Microbial protein structure

Microbial protein structure

My group is presently working on three projects:

  1. Analysis of the structure and the catalytic mechanism of [Fe]-hydrogenase (Hmd) (Pilak et al. 2006; Korbas et al. 2006; Vogt et al. 2008; Guo et al. 2008; Pilak et al. 2008; Shima & Thauer 2007)
  2. Crystal structure analysis of tetrahydromethanopterin (H4MPT)- and F420-dependent enzymes (Acharya et al. 2006; Seedorf et al. 2007)
  3. Characterization of enzymes involved in anaerobic oxidation of methane (Kahnt et al. 2007; Shima & Thauer 2006; Thauer & Shima 2006; Thauer & Shima 2007; Thauer & Shima 2008; and see R. Thauer's report).

We intensively collaborate with other groups for biophysical experiments: Ulrich Ermler (X-ray crystallography, MPI for Biophysics, Frankfurt), Eckhard Bill (Mössbauer spectroscopy, MPI for Bioinorganic Chemistry, Mulheim) and Wolfram Meyer-Klaucke (X-ray absorption spectroscopy, EMBL Outstation Hamburg). The results of the analysis of [Fe]-hydrogenase are described in detail below.

Molecular hydrogen and proton are used as electron donor or acceptor, respectively, in the metabolism of many anaerobic microorganisms. Enzymes named hydrogenases have evolved that catalyse the reversible reaction

H2 = 2 H+ + 2 e-.

The most prominent hydrogenases are [NiFe]-hydrogenase (in Bacteria and Archaea) and [FeFe]-hydrogenase (in Bacteria and Eukarya). The structures of their binuclear metal active sites are pictured in Fig.1a and Fig.1b. In addition to the binuclear metal centre both types of hydrogenases harbour at least one essential [4Fe4S] cluster oxidized and reduced in the catalytic cycle.

Not related to either [NiFe]-hydrogenase or [FeFe]-hydrogenase is a third type of hydrogenases, the [Fe]-hydrogenase or formerly named iron-sulphur-cluster free hydrogenase, with a mononuclear iron centre cartooned in Fig.1 c on the basis of the crystal structure.

This enzyme is only found in some hydrogenotrophic methanogenic archaea and catalyzes the reversible dehydrogenation of methylene-H4MPT to methenyl- H4MPT+, which is a step involved in CO2 reduction to methane.

H2 + Methenyl-H4MPT+ = Methylene-H4MPT + H+ (ΔGo' = - 5.5 kJ/mol)

The homodimeric enzyme contains per subunit one iron, which is not redox active, and no iron-sulphur clusters. The single iron in [Fe]-hydrogenase is associated with an iron guanylyl pyridone cofactor (FeGP-cofactor). IRspectroscopic analysis of the cofactor and of the holoenzyme revealed the presence of two CO bound to iron at an angle of 90°. Mössbauer spectroscopy showed that the iron is low spin, either Fe(0) or Fe(II). X-ray absorption spectroscopy indicated the presence of two CO, one sulphur and one or two N/O ligands at coordination distance.

After more than 10 years of struggle we finally achieved to obtain a crystal structure of [Fe]-hydrogenase (holoenzyme) from Methanocaldococcus jannaschii (reconstituted from the heterologously produced apoenzyme and FeGP-cofactor) at 1.75 Å resolution. The overall structure of the holoenzyme is very similar to that of the recently solved apoenzyme.

The FeGP-cofactor - unambiguously found in the electron density - is positioned in front of the C-terminal end of the parallel b-sheet of both Rossmann-fold like peripheral units such that the catalytically relevant iron is directed towards the inter-subunit clefts. Due to the high occupancy of FeGP its electron density is well-shaped except for the carboxyl group of the carboxymethyl group of the pyridone which is partly disordered.

The iron is ligated by the pyridone nitrogen, the sulphur of Cys176, two diatomic ligands, which are most certainly CO, and an unknown ligand. Together with the fifth unknown ligand they form a slightly distorted square pyramid (or an octahedron with a site vacant) with the iron in the plane of the square and the pyridone nitrogen at the top of the pyramide. The two CO form an angle of 90° as predicted from the IR spectrum of the holoenzyme. The iron, one of the CO and the unknown ligand are located exactly in the plane of the pyridone ring indicating that it is in its 2-pyridinol tautomeric form. The pyridinol hydroxyl group is not involved in iron ligation. The electron density of the unknown ligand cannot definitely be assigned as monatomic or diatomic ligand although clearly connected with that of the iron and of relatively high occupancy (ca. 60% for a tentatively modelled water molecule).

The vacant sixth coordination site of the iron contains a spherical electron density interpreted as monatomic solvent molecule that is, however, at a distance of 2.7 Å too far away to be considered as a ligand. We predict this site to be the binding position of extrinsic CO known to inhibit [Fe]-hydrogenase as indicated by the IR spectrum. Since CO is a competitive inhibitor with respect to H2 the latter most likely also binds to this site (Fig.1).

The structures of the [Fe]-hydrogenase, of the [FeFe]- hydrogenase and of the [NiFe]-hydrogenase are completely different on the primary, secondary, tertiary and quaternary level but share features in their active site which can only have evolved convergently. All three have in common a low spin iron ligated by thiolate(s), CO and cyanide or a cyanide functional analogue which acts together with a partner in the heterolytic cleveage of H2. This partner is in the case of the [Fe]-hydrogenase a carbocation (C14a of methenyl-H4MPT+), in the case of [FeFe]-hydrogenase the iron distal to the [4Fe4S]- cluster and in the case of [NiFe]-hydrogenase the Ni (Fig.1). Model complexes to be constructed on the basis of the iron centre of [Fe]-hydrogenase should give further insight into the essential but not yet understood function of the low spin iron with its unique ligands in H2 activation.

Other Interesting Articles

Go to Editor View