Oxygen‐18 Labeling Reveals a Mixed Fe−O Mechanism in the Last Step of Cytochrome P450 51 Sterol 14α‐Demethylation

Abstract

The 14α‐demethylation step is critical in eukaryotic sterol biosynthesis, catalyzed by cytochrome P450 (P450) Family 51 enzymes, e.g. with lanosterol in mammals. This conserved 3‐step reaction terminates in a C‐C cleavage step that generates formic acid, the nature of which has been controversial. Proposed mechanisms involve roles of P450 Compound 0 (ferric peroxide anion, FeO2¯) or Compound I (perferryl oxygen, FeO3+) reacting with either the aldehyde or its hydrate, respectively. Analysis of 18O incorporation into formic acid from 18O2 provides a means of distinguishing the two mechanisms. Human P450 51A1 incorporated 88% 18O (one atom) into formic acid, consistent with a major but not exclusive FeO2¯ mechanism. Two P450 51 orthologs from amoeba and yeast showed similar results, while two orthologs from pathogenic trypanosomes showed roughly equal contributions of both mechanisms. An X‐ray crystal structure of the human enzyme showed the aldehyde oxygen 3.5 Å away from the heme iron. Experiments with human P450 51A1 and H218O yielded primarily one 18O atom but 14% of the formic acid product with two 18O atoms, indicative of a minor contribution of a Compound I mechanism. LC‐MS evidence for a Compound 0‐derived Baeyer‐Villiger reaction product (a 14α‐formyl ester) was also found.

Oxygen-18 Labeling Defines a Ferric Peroxide (Compound 0) Mechanism in the Oxidative Deformylation of Aldehydes by Cytochrome P450 2B4

“…Although it has not been reported in the case of P450 2B4, a Baeyer–Villiger-type reaction yielding a one-carbon homologated ester ( 4 in Scheme C) might be possible when Compound 0 is acting as a reactive iron species. Indeed, some P450s have been proposed to involve a Baeyer–Villiger mechanism. , In a Compound I mechanism, one hydrogen atom of a gem -diol is abstracted (but not from the alcohol). Electron transfer yields the carbocation, which then rearranges to release formic acid and form an olefin product. , On the basis of these proposed mechanisms, Compound I vs Compound 0 reactions can be distinguished by identifying the source of oxygen incorporated into the formic acid product.…”

“…Rates of P450 2B4-catalyzed deformylations are relatively slow (0.1–1 min –1 ), which is problematic in obtaining an adequate product for the analysis. To improve the sensitivity, we revised the procedures for extraction and derivatization of formic acid using the diazo reagent 1-diazo-3-(3-pyridinyl)propane ( 5 ) , as detailed in the Supporting Information. Additionally, a P450 17A1-catalyzed 17α-hydroxylation of progesterone reaction was included in all of the samples as an internal control to accurately calculate the percentage of 18 O incorporation in each sample (i.e., from 18 O analysis of the product 17α–OH progesterone), thereby correcting for any artifactual loss of 18 O 2 (e.g., gas leaks).…”

“…As mentioned, the possibility existed that the Compound 0 mechanism could include a Baeyer–Villiger rearrangement prior to C–C scission. , To address this mechanistic question, we synthesized the proposed Baeyer–Villiger intermediates of each of the three substrates ( 4 , 10 , and 11 ). Because the putative intermediate expected from citronellal ( 4 ) displayed poor ionization character and minimal UV activity, our analysis was based on experiments with only the phenylpropionaldehyde model substrates.…”

“…The 18 O labeling approach that we have employed here is applicable to other P450-catalyzed oxidative deformylations, e.g., P450 19A1-catalyzed androstenedione aromatization (vide supra) and P450 51-catalyzed lanosterol 14α-demethylation. , Although Ortiz de Montellano and co-workers employed 18 O 2 to study an oxidative terminal side chain carbon cleavage reaction of cholesterol by bacterial P450 125A1, direct evidence of Compound 0 contribution was not presented, and this question could also be addressed by employing deuterium-labeled formic acid analysis. In principle, the approach can be applied to C–C bond scissions that yield acetic acid (e.g., P450 17A1-catalyzed 17α,20-lyase reaction and P450 1A2-catalyzed side-chain cleavage of nabumetone), , but with α-ketols, the 18 O labeling results are not unambiguous. ,, …”

“…We attribute the differences in utilization of Compound 0 and Compound I mechanisms (Scheme ) to the relative rates for the ferric peroxy anion to be protonated versus nucleophilic attack on an electrophilic carbonyl group, but the roles of individual amino acid residues are only a matter of speculation. In the case of human and rat P450 51A1, there is evidence for a Baeyer–Villiger rearrangement (Scheme C), , but this product was not observed here for these P450 2B4 reactions. We are also unable to define the elements of the P450s that control the Baeyer–Villiger rearrangement at this time.…”