Book Description

The foci of this dissertation are: 1) combined use of spectroscopies for mechanistic understanding of the oxygen reactions of various non-heme iron enzymes and related model complexes, and 2) the development of the recently described nuclear vibrational resonance spectroscopy (NRVS) coupled with density functional calculations (DFT) for characterization of non-heme iron enzyme intermediates. Binuclear non-heme iron enzymes are involved in many medically and industrially important processes such as DNA synthesis by ribonucleotide reductase (RNR), conversion of methane to methanol by methane monooxygenase (MMO), fatty acid desaturation by [Delta]9 desaturase, iron storage and homeostasis by ferritins, degradation of aromatic compounds by various bacterial monooxygenases (ToMO, T4MO, etc.) and antibiotic biogenesis by p-aminobenzoate N-oxygenase (AurF), etc. Interestingly, these diverse reactions typically begin with O2 reacting with a biferrous active site, coordinated by highly conserved protein ligands (ExxH motifs) in four [Alpha]-helix bundles. Moreover, spectroscopically and chemically similar intermediates can be detected in many of the enzyme systems. The best studied in this family are RNRs, where biferric peroxo intermediates (P and P'), and the high-valent Fe(III)Fe(IV) intermediate X have been stabilized and spectroscopically characterized in wt and numerous variants. De novo designed four [Alpha]-helix bundles have been synthesized (the ~140 amino acid dui ferri (DF) peptide family) and are good models for binuclear non-heme iron enzymes. These systems provide a protein environment and can be viewed as a bridge between inorganic model complexes and native proteins. The pseudo-symmetric single chain version (DFsc) coordinates two ferrous ions by two His and four Glu amino acid residues. Circular dichroism (CD), magnetic CD (MCD) and variable-temperature variable-field MCD (VTVH MCD) show that this "active site" in DFsc has a 4-coordinate and 5-coordinate (4C+5C) geometry that is weakly antiferromagnetically coupled (J [approximately equal to] --2 cm-1) indicative of [Mu]1,3 carboxylate bridges, highly similar to RNR biferrous structures. Extended x-ray absorption fine structure (EXAFS) data are consistent with this assignment and show that one terminal carboxylate residue coordinates in a bidentate fashion. Changes in the CD/MCD/VTVH MCD and EXAFS spectra in the Y51L and E11D variants show that the 4C site is proximal to (but not bound by) Y51 and the bidentate carboxylate is coordinated to the 5C iron. Open coordination positions on both irons allow for dioxygen to react rapidly with the biferrous site. The reaction of biferrous DFsc with dioxygen yields a 520 nm ([Epsilon] = [weak approximation to]1200 M-1cm-1) species with a formation rate of 2 s-1, again similar to RNR (the Class Ia RNR from Escherichia coli has a dioxygen reaction rate of ~1 s-1, however the first species formed (intermediate P) has [Lambda]max = 700 nm). The resonance Raman (rR) spectrum obtained by excitation into the 520 nm feature in DFsc (and the E11D variant) proves this chromophore arises from a Tyr to ferric charge transfer (CT) transition. The 520 nm feature is lost by substitution of Y51 but not Y18, thus Y51 binds to the site after reaction with dioxygen. Subsequent binding of Y51 functions as an internal spectral probe of the dioxygen reaction and as a proton source that would promote loss of hydrogen peroxide. Coordination by a ligand that functions as a proton source could be a structural mechanism used by natural binuclear iron enzymes to drive their reactions past peroxo biferric level intermediates. RNR's can be divided into 3 major classes based on the radical generating machinery. Class I RNR's utilize a dimetal cofactor that reacts with dioxygen and can be subdivided into Classes Ia, Ib and Ic based on sequence homology and metal dependency. Class Ia enzymes are the best studied an present in higher organisms including human (host) while Class Ib enzymes are typically found in pathogens. CD, MCD and VTVH MCD data on biferrous loaded Class Ib RNR from Bacillus cereus allow assignment of the active site as 4C+5C in solution, resolving discrepancies from available crystal structures. Differences in the zero-field splitting parameters (D and E) and magnetic coupling extracted from fits to the VTVH MCD data can be ascribed to differences in the bridging carboxylate conformations. FeII loading, monitored by CD, shows cooperative binding with Kd 100 mM, significantly stronger that the metal binding in Class Ia. This provides the pathogen a competitive advantage relative to host in physiological, iron-limited environments Returning to Class Ia, the recently discovered intermediate P' notably lacks structural definition. This is mainly due to the lack of spectroscopic handles from which to obtain the needed experimental data. What is know, however, is that this species directly forms intermediate X and is directly derived from the well-defined intermediate P. Spectroscopically, P' has Mössbauer isomer shifts ([lowercase Delta] = 0.52 and 0.45 mm/s) that are significantly lower than the cis-[Mu]1,2 peroxo P ([lowercase Delta] = 0.63 mm/s) and lacks the ~700 nm peroxo to ferric CT suggesting some change in coordination mode or protonation may be involved in P -- P'. Comparisons of the reduced and oxidized crystal structures show differences in carboxylate coordination modes and water binding that must occur at some stage along the reaction coordinate. All of these potential structural perturbations were systematically incorporated into computational models of the intermediate site and correlated with experimental data using density functional theory (DFT). Two potential reaction pathways consistent with available experimental data were found. The first involves water addition to Fe1 of the cis-[Mu]-1,2 peroxo intermediate P causing opening of a bridging carboxylate to form intermediate P' which has an increased electron affinity and is activated for proton-coupled electron transfer to form the Fe(III)Fe(IV) intermediate X. While the second, more energetically favorable pathway, involves addition of a proton to a terminal carboxylate ligand in the site which increases the electron affinity and triggers electron transfer to form X. Vibrational characterization could, in principle, distinguish these pathways. However, the lack of a reasonably intense chromophore precludes rR experiments. The recently available method of nuclear vibrational resonance spectroscopy (NRVS) does not have these chromophoric constraints and can provide the needed vibrational data for P'--and many other "spectroscopically challenged" intermediates in non-heme iron biochemistry. The vibrations enhanced in NRVS are typically lower in energy and differ from those observed in rR, thus studies on well defined model complexes are needed prior to intermediate studies. A series of mononuclear Fe(IV)=O have been characterized by NRVS coupled with DFT calculations to define NRVS spectral assignments and set a foundation for vibrational characterization of non-heme iron enzyme intermediates. These studies show that the NRVS spectrum is rich in structural information. Of the four Fe(IV)=O models, supported by the 1, 4, 8, 11-tetramethyl-1,4,8,11-tetraazacyclotetradecane (TMC); N, N-bis(2-pyridylmethyl)-N-bis(2-pyridyl) methylamine (N4Py); N-benzyl-N, N', N'-tris(2-pyridylmethyl)-1,2-diaminoethane (BnTPEN); and 1,1,1-tris{2-[N(2)-(1,1,3,3-tetramethylguanidino)]ethyl}amine (TMG3tren) ligand sets, only the trigional bipyramidal geometry (relative to the 6C approximatly C4v geometry of TMC, N4Py and BnTPEN) enforced by the TMG3tren ligand affords a high-spin species. Isotope sensitive Fe-O stretches are observed for all complexes at 820 to 831 cm-1. However, at lower energy (