Structure/Function Studies on a S-Adenosyl-l-methionine-dependent Uroporphyrinogen III C Methyltransferase (SUMT), a Key Regulatory Enzyme of Tetrapyrrole Biosynthesis

The crystallographic structure of the Pseudomonas denitrificans S-adenosyl-l-methionine-dependent uroporphyrinogen III methyltransferase (SUMT), which is encoded by the cobA gene, has been solved by molecular replacement to 2.7 Å resolution. SUMT is a branchpoint enzyme that plays a key role in the biosynthesis of modified tetrapyrroles by controlling flux to compounds such as vitamin B₁₂ and sirohaem, and catalysing the transformation of uroporphyrinogen III into precorrin-2. The overall topology of the enzyme is similar to that of the SUMT module of sirohaem synthase (CysG) and the cobalt-precorrin-4 methyltransferase CbiF and, as with the latter structures, SUMT has the product S-adenosyl-l-homocysteine bound in the crystal. The roles of a number of residues within the SUMT structure are discussed with respect to their conservation either across the broader family of cobalamin biosynthetic methyltransferases or within the sub-group of SUMT members. The D47N, L49A, F106A, T130A, Y183A and M184A variants of SUMT were generated by mutagenesis of the cobA gene, and tested for SAM binding and enzymatic activity. Of these variants, only D47N and L49A bound the co-substrate S-adenosyl-l-methionine. Consequently, all the mutants were severely restricted in their capacity to synthesise precorrin-2, although both the D47N and L49A variants produced significant quantities of precorrin-1, the monomethylated derivative of uroporphyrinogen III. The activity of these variants is interpreted with respect to the structure of the enzyme.

Introduction

The modified tetrapyrroles encompass a family of structurally related metallo-prosthetic groups including haem, chlorophyll, vitamin B₁₂ (cobalamin), sirohaem and coenzyme F₄₃₀.¹ These molecules are synthesised along a branched biosynthetic pathway via the common primogenitor uroporphyrinogen III. Decarboxylation of the acetic acid side-chains of this asymmetric macrocyclic intermediate by the enzyme uroporphyrinogen III decarboxylase generates coproporphyrinogen III,² which is then further metabolised into either haem or chlorophyll by peripheral modification of the molecule's side-chains and macrocyclic ring oxidation.³ Alternatively, C-methylation of uroporphyrinogen III at positions 2 and 7 by the enzyme S-adenosyl-l-methionine uroporphyrinogen III methyltransferase (SUMT) generates precorrin-2 (dihydrosirohydrochlorin),⁴ an intermediate that can then be further transformed into vitamin B₁₂, sirohaem, coenzyme F₄₃₀⁵ and, in certain organisms, even into haem⁶ (Figure 1(a)). These two branchpoint enzymes, uroporphyrinogen III decarboxylase and SUMT, are key regulatory enzymes that play an important role in the metabolic flux through this branched biosynthetic pathway. Much attention has focussed previously on the structure–function relationship of uroporphyrinogen III decarboxylase since mutations in its gene or chemical inactivation of the enzyme can cause porphyria cutanea tarda,⁷ a disease associated with photosensitivity and dermatological lesions due to the accumulation of phototoxic uroporphyrin III (oxidised uroporphyrinogen III) within the body. In contrast, comparatively little is known about SUMT, mainly due to the difficulty in assaying the enzyme.

SUMT was first isolated from Pseudomonas denitrificans and was found to consist of 280 amino acid residues.⁴ The P. denitrificans SUMT, which is encoded by cobA as part of a cobalamin biosynthetic operon,⁸ has a molecular mass of 28 kDa, and has been shown to exist as a homodimer. Kinetic studies revealed that the enzyme methylates uroporphyrinogen III initially at position 2 and subsequently at position 7 (Figure 1(b)). The enzyme is likely to proceed via a ternary complex, with S-adenosyl-l-methionine (SAM) and uroporphyrinogen III binding to the enzyme prior to the release of precorrin-1 and S-adenosyl-l-homocysteine (SAH).⁴ After binding a second molecule of SAM and precorrin-1, precorrin-2 is released with another molecule of SAH.⁴ The enzyme has a comparatively low k_cat of 38 h⁻¹, and micromolar K_m values for co-substrates SAM (6.3 μM) and uroporphyrinogen III (1.0 μM). SUMT displays both powerful feedback inhibition (by product SAH, K_i=0.32 μM)⁴ and substrate inhibition (by uroporphyrinogen III at concentrations over 2 μM), which is thought to act as an important regulatory mechanism in cobalamin biosynthesis.⁴

Subsequently, SUMT activity was described in CysG (sirohaem synthase),⁹ a 457 amino acid residue multifunctional enzyme that is able to transform uroporphyrinogen III into sirohaem by first methylating uroporphyrinogen III to generate precorrin-2, then catalysing the dehydrogenation of precorrin-2 into sirohydrochlorin and finally chelating ferrous iron into the central cavity of the macrocycle to give sirohaem¹⁰ (Figure 1(a)). However, it is now clear that CysG represents a fusion between a SUMT (amino acid residues 202–457) and a dehydrogenase/chelatase (amino acid residues 1–201), since the SUMT module can be dissected out of CysG in a fully functional form.11, 12

Overproduction of SUMT is thought to be key to the generation of good vitamin B₁₂ producing strains.¹³ Related to this have been efforts to detect a SUMT that lacks substrate inhibition, since an enzyme that lacks such an inhibitory property is likely to allow greater flux through the biochemical pathway.¹⁴ In a quite unrelated application, the strong fluorescence of the product of the reaction means that enzyme activity can easily be identified in a range of cell types. Thus the SUMT gene has been used as a marker similar to β-galactosidase to detect cloning of DNA into plasmids, where inserts are detected in clones lacking fluorescence.¹⁵ The gene has also been used as a red fluorescent transcriptional reporter for Escherichia coli, yeast, and mammalian cells, with enhanced properties over green fluorescent protein.¹⁶

Here we report on the three-dimensional structure of SUMT from P. denitrificans and analyse the activity of a number of mutant variants of the enzyme. This has allowed an insight into how SUMT is able to accommodate two distinct regiospecific methylations of the porphyrinogen template within the one active site.