`
`J. CHEM. SOC. PERKIN TRANS. I1 1983
`
`621
`
`Model Systems for Cytochrome P450 Dependent Mono-oxygenases.
`Part 2.' v2 Kinetic Isotope Effects for the Oxidative Demethylation of
`Anisole and [Me-2H,]Anisole by Cytochrome P450 Dependent
`Mono-oxygenases and Model Systems
`John R. Lindsay Smith * and Paul R. Sleath
`Department of Chemistry, The University of York, York YO1 5DD
`
`Anisole, [Me-2H3]anisole, and [ 1 8 0 J a n i ~ ~ l e have been used as substrates to study the mechanisms of
`oxidative demethylation by model systems for the cytochrome P450 dependent mono-oxygenases. The
`size of the kinetic isotope effect for the demethylation can be used as a sensitive probe of the oxidation
`mechanism and as a method for classifying the chemical systems. By this procedure 17 mono-oxygenase
`models and four microsomal systems have been examined. Two systems involving iron (111) porphyrins
`and iodosylbenzene show large kinetic isotope effects comparable with those from the microsomal oxid-
`ations and can be considered as good models for the biological process. The remaining systems exhibit
`smaller isotope effects, kH/kD 1-3.4. Alternative mechanisms for the oxidative demethylations are dis-
`cussed and the major routes are shown to be either radical ipso-substitution or attack on the C-H bond of
`the methoxy-group.
`
`Oxidative 0-demethylation is one of the several types of
`oxidation brought about by the cytochrome P450 dependent
`
`m ~ n ~ - ~ x y g e n a ~ e ~ . ~ For phenolic ethers the generally accepted
`mechanism involves hydroxylation of the a-carbon to give a
`hemi-acetal (1) which subsequently breaks down to a phenol
`and an aldehyde or ketone 3a [reaction (l)]. Analogous schemes
`have been proposed for enzymic N- and S-oxidative dealkyl-
`ati0n.3~9~
`Support for this mechanism comes from [180] labelling and
`kinetic isotope effect studies. In the oxidative 0-demethylation
`of 4-methoxyacetanilide with rat liver microsomes in the
`presence of lSO2 or H2180 none of the label is incorporated in
`the product 4-hydroxya~etanilide.~ These results show that
`cleavage of the 0-alkyl bond must occur in the reaction and
`that the alkoxy-group is not displaced by the incoming hydr-
`oxy in an ipso-substitution.
`In vitro studies with aryldeuteriomethyl ethers have shown
`that the oxidative demethylation proceeds with a kinetic
`isotope effect (kH/kD 2-10).6 Although there is a wide range of
`values for the isotope effect the data clearly indicate that C-H
`bond cleavage is occurring during enzymic 0-dealkylation.
`In contrast with the biological reactions above, pulse
`radiolysis studies suggest that the demethylation of 1,4-
`dimethoxybenzene by the hydroxyl radical proceeds by @so-
`subs ti t u t ion via t he in termed iat e h ydr ox ycyclo hexadienyl
`radical (2; R = OCH3) [reaction (2)].'
`This difference in behaviour towards arylmethyl ethers by
`the hydroxyl radical and the active oxidant in the cytochrome
`P450 dependent mono-oxygenases (this is currently thought to
`led us to examine the oxidative 0-
`act as an oxyl radical 3g98)
`demethylation of anisole, [Me-'H3]anisole, and ['80]anisole
`by a selection of model systems. We report here the results
`from this study and how measurement of the kinetic isotope
`effect for demethylation provides a simple method for classify-
`ing the model systems.
`
`Results
`Measurement of the Kinetic Isotope efect for the Oxidative
`Demethylation of Anisole and [Me-2H3]Anisole.-(a) Rat liver
`microsomes. The major products from the oxidation of anisole
`with rat liver microsomes in the presence of NADPH t and
`dioxygen are phenol and 4-methoxyphenol as reported by
`Jerina et al. (Table l).9 Neither product was obtained from
`
`Ar OC H R1 R2
`
`ArOCR1R2
`
`I OH
`
`ArOH + R1R2C=0
`
`( 1 1
`
`( 1 1
`Table 1. Yields of phenol and 4-methoxyphenol from the in v i m
`metabolism of anisole and [Me-ZH3]anisole with rat liver micro-
`somes-N ADPH-dioxy gen
`
`Substrate
`Anisole
`[Me-2H3]Anisole
`Anisole
`[Me-ZH3]Anisole
`An is ole
`[Me-2H3]Anisole
`Anisole
`[Me-2H3]Anisole
`Anisole
`[Me-2H3]Anisole
`
`Yield of
`Incubation
`time (min) phenol (pmol)
`15
`0.183
`15
`0.025
`30
`0.214
`0.028
`30
`0.235
`60
`0.034
`60
`0.221
`60
`60
`0.027
`0.244
`60
`0.036
`60
`
`Yield of
`4-methoxyphenol
`(pmol)
`0.498
`0.714
`0.615
`0.900
`0.609
`0.890
`0.698
`0.906
`0.629
`0.865
`
`incubations performed at 0 "C, or with heat denatured micro-
`somes or when anisole was omitted from the reaction. The
`results show that after 30 min incubation the rate of oxidation
`becomes very slow. The 60 min incubation was repeated with
`three separate batches of microsomes and the results show
`good reproducibility. Pairs of identical reaction mixtures, one
`containing anisole and the other [Me-2H3]anisole, were used
`to obtain the kinetic isotope effects. The yield of phenol from
`the deuterioanisole is markedly less than that from anisole
`revealing a large kinetic isotope effect for demethylation.
`
`f The following abbreviations are used in this paper: NADP and
`NADPH for nicotinamide adenine dinucleotide and the reduced co-
`enzyme, EDTA, ethylenediaminetetra-acetic acid disodium salt,
`Fe' "TPPCl tetraphenylporphinatoiron(u1) chloride, and Fell'-
`TFPPCl tetrakis(pentafluorophenyl)porphinatoiron(m) chloride.
`
`Auspex Exhibit 2008
`Apotex v. Auspex
`IPR2021-01507
`Page 1
`
`
`
`View Article Online
`
`.I. CHEM. SOC:. PI'RKIN TRANS. I t 1983
`
`R
`
`R
`( 2 1
`
`I
`
`R
`
`R
`
`Furthermore the yield of 4-methoxyphenol from [MP-~H&
`anisole is significantly larger than that from the undeuteriated
`substrate.
`The magnitude of the kinetic isotope effect obtained from a
`comparison of the absolute yields of phenol from each pair of
`reactions is independent of the incubation time, k H / k , 7.4.
`Alternatively, the isotope effect can be calculated from the
`yield of phenol relative to that of 4-methoxyphenol from a
`pair of reactions. The latter method assumes that deuteriation
`of the methoxy-group of anisole has no influence on the yield
`of 4-methoxyphenol and uses the yield of this product as an
`internal standard for the reaction. It was assumed that the
`latter method, which should eliminate minor differences in
`enzyme activity, would give a better measure of the kinetic
`isotope effect, k,,/k, 10.6. However, in view of the observed
`increase in 4-hydroxylation with methoxy-deuteriation this
`assumption may be invalid.
`The oxidations with rat liver microsomes were repeated
`with added oxidants (iodosylbenzene, t-butyl hydroperoxide,
`or 3-chloroperbenzoic acid) in the absence of NADPH. As
`described above, the experiments were carried out in pairs,
`one containing anisole and the other [Me-*H,]anisole. The
`initial rates of oxygenation with these systems are higher than
`those with NADPH and dioxygen but the oxidations are only
`sustained for a short time. Thus the yields are low and give
`rise to larger errors in the product analyses. Indeed the quan-
`tity of phenolic products from the iodosylbenzene supported
`system was too low to quantify the isotope effect. The results
`from these modified microsomal systems, calculated as des-
`cribed above, are the same within experimental error as those
`from the NADPH-dioxygen supported system (Table 2). The
`modified systems also show the effect of increased yield of 4-
`methoxyphenol when [M~-~H,]anisole is used in place of
`anisole.
`{ b) Chemical model systems fbr. cytuchromr P450 ckpendetit
`mono-oxy~enases. An i sole and [ Mc-'H3]an i so le were ox i d i sed
`with chemical model systems and the kinetic isotope effects
`for dernethylation were obtained. Since deuteriation of the
`methoxy-group of anisole should not influence the rate of 2-
`hydroxylation we used the yield of 2-methoxyphenol as an
`internal standard for each oxidation. Thus the kinetic isotope
`effects were obtained by comparing the ratio of the yields of
`phenol (demethylation) and 2-methoxyphenol (hydroxylat ion)
`from equivalent oxidations of the two substrates. This method
`of analysis is simple and convenient since the two oxidation
`products have similar retention times using g.c. analysis and
`it minimises errors arising from variable yields of oxidation.
`We investigated 17 model systems and kinetic isotope effects
`from oxidations by 13 of these are given in Table 3. Each
`value of li,,/ku is the average of all the analyses from at least
`two experiments and the quoted errors represent the spread of
`the results from these analyses. The large error for Groves'
`system and the approximate value for the modified system
`with Fe'"TFPPC1 reflect the low yield of phenolic products
`from these oxidations. This, coupled with a large kinetic
`isotope etl'e'ect, gives low yields of phenol from the deuteriated
`anisole and poor quantitation of the results.
`Cirilike the enzymic systems, none of the chemical models
`shows an increase in the yield of methoxyphenols when [Me-
`2H3]aniwle is the substrate in place of anisole.
`
`Table 2. Kinetic isotope cfTect\ for the oxidative demethylation of
`anisole and [Me-*H3]anisole with rat liver inicrosoines siipported by
`a selection of oxidants
`
`X t,/X
`from
`ratio of
`X,,/kD from
`relative yield
`ratio of
`Incubation
`of phenol to
`time (min) phenol yields 4-inethoxyphenol
`10.5 k 0.7
`7.3 t 0.9
`15
`11.2 1 0.7
`7.6 f 0.9
`30
`10.0 4 0.7
`60
`7.3 f 0.9
`6.0 I 2.0
`12.3 i 2.5
`I0
`5.8 t 1.8
`
`8.4
`
`3.0
`
`OxIdalit
`NADPH-0,
`NADPH-02
`NADPH- 0,
`t-Butyl
`h yd roperox icie
`3-Chloroperbenzoic
`acid
`
`I 0
`
`Table 3. Kinetic isotope di'cctb for the damethylation of aniwle and
`
`[ Me-2t-I,]anisolc hy chemical model ~ y ~ t e i i i s f o r thc cytochronie
`P450 dcpenden t niono-oxygenases
`0 Kid isi ng syslern
`Fe" --HzOz (Fenton's reagent) "'
`Fez * -H2O2-CH.%CN (Non-aqueous Fcntun's
`reagent)
`Fe"' -H,O,-catechol (Hamilton's system) ''
`Fez + - EDTA-ascorbic acid Or (Udenfriend's systeni) I '
`Fez+ EDTA -02 l4
`Fe' + - N-benzyl-l,4-dihydronicotinamide-Oz l 4
`Reduced tlavin niononucleotide O2 l5
`Fez + -2-mercaptobenzoic acid-02 (Ullrich's system) lh
`Trialkylphosphite-hv-02 l7
`SnZ + -pyrophosphate-02 l 8
`Diuofluorene-hv - 0, l 9
`Tetraphenylporphinatoiron(lri) chloride-PhlO
`(Groves' system)
`Tetra kis(pentafluoropheny1 )porphinatoiron(rii 1
`chloride-PhlO 21
`" The kinetic isotope effect was calculated from yields ot phenol and
`2-methoxyphenol (see text).
`
`11/kD
`1.0 i 0.1
`I .o j- 0.3
`1.3 1 0.1
`1.2 i 0.1
`1.0 t 0.1
`1.2 i 0.1
`1.0 c 0.1
`2.2 f 0.1
`2.1 f 0.2
`7.1 f 0.1
`3.4 * 0.4
`9.0 I 3.0
`
`8 .O
`
`Four of the model systems were not amenable to the analy-
`sis described above. (i) Trifluoroperacetic acid " hydroxy-
`lates anisole in high yield but does not bring about demethyl-
`ation. (ii) The photoactivation of pyridine N-oxide gives both
`phenol and methoxyphenols with anisole in aqueous solution
`as reported by Jerina et al." However, the low pressure U.V.
`source needed for the photoactivation also brings about the
`photodemethylation of anisole. The phenol from this reaction
`of anisole accounts for most, if not all, of the phenol yield from
`the model system. (iii) The oxidation of anisole in the vapour
`phase with triplet oxygen atoms, O(,P), from the mercury-
`photosensitised decomposition of nitrous oxide 24 gives phenol
`and methoxyphenols but, as with the N-oxide system above,
`the fcwrner product comes predominantly, if not entirely, from
`the photodecomposition of anisole. (iv) Attempts to oxidise
`anisole with a carbonyl oxide generated by the photosensitised
`formation of singlet dioxygen, lo2, in the presence of diazo-
`fluorene 25 were unsuccessful. No phenolic products were
`obtained from anisole.
`
`Auspex Exhibit 2008
`Apotex v. Auspex
`IPR2021-01507
`Page 2
`
`
`
`J. CHEM. soc. PERKIN TRANS. 11 1983
`
`Table 4. [ 1 8 0 ] Content of Zmethoxyphenol and phenol produced
`from the oxidation of ['80]anisole (3.19 f 0.2%) by model systems
`for the cytochrome P450 dependent mono-oxygen-
`Loss of
`['*O] in
`oxidative
`demethylation
`(%)
`91
`
`["O] content
`of
`[ 1 8 0 ] content
`2-methoxyphenol of phenol
`(%)
`(%I
`0.3
`3.1
`
`Model system
`Fe2 + -Hz02
`(Fenton's
`reagent)
`Fe2+-EDTA-
`ascorbic acid (Uden-
`friend's system)
`Fe3+-H202-
`catechol
`(Hamil ton's
`system)
`Sn2 + -
`py rophosp ha te-
`0 2
`
`3.0
`
`3.2
`
`3.2
`
`0.0
`
`0.7
`
`2.4,
`
`100
`
`78
`
`4,
`
`
`
`Demethylatiori of ["O] Ariisule with Model System.-The
`origin of the oxygen in the phenol frbm demethylation of
`anisole by four of the model systems namely, Fenton's reagent
`and Udenfriend's, Hamilton's, and the tin(i1)-pyrophosphate-
`dioxygen systems was determined by the use of [180]anisole as
`substrate. These systems which were selected as representa-
`tives of oxidants exhibiting small or medium-sized isotope
`effects, only require small amounts of substrate to produce
`sufficient phenol for g.c.-m.s. analysis. Table 4 shows that no
`[180] label is lost during the formation of 2-methoxyphen-
`01. However, for three of the systems most, if not all, the label
`is removed during demethylation. With the fourth model,
`tin(I1)-pyrophosphate-dioxygen, 91 of the isotopic label is
`retained in the phenol.
`
`Discussion
`Although it is generally accepted that 0-demethylation of
`arylmethyl ethers by cytochrome P450 dependent mono-
`oxygenase occurs via a hemi-acetal intermediate (l), the
`mechanism of the latter oxygenation remains unclear. The
`formation of phenol from anisole by model systems has been
`noted but the mechanism of this reaction has received very
`limited at tent ion.
`Six possible routes for the oxidative 0-demethylation are
`given below [reactions (3)-(5)
`and (7)-+9)]. All these need
`to be considered for the model systems; however, only reac-
`tions (3), (4), and (7) involve an intermediate hemi-acetal and
`are possible alternatives for the biological process.
`
`is inserted directly into the methoxy
`Reaction (3).-Oxygen
`C-H bond by a singlet oxenoid species to give the hemiacetal
`(1; R' = R2 = H).
`is abstracted from the methoxy-
`Reactiuri (4).-Hydrogen
`group (most probably as a hydrogen atom) followed by
`hydroxylation of the aryloxymethyl radical (3).
`
`Reaction (5).-ipso-Hydroxylation by an oxy-radical results
`in a cyclohexadienyl intermediate (4) which is aromatised to
`give phenol via the phenoxyl radical.
`
`abstraction gives the anisole radical
`Reaction (6).-Electron
`cation (5) which might be demethylated by one of three mech-
`an isms [react ions (7)-(9)].
`
`View Article Online
`
`623
`involves loss of aproton from the
`Reuctioii (7).-This
`methoxy-group followed by hydroxylation as in reaction (4).
`Reaction (8).-ipsu-Hydroxylation of the radical cation gives
`a hydroxycyclohexadienyl radical (1 ; R' = R2 = H) which is
`aromatised by loss of methanol to give the phenoxyl radical.
`
`Reaction (9).-Nucleophil
`ic substitution on the methyl
`group of the radical cation gives the phenoxyl radical directly.
`This last process is analogous to the non-oxidative dealkyl-
`ation of arylmethyl ethers by nucleophiles with acid catalysis.
`Oxidative N- and S-demethylations by electron-transfer
`pathways analogous to reactions (6) and (7) are well docu-
`mented for model systems and have been proposed for the
`biological processes ~ I S O . ~ * ~ ' * ~ ' However, these oxidations
`should occur more easily for tertiary amines and sulphides
`which have lower oxidation potentials than the corresponding
`ethers. The hydroxylation of an aromatic compound by reac-
`tion of its radical cation with water [reaction (lo), analogous
`to the first step of reaction (S)] has been observed for the
`reactions of benzene and toluene with strong oxidants.z8 Very
`recently Torii et aLf9 proposed @so-hydroxylation of the 4-
`methoxytoluene radical cation leading to the 4-methyl-
`phenoxyl radical [cf. reaction (8)] as a pathway in the oxid-
`ation of 4-methoxytoluene by cerium(rv) ammonium nitrate
`in aqueous methanol or acetic acid. However, Eberhardt
`reports that the persulphate radical anion does not hydroxyl-
`ate anisole 28cl and he suggests that the radical cation from
`anisole is not susceptible to nucleophilic attack by water.
`O'Neill et al.30 who reach a similar conclusion about the
`hydration of the anisole radical cation from pulse radiolysis
`studies, also suggest that the removal of a proton from the
`radical cation to give the phenoxymethyl radical [reaction
`(7)], in an analogous manner to the radical cations of methyl-
`benzenes, is an unfavourable process. We conclude that it is
`unlikely that anisole is oxidised by an electron-transfer pro-
`cess either in the biological system, in agreement with Oae
`and his co-workers,6C or in the model systems.
`We argued that the remaining three mechanisms [reactions
`(3)-(5)] might be distinguishable by a combination of [180]
`labelling and kinetic isotope effect studies. Thus, for the direct
`insertion and hydrogen abstraction processes the phenolic
`oxygen is that in the starting anisole but the latter process
`might show a larger kinetic isotope effect, the value depending
`on the extent of C-H bond breakage in the transition ~ t a t e . ~ '
`By analogy with C-H insertion with singlet carbenes the
`kinetic isotope effect for an oxene insertion would be small,
`kH/kD ca. 1-2.5.3L The ipso-substitution should not show a
`kinetic isotope effect and the oxygen of the anisole would be
`lost in forming phenol.
`Microsomal Oxidative 0 - Demeth ylat ior i . -T he mi c r oso m al
`demethylatiohs, whether supported by NADPH-dioxygen, t-
`butyl hydroperoxide, or 3-chloroperbenzoic acid, show large
`kinetic isotope effects. The effect is independent of incubation
`time and within experimental error, which is large for the
`peroxide and peroxyacid supported systems, it is also inde-
`pendent of the oxygen source.
`The value obtained from a direct comparison of the yield of
`phenol from separate experiments with anisole and [Me-2H3]-
`anisole (kH/kD 7.4) is an intermolecular isotope effect. This
`value is larger than the intermolecular isotope effects for 0-
`demethylation reported for a selection of arylmethyi ethers
`(kH/kD ca. 2 ) . 6 0 9 b However, recently Watanabe et a1.6' ob-
`tained a value of 5.1 for the monodemethylation of 1,4-
`dimethoxybenzene and [Me-'H6]-1 ,4'dimethoxybenzene with
`rabbit liver microsomes. The absence of an observable kinetic
`isotope effect or its suppression in intermolecular competition
`
`Auspex Exhibit 2008
`Apotex v. Auspex
`IPR2021-01507
`Page 3
`
`
`
`624
`
`View Article Online
`
`J. CHEM. SOC. PERKIN TRANS. I1 1983
`
`Fern ___)
`
`OH 6 + HCHO
`
`FdY- OH
`
`OC H3
`
`CH3O
`
`OFe"
`
`0'
`
`OH
`
`PhOCH3 +
`
`( 4 )
`
`FeI'O'
`
`+
`
`FeyO
`
`PhOCH3
`+*
`
`( 5 )
`
`S)CH20H
`-HCHO d
`
`HO
`
`OCH3
`
`+ ( H ' )
`___c
`
`( 5 )*
`
`( 7 )
`
`( 8 )
`
`( 9 )
`
`(10)
`
`( 5 )
`
`Q -
`
`- e-
`
`- e'
`__t
`- H +
`
`R
`* Reactions (3)--(6) are illustrated with the iron-oxy-species considered to be the active oxidant in the cytochrome P450 mono-
`oxygenase.' This is abbreviated to Fev=O and Fe'"-O.
`
`experiments with mono-oxygenases can be attributed to the
`oxidation being part of a multistep process in which the rate-
`determining step may not be the C-H bond cleavage of the
`substrate .33
`The kinetic isotope effect obtained by comparing the ratio
`of the yields of phenol and 4-methoxyphenol from each sub-
`strate is less easily defined. It is an intermolecular effect if
`demethylation and hydroxylation are brought about by two
`different enzymes and intramolecular if the same enzyme(s)
`is(are) involved in both oxidations. The results from this
`
`study favour the latter explanation with anisole as substrate.
`Thus once the substrate is bound to the enzyme it is committed
`to oxidation, so that when demethylation is made less favour-
`able by deuterium substitution ring hydroxylation is enhanced.
`This explanation accounts for the higher yield of 4-methoxy-
`phenol from [Me-*H,]anisole than from anisole and for the
`observation that the sum of the yields of phenol and 4-
`met hoxyphenol from equivalent experiments is independent
`of deuterium substitution. Mitoma et u Z . ~ ~ reported a similar
`effect when studying the influence of deuterium substitution on
`
`Auspex Exhibit 2008
`Apotex v. Auspex
`IPR2021-01507
`Page 4
`
`
`
`View Article Online
`
`625
`0-demethylation by any of them involves a rate-determining
`cleavage of the methoxy C-H bond. This result was not un-
`expected for Fenton's reagent in which the active oxidant is
`the hydroxyl radical since, as described above, the hydroxyl
`radical has been shown to demethylate 1,4-dimet hoxybenzene
`by ipso-substitution and not by hydrogen-atom abstraction
`from the methoxy-group.7 This mechanism was confirmed by
`the complete loss of
`label in the conversion of
`anisole to phenol. It is probable that for the oxidising systems
`with kH/kD ca. 1.0 all, or almost all, the demethylation is by
`an oxyl radical ipso-substitution. This was confirmed for
`Udenfriend's system with
`['*O]anisole. However, with
`Hamilton's system ["O] labelling studies suggest that although
`@so-substitution is the major pathway (ca. 80"/,), possibly
`involving the hydroxyl radical, a small proportion of the reac-
`tion may take place with retention of the oxygen label. If 207;
`of the oxidation occurs by hydrogen-atom abstraction this
`could account for the kinetic isotope effect being greater than
`unity. A kinetic isotope effect of 8.0 for the minor reaction
`would give an overall isotope effect of 1.2.
`
`B CXJ,V. soc. PERKIN TRANS. TI 1983
`I i:t: vields of hydroxylated products from microsomal oxid-
`,ition of 4-nitrophenyl propyl ether and concluded the same
`tnryme(s) was(were) responsible for the formation of all the
`products. Relevant to this study is the recent work of Gelb et
`d3' who show that cytochrome P45OCAM mono-oxygenase can
`remove the 4-exo- or 4-endu-hydrogen of camphor to give 5-
`P yo-hydroxycamphor. They discuss at length the influence of
`the geometrical selection of the mono-oxygenase for the 5-
`endo- and 5-em-positions on the observed isotope effect.
`However, studies with competitive inhibitors suggest that
`aromatic hydroxylation and dimethylation may be mediated
`by different cytochrome P450 haem~proteins.~~
`Despite the uncertainties described above, it is clear from
`the large isotope effects that C-H bond cleavage is occurring
`in the demethylation of anisole by microsomal enzymes. This
`conclusion is fully in agreement with the generally accepted
`mechanism for these oxidative 0-demethylations.
`The kinetic isotope effects from this study are comparable
`with the intramolecular effects for monodemethylation of 1,4-
`dimethoxybenzene 6b and for alkane hydro~ylation.~~ The size
`of these effects is in agreement with these processes occurring
`by hydrogen-atom abstraction by the active
`The similarity of the isotope effects exhibited by the NADP-
`H-dioxygen system and those using t-butyl hydroperoxide or
`3-chloroperbenzoic acid suggests that the same or similar
`active oxidants are present in all three systems. Ullrich 37b also
`reached this conclusion from the intramolecular kinetic iso-
`tope efkcts in the hydroxylation of [2Hll]cyclohexane. He
`obtained values of kH/kD of 7-8.6
`for liver microsomal hy-
`droxylat ion supported by NADPH-dioxygen, hydrogen
`peroxide, iodosylbenzene, 3-chloroperbenzoic acid, or cum-
`m e hydroperoxide. Likewise Groves ef al. obtained k,/k,
`for the hydroxylation of deuteriocyclohexenes by a re-
`4-6
`constituted mono-oxygenase with a similar range of oxygen
`donors."' Currently the nature of the active oxidant in the
`rnicrosomal and modified microsomal systems is an active
`area of research and
`
`Model Systems with k,/k, 2-3.4.-The
`four systems that
`show medium-sized isotope effects may bring about demethyl-
`ation by a combination of @so-substitution and side-chain
`attack or alternatively solely by side-chain attack by a
`mechanism exhibiting a medium-sized isotope effect. From
`["O] labelling experiments it is clear that for the tin(1r)-
`pyrophosphate-dioxygen system the retention of the oxygen
`from anisole in the phenol is only compatible with the latter
`explanat ion.
`The kinetic isotope effects in this study for the tin(1r)-
`pyrophosphate-dioxygen and diazoalkane-hv-dioxygen sys-
`tems agree well with values reported for the hydroxylation of
`cyclohexane and deuteriated cyclohexane. Ullrich obtained an
`isotope effect for the former system of 1.9 37b and Hamilton
`and Giacin a value of 4.6 for the latter l9 as compared with 2.1
`and 3.4, respectively from this study.
`The side-chain reaction could involve hydrogen-atom
`abstraction by a radical or oxene insertion into the methoxy
`Moclrl Systems.-With
`the requirements of the model
`C-H bond. In this respect it is noteworthy that three of the
`system clearly delineated from the results with the microso-
`systems, namely Ullrich's 16*41 and those using trialkyl phos-
`ma1 syrtems, we investigated a wide range of model systems
`phite-hv-dioxygen l7 and diazoalkane-hv-dioxygen l 9 are
`to examine (i) which could bring about both oxidative
`thought to oxidise organic compounds by radical mechanisms.
`0-demethylation and aromatic hydroxylation of anisole, and
`However, Ullrich and Staudinger have proposed an oxenoid
`mechanism for the tin(rr)-pyrophosphate-dioxygen system.'8
`( i i ) which would show a large kinetic isotope effect for the
`deme t hylation.
`The size of the kinetic isotope effect to be expected in an
`Trifluoroperacetic acid, a typical peroxycarboxylic acid,
`oxene insertion into a C-H bond is uncertain 42 but by analogy
`did not demethylate anisole. So that although peroxycar-
`with equivalent carbene insertions it is unlikely to be >2.5.
`boxylic acids epoxidise alkenes stereospecifically and bring
`For hydrogen-atom abstraction it will depend on the position
`about aromatic hydroxylation with large values of the NIH
`of the transition state on the reaction profile which in turn
`shift zzu it seems unlikely that they are good models for cyto-
`defines the extent of C-H bond cleavage in the transition
`chrome P450 dependent mono-oxygenases.
`For early or late transition states kH/kD should be
`Two photochemical systems requiring U.V.
`light (aqueous
`small. The kinetic isotope effect should be maximal for a
`pyridine N-oxide-hv and Hg-NzO-hv) could not be tested
`thermoneutral process with a synvnetrical transition state in
`because the anisole was photochemically demethylated in the
`which the C-H bond would be approximately h a l f - b r ~ k e n . ~ ' * ~ ~
`absence of the oxidant. It is likely that the excited anisole
`Thus radical chlorination of toluene by the reactive chlorine
`loses a hydrogen atom from the methoxy-group to give a
`atom has an early transition state with little C-H bond cleav-
`phenoxymethyl radical which reacts further to give
`age and exhibits a small kinetic isotope effect (kEi/kD 1.3 at
`In our hands a fourth system, that is reported to generate a
`77 0C).44 However, the equivalent bromination has a larger
`carbonyl oxide from singlet dioxygen and a d i a z ~ a l k a n e , ~ ~ ~
`value (kH/kD 4.9 at 77 "C) as would be predicted for this
`did not oxidise anisole. This result was unexpected since the
`nearly thermoneutral process.44 The majority of kinetic isotope
`diphenyldiazomethane-hv-singlet dioxygen system is reported
`effects for C-H bond cleavage by hydrogen-atom abstraction
`to hydroxy late naphthalene .40
`in solution have kH/kD values between 2 and 8 . 4 3 p 4 s
`If these oxidative 0-demethylations are initiated by hydro-
`gen-atom abstraction from the methoxy-group it is likely that
`the transition state is early on the reaction profile. However,
`although the authors are in favour of radical mechanisms for
`these oxidations, the present data cannot dlstiryish con-
`
`Model Systems with kH/kD ca. 1.0.-Seven of the oxidising
`systems give kinetic isotope effects near unity and consequently
`cannot be considered to be good models for the cytochrome
`P450 dependent mono-oxygenases. It is unlikely that oxidative
`
`Auspex Exhibit 2008
`Apotex v. Auspex
`IPR2021-01507
`Page 5
`
`
`
`View Article Online
`
`clusively between such mechanisms and C-H
`singlet oxenoid species.
`
`insertion by
`
`the two model
`Model Systems with kH/ko >6.0.--0nly
`systems consisting of iron(1Ir) porphyrins and iodosylbenzene
`show isotope effects comparable with those of the cytochrome
`P450 dependent mono-oxygenases. The values, which are close
`to the maximum value to be
`indicate that for these
`systems methoxy C-H bond breakage must be occurring in
`the rate-determining step. It is most likely, as discussed above
`for the microsomal systems, that these oxidations proceed by
`hydrogen-atom abstraction via a linear transition state [reac-
`tion (4)].
`In conclusion, the measurement of the size of the kinetic
`isotope effect for oxidative demethylation of anisole and [Me-
`2H3]ani~~le provides a new and simple method for classifying
`model systems for the cytochrome P450 dependent mono-
`oxygenases. Only systems using iron(1rr) porphyrins and iodo-
`sylbenzene show isotope effects comparable with those of the
`microsomal systems and can be considered as good models.
`The combination of kinetic isotope effect and [180] labelling
`studies reveal some interesting information about the mechan-
`isms of demethylation of anisole.
`
`Experiment a1
`the materials were commercial reagent
`Materials.-All
`grade unless otherwise stated and were obtained from Aldrich
`Chemical Co. Ltd., Fisons Scientific Apparatus Ltd., or
`Koch-Light Ltd. [2H4]Methanol (Aldrich Chemical Co. Ltd.)
`was 99.5% deuteriated and [180]water (British Oxygen Co.)
`was 5% enriched. The nitrogen (white spot grade) and nitrous
`oxide were from British Oxygen Co.
`Anisole was purified by washing with 20% aqueous sodium
`hydroxide followed by distillation. Acetonitrile was purified
`following O'Donnell et aL4' and stored in the dark over 3A
`molecular sieves. Iron(1r) perchlorate was prepared by the
`action of cold aqueous perchloric acid (30%, w/v) on iron
`powder. The salt was recrystallised from water and stored in
`the dark over phosphorus pentaoxide. The iron(rr1) porphyrins
`and iodosylbenzene were obtained as described previously.'
`Diazofluorene was prepared following Miller 48 and had m.p.
`"C (decomp.) (lit.,"9 99 "C). N-Benzyl-l,4-dihydro-
`98-100
`nicotinamide was prepared
`from N-benzylnicotinamide
`chloride by reduction with sodium dithionite following
`Mauzerall and Westheimer 50 and had m.p. 122-123 "C
`(lit.,50 120-122
`"C). [2H3]Iodomethane for the synthesis of
`[Me-2H3]anisole was prepared from [2H4]methanol by a
`standard procedure" and had b.p. 42-43 "C (lit.?l 42-
`42.5 "C). The preparation of [Me-2H3]anisole was a modific-
`ation of the method of Dalton et aLS2 Sodium phenoxide,
`prepared from phenol (47 g) and sodium ethoxide, was dis-
`solved in dioxan (600 cm3) before [2H3]iodomethane (72.5 g)
`was added and the mixture was heated at 85 "C overnight.
`Water (800 cm3) containing sodium chloride (40 g) and sodium
`hydroxide (8 g) was added and the mixture was worked up by
`extraction with diethyl ether (4 x 100 cm3). The combined
`ether extracts were washed with 10% aqueous sodium hy-
`droxide, dried (MgSO,), concentrated under vacuum and
`distilled to give [Me-2H3]anisole in 31% yield based on
`"C),
`152-153
`[2H4]methanol, b.p. 150-153
`"C
`G,(CDCI:,) 7.42-6.68, m/z 111 (M').
`[180]Anisole was synthesised from ['80]phenol obtained
`from the decomposition of benzenediazonium ion in [180]
`enriched water by a modification of a standard procedure.53
`Aniline (3.75 cm3) was dissolved in a solution of sulphuric
`acid (4.5 cm3) in [180] enriched water (15 cm3). The mixture
`was cooled to - 10 "C and a solution of sodium nitrite (3 g) in
`
`J. CHEM. SOC. PERKIN TRANS. I1 1983
`[180] enriched water (5 cm3) was added dropwise while the
`solution was maintained at -10 "C. The diazonium ion was
`allowed to decompose very slowly <O "C and was warmed
`to room temperature after the evolution of nitrogen ceased.
`Extraction with diethyl ether (3 x 50 cm3), drying of the com-
`bined extracts (MgSO,), and concentration under vacuum
`gave a black viscous residue. Distillation of the residue gave a
`yellow liquid consisting largely of phenol (g.c. analysis) which
`was methylated with dimethyl sulphate without further
`purification. The product was extracted into diethyl ether,
`washed with 10% aqueous sodium hydroxide, and distilled to
`give [180]anisole (450 mg) which was found to be pure by g.c.
`analysis. Mass spectrometr