Ryan D. Bethel

Biomimetic chemistry: Merging the old with the new

The classic organometallic compound ferrocene has been combined with a unique diiron unit in the latest synthetic analogue of an enzyme active site, achieving the three functionalities needed for a working model of diiron hydrogenase, itself of ancient origin.

Carbon monoxide induced reductive elimination of disulfide in an N-heterocyclic carbene (NHC)/thiolate dinitrosyl iron complex (DNIC).

Dinitrosyliron complexes (DNICs) are organometallic-like compounds of biological significance in that they appear in vivo as products of NO degradation of iron-sulfur clusters; synthetic analogues have potential as NO storage and releasing agents. Their reactivity is expected to depend on ancillary ligands and the redox level of the distinctive Fe(NO)2 unit: paramagnetic {Fe(NO)2}(9), diamagnetic dimerized forms of {Fe(NO)2}(9) and diamagnetic {Fe(NO)2}(10) DNICs (Enemark-Feltham notation). The typical biological ligands cysteine and glutathione themselves are subject to thiolate-disulfide redox processes, which when coupled to DNICs may lead to intricate redox processes involving iron, NO, and RS(-)/RS•. Making use of an N-heterocyclic carbene-stabilized DNIC, (NHC)(RS)Fe(NO)2, we have explored the DNIC-promoted RS(-)/RS• oxidation in the presence of added CO wherein oxidized {Fe(NO)2}(9) is reduced to {Fe(NO)2}(10) through carbon monoxide (CO)/RS• ligand substitution. Kinetic studies indicate a bimolecular process, rate = k [Fe(NO)2](1)[CO](1), and activation parameters derived from kobs dependence on temperature similarly indicate an associative mechanism. This mechanism is further defined by density functional theory computations. Computational results indicate a unique role for the delocalized frontier molecular orbitals of the Fe(NO)2 unit, permitting ligand exchange of RS• and CO through an initial side-on approach of CO to the electron-rich N-Fe-N site, ultimately resulting in a 5-coordinate, 19-electron intermediate with elongated Fe-SR bond and with the NO ligands accommodating the excess charge.

Bioinorganic chemistry: Enzymes activated by synthetic components

Synthetic analogues of the catalytic subsite of the hydrogen-producing enzyme HydA1 have been disappointingly inactive. The incorporation of such analogues into the enzymes active site reveals the requirements for activity. See Letter p.66 [FeFe]-hydrogenases are metalloenzymes involved in microbial energy metabolism in various bacteria and algae, capable of extremes of catalytic performance that would be extremely useful if translated into a means of producing and using hydrogen in fuel cells. In these enzymes catalysis occurs at a unique di-iron centre containing a bridging dithiolate ligand, three CO ligands and two CN− ligands. In this manuscript, the authors show that three synthetic mimics of this di-iron centre can be loaded onto the [FeFe]-hydrogenase maturation protein HydF and then transferred to the algal variant apo-HydA1. Full activation of HydA1 was achieved only with the HydF hybrid protein that contained the mimic with an azadithiolate bridge, confirming the presence of this ligand in the active site of native [FeFe]-hydrogenases. This is the first example of controlled metalloenzyme activation using the combination of a specific protein scaffold and active-site synthetic analogues.

The modular assembly of clusters is the natural synthetic strategy for the active site of [FeFe] hydrogenase.

The recognition of the presence of iron–sulfur clusters and their electron shuttling roles in redox-active enzymes is one of the giant steps in a march of almost 80 years towards understanding the enzymes that control hydrogen metabolism in microorganisms of ancient origin: hydrogenases. The versatility of the Fe/S/SR combination in structural and physical properties has been well-established through synergistic studies evolving from the laboratories of chemists, biochemists, and biophysicists. 3] The remarkable coincidences of FeS cluster reactivities in vitro (using synthetic analogues) and in vivo (in FeS cluster-containing proteins) impress regarding the minor role of the protein in determining the existence of the FeS clusters, and have led to proposals that small chunks of iron sulfide minerals might have been the first catalysts on planet earth. Their eventual incorporation into proteins led to such sophisticated constructs as are found in the inorganic/organometallic natural products shown in Scheme 1. 5–8] Clearly, the presence and alignment of multiple FeS clusters observed in the protein crystal structures of [NiFe] and [FeFe] hydrogenases (H2ases) can only be interpreted as the electron-transfer routes that connect the active sites to the electron-donor or -acceptor unit docked into the exterior of the protein. In the [FeFe] H2ase active site (Scheme 1e), one typical 4Fe4S cluster is “hard-wired” or directly attached to an unusual 2Fe subsite through a cysteinyl bridge. In this way, the composition of the “H-cluster”, the hydrogenproducing cluster of [FeFe] H2ase, resembles that of sulfite reductase (Scheme 1g) or acetyl-CoA synthase (Scheme 1 f); the 4Fe4S cluster has been called upon to serve as a redoxvariable metallothiolate ligand via its cysteinyl sulfur that bridges to the 2Fe subsite. That the 2Fe portion of the Hcluster is a genuine organometallic species, replete with carbon monoxide, cyanide, and a previously biologically unknown dithiolate cofactor, coupled with the impressive rate of the [FeFe] H2ase catalysis of H2 production from mild potential electrons and water as proton source, has brought global attention of chemists in search of an optimal synthetic analogue of the active site, without protein, as prospective molecular electrocatalysts for hydrogen production. An engaging and difficult challenge has been biosynthesis issues: How does nature generate and manage CN and CO, known to poison metal sites if uncontrolled? How is the azadithiolate that connects the irons within the 2Fe subsite made? How is the H-cluster assembled? Does a 6Fe supercluster precede and extrude the 2Fe subsite, or is the assembly modular? Insights into the first two questions have been gained in recent discoveries of gene products utilizing radical SAM (S-adenosyl methionine) pathways that result in degradation of tyrosine into p-cresol and the diatomic ligands, CO and CN , the latter presumably though a glycyl radical. Guidance to answers to the latter two questions is the focus of this Highlight on a structural report from Mulder, Peters, Broderick et al. , and additional biosynthetic and spectroscopic results on the nature of the 2Fe2S subsite precursor. 14] As to the question “How is the H-cluster assembled?”, the trivial answer of “very carefully” is without a doubt correct. Mulder et al. have been able to obtain the [FeFe] H2ase (also known as HydA) protein as expressed in the absence of the HydE, HydF, and HydG proteins required for the synthesis of the 2Fe subsite and the maturation of the enzyme into active form. The immature protein, produced without the accessory proteins and known as HydA, was derived from the Chlamydomonas reinhardtii green alga and expressed in E. coli ; its X-ray crystal structure was determined and compared to those of the holoprotein crystallized from C. pasteurianum and Desulfovibrio desulfuricans. Whereas both latter structures show the full H-cluster in the form of cysteine-bridged subsites, that is, 4Fe4S(m-SCys)2Fe, only the 4Fe4S cluster is found in HydA. The structure of HydA shows the already-present 4Fe4S cluster resides in a cavity at the end of a channel (8–15 wide and 25 long; Figure 1). Overlays of the structures of the immature or apoprotein that lacks the 2Fe subsite with the complete or [*] R. D. Bethel, M. L. Singleton, Prof. M. Y. Darensbourg Department of Chemistry, Texas A&M University College Station, TX 77843 (USA) Fax: (+ 1)979-845-0158 E-mail: marcetta@chem.tamu.edu Homepage: http://www.chem.tamu.edu/rgroup/marcetta/

Regioselectivity in ligand substitution reactions on diiron complexes governed by nucleophilic and electrophilic ligand properties.

The discovery of a diiron organometallic site in nature within the diiron hydrogenase, [FeFe]-H2ase, active site has prompted revisits of the classic organometallic chemistry involving the Fe-Fe bond and bridging ligands, particularly of the (μ-SCH2XCH2S)[Fe(CO)3]2 and (μ-SCH2XCH2S)[Fe(CO)2L]2 (X = CH2, NH; L = PMe3, CN(-), and NHCs (NHC = N-heterocyclic carbene)), derived from CO/L exchange reactions. Through the synergy of synthetic chemistry and density functional theory computations, the regioselectivity of nucleophilic (PMe3 or CN(-)) and electrophilic (nitrosonium, NO(+)) ligand substitution on the diiron dithiolate framework of the (μ-pdt)[Fe(CO)2NHC][Fe(CO)3] complex (pdt = propanedithiolate) reveals the electron density shifts in the diiron core of such complexes that mimic the [FeFe]-H2ase active site. While CO substitution by PMe3, followed by reaction with NO(+), produces (μ-pdt)(μ-CO)[Fe(NHC)(NO)][Fe(CO)2PMe3](+), the alternate order of reagent addition produces the structural isomer (μ-pdt)[Fe(NHC)(NO)PMe3][Fe(CO)3](+), illustrating how the nucleophile and electrophile choose the electron-poor metal and the electron-rich metal, respectively. Theoretical explorations of simpler analogues, (μ-pdt)[Fe(CO)2CN][Fe(CO)3](-), (μ-pdt)[Fe(CO)3]2, and (μ-pdt)[Fe(CO)2NO][Fe(CO)3](+), provide an explanation for the role that the electron-rich iron moiety plays in inducing the rotation of the electron-poor iron moiety to produce a bridging CO ligand, a key factor in stabilizing the electron-rich iron moiety and for support of the rotated structure as found in the enzyme active site.

A reduced 2Fe2S cluster probe of sulfur-hydrogen versus sulfur-gold interactions.

The Ph3 PAu(+) cation, renowned as an isolobal analogue of H(+) , was found to serve as a proton surrogate and form a stable Au2 Fe2 complex, [(μ-SAuPPh3 )2 {Fe(CO)3 }2 ], analogous to the highly reactive dihydrosulfide [(μ-SH)2 {Fe(CO)3 }2 ]. Solid-state X-ray diffraction analysis found the two SAuPPh3 and SH bridges in anti configurations. VT NMR studies, supported by DFT computations, confirmed substantial barriers of approximately 25 kcal mol(-1) to intramolecular interconversion between the three stereoisomers of [(μ-SH)2 {Fe(CO)3 }2 ]. In contrast, the largely dative SAu bond in μ-SAuPPh3 facilitates inversion at S and accounts for the facile equilibration of the SAuPPh3 units, with an energy barrier half that of the SH analogue. The reactivity of the gold-protected sulfur atoms of [(μ-SAuPPh3 )2 {Fe(CO)3 }2 ] was accessed by release of the gold ligand with a strong acid to generate the [(μ-SH)2 {Fe(CO)3 }2 ] precursor of the [FeFe]H2 ase-active-site biomimetic [(μ2 -SCH2 (NR)CH2 S){Fe(CO)3 }2 ].