Reports: AC3
48676-AC3 Organometallic Chemistry of Methanogenesis: Modeling the Fe(CO)2-GP Cofactor in the Enzyme Hmd
One of the great surprises in bioinorganic chemistry was the discovery of the iron carbonyl sites in the [NiFe]- and the [FeFe]-hydrogenases. A third hydrogenase, Hmd (methylenetetrahydromethanopterin dehydrogenase), formerly thought to be “metal-free,” was recently shown to also contain an iron carbonyl center. Hmd represent possibly the last hydrogenase to be discovered: it is expressed only under special conditions (Ni-deficiency) and only by methanogenic archaea. The fact that three genetically independent hydrogenases evolved similar catalytic motifs underscores the versatility of the Fe-S-CO compounds for reactions involving hydrogen.
Synthetic modeling of the active sites of the hydrogenases, in parallel with biophysical studies, provides mechanistic insights. Good progress is being made in the biomimicry of the active sites of the [NiFe]- and [FeFe]-hydrogenases. The active site of Hmd consists of a dicarbonyl iron center bound to a thiolate and the pyridine-like nitrogen of the guanylylpyridinol (GP) cofactor. This ensemble, the Fe-GP cofactor, is extractable from the protein via transthiolation using mercaptoethanol. Because it is bound to the protein via this single exchangeable residue, this active site is particularly ripe for modeling, and indeed models have already been described. Recently, however, the structural assignment of the active site has been significantly revised. The new analysis indicates that Fe is bound also to an acyl ligand, provided by the GP cofactor. Acyl ligands are rarely encountered in bioinorganic chemistry, and their coexistence with thiolato ligands defines a novel platform from the perspective of homogeneous catalysis. The new structural information provided sufficient information to enable the design of a first generation model for this active site.
In terms of retrosynthetic analysis, the structure of the Fe center at the active site suggested that thioesters would oxidatively add to Fe(0) reagents. The interaction of thioesters with metal complexes has been intermittently investigated, including studies suggesting that this interaction is of prebiotic significance. The oxidative addition of a thioester has been established for rhodium(I) complexes. Our approach focused on the use of a donor-functionalized thioester, which upon oxidative addition would simultaneously deliver the Lewis base, thiolate, and acyl groups. To simplify the analysis of the synthetic studies, we chose to use a phosphine in place of the nitrogen heterocycle, since phosphorus NMR analysis provides a convenient means to monitor reactive intermediates. The probe thioester was efficiently generated by condensation of 2-diphenylphosphinobenzoic acid and benzenethiol. Related phosphine thioesters are known but have been only lightly studied.
We found that diiron nonacarbonyl, a source of the reactive iron tetracarbonyl entity, reacted readily with the phosphine thioester to give the targeted acyliron(II) thiolate. This complex readily and reversibly decarbonylates to give the diiron dithiolato diacyl tricarbonyl. This result highlights the advantages of the protein sheath that suppresses dimerization.
In summary, through this project, we have shown that donor-functionalized thioesters oxidatively add to Fe(0) to afford acyl ferrous thiolates, which provide models for the inhibited active site of Hmd. In view of the assembly strategy described here, functional modeling of this active site appears highly feasible. More broadly, the work shows that thioesters are precursors to versatile organometallic platforms.
