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

Aerobic synthesis of vitamin B12: ring contraction and cobalt chelation

D. Heldt, A.D. Lawrence, M. Lindenmeyer, E. Deery, P. Heathcote, S.E. Rigby, M.J. Warren


The aerobic biosynthetic pathway for vitamin B12 (cobalamin) biosynthesis is reviewed. Particular attention is focused on the ring contraction process, whereby an integral carbon atom of the tetrapyrrole-derived macrocycle is removed. Previous work had established that this chemically demanding step is facilitated by the action of a mono-oxygenase called CobG, which generates a hydroxy lactone intermediate. This mono-oxygenase contains both a non-haem iron and an Fe-S centre, but little information is known about its mechanism. Recent work has established that in bacteria such as Rhodobacter capsulatus, CobG is substituted by an isofunctional protein called CobZ. This protein has been shown to contain flavin, haem and Fe-S centres. A mechanism is proposed to explain the function of CobZ. Another interesting aspect of the aerobic cobalamin biosynthetic pathway is cobalt insertion, which displays some similarity to the process of magnesium chelation in chlorophyll synthesis. The genetic requirements of cobalt chelation and the subsequent reduction of the metal ion are discussed.

  • biosynthesis
  • aerobic synthesis
  • cobalamin (vitamin B12)
  • cobalt chelation
  • precorrin
  • ring contraction

Vitamin B12 (cobalamin) is a cobalt-containing modified tetrapyrrole that belongs to the same family of metallo-prosthetic groups as haem, chlorophyll, sirohaem and coenzyme F430 [1,2]. It is generally found in one of the two biologically active forms, adenosylcobalamin and methylcobalamin, which are involved in rearrangement and methylation reactions respectively [3]. Cobalamin biosynthesis represents one of Nature's most complex metabolic pathways, requiring approx. 30 enzymes for its complete de novo synthesis. There are two known routes for cobalamin construction: (i) an aerobic pathway that requires molecular oxygen and where cobalt is inserted late-on in the pathway [15] and (ii) an anaerobic pathway where there is no requirement for molecular oxygen [1,2,6,7] and where cobalt insertion represents the first committed step towards cobalamin synthesis [8]. The anaerobic pathway is covered in the accompanying article [8a]. This report will concentrate on aspects of the aerobic pathway for corrin synthesis.

The aerobic pathway was elucidated about a decade ago with major contributions from three laboratories – the Rhone Poulenc industrial group that pioneered the molecular genetics and biochemistry of the pathway [5,9], and the bioorganic approach adopted by Professor S.R. Battersby (University of Cambridge, Cambridge, U.K.) [4] and Professor A.I. Scott (Texas A&M University, College Station, TX, U.S.A.) [1]. Cobalamin is derived from the common tetrapyrrole primogenitor uroporphyrinogen III (uro'gen III) [10] (Figure 1). The transformation of uro'gen III into cobalamin requires a contraction of the macrocyclic ring structure (such that C-20 is removed), peripheral methylation, decarboxylation, cobalt insertion, amidation of a number of the acetic and propionic acid side chains, adenosylation of the cobalt ion and the assembly and attachment of the lower nucleotide loop. The aerobic biosynthesis is outlined in Figure 2. In the present study, we shall concentrate on two aspects of the aerobic pathway, namely the steps associated with ring contraction and cobalt insertion.

Figure 1 Comparison of the structures of uroporphyrinogen III and vitamin B12
Figure 2 Transformation of uroporphyrinogen III into adenosylcobyric acid

The Figure outlines the genetic requirements and intermediates involved in the aerobic biosynthesis of vitamin B12 in Ps. denitrificans. In R. capsulatus, cobG is replaced by cobZ.

The key step for the ring contraction process in the aerobic pathway involves the synthesis of precorrin-3B, the hydroxylated derivative of precorrin-3A (Figure 2). In Pseudomonas denitrificans, the bacterium in which the aerobic pathway was elucidated, this reaction is mediated by an enzyme called CobG [9,11] that displays some similarity to sulphite reductase. However, unlike the six-electron reduction catalysed by sulphite reductase, CobG mediates a two-electron oxidation of precorrin-3A into precorrin-3B. Elegant labelling studies revealed that CobG acts as a mono-oxygenase where the oxygen is incorporated into the C-20 position before lactone ring formation [12]. Despite the initial characterization of the Ps. denitrificans CobG, there are a number of important mechanistic questions that still need to be addressed including how the enzyme binds molecular oxygen and the nature and type of the predicted Fe-S centre within the protein. Our initial characterization of the Ps. denitrificans CobG by EPR analysis (A.D. Lawrence, S.E. Rigby and M.J. Warren, unpublished work) has confirmed that CobG contains a single Fe4-S4 centre in addition to a mononuclear non-haem iron site. Work is on-going to elucidate the roles these groups play within the catalytic cycle of the enzyme.

In another development concerning ring activation before contraction, we had observed that some bacteria that harboured the genes for the aerobic pathway were missing the gene encoding CobG [13]. Significantly, these organisms contained another gene for which no function was ascribed. From one of these organisms (Rhodobacter capsulatus), we cloned this gene and called it cobZ [14]. A bioinformatics study indicated that the CobZ protein contained two distinct entities – an N-terminal region that bore strong similarity to some flavin-containing proteins such as succinate dehydrogenase and a C-terminal region that was predicted to contain a number of transmembrane helices and which showed similarity to a protein involved in tricarballyate metabolism [15]. The two regions were joined by a central sequence that was rich in cysteine residues suggesting the presence of Fe-S centres.

Recombinant production of CobZ in Escherichia coli led to the isolation of CobZ in two forms [14]. The N-terminal region of the protein could be isolated in a soluble form after proteolysis. This truncated protein was purified to homogeneity and was yellow in colour, consistent with the presence of a flavin. The flavin was subsequently characterized as FAD. The full-length protein could be purified to homogeneity after detergent extraction and appeared much darker in colour than the truncated N-terminal form. A UV–visible spectrum of the full-length protein was dominated by the Soret band of a haem group. This suggested that the C-terminal membrane-bound region of CobZ was a novel haemoprotein. EPR analysis of the holoprotein also revealed the presence of two closely coupled Fe4-S4 centres. To demonstrate that CobZ is involved in cobalamin synthesis, we showed that CobZ had the ability to complement a cobG-deficient strain. This indicated that CobZ and CobG are isofunctional and catalyse the same reaction, i.e. CobZ must also act as a mono-oxygenase in the synthesis of precorrin-3B. Based on the characterization of CobZ to date, we have proposed a reaction mechanism outlined in Figure 3, where the reduced flavin binds oxygen and where the Fe-S centres and haem are used to reduce the flavin during the catalytic cycle [14].

Figure 3 The catalytic cycle of CobZ

The reduced flavin reacts with molecular oxygen to generate a peroxide intermediate, which then combines with the precorrin substrate to give the precorrin-hydroxy lactone product and oxidized flavin. The oxidized flavin is subsequently reduced, probably by electrons channelled through the haem group and Fe-S centres.

The other area in the aerobic pathway that has caught our attention is the cobalt insertion and reduction steps [16,17]. This is an incredibly energy-demanding stage in the biosynthesis of cobalamin. Cobalt is inserted into an intermediate called hydrogenobyrinic acid a,c-diamide by a chelatase that requires three subunits for activity [17] (Figure 2). These subunits are referred to as CobN, CobS and CobT, where at least CobS and CobT are known to form a complex. Cobalt chelation also requires ATP hydrolysis although the stoichiometry of ATP hydrolysis to cobalt insertion has not been determined [17]. The aerobic cobaltochelatase system is very similar to the magnesium chelatase system that is found in the biosynthesis of chlorophyll (and bacteriochlorophyll) [18,19]. Indeed, the magnesium chelatase subunits, ChlH, ChlI and ChlD, display varying degrees of sequence similarity to their respective cobaltochelatase counterparts, indicating that they are likely to be derived from a common ancestor. Magnesium chelation has been more extensively studied than the cobaltochelation reaction and kinetic analysis has shown that it is the ChlI subunit that is responsible for the ATPase activity during chelation [2023]. More detailed studies have shown that about 15 ATPs are required for the insertion of every magnesium ion [24] and it is likely that a similar number will be required for cobalt insertion into the corrin ring.

We have now overproduced recombinant forms of CobN, CobS and CobT (D. Heldt and M.J. Warren, unpublished work). We have shown that CobS harbours ATPase activity and, by EPR, have shown that CobS, like ChlI [24], contains a separate magnesium-binding site in close proximity to the nucleotide-binding site (D. Heldt, S.E. Rigby and M.J. Warren, unpublished work). Our ability to produce comparatively large quantities of intermediates such as hydrogenobyrinic acid through metabolic engineering in E. coli means that we shall be able to characterize further this cobaltochelatase complex [14]. We have also purified a recombinant form of CobW, a cobalamin biosynthetic protein for which no function had been ascribed [5,25]. However, CobW does display significant similarity to the NHase (nitrile hydratase) activase protein of the Fe2+-containing NHase [26], and is proposed to play a role in metal delivery to the enzyme complex. Recent findings suggest that the NHase activator mainly participates in iron trafficking in NHase biogenesis as an iron-type metallochaperone [27]. It is feasible that CobW may play a similar role in the delivery and presentation of cobalt to CobN. We have shown that recombinant CobW, even in the absence of any His tag, can be purified on a cobalt column, clearly indicating that the protein can bind to cobalt (D. Heldt and M.J. Warren, unpublished work).

Once cobalt is inserted into hydrogenobyrinic acid a,c-diamide, the chelated metal ion is reduced to the CoI form by a cobalt reductase (Figure 2). Although the enzyme catalysing this reaction was purified to homogeneity and characterized, no genetic sequence encoding this enzyme has ever been reported [16]. We have recently identified two potential genes that may encode the cobalt reductase (which we have named CobR, see Figure 2) (E. Deery, A.D. Lawrence and M.J. Warren, unpublished work). Recombinant overproduction of the encoded proteins has led to the purification of flavoproteins with similar properties to those initially reported by the Rhone Poulenc scientists [16]. Work is now under way to investigate how these flavoproteins perform a single electron reduction of the centrally chelated cobalt ion.

In summary, a multidisciplinary approach to the study of the aerobic cobalamin biosynthetic pathway has provided major insights into the step-by-step synthesis of this remarkable coenzyme. The pathway elucidation is being complemented by structural studies as the mechanisms of individual enzymes are being investigated, providing molecular details on this biosynthetic route. There are still many interesting questions to address and the role of a number of proteins such as CobE and CobW [5] in the control and regulation of the pathway requires further investigation.


  • 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: NHase, nitrile hydratase


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 8a.
  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.
View Abstract