The Mechanism of the Acyl-Carbon Bond Cleavage Reaction Catalyzed by Recombinant Sterol 14α-Demethylase of Candida albicans (Other Names Are: Lanosterol 14α-Demethylase, P-45014DM, and CYP51)

The Candida albicans sterol 14α-demethylase gene (P-45014DM, CYP51) was transferred to the yeast plasmid YEp51 placing it under the control of the GAL10 promoter. The resulting construct (YEp51:CYP51) when transformed into the yeast strain GRF18 gave a clone producing 1.5 μmol of P-450/liter of culture, the microsomal fraction of which contained up to 2.5 nmol of P-450/mg of protein. Two oxygenated precursors for the 14α-demethylase, 3β-hydroxylanost-7-en-32-al and 3β-hydroxylanost-7-en-32-ol, variously labeled with 2H and 18O at C-32 were synthesized. In this study the conversion of [32-2H,32-16O]- and [32-2H,32-18O]3β-hydroxylanost-7-en-32-al with the recombinant 14α-demethylase was performed under 16O2 or 18O2 and the released formic acid analyzed by mass spectrometry. The results showed that in the acyl-carbon bond cleavage step (i.e. the deformylation process) the original carbonyl oxygen at C-32 of the precursor is retained in formic acid and the second oxygen of formate is derived from molecular oxygen; precisely the same scenario that has previously been observed for the acyl-carbon cleavage steps catalyzed by aromatase (P-450arom) and 17α-hydroxylase-17,20-lyase (P-450,CYP17). In the light of these results the mechanism of the acyl-carbon bond cleavage step catalyzed by the 14α-demethylase is considered.


O at C-32 were synthesized. In this study the conversion of [32-2 H,32-16 O]-and [32-2 H,32-18 O]3␤-hydroxylanost-7-en-32-al with the recombinant 14␣-demethylase was performed under 16 O 2 or 18 O 2 and the released formic acid analyzed by mass spectrometry. The results
showed that in the acyl-carbon bond cleavage step (i.e. the deformylation process) the original carbonyl oxygen at C-32 of the precursor is retained in formic acid and the second oxygen of formate is derived from molecular oxygen; precisely the same scenario that has previously been observed for the acyl-carbon cleavage steps catalyzed by aromatase (P-450 arom ) and 17␣-hydroxylase-17,20-lyase (P-450 17␣ ,CYP17). In the light of these results the mechanism of the acyl-carbon bond cleavage step catalyzed by the 14␣-demethylase is considered.
Our earlier studies on the removal of C-19 of androgens in the formation of estrogens (1)(2)(3) and of the 14␣-methyl group of lanosterol during sterol biosynthesis (Scheme I, Conversion 134) (4) raised the possibility that these seemingly unrelated conversions may occur by closely related mechanisms involving three steps as shown in Reaction 1.
These studies also indicated that in each case the same catalyst was responsible for all three reactions, and this feature was firmly established through genetic studies and purification to homogeneity of the two enzymes, aromatase (P-450 arom ) (5,6) and lanosterol 14␣-demethylase (P-450 14DM ) (7,8). The third step in estrogen biosynthesis has aroused much interest (9,10) and the current view of the mechanism is influenced by our 18 O labeling experiments (2,3), which highlighted the novel nature of the process, leading to the proposal that the reaction involves an acyl-carbon cleavage represented by Reaction 2 (9).
Although all the experimental findings available to date on the C-C bond cleavage step in 14␣-demethylation, for example the requirement for NADPH plus O 2 for the reaction and release of the C 1 unit as formate, could be explained (3,9) by the reaction of Reaction 2, the direct scrutiny of the hypothesis has not been possible hitherto due to the unavailability both of appropriately labeled 18 O substrates and an enzyme preparation that produced sufficient formic acid for accurate 18 O isotope analysis.
The present paper describes a satisfactory resolution of these difficulties and reports on the status of oxygen during the C-C bond cleavage step catalyzed by lanosterol 14␣-demethylase (334).

EXPERIMENTAL PROCEDURES
Materials-Isotopically enriched 18 O 2 (97%) admixed with 2 volumes of argon was obtained from Isogas Limited, Croydon, Surrey and H 2 18 O was from MSD Isotopes, Montreal, Canada. Dry redistilled solvents were used, and the petroleum ether used was that with a boiling range of 60 -80°C. Diazotoluene was prepared from N-benzyl-N-nitrosotoluene-4-sulfonamide (11). The phrase "in the usual manner" indicates that the reaction mixture was poured into water, the product extracted with ethyl acetate, the combined organic extracts washed with water, dried over anhydrous sodium sulfate, and the solvent removed under reduced pressure. All the intermediates used for the synthesis of 3␤acetoxylanost-7-en-32-onitrile (5) gave expected melting points, R F values, IR spectra, as well as mass spectrometric data.
Gas Chromatography-Mass Spectrometry-Isotopic distributions in the labeled substrates were determined by direct introduction probe mass spectrometric analyses of either the underivatized materials or their trimethylsilyl derivatives and are corrected for 13 C natural abundance. All the mass spectra were recorded in the electron impact positive ion mode.
The analysis of benzyl formate, prepared from the enzymatically produced formic acid, was performed by gas chromatography-mass spectrometry using a Hewlett-Packard 5890/VG TS-250 and a 30 m ϫ 0.32 mm inner diameter column of DB17 with splitless injection (12).
The experimentally determined value for 13 C natural abundance (12.4%) for benzyl formate was used to correct all the peaks between m/z 137-141, due to other isotopomers. The distribution of the isotopomers in benzyl formate was measured by comparison of the normalized ion signal areas determined by selected ion recording and corroborated by recording the full spectrum.
Preparation of Microsomes from Pig and Rat Liver-Pig or rat livers were cut up into small pieces and suspended in approximately 2.5 times their volume of 100 mM potassium phosphate (also containing 2 mM glutathione, 1 mM EDTA, 4 mM magnesium chloride, 0.25 M sucrose, 0.25 mM phenylmethylsulfonyl fluoride, pH 7.4) and homogenized. The homogenate was centrifuged at 10,000 ϫ g for 30 min. The supernatant was subsequently spun at 10,5000 ϫ g for 1.5 h twice. The resulting microsomal pellet was resuspended in 0.1 M potassium phosphate buffer (1 mM glutathione, 0.1 mM EDTA, pH 7.4) to give a final concentration of 40 -60 mg ml Ϫ1 protein.
Recombinant DNA Manipulations-Our previous studies have employed a yeast expression system to express the Candida albicans CYP51 using the Saccharomyces cerevisiae phosphoglycerate kinase promotor in vector pW91P (15). Higher level expression was achieved by recombinant PCR 1 to allow transfer of CYP51 to YEp51 on a SalI/ HindIII fragment and expression from the GAL10 promotor (Fig. 1). The following oligonucleotides were used as outside primers: 1, 5Ј-AAACTCGACAATATGGCTATTGTTGAAACTG-3Ј annealing to positions 1-21 of the C. albicans CYP51 (16) and containing a 5Ј-added SalI site prior to the initiating methionine and 2, 5Ј-TGGCATATGCATTCT-GAGAGTTTCCTT-3Ј annealing to the 3Ј end at position 1098 -1125 of the CYP51 and containing the endogenous NsiI site present in the gene.
Recombinant PCR was used to replace the triplet 263 (CTG) with one encoding serine (TCT) in the S. cerevisiae host. Inside primers used in the PCR mutagenesis were: 1, 5Ј-AAAGAAATTAAATCTAGAA-GAGAA-3Ј and 2, 5Ј ACGTTCTCTTCTAGATTT AATTTCTTT-3Ј. In a first step two separate PCR reactions were performed using outside primer 1/inside primer 2 and inside primer 1/outside primer 2, respectively. The partially overlapping DNA fragments obtained were purified, mixed, and recombined in a subsequent PCR step using outside primers 1 and 2. PCR reactions were performed on a Perkin-Elmer DNA thermal cycler; conditions consisted of an initial 5 cycles of 1-min denaturation at 94°C, an annealing step for 4 min at 48°C, and an extension step for 3 min at 70°C, followed by 25 cycles of a denaturation step for 1 min at 94°C, an annealing step for 2 min at 55°C, and an extension step for 3 min at 72°C. PCR was undertaken using Pfu polymerase (Promega). Introduction of the mutation and maintenance of the authentic sequence was corroborated by DNA sequencing using Sequenase 2 (Amersham Corp.) after cloning the mutant SalI/NsiI fragment into YEp51 with ligation to the NsiI/HindIII fragment containing the C terminus and terminator regions of C. albicans CYP51. The restored CYP51 fragment was cloned directly into SalI/HindIII digested vector. All restriction enzymes and T4 DNA ligase were ob- 1 The abbreviation used is: PCR, polymerase chain reaction. SCHEME I. The sequence of reactions catalyzed by sterol 14␣demethylase. Although lanosterol is the physiological substrate for the enzyme from most sources, dihydrolanosterol (reduced at the 24,25double bond) as well as lanostane derivatives containing a ⌬ 7 -double bond also serve as substrates. With ⌬ 7 -substrates, 7,14-diene is formed following the C-C bond cleavage step. REACTION 1 REACTION 2 tained from Promega and the recommended conditions for use were applied.
Growth of Yeast for Heterologous Expression-Yeast transformants were grown at 28°C, 250 rpm with 250 ml of culture in 500-ml flasks. The medium used consisted of Difco yeast nitrogen base without amino acids supplemented with 100 mg/liter histidine and 2% (w/v) glucose as initial carbon source. Heterologous expression was induced when the glucose was exhausted at a cell density of approximately 10 8 cells/ml. The culture was left a further 4 h before the concentration of galactose was raised to 3% (w/v). After 20-h induction cells were harvested by centrifugation, resuspended in buffer containing 0.4 M sorbitol, 50 mM Tris-HCl, pH 7.4, and broken using a Braun disintegrator (Braun GmbH, Mesungen, Germany) with four bursts of 30 s together with cooling from liquid carbon dioxide. Cell debris was removed by centrifugation at 1500 ϫ g for 5 min using a bench centrifuge. The resulting supernatant was centrifuged twice at 10,000 ϫ g for 20 min to remove mitochondria and then at 100,000 ϫ g for 90 min to yield the microsomal pellet. This was resuspended using a Potter-Elvehjeim glass homogenizer at about 10 mg of protein/ml in the same buffer described above. P-450 concentration by reduced carbon monoxide difference spectroscopy was measured according to Ref. 17 and protein using a Sigma BCA kit.
Determination of Sterol 14␣-Demethylase Activity of the Microsomes-A solution of NADP ϩ (2 mg), glucose 6-phosphate (5 mg), and glucose-6-phosphate dehydrogenase (3 units) in 100 mM potassium phosphate buffer containing 0.1 mM EDTA, 1 mM glutathione, and 20% v/v glycerol (0.5 ml, pH 7.4) was incubated at 37°C for 20 min. To this was added microsomal protein (approximately 10 mg) and the volume made up to 1 ml with the above buffer. Following the addition of the 32-tritiated substrate (52 g, 1.62 Ci in 10 l of dimethylformamide), aliquots (0.1 ml) were removed (at intervals of 0, 5, 10, 30, and 60 min) and added to a mixture of dichloromethane (0.5 ml) and water (0.5 ml). The mixtures were immediately shaken and then centrifuged. The organic layer was discarded and further dichloromethane (2 ϫ 0.5 ml) was added and the above procedure repeated. To the resulting aqueous phase was added charcoal, the suspension shaken, left at 4°C for ϳ1 h, and finally centrifuged to remove the charcoal. The radioactivity of the aqueous layer was measured by liquid scintillation counting.
Large Scale Incubation, under Air, for the Isolation of C-32 as Formate-A solution of NADP ϩ (5 mg), glucose 6-phosphate (10 mg), and glucose-6-phosphate dehydrogenase (5 units) in 100 mM potassium phosphate buffer containing 0.1 mM EDTA, 1 mM glutathione, and 20% glycerol (3.2 ml, pH 7.4) was incubated at 37°C for 20 min. To the incubation mixture was then added yeast microsomes (20 mg of protein in 0.8 ml of the buffer) and 0.4 mg of appropriately labeled 3␤-hydroxylanost-7-en-32-al (6) or 3␤-hydroxylanost-7-en-32-ol (7), admixed with tracer amounts of the 32-tritiated derivative (2.25 Ci), in dimethylformamide (40 l). The mixture was shaken in air at 37°C for 50 min, and following acidification with 10% phosphoric acid (0.4 ml) the volatile fraction was collected by freeze-drying. The volatile fraction containing the biosynthetic formic acid was neutralized with 0.4 M sodium hydroxide (30 l) and the solution again subjected to freeze-drying. The residue containing sodium formate was dissolved in water (3 ϫ 100 l) and transferred to a small vial. After the removal of water by freeze-drying, the residue containing 0.05-0.15 mol of sodium formate (determined by the measurement of 3 H) was converted to benzyl formate and analyzed by gas chromatography-mass spectrometry as described previously (12).
Incubations under 18 O 2 Gas-A procedure similar to that given immediately above was employed except that 18 O 2 gas was used instead of air in the following manner. The vessel containing all the components but the substrate was evacuated, using a water pump, and flushed with argon. After two rounds of the preceding operation a solution of the substrate was added and the vessel immediately evacuated, flushed with 18 O 2 (isotopic purity: 97%)/argon (1:2 ratio by volume) and after closing the tap the incubation was performed as above.

Synthesis of 32-Labeled Precursors for Sterol 14␣-Demethylase and Preliminary Enzymic Studies-3␤-Acetoxylanost-7-
en-32-onitrile (5), obtained by a lengthy 10-step sequence as described by Barton and co-workers (13, 14), was the crucial intermediate used for the preparation of four isotopomers each of the 32-oxo (6) and 32-hydroxy (7) substrates. The main methodological improvement made in the synthetic protocol (Scheme II) was the use of diisobutylaluminum hydride, instead of LiAlH 4 in the original work for the conversion of the nitrile (5) into aldehyde (6), which decreased the reaction time from 72 to less than 0.5 h.
The two tritiated substrates (6e) and (7d) were used for the assay of the 14␣-demethylase activity by monitoring the release, in the medium, of 3 HCOOH from the aldehyde (6e) or 3 HCOOH plus 3 H 2 O from the hydroxy compound (7d). In the metabolism of the hydroxy compound (7d), tritium is released in water during the oxidation of the hydroxy into the aldehyde group and in formic acid during the subsequent C-C bond cleavage step converting the 32-oxo derivative (6) to the 7,14diene (see Scheme III, structure of the type 11). An oxidative activity in most preparations of 14␣-demethylase converts the initially produced formate into CO 2 and H 2 O. Our projected mechanistic experiments required an improved enzyme activity, free from the above oxidation reaction, in order to provide at least 4 g of formic acid for MS analysis.
Heterologous Expression of Sterol 14␣-Demethylase of C. albicans-The requirement for an improved source of enzyme for the projected study and the importance of sterol 14␣-demethylase as a target for the development of antifungal agents prompted experiments on the expression of the enzyme. In our previous studies the vector pW91P containing phosphoglycerate kinase promoter was used for the expression of C. albicans CYP51 gene in S. cerevisiae, and about 100 pmol of enzyme/mg of microsomal protein were obtained (18). Further improvement has now been achieved using GAL10 promoter (19) of the vector YEp51 in conjunction with the yeast strain GRF18 (20). Under the conditions of growth used in the present study, GRF18 harboring the expression vector without the insert gave undetectable levels of P-450; however, the cells still synthesized ergosterol, indicating a low level of endogenous expression. Although the phosphoglycerate kinase expression system had indicated that functional C. albicans CYP51 is produced from the native gene (18), we rectified the mutation that will occur on expression in S. cerevisiae due to the deviation in the genetic code discovered in C. albicans (21). In the latter organism, CTG, the triplet for Leu, is used for the incorporation of Ser. The alteration of the CTG triplet at position 263 to TCT by recombinant PCR was undertaken to allow a Ser to be inserted in this position, as occurs in C. albicans, when the protein is expressed in S. cerevisiae instead of Leu. The cloning strategy is illustrated in Fig. 1 and other details are described under "Experimental Procedures." Transformation of the yeast strain GRF18 with YEp51: CYP51 produced 1.5 mol of the demethylase/liter of culture, while the derived microsomal fraction was found to contain up to 2.5 nmol of P-450/mg of protein. The level of expression is SCHEME II. Structure of the key synthetic intermediate (5) and various isotopomers of the 32-oxo (6) and 32-hydroxy derivatives (7). higher than has been reported for other P-450 in yeast or E. coli, suggesting that the availability of heme is not limiting. This productivity was not dependent on the CTG to TCT mutation undertaken. Molecular modelling studies predict that the residue at position 263 is on the surface of the protein, thus explaining the absence of effect on the activity of the enzyme when the unmodified gene was expressed previously. 2 Table I shows that the specific activity of the enzyme in microsomes from recombinant vector, based on release of formic acid from the 3 H-labeled 32-oxo derivative (6e), was 0.1-0.25 nmol/nmol of P-450/min, and, as expected, no activity was detectable in the host strain harboring the parent vector. The specific activity of the cloned enzyme remained unchanged when it was purified to homogeneity and reconstituted with NADPH-cytochrome P-450 reductase from pig liver. The activity is similar to that found for homogeneous CYP51 obtained from a wild type strain of C. albicans (22), but lower than those reported from other sources (8,(23)(24)(25). The reason for the low specific activity of C. albicans CYP51 is not known but the possibility has been considered that the physiological substrate for this enzyme may be 24-methylene dihydrolanosterol rather than the lanosterol derivatives used in in vitro assays by us and others (22).
Mass Spectrometric Analysis of Formate Released from the 32-Oxo Derivatives (6) Variously Labeled at C-32 with 2 H and 18 O-Before dealing with the results of the enzymic incubations, attention is drawn to the fact that a quantitative determination of the 18 O content of formate produced biosynthetically is fraught with a number of problems. The one that should SCHEME III. Postulated mechanism for the acyl-carbon bond cleavage reaction catalyzed by sterol 14␣-demethylase using 6a as the substrate. The reactions in the sequence are: (i) adduct formation using the Fe III -OOH species, which is formed from the resting state of the enzyme, 2e, O 2 , and H ϩ ; (ii) homolytic cleavage; (iii) fragmentation; and (iv) disproportionation.
FIG. 1. Strategy for the cloning of the modified CYP51 gene of C. albicans. PCR mutagenesis to change the triplet at position 263 from CTG to TCT was performed using the four primers as described under "Experimental Procedures." The SalI-NsiI fragment coding for the N terminus of the protein and containing the mutation was then ligated to the C terminus encoding NsiI-HindIII fragment from pW91P, and the modified gene was inserted into SalI-HindIII cut YEp51. be mentioned from the very outset is the dilution of the biosynthetic formate with the unlabeled species ubiquitously present in reagents, solvents, and enzymic preparations. We have estimated that 5-10 g of formate is present per ml of a typical incubation, and to circumvent this interference, the precursors used in this study were labeled with deuterium at the position of interest, C-32, so that the 18 O content of only deuterium containing isotopomers of formic acid analyzed as benzyl formate (HCOO-Bzl) was used in the quantitation. Even this approach is not without its problems, since the peak at m/z 137 due to 2 HCOO-Bzl also contains substantial contributions from the 13 C isotopic species present in the unlabeled material. The intensity of the m/z 137 peak and also of peaks with higher masses could be corrected for 13 C natural abundance, but with limited accuracy. In the experiments reported in this paper, the biosynthetically produced formate was between 5 and 7% of the total formate present in the sample, and the corrected intensities of peaks are subject to a standard deviation of Ϯ10.
The mass spectrometric analysis, using either full scan or selected ion recording, of benzyl formate obtained from the incubation of the 2 H-labeled aldehyde (6b) under 16 O 2 gave a single isotopic peak at m/z 137 due to 2 HCOO-Bzl (entry 1, Table II). The absence of molecular ions due to higher masses m/z 138 -139 established that the important region of the spectrum was free from background noise. Benzyl formate from the incubation of the 2 H-labeled aldehyde (6b) under 18 O 2 gave a major peak at m/z 139 ascribable to 2 HC 18 OO-Bzl (entry 2, Table II). The intensity of the latter peak was at least 50% of the combined intensities of the peaks due to all the 2 H-containing isotopomers of benzyl formate. This result shows that during the cleavage of the C-14 -C-32 bond of the aldehyde (6) one atom of oxygen from 18 O 2 is incorporated into the released formate. Now the complementary experiment was performed in which the doubly labeled aldehyde (6c), containing 2 H as well as 18 O at C-32, was used as the substrate and incubated under 16 O 2 . Under these conditions, a predominant peak at m/z 139 was observed (entry 3, Table II), showing the transfer of the carbonyl oxygen of the aldehyde into the formate. The most significant feature of the experiment in which the same doubly labeled precursor (6c) was deformylated under 18 O 2 was the presence of a peak at m/z 141 for the isotopomer in which the deuterium containing benzyl formate contained two atoms of 18 O. In essence, this experiment (entry 4, Table II) represents the summation of the results of entries 2 and 3, showing that in the cleavage of the C-C bond of the aldehyde by the 14␣demethylase the original aldehydic oxygen of the substrate is retained in the released formic acid, while its second oxygen is derived from molecular oxygen.
Attention is drawn to the presence of substantial amounts of deuteriated benzyl formate containing either one or no 18 O in the experiment of entry 4 (Table II). The formation of these species is attributed to the loss of the aldehydic oxygen by exchange with the oxygen of H 2 O during the incubation. The extent of the exchange increases in the experiments performed under 18 O 2 because of the need to perform time-consuming manipulations for replacing air with 18 O gas. Another adverse consequence of this operation is some denaturation of the enzyme resulting in the production of lower amounts of formate.  O 2 Incubations were performed and processed as described under "Experimental Procedures." Only the deuterium and 18 O containing isotopomers of benzyl formate are included in the calculation of the data. The unlabeled benzyl formate originating from formic acid ubiquitously present in reagents and enzyme preparation has the molecular iron peak at m/z 136 the intensity of which was used to correct for 13 C natural abundance from all the peaks between m/z 137-141. D ϭ 2 H; F ϭ 18 O and U denotes the unlabeled species.

DISCUSSION AND CONCLUSIONS
The notion that the C-C bond cleavage reaction in the multistep process catalyzed by 14␣-demethylase occurs by the same generic reaction that has previously been found to operate for aromatase (CYP19) (2, 3) and 17␣-hydroxylase-17,20lyase (CYP17) (12) is supported by the present study. The fission process corresponding to an acyl-carbon cleavage is reducible to the stoichiometry of Reaction 2. The two main predictions of Reaction 2, that in the overall process the carbonyl oxygen atom of the substrate, together with an atom of oxygen from O 2 , are incorporated in the expelled formate, have been validated experimentally. We have advocated that the cleavage process may be rationalized by assuming that in the catalytic cycle of P-450s, the Fe III -OOH species, which is normally directed to produce an iron-monooxygen species involved in the hydroxylation reaction, may be trapped to give an adduct when the substrate skeleton contains a properly juxtapositioned electrophilic center (26,27).
The intermediacy of a peroxide adduct in acyl-carbon bond cleavages (Reaction 2) has been inferred from a range of observations, either described or reviewed in previous publications (10,(27)(28)(29)(30). Several mechanistic alternatives are possible through which the products of the reaction of Reaction 2 may be formed from the peroxide-adduct of the type 8, Scheme III. For example, in the case of CYP17 evidence has been presented to show that certain acyl-carbon bond cleavage reactions catalyzed by the enzyme occur by a homolytic fission route producing a carbon radical that either undergoes a disproportionation process producing an olefin or an oxygen rebound reaction forming a hydroxy compound (29). A similar scenario may be envisaged for the related acyl-carbon bond cleavage reaction, promoted by 14␣-demethylase, as shown in Scheme III. When the intermediacy of a peroxide-adduct was originally proposed, the possibility was considered that it may rearrange by a Baeyer-Villiger process to produce a formate ester which then, through an elimination reaction, creates the double bond in the final product (2). Such a possibility was, however, excluded for aromatase by showing that 10␤-hydroxyestr-4-ene-3,17-dione formate was not aromatizd by the enzyme (2,9). O-Acyl derivatives have been isolated during the reactions catalyzed by 14␣-demethylase (31) and CYP17 (32), but further work is required to establish whether these are bona fide intermediates in the acyl-carbon cleavage reaction or are formed merely as side products. The problem posed by a mechanism for the reaction of Reaction 2, which operates through the intermediacy of an O-acyl derivative is that it requires a single enzyme, not only to possess the activity for three different oxidative reactions, but also a fourth activity to promote the difficult removal of the elements of formic acid.
In the light of these considerations, we favor the mechanism shown in Scheme III for the conversion of the 32-oxo derivative (6) into the 7,14-diene (11). In principle the initially formed peroxy adduct (8) may cleave by an ionic or a radical process. The latter cleavage mode, however, has the advantage that it gives an intermediate alkoxy radical (9), which is ideally suited to undergo fragmentation producing formate, and the substrate radical (10), which can be conveniently converted into the product (11). Furthermore, the mechanism is based on a precedent from an equivalent acyl-carbon cleavage reaction catalyzed by CYP17 for which evidence for a radical process has been obtained (29).