) and subsequently precorrin-2 (Figure 1) [11]. These are the first two of a total of eight methyl groups that are added to the tetrapyrrole framework during corrin synthesis. Indeed, the names of the intermediates along the early stages of the cobalamin biosynthetic pathway reflect the number of methyl groups that have been added to the macrocyclic framework [12]. Thus precorrin-2 is an intermediate that has had two methyl groups added to the periphery of the macrocycle. The SUMT enzyme is unique in that it appears to be the only corrin biosynthetic enzyme that methylates at two separate positions [7]. The remaining six methyl groups would appear to be added by six separate methyltransferases [2].

Figure 1 Transformation of uro'gen III into cobalt-factor II

The numbering and labelling of the tetrapyrrole ring is shown for uro'gen III. Initially, uro'gen III is transformed into precorrin-2, which is oxidized to factor II and finally chelated with cobalt. A, acetic acid side chain; P, propionic acid side chain.

Biosynthesis of precorrin-2 is common to both the aerobic and anaerobic pathways. At this stage, the two pathways divide, rejoining only after the synthesis of adenosylcobyric acid [2]. The anaerobic pathway was thought to proceed by the insertion of cobalt into precorrin-2 [4,13,14]. However, recent evidence has shown that cobalt is inserted into the oxidized form of precorrin-2, factor II (sirohydrochlorin) [9]. Thus precorrin-2 is oxidized into factor II in a reaction that is catalysed by SirC and requires NAD+ (Figure 1) [15]. Factor II then acts as the substrate for chelation. There are two distinct cobalt-inserting enzymes found in the anaerobic pathway called either CbiK (e.g. as found in Salmonella enterica) [16] or CbiX (e.g. as found in Bacillus megaterium) [9]. It is likely that these two enzymes are derived from a common ancestral chelatase but their level of sequence similarity is very low. These two cobaltochelatases (CbiK and CbiX) belong to the class II chelatases as they are homomeric enzymes that do not require ATP for activity [17]. Structural studies have also revealed that CbiK is clearly related to the ferrochelatase of haem synthesis [16,18]. Another interesting similarity with the haem ferrochelatase is that some of the CbiX-cobaltochelatases contain an Fe-S centre, although in cobalamin synthesis this is an Fe4-S4 centre whereas the redox groups generally found in protoporphyrin IX ferrochelatases are Fe2-S2 centres [9,19]. No role for these centres has yet been described, although it has been suggested that the Fe4-S4 centre present on CbiX may be involved in maintaining the oxidation state of the chelated cobalt ion [9].

The insertion of cobalt (CoII) into factor II generates (sirohydrochlorin) CoII-factor II (CoII-sirohydrochlorin) (Figure 1), a result confirmed by EPR analysis of the reaction product [9]. The next enzyme in the pathway is CbiL [20]. This enzyme adds a SAM-derived methyl group to the C-20 position of the macrocycle (Figure 2). In subsequent steps, both the methyl group and the C-20 carbon are lost during the ring contract process, extruded as acetaldehyde [21]. The S. enterica CbiL has been shown to methylate both cobalt-precorrin-2 and cobalt-factor II, in both the CoII and CoIII oxidation states [20]. However, no yields, rates or specificity have been reported for this reaction. More recently, we have cloned cbiL from Methanothermobacter thermoautotrophicus, overproduced the enzyme recombinantly in Escherichia coli and purified the protein to homogeneity (S. Frank and M.J. Warren, unpublished work). We have devised a simple assay and have shown that M. thermoautotrophicus CbiL has a strong preference for CoII-factor II over CoII-precorrin-2. Moreover, the protein has been crystallized, where the crystals have the space group P62 and contain a homodimer in the asymmetric unit. They were found to diffract with 2.1 Å (1 Å=0.1 nm) resolution at the European Synchrotron Radiation Facility (Grenoble, France) and a model of CbiL was built to an experimentally phased map at 2.4 Å resolution calculated using two isomorphous derivatives (mercury and samarium) (S. Frank, R.W. Pickersgill and M.J. Warren, unpublished work).

Figure 2 Reaction catalysed by CbiL

Cobalt-factor II is methylated at the C-20 position to give cobaltfactor III. A, acetic acid side chain; P, propionic acid side chain.

Sequence analysis of CbiL suggests that it belongs to the same family of enzymes that methylate the corrin ring structure at C-2 and C-7 (CobA/SUMT), C-17 (CbiH), C-12 (CbiF) and C-5 (CbiE) [22]. The remaining methylation at C-1 (CbiD) and C-15 (CbiT) would appear to be catalysed by enzymes that belong to distinct families of methyltransferases [23,24]. The structure of CbiL will complement the known structures of SUMT (CobA) [7,25] and CbiF, and provide an insight into how these enzymes have evolved their regiospecificity. As a family of enzymes, they are also of interest as they may have arisen through retrograde pathway evolution [26].

The product of the CbiL reaction is CoII-factor III (Figure 2). Previous work has shown that CbiH, the subsequent pathway enzyme, catalyses the transformation of cobalt-precorrin-3 into cobalt-precorrin-4 [27,28], although neither the oxidation of the product nor the oxidation state of the cobalt could be verified due to the comparatively low yield of the reaction. The striking feature of cobalt-precorrin-4, which was isolated as cobalt-factor IV, is that it has undergone ring contraction. In this respect, CbiH catalyses both the ring contraction of the macrocycle and the methylation at C-17. If cobalt-precorrin-3 is the substrate for CbiH then the product of the CbiL reaction, cobalt-factor III, has to undergo a two-electron reduction. In this respect, it is interesting to note that the M. thermoautotrophicus CbiH enzyme contains several Fe-S centres (A.A. Brindley and M.J. Warren, unpublished work). However, these redox groups are not present on the S. enterica orthologue suggesting that if a redox change is involved then at least another protein would have to be involved. Alternatively, as it has not been possible to rigorously identify the oxidation of cobalt-precorrin-4 [27,28], the possibility exists that it may act upon cobalt-factor III and generate cobalt-factor IV. There is thus a need to gain further insight into the reaction catalysed by CbiH (Figure 3).

Figure 3 Possible routes for the transformation of cobalt-factor III into vitamin B12

There is evidence that CbiH can convert either cobalt-factor III or cobalt-precorrin-3 into cobalt-factor IV or cobalt-precorrin-4 respectively. There are no known intermediates between the precorrin-4 stage and cobyrinic acid.

Nonetheless, painstaking work by the group in Texas has shown that the intermediate synthesized by the S. enterica CbiH is converted by cell free extracts into cobyrinic acid, demonstrating that whatever the oxidation state, it is a true intermediate [28]. However, between cobalt-precorrin-4 and cobyrinic acid no other intermediates have been isolated (Figure 3). This transformation requires methylation at C-11 (CbiF), methylation at C-1 (CbiD), loss of the extruded methylated C-20 as a C-2 fragment, methylation at C-5 (CbiE), methylation at C-15 (CbiT), decarboxylation of the acetate side chain on ring C (CbiT), and finally, rearrangement of the methyl group from C-11 to C-12 (CbiC) [2]. The role of CbiD as the C-1 methyltransferase was demonstrated recently in a recombinant strain of E. coli harbouring all the genes to make cobyrinic acid a,c-diamide except for CbiD [24]. This gene-depleted strain was found to accumulate 1-desmethylcobyrinic acid a,c-diamide. Circumstantial evidence was also provided to suggest that some of the corrin biosynthetic enzymes may form a complex, helping to explain why it may be difficult to isolate later intermediates in the pathway [24].

Cobyrinic acid is transformed into cobyric acid by the action of CbiA and CbiP, which amidate the propionic and acetic acid side chains [2]. At some stage, the molecule is adenosylated, after reduction of the central cobalt ion to a CoI oxidation state, although the exact timing of this adenosylation reaction has not been determined [2]. The lower nucleotide loop is then assembled and attached by a number of enzymes including CobD, CbiB, CobU, CobT, CobC and CobS to generate adenosylcobalamin [2].

The anaerobic pathway towards cobalamin biosynthesis is slowly and steadily being unravelled through a combination of molecular genetics, bioorganic chemistry and biochemistry. The exquisite sensitivity of many of the intermediates to molecular oxygen has made this process a painstaking one, but one where the development of rigorous anaerobic techniques and improved detection techniques is allowing progress to be made. The combination of this work with structural biology is a unique opportunity to observe how Nature has evolved the synthesis of this highly complex vitamin.


  • Coenzymology: the biochemistry of vitamin biogenesis and cofactor-containing enzymes: Independent Meeting held at King's College, Cambridge, U.K., 4–7 April 2005. Organized and Edited by A.G. Smith (Cambridge, U.K.) and A.W. Munro (Leicester, U.K.).

Abbreviations: uro'gen III, uroporphyrinogen III; SUMT, S-adenosyl-L-methionine uro'gen III methyltransferase


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 7a.
  9. 8.
  10. 9.
  11. 10.
  12. 11.
  13. 12.
  14. 13.
  15. 14.
  16. 15.
  17. 16.
  18. 17.
  19. 18.
  20. 19.
  21. 20.
  22. 21.
  23. 22.
  24. 23.
  25. 24.
  26. 25.
  27. 26.
  28. 27.
  29. 28.
View Abstract