A Highly Reactive P450 Model Compound I

The substrate oxygenation reactions mediated by cytochrome P450 proteins comprise many key transformations in steroid biosynthesis and drug metabolism. Significant advances in heme-protein enzymology,¹ the characterization of synthetic model compounds² and computational approaches³ have pointed to a reactive oxoiron(IV)porphyrin π-radical-cation in the consensus mechanism. Compound I model compounds, such as the thoroughly characterized [OFe^IVTMP⁺],^4,5 have exhibited a full range of oxygen transfer reactions.⁶ However, the low kinetic reactivity of [OFe^IVTMP⁺] has left unanswered just how the protein manages to generate a sufficiently reactive intermediate and what that reactive species is.^7–10 We have shown that cationic oxomanganese(V) porphyrin complexes such as OMn^V-4-TMPyP display very fast rates for oxygen transfer reactions in aqueous solution.^11,12 Here we report the detection and kinetic characterization of [OFe^IV-4-TMPyP⁺], which shows extraordinary rates for C-H hydroxylations.

The reaction of Fe^III-4-TMPyP with mCPBA was analyzed by stopped-flow spectrophotometry, as we have previously described.^11,13 As can be seen in the UV-Vis spectral time course (Figure 1), the starting iron(III) porphyrin spectrum (λ_max 421 nm) was completely transformed within 20 ms into that of a new intermediate, 1, displaying a weak, blue-shifted Soret band at 402 nm and an absorbance at 673 nm, typical of a porphyrin π-radical-cation.^4,14 Accordingly, we assign the [OFe^IV-4-TMPyP⁺] formulation to 1. Global analysis afforded a second-order rate constant k₁ = 1.59(±0.06)×10⁷ M⁻¹s⁻¹ for the appearance of 1. In a slower, subsequent phase of the reaction (Scheme 1), 1 transformed into the well-characterized OFe^IV-4-TMPyP (λ_max 427 nm; k₂=8.8(±0.1) s⁻¹).^15–17 Isosbestic points for the conversion of 1 to OFe^IV-4-TMPyP were apparent at 413 and 570 nm. This transformation was more rapid at higher pH, in contrast to the behavior of the corresponding OMn^V-4-TMPyP complex.^11,18

Figure 1.

A. UV-vis transients observed upon mixing of 10 μM Fe^III-4-TMPyP with 3 equiv mCPBA at pH 4.7, 10.0 °C. B. Data obtained at (a) 402 nm and (b) 427 nm and the superimposed global fit.

Scheme 1.

Kinetic partitioning of [OFe^IV-4-TMPyP⁺] (1) between spontaneous decay to OFe^IV-4-TMPyP (k₂) and substrate hydroxylation (k₃) resulting in the formation of Fe^III-4-TMPyP.

Intermediate 1 was found to be very highly reactive toward C-H bond hydroxylations. Slow addition of mCPBA to a solution of 5 mM xanthene and 50 μM Fe^III-4-TMPyP was monitored by ¹H- and ¹³C-NMR. The major product, 9-xanthydrol (H-9, δ 5.75; C-9, 62 ppm), was formed in 90% yield, based on added oxidant (Figure S1). A small amount of xanthone was also formed. The Fe^III catalyst was recovered with negligible decomposition. Significantly, when the oxidation was carried out in 47.5% ¹⁸OH₂, ESI mass spectrometry of the product 9-xanthydrol indicated 21% ¹⁸O-incorporation (m/e 199/197 = 0.27). This partial exchange is consistent with an oxygen rebound reaction scenario with ~50% dilution of the ¹⁶O-peroxyacid-derived oxygen with ¹⁸Owater occupying the second axial ligation site.^19,20 Xanthene:xanthene-d₂ (1:1) revealed a modest kinetic isotope effect, k_H/k_D = 2.1 (m/e 197/198).

Since the oxidation of Fe^III-4-TMPyP by mCPBA was so fast, we could conveniently monitor the kinetics of the reaction of 1 with xanthene using a single-mixing stopped-flow experiment (Figure 2A). Global analysis of the spectral data produced a rate constant k₃ = 3.6(±0.3)×10⁶ M⁻¹s⁻¹. A similar value was obtained by monitoring 673 nm (Figure 2A). Control experiments showed that OFe^IV-4-TMPyP had negligible reactivity toward xanthene over this time course.

Figure 2.

A. Observed rate, k_obs at 673 nm, vs. [xanthene] for its oxidation by 1 at pH 4.7, 14.5 °C; k₃ = 2.8 (±.3)×10⁶ M⁻¹s⁻¹. B. Correlation of k₃ vs. C-H bond dissociation energy.

Reactions of fluorene-4-carboxylic acid, 4-isopropyl-, and 4-ethylbenzoic acid with 1 also gave very high rates for C-H hydroxylation with second-order rate constants of 1.80(±.06)×10⁵ M⁻¹s⁻¹, 9.0(±.2)×10³ M⁻¹s⁻¹ and 7.5(±.1)×10³ M⁻¹s⁻¹, respectively. The corresponding alcohol products were formed in each case. These reaction rates are remarkably fast and, as can be seen in Figure 2B, the rates correlated well with the scissile C-H bond energy, indicating a homolytic hydrogen abstraction transition state.²¹ Reaction of 1 with bromide ion afforded hypobromite with k₃=3.14 (±.04) × 10⁵ M⁻¹s⁻¹.¹³

The results show that [OFe^IV-4-TMPyP⁺] (1), which is accessed chemically from a peroxide precursor, is orders of magnitude more reactive than the well-studied tetramesityl analogs.^4,6,7,9 By mapping the rate constants observed here for C-H bond cleavage by 1 onto the Brønsted-Evans-Polanyi relationship^21,22 for similar substrates (TOC graphic), the BDE of H-OFe^IV-4-TMPyP can be estimated to be ~100 kcal/mol, much stronger than the values estimated for H-OFe^IVTMP (92 kcal/mol)⁷ and H-OFe^IIITPFPP (~86 kcal/mol).²³ Indeed, 1 showed a rate constant for ethylbenzoic acid hydroxylation similar to that reported recently by Newcomb, et al. for ethyl benzene with CYP119 and chloroperoxidase compound I.²⁴

The interesting question is why should synthetic OFe^IV-porphyrin π-radical-cations have such high variability in their intrinsic reactivities. We have shown that electron-withdrawing meso-substituents dramatically decrease the reactivity of oxoMn^V porphyrins.¹¹ The effects have been traced to stabilization of the singlet d² ground state with respect to the triplet and quintet manifolds, stabilization of the porphyrin HOMO and the unique centrosymmetric trans-dioxo ligand arrangement.²⁵ Clearly, the situation for iron described here is markedly different. A relatively high redox potential for porphyrin ring oxidation is expected for Fe^III-4-TMPyP. Further, efficient transmission of the electron deficit to the metal core is assured by the large orbital coefficients on the meso carbon and the pyrrole nitrogens in the a_2u porphyrin HOMO. Indeed, the pK_a of the axial water in diaquaFe^III-4-TMPyP is only 5.5 while the corresponding acidity of diaquaFe^IIITMPS is 7.5.²⁶ DFT calculations have shown that reaction barriers in oxo-transfer reactions may be lowered by decreasing the energy gaps to higher spin configurations for both manganese^25,27 and iron.^22,28 A decrease in the σ-donor properties of the coordinating pyrrole nitrogens due to the meso-substituent would be expected to decrease the energy level of the iron d_(x²−y²) orbital. The high-spin state of an unsubstituted oxoFe^IV porphyrin radical-cation, with the iron d_(x²−y²) orbital singly occupied, has been calculated to be only 14 kcal/mol above the nearly degenerate doublet and quartet ground state structures.²⁹ We suggest that the high kinetic reactivity observed here experimentally for [OFe^IV-4-TMPyP⁺] (1) may result from both a low-lying a_2u porphyrin HOMO and facilitated spin state crossing phenomena in the course of reaction. More generally, the results suggest that even subtle charge modulation at the heme active site of P450 to accomplish a similar tuning of state crossing energies or lower stability of the porphyrin π-radical-cation could result in high reactivity of a cytochrome P450 compound I.