Cis-trans isomerization of a cyclopropyl radical trap catalyzed by extradiol catechol dioxygenases: Evidence for a semiquinone intermediate

Article abstract of DOI:10.1021/ja9607704

Substrate analogues cis- and trans-2-(2,3-dihydroxyphenyl)cyclopropane-1-carboxylic acid were synthesized as probes for a semiquinone radical intermediate in the (2,3-dihydroxyphenyl)propionate 1,2-dioxygenase reaction. These analogues were found to be substrates for oxidative cleavage by extradiol dioxygenases from Escherichia coli and Alcaligenes eutrophus. The stereochemistry of the ring fission products was analyzed by conversion to cyclopropane-1,2-dicarboxylic acids using the ensuing hydrolase enzyme MhpC, followed by GCMS analysis. This analysis revealed 85-94% trans product and 6-15% cis products, implying that cis/trans isomerization of the cyclopropyl ring substituents had taken place during the enzymatic conversion. These results are consistent with a reversible opening of the cyclopropyl ring, and hence consistent with the intermediacy of a semiquinone radical intermediate in the extradiol catechol dioxygenase reaction.

Full text of DOI:10.1021/ja9607704

8336
J. Am. Chem. Soc. 1996, 118, 8336-8343
Cis-Trans Isomerization of a Cyclopropyl Radical Trap
Catalyzed by Extradiol Catechol Dioxygenases: Evidence for a
Semiquinone Intermediate
Emma L. Spence, G. John Langley, and Timothy D. H. Bugg*
Contribution from the Department of Chemistry, UniVersity of Southampton,
Highfield, Southampton SO17 1BJ, U.K.
ReceiVed March 11, 1996X
Abstract: Substrate analogues cis- and trans-2-(2,3-dihydroxyphenyl)cyclopropane-1-carboxylic acid were synthesized
as probes for a semiquinone radical intermediate in the (2,3-dihydroxyphenyl)propionate 1,2-dioxygenase reaction.
These analogues were found to be substrates for oxidative cleavage by extradiol dioxygenases from Escherichia coli
and Alcaligenes eutrophus. The stereochemistry of the ring fission products was analyzed by conversion to
cyclopropane-1,2-dicarboxylic acids using the ensuing hydrolase enzyme MhpC, followed by GCMS analysis. This
analysis revealed 85-94% trans product and 6-15% cis products, implying that cis/trans isomerization of the
cyclopropyl ring substituents had taken place during the enzymatic conversion. These results are consistent with a
reversible opening of the cyclopropyl ring, and hence consistent with the intermediacy of a semiquinone radical
intermediate in the extradiol catechol dioxygenase reaction.
Introduction
cofactor in both families of enzyme binds the catecholate
oxygens and also binds molecular oxygen.7,8 Mechanistic
proposals for the subsequent steps of the extradiol catechol
dioxygenase mechanism have invoked a Criegee rearrangement
of a peroxy intermediate.8-10 Experimental evidence for such
a rearrangement has been obtained from 18O-labeling studies
on the extradiol enzyme (2,3-dihydroxyphenyl)propionate 1,2-
dioxygenase (MhpB) from Escherichia coli, for which the
mechanistic scheme shown in Figure 1 has been proposed.10
Coordination of substrate 1 to the non-heme iron(II) cofactor
is followed by electron transfer to form a semiquinone inter-
mediate (2). Recombination with superoxide could form one
of two possible peroxy intermediates, either of which could
undergo a Criegee rearrangement to give an unsaturated lactone
intermediate (3). The hydrolysis of lactone 3 by enzyme-bound
iron(II) hydroxide is supported experimentally by the MhpB-
catalyzed hydrolysis of a saturated lactone analogue.10
The rapid ring opening of cyclopropylmethyl radicals has
been employed in the design of cyclopropyl radical traps to study
the existence of radical intermediates in a number of enzyme-
catalyzed reactions.1 Evidence for radical intermediates in the
reactions of several iron-dependent oxygenase enzymes has been
obtained by the observation of products in which a cyclopropyl
ring has been opened.2 In other cases enzyme inactivation has
been observed by cyclopropyl substrates, via covalent enzyme
modification by a ring-opened butenyl radical.3
The catechol dioxygenases comprise two families (intradiol
and extradiol) of non-heme iron-dependent enzymes which
catalyze the oxidative cleavage of the aromatic ring of catechol
substrates. These enzymes play an important role in the
biodegradation of a wide range of man-made industrial pollut-
ants containing aromatic rings.4 High-resolution structural
information is now available for both the iron(III)-dependent
intradiol dioxygenases5 and the iron(II)-dependent extradiol
dioxygenases.6
Precedent for the involvement of semiquinone intermediates
in these reactions is provided by model catecholate complexes
of redox-active metal ions. Exposure of a rhodium(III) cate-
cholate complex to oxygen gave a semiquinone-metal-
superoxide complex (5),11 while the corresponding iridium(III)
complex gave a peroxy adduct (6) presumably formed by
recombination of semiquinone with superoxide.12
There is considerable evidence from ESR spectroscopic
studies on the catechol dioxygenases that the non-heme iron
X Abstract published in AdVance ACS Abstracts, August 15, 1996.
(1) Suckling, C. J. Angew. Chem., Int. Ed. Engl. 1988, 27, 537-552.
Nonhebel, D. C. Chem. Soc. ReV. 1993, 348-359.
(2) (a) Cytochrome P450: Atkinson, J. K.; Ingold, K. U. Biochemistry
1993, 32, 9209-9214. Atkinson, J. K.; Hollenberg, P. F.; Ingold, K. U.;
Johnson, C. C.; Le Tadic, M. H.; Newcomb, M.; Putt, D. A. Biochemistry
1994, 33, 10630-10637. Newcomb, M.; Le Tadic, M. H.; Putt, D. A.;
Hollenberg, P. F. J. Am. Chem. Soc. 1995, 117, 3312-3313. Newcomb,
M.; Le Tadic-Biadatti, M. H.; Chestney, D. L.; Roberts, E. S.; Hollenberg,
P. F. J. Am. Chem. Soc. 1995, 117, 12085-12091. (b) Methane monooxy-
genase: Liu, K. E.; Johnson, C. C.; Newcomb, M.; Lippard, S. J. J. Am.
Chem. Soc. 1993, 115, 939-947. (c) Isopenicillin N synthase: Baldwin,
J. E.; Adlington, R. M.; Domayne-Hayman, B. P.; Knight, G.; Ting, H. H.
J. Chem. Soc., Chem. Commun. 1987, 1661-1663.
(3) (a) Monoamine oxidase: Silverman, R. B.; Zhou, J. P.; Eaton, P. E.
J. Am. Chem. Soc. 1993, 115, 8841-8842. (b) Acyl CoA dehydrogenase:
Lai, M. T.; Liu, L. D.; Liu, H. W. J. Am. Chem. Soc. 1991, 113, 7388-
7397. (c) γ-Butyrobetaine hydroxylase: Ziering, D. L.; Pascal, R. A. J.
Am. Chem. Soc. 1990, 112, 834-838.
(6) Han, S.; Eltis, L. D.; Timmis, K. N.; Muchmore, S. W.; Bolin, J. T.
Science 1995, 270, 976-980. Senda, T.; Sugiyama, K.; Narita, H.;
Yamamoto, T.; Kimbara, K.; Fukuda, M.; Sato, M.; Yano, K.; Mitsui, Y.
J. Mol. Biol. 1996, 255, 735-752.
(7) Que, L., Jr. J. Chem. Educ. 1985, 62, 938-943. Arciero, D. M.;
Lipscomb, J. D.; Huynh, B. H.; Kent, T. A.; Mu¨nck, E. J. Biol. Chem.
1983, 258, 14981-14991. Mabrouk, P. A.; Orville, A. M.; Lipscomb, J.
D.; Solomon, E. I. J. Am. Chem. Soc. 1991, 113, 4053-4062.
(8) Arciero, D. M.; Lipscomb, J. D. J. Biol. Chem. 1986, 261, 2170-
2178.
(9) Shu, L.; Chiou, Y. M.; Orville, A. M.; Miller, M. A.; Lipscomb, J.
D.; Que, L., Jr. Biochemistry 1995, 34, 6649-6659.
(10) Sanvoisin, J.; Langley, G. J.; Bugg, T. D. H. J. Am. Chem. Soc.
1995, 117, 7836-7837.
(11) Bianchini, C.; Frediani, F.; Laschi, F.; Meli, A.; Vizza, F.; Zanello,
P. Inorg. Chem. 1990, 29, 3402-3409.
(4) Dagley, S. Essays Biochem. 1975, 11, 81-135. Feig, A. L.; Lippard,
S. J. Chem. ReV. 1994, 94, 759-805.
(5) Ohlendorf, D. H.; Lipscomb, J. D.; Weber, P. C. Nature 1988, 336,
403-404.
(12) Barbaro, P.; Bianchini, C.; Mealli, C.; Meli, A. J. Am. Chem. Soc.
1991, 113, 3181-3183. Barbaro, P.; Bianchini, C.; Frediani, P.; Meli, A.;
Vizza, F. Inorg. Chem. 1992, 31, 1523-1529.
S0002-7863(96)00770-6 CCC: $12.00 © 1996 American Chemical Society

Cis-Trans Isomerization of a Cyclopropyl Radical Trap
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8337
demethylated using boron tribromide, which gave the desired
2,3-dihydroxycinnamic acid and a 15-40% yield of 8-hydroxy-
coumarin (8). The methoxymethyl (MOM) ether was found to
be the most suitable protecting group for the phenolic hydroxyl
groups in this route, with bulkier protecting groups at C-2
hindering the cyclopropanation reaction. The 2,3-MOM-
protected cinnamate ethyl ester (9) was reacted with trimethyl-
oxosulfonium ylide,13 affording selectively the trans-cyclopropyl
ester (10) in 52% yield. Alkaline hydrolysis of the ethyl ester,
followed by acidic removal of the MOM groups, gave the trans-
cyclopropyl analogue (7a) in 62% yield.
The cis-cyclopropyl analogue was prepared in similar fashion
(see Scheme 1) from 8-hydroxycoumarin (8). MOM protection
of 8 proceeded smoothly in 97% yield to give the MOM-
protected lactone (11). However, in this case cyclopropanation
of 11 did not proceed to completion, and at no stage could the
cis-cyclopropyl lactone (12) be separated from the lactone
precursor (11).29 Thus the cis-cyclopropyl analogue (7b) was
isolated as a 4:1 mixture with 8-hydroxycoumarin (10), which
was shown to be neither a substrate nor an inhibitor for
dioxygenase MhpB.
Figure 1. Proposed mechanistic scheme for (2,3-dihydroxyphenyl)-
propionate 1,2-dioxygenase (see ref 10).
1H-NMR spectroscopic analysis of 7a and 7b revealed
characteristic signals for the CH-Ar proton at 2.71 and 2.55
ppm, respectively. Close inspection revealed that both 7a and
7b were diastereomerically pure, to the limits of detection
(>99%).
Processing of Cyclopropyl Analogues by Extradiol Di-
oxygenases. Both cis- and trans-cyclopropyl analogues (7a and
7b) were tested as substrates for Escherichia coli (2,3-
dihydroxyphenyl)propionate 1,2-dioxygenase (MhpB), purified
as previously described.14 The trans-analogue (7a) was sub-
sequently assayed as a substrate for Alcaligenes eutrophus
dioxygenase MpcI, purified to >50% homogeneity from over-
expressing strain JM109/pAE166.15 The A. eutrophus enzyme
shares significant sequence similarity with the E. coli enzyme,
and processes a similar range of substrates with comparable
catalytic efficiency.14,15
In each case the rapid formation of a UV absorption at 400-
405 nm was observed corresponding to the meta-ring-fission
product, indicating that these analogues were substrates for
oxidative meta cleavage. Apparent Km and kcat values of 100
µM and 18 s-1 were measured for the processing of 7a by
MhpB, which upon comparison with the natural substrate (1)
reveal that 7a is processed at 16% of the catalytic efficiency of
1 (see Table 1).29 7a is also processed efficiently by A.
eutrophus MpcI (see Table 1), suggesting that these substrates
adopt similar orientations to the natural substrates at the
respective active sites. Extended incubation with 7a or 7b gave
no detectable enzyme inactivation, compared with control
experiments.
Figure 2. Anticipated outcomes for reaction of cyclopropyl analogue
7 with dioxygenase MhpB.
In order to probe the existence of semiquinone 2, we
anticipated that analogues containing a cyclopropyl group in
the propionate side chain might act as radical traps for this
enzyme. Analogue 7 was designed, in which the introduction
of a cyclopropyl group is achieved with only minor modification
of the propionate side chain. Upon formation of the semi-
quinone radical intermediate, we anticipated that the cyclopropyl
ring of 7 would open, leading either to a novel enzymatic
product or to enzyme inactivation (Figure 2). We report the
preparation of cis and trans diastereoisomers of 7 and analysis
of their reaction with the Escherichia coli dioxygenase and with
a related extradiol dioxygenase from Alcaligenes eutrophus.
Attempts were made to isolate and characterize the ring-
fission products 13a and 13b from enzymatic conversion of 7a
and 7b. The isolated products showed no characterizable signals
1
by H NMR spectroscopy other than 7a and 7b, in contrast to
the natural ring fission product (4) which has been previously
characterized.16 It appears that the presence of the cyclopropyl
substituent in the side chain renders the ring-fission products
13a and 13b chemically unstable. No additional signals were
detected which might correspond to products in which the
cyclopropyl ring has been opened.
(13) Corey, E. J.; Chaykovsky, N. J. Am. Chem. Soc. 1965, 87, 1353-
1364.
Results
(14) Bugg, T. D. H. Biochim. Biophys. Acta 1993, 1202, 258-264.
(15) Kabisch, M.; Fortnagel, P. Nucleic Acids Res. 1990, 18, 3405-
3406. Spence, E. L.; Kawamukai, M.; Sanvoisin, J.; Braven, H.; Bugg, T.
D. H. J. Bacteriol. 1996, in press.
Preparation of Cyclopropyl Analogues. trans-2-(2,3-Di-
hydroxyphenyl)cyclopropane-1-carboxylic acid (7a) was pre-
pared as shown in Scheme 1. 2,3-Dimethoxycinnamic acid was

8338 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Spence et al.
Scheme 1. Synthetic Route for Cyclopropyl Analoguesa
a (a) BBr3, CH2Cl2, -78 °C. (b) H2SO4, EtOH, reflux. (c) CH3OCH2Cl, iPr2NEt, CH2Cl2. (d) Me3SO+I-, NaH, DMSO. (e) 0.5 M NaOH/H2O.
(f) 2 M HCl/H2O. Yields given in the Experimental Section.
Table 1. Steady State Kinetic Parameters for E. coli MhpB and A. eutrophus MpcIa
E. coli MhpB
Km(app) (µM)
kcat (s-1
A. eutrophus MpcI
substrate
λmax(RFP) (nm)
)
kcat/Km(app) (M-1 s-1
)
Km(app) (µM)
Vrel
(2,3-dihydroxyphenyl)propionic acid (1)
trans-cyclopropyl analogue 7a
2,3-dihydroxycinnamic acid
394
405
452
26
100
36
29
18
19
6
1.1 × 106
1.8 × 105
5.3 × 105
2.0 × 104
7.6
40
6.9
ND
1.0
0.85
0.67
0.55
2,3-dihydroxyphenoxyacetic acid
300
a Kmapp values measured under saturating concentrations of oxygen. Vrel values measured for A. eutrophus MpcI since the enzyme was not
homogeneous and is prone to time-dependent inactivation.15 ND ) not determined.
after MhpB/MhpC treatment, again suggesting that no stable
rearrangement product was formed from cyclopropyl ring
opening.
Stereochemical Analysis of Enzymatic Products. We
hypothesized that upon formation of the semiquinone intermedi-
ate in the MhpB reaction, a reversible opening of the cyclopropyl
ring might be taking place, leading to cis/trans isomerization
of the cyclopropyl ring substituents and the production of cis
(14b) and trans (14a) isomers of cyclopropane-1,2-dicarboxylic
acid (see Figure 3). In order to test this hypothesis authentic
standards of cis- and trans-dimethylcyclopropane-1,2-dicar-
boxylate were synthesized by the method of Saal et al.18
Analysis of the standards by GC revealed that these diastereo-
isomers could be clearly resolved, and GCMS analyis under
electron impact conditions gave distinct retention times and
distinct fragmentation patterns for the two isomers (see Figure
4). The trans isomer 15a at retention time 1071 s showed a
fragment ion at m/z 127 (100%) and a major fragment ion (72%)
at m/z 98, whereas the cis isomer 15b at retention time 1105 s
showed a fragment ion at m/z 127 (100%) and a minor fragment
ion (32%) at m/z 99.
Incubation of trans-cyclopropyl analogue 7a with MhpB and
MhpC was then carried out, and the acidic products were
Figure 3. Mechanism for cis-trans isomerization of cyclopropyl ring
derivatized with diazomethane (see Scheme 2). Analysis by
GCMS revealed the presence of trans-product 15a and a small
peak corresponding to cis product 15b, both with identical
fragmentation patterns to authentic standards. The ratio of 15a:
15b was estimated at 94:6. Observation of 6% cis product
implied that cis-trans isomerization was taking place during
the enzymatic conversion of 7a.
Processing of cis-cyclopropyl analogue 7b with MhpB and
MhpC followed by methylation and GCMS analysis also gave
a mixture of trans 15a and cis 15b, with the trans product again
the major product. The ratio of 15a:15b was 90:10 in this case,
indicating that substantial cis-trans isomerization is taking
place. The similar ratios of 15a:15b obtained starting from
via reversible ring opening of a semiquinone radical intermediate.
In order to confirm the identity of the cyclopropyl-containing
ring-fission products 13a and 13b, they were further treated with
hydrolase enzyme MhpC. This enzyme catalyzes the next step
of the phenylpropionate pathway, namely the hydrolytic cleav-
age of the C-5/C-6 bond (see Figure 3).16 Treatment with E.
coli MhpC, purified as described elsewhere,17 led to the
disappearance of the UV absorptions corresponding to 13a and
13b, verifying that they correspond to authentic meta-ring-fission
products. No additional UV absorbent materials were observed
(16) Lam, W. W. Y.; Bugg, T. D. H. J. Chem. Soc., Chem. Commun.
1994, 1163-1164.
(17) Lam, W. W. Y.; Bugg, T. D. H. Submitted for publication.
(18) Van der Saal, W.; Reinhardt, R.; Seidenspinner, H. M.; Stawitz, J.;
Quast, H. Liebig’s Ann. Chem. 1989, 703-721.

Cis-Trans Isomerization of a Cyclopropyl Radical Trap
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8339
Figure 4. GCMS data for 15a and 15b. (A) Authentic standards, synthesized as described in the Experimental Section. (B) Enzymatic products,
generated from treatment of 7a with Alcaligenes eutrophus MpcI and E. coli MhpC, followed by treatment with diazomethane.
Table 2. Ratios of trans-15a:cis-15b Products Obtained from
Enzymatic Conversion of trans-7a and cis-7b Substrate Analoguesa
Scheme 2. Cis and Trans Isomers of Dimethyl
Cyclopropane-1,2-dicarboxylate Produced Enzymatically and
Synthetically
E. coli MhpB + MhpC
A. eutrophus MpcI + MhpC
substrate
15a:15b
15a:15b
7a
7b
94:6
90:10
85:15
ND
a Ratios determined by weighing of peaks obtained from GCMS
analysis. ND ) not determined.
dependent on the structure of the enzyme active site, presumably
being affected by the relative fit of cis vs trans isomers and the
freedom of rotation of the ring-opened radical. A summary of
the data obtained is given in Table 2, and the GCMS data
obtained from the A. eutrophus MpcI conversion are illustrated
in Figure 4.
Further Control Experiments. The observed cis-trans
isomerization could in theory be explained by other epimeriza-
tion mechanisms, in particular deprotonation/reprotonation of
the cyclopropyl protons of 7, 13, or 14. trans-Cyclopropyl
substrate 7a and cis-cyclopropane-1,2-dicarboxylic acid 14b
either 7a or 7b suggest that the rate of isomerization of the
cyclopropyl substituents is substantially faster than the rate of
the enzymatic reaction.
Processing of trans-analogue 7a with Alcaligenes eutrophus
dioxygenase MpcI, followed by treatment with E. coli MhpC
and derivatization as before, also gave both 15a and 15b by
GCMS analysis (see Figure 4), this time with a ratio of 15a:
15b of 85:15. The increased percentage of 15b obtained in
this case suggests that the ratio of cis to trans products is
1
were incubated in NaOD/D2O, and examined directly by H
2
NMR spectroscopy. No H exchange or epimerization was
observed after 24 h, implying that the protons attached to the
cyclopropyl ring are not appreciably acidic.
However, there is a possibility that hydrogen atom abstraction
is occurring at the stage of the meta-ring-fission products 13a

8340 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Spence et al.
Scheme 3. Other Examples of Cyclopropane Isomerization
Reactions Occurring via Free Radical-Induced Reversible
Ring Openings (See ref 21 and 22)
hydroquinone was isolated.22 The latter example also involves
the reversible opening of a cyclopropyl ring adjacent to a
semiquinone radical, similar to our observations.
We have established that there is no exchange of the protons
attached to the cyclopropyl ring during the MhpB- and MhpC-
catalyzed reactions, nor does exchange occur non-enzymatically.
Therefore, the only reasonable explanation for the observed cis-
trans isomerization is that an adjacent radical is formed during
the enzymatic processing of 7. Since the MhpC reaction
involves neither cofactors nor redox chemistry the radical
intermediate must be in the extradiol dioxygenase reaction.
Hence this behaviour is consistent with the existence of a
transient semiquinone intermediate (2) in the MhpB reaction.
The reversibility of radical-induced cyclopropyl ring openings
in chemical reactions has been demonstrated in several
instances.23-25 In these cases the equilibrium constant for
cyclopropyl ring opening is governed by the relative stability
of the ring-closed and ring-opened radicals. Thus, for example,
Bowry et al. have reported that a reversible opening of the
R-cyclopropylbenzyl radical can take place, although none of
the ring-opened radical could be detected, due to the high
stability of a benzylic radical relative to an alkyl radical.24 The
reversibility of MhpB-catalyzed opening of 7a and 7b can
therefore be explained in this case by the high stability of the
initial semiquinone radical.26 Since no re-arranged product in
which the cyclopropyl ring was opened was detected by NMR
spectroscopy or HPLC the equilibrium constant for the cyclo-
propyl ring opening would appear to lie strongly in favor of
the semiquinone radical. The similar ratios of 15a:15b obtained
above from enzymatic processing of 7a or 7b imply that the
rates of opening and reclosing of the cyclopropyl ring system
are fast compared with the rate of reaction of the semiquinone
radical with superoxide in the enzymatic reaction. However,
we have at present no insight into the absolute values of these
rate constants.
The existence of a semiquinone radical supports the proposed
mechanism for the MhpB reaction, involving binding of the
catechol substrate and dioxygen to iron(II), followed by electron
transfer to generate the semiquinone-iron(II)-superoxide
intermediate.10 However, the observation of radical character
at C-1 does not allow us to differentiate between subsequent
recombination of the semiquinone with superoxide at C-1 vs
C-2, since radical character can be found on every carbon center
in the semiquinone system via either electron delocalization or
iron-oxygen electron transfer (see Figure 5).
The enzymatic reaction could thence proceed either via
formation of a C-1 peroxy adduct followed by acyl migration
to give lactone 3 or by formation of a C-2 peroxy adduct
followed by alkenyl migration to give 3 (see Figure 5).10 Further
studies are required in order to distinguish between these
possibilities; however, we have found that 2,3-dihydroxycin-
namic acid (side chain -CHdCHCO2H) and 2,3-dihydroxy-
and 13b or during the MhpB-catalyzed reaction. Since the rate
of isomerization is apparently fast compared with the rate of
enzymatic processing, hydrogen atom abstraction should lead
to significant deuterium exchange from D2O into 15a and 15b.
Accordingly, the enzymatic processing of 7a by MhpB/MhpC
was carried out 98% D2O/2% H2O, and the derivatized products
examined by GCMS.29 The observed mass spectra of 15a and
15b again showed peaks at m/z 127, but in neither case was
any increase in intensity observed at m/z 128 corresponding to
deuterium incorporation. This experiment implies that there is
no significant hydrogen atom exchange on the cyclopropyl ring
during the MhpB and MhpC reactions. The only reasonable
explanation for the cis/trans isomerization is therefore a
reversible opening of the cyclopropyl ring.
Cis/trans isomerization could also occur prior to conversion
by dioxygenase MhpB. In order to test this 7a was treated on
a larger scale (10 mg) with MhpB (10 units), and the residual
substrate isolated and analyzed by 1H NMR spectroscopy.
Signals corresponding to residual 7a were clearly visible, but
no trace (<1%) of the clearly resolvable R proton of 7b was
observed, implying that the 6% trans-to-cis isomerization occurs
after an irreversible step in the enzymatic reaction.
Generation of a free semiquinone species from 7a was
attempted chemically and enzymatically, in order to examine
whether cis/trans isomerization could take place in solution.
Air oxidation of catechols under basic conditions and oxidation
by commercially available tyrosinase are both precedented to
involve semiquinone generation,19,20 and in both cases treatment
of catechol gave a characteristic EPR spectrum of a semiquinone
radical species. However, upon incubation of 7a in either
1
system, re-isolation and H-NMR spectroscopic analysis re-
vealed that no isomerization to 7b had taken place. Thus it
appears that the isomerization of 7a occurs only at the
dioxygenase active site: perhaps bidentate iron coordination by
the non-heme cofactor is also required.
Discussion
Cis-trans isomerization of a cyclopropyl radical trap has to
our knowledge not been observed in an enzymatic reaction.
However, at least two non-enzymatic examples have been
reported, as shown in Scheme 3. Tanner et al. observed upon
single electron reduction of ketone 16 cis/trans isomerization
of the cyclopropyl substituents, via a reversible opening of a
highly stabilized cyclopropylmethyl radical.21 Recently Zeng
et al. have reported that upon single electron reduction of cis-
substituted cyclopropylbenzoquinone 17 70% of the trans-
(22) Zeng, Z.; Cartwright, C. H.; Lynn, D. G. J. Am. Chem. Soc. 1996,
118, 1233-1234.
(23) Halgren, T. A.; Howden, M. E. H.; Medof, M. E.; Roberts, J. D. J.
Am. Chem. Soc. 1967, 89, 3051-3052.
(24) Bowry, V. W.; Lusztyk, J.; Ingold, K. U. J. Chem. Soc., Chem.
Commun. 1990, 923-925.
(25) Newcomb, M. Tetrahedron 1993, 49, 1151-1176.
(26) Depew, M. C.; Wan, J. K. S. The Chemistry of Quinonoid
Compounds; Patai, S., Rappoport, Z., Eds.; Wiley Press: New York, 1988;
Vol II, pp 963-1018.
(27) Christiansen, W. G. J. Am. Chem. Soc. 1926, 48, 1358-1365.
(28) Bo¨hme, H.; Severin, T. Arch. Pharm. 1957, 290, 405-412.
(29) Conversions with MpcI, kinetic analysis, and further control
experiments were not carried out with 7b due to its instability toward
lactonization and problems experienced in its re-synthesis. Attempts to
re-synthesize 7b gave lower conversions for the cyclopropanation of 11,
yielding impure samples of 7b.
(19) Beck, R.; Nibler, J. W. J. Chem. Educ. 1989, 66, 263-266.
(20) Kalyanaraman, B. Meth. Enzymol. 1990, 186, 333-343. Felix, C.
C.; Sealy, R. C. J. Am. Chem. Soc. 1981, 103, 2831-2836.
(21) Tanner, D. D.; Chen, J. J.; Luelo, C.; Peters, P. M. J. Am. Chem.
Soc. 1992, 114, 713-717.

Cis-Trans Isomerization of a Cyclopropyl Radical Trap
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8341
combined organic portions were initially extracted with 5% NaHCO3
solution (6 × 40 mL) to separate trans-2,3-dihydroxycinnamic acid
from its corresponding cis lactone 8-hydroxycoumarin. The remaining
organic layer was extracted further with K2CO3 (3 × 40 mL) in order
to obtain the cis lactone. Both of the aqueous layers were acidified to
pH 1-2 and re-extracted into ether (6 × 30 mL). Drying over MgSO4
and evaporation of the ether portions furnished an orange solid (3.8 g,
62%), identified as trans-2,3-dihydroxycinnamic acid and a cream solid
(1.7 g, 31%) which was identified as 8-hydroxycoumarin. The crude
products were recrystallized (ether/petroleum ether) to afford both trans-
2,3-dihydroxycinnamic acid and 8-hydroxycoumarin as white crystalline
solids.
trans-2,3-Dihydroxycinnamic acid: mp 201-203 °C [lit.28 mp 202
°C]; IR (Nujol) 3460, 3283, 3104, 1681, 1625, 1587, 1485, 1273, 1255
cm-1; 1H NMR (300 MHz, d6-acetone) 8.00 (1H, d, J ) 15.8 Hz, CH-
Ar), 7.06 (1H, dd, J ) 7.8, 1.4 Hz, o-Ph), 6.91 (1H, dd, J ) 7.7, 1.5
Hz, p-Ph), 6.69 (1H, t, J ) 7.9 Hz, m-Ph), 6.55 (1H, d, J ) 16.2 Hz,
CH-CO2H) ppm; 13C NMR (75.5 MHz, d6-acetone) 168.20, 145.30,
145.27, 140.43, 121.49, 119.50, 119.25, 118.05, 116.55 ppm; m/z 180
(M+, 58%), 162 (M - H2O, 99%), 134 (M - CO2H - H, 100%), 78
(31%).
8-Hydroxycoumarin (8): mp 161-162 °C [lit.28 mp 160 °C]; IR
(Nujol mull) 3384, 1711, 1603, 1578, 1273 cm-1; 1H NMR (300 MHz,
d6-acetone) 7.92 (1H, d, J ) 9.6 Hz, CH-Ar), 7.13 (3H, m, Ph), 6.39
(1H, d, J ) 9.6 Hz, CH-CO2R) ppm; 13C NMR (75.5 MHz, d6-acetone)
160.13, 145.19, 144.80, 143.22, 125.11, 120.47, 119.44, 119.00, 116.91
ppm; m/z 162 (M+, 100%), 134 (M - CO, 66%); HRMS M+ 162.0326,
C9H6O3 requires 162.0317.
Figure 5. Delocalization of radical character in the semiquinone
intermediate 2 and possible subsequent enzymatic reactions at C-1 or
C-2.
Ethyl 2,3-Dihydroxycinnamate. To a solution of 2,3-dihydroxy-
cinnamic acid (2.0 g, 11.1 mmol) in ethanol (50 mL) was added
concentrated H2SO4 (5 mL) and the resultant mixture was refluxed over
night. The mixture was then diluted with water (50 mL), and the
product extracted into ether (3 × 40 mL), washed with saturated 5%
NaHCO3 solution (2 × 30 mL), dried over MgSO4, and concentrated
in Vacuo to give an orange solid. Recrystallization (ether/petroleum
ether) furnished ethyl 2,3-dihydroxycinnamate as white crystals (2.04
g, 88%). Mp 137-139 °C; IR (Nujol mull) 3593, 3280, 1686, 1632,
phenoxyacetic acid (side chain -OCH2CO2H) are both efficient
substrates for oxidative cleavage by MhpB (see Table 1). If a
C-1 peroxy intermediate were formed, then one might expect
the introduction of electron-withdrawing and electron-donating
groups into the side chain to have a significant effect on the
rate of the Criegee rearrangement, and hence on kcat and Km.
The absence of such an effect appears more consistent with a
C-2 adduct, whose structure is precedented by the iridium(III)
model complex (6) mentioned above.12
The reversible ring opening of 7a and 7b provides further
evidence for the existence of stabilized radical intermediates in
iron-dependent oxygenase-catalyzed reactions.2 It is interesting
to speculate whether other cyclopropyl radical trap studies of
enzymatic reactions might also have involved reversible ring
openings which have escaped detection.
1
1590, 1506, 1323, 1195 cm-1; H NMR (300 MHz, d6-acetone) 7.99
(1H, d, J ) 16.2 Hz, CH-Ar), 7.10 (1H, dd, J ) 8.1, 1.5 Hz, o-Ph),
6.91 (1H, dd, J ) 8.1, 1.5 Hz, p-Ph), 6.72 (1H, t, J ) 7.7 Hz, m-Ph),
6.60 (1H, d, J ) 16.2 Hz, CH-CO2Et), 4.20 (2H, q, J ) 7.4 Hz,
CH2CH3), 1.29 (3H, t, J ) 7.3 Hz, CH2CH3) ppm; 13C NMR (75.5
MHz, d6-acetone) 166.76, 145.41, 145.04, 139.85, 121.64, 119.66,
119.57, 118.20, 116.53, 59.73, 13.82 ppm; m/z 208 (M+, 35%), 162
(M - C2H5OH, 100%), 134 (M - HCO2Et, 60%); HRMS M+
208.0725, C11H12O4 requires 208.0736.
Ethyl 2,3-Bis(methoxymethyleneoxy)cinnamate (9). To an ice-
cooled solution of ethyl 2,3-dihydroxycinnamic acid (0.5 g, 2.41 mmol)
in CH2Cl2 (25 mL) and N,N-diisopropylethylamine (dried over CaH2)
(1.17 g, 1.57 mL, 9.02 mmol) was added, dropwise, MOMCl (0.73 g,
0.70 mL, 9.02 mmol). The reaction mixture was stirred on ice for 1 h
before it was warmed to room temperature and stirred for a further 20
h. The resulting mixture was diluted with water (50 mL), extracted
into ether (3 × 30 mL), washed (10% NaHCO3, saturated NaCl), dried
over MgSO4, and concentrated in Vacuo to afford 9 as a yellow oil
(0.65 g, 90%). IR (liquid) 2980, 1711, 1635, 1578, 1477, 1441, 1262
cm-1; 1H NMR (300 MHz, CDCl3) 8.11 (1H, d, J ) 16.2 Hz, CH-Ar),
7.24 (1H, dd, J ) 7.4, 1.5 Hz, o-Ph), 7.18 (1H, dd, J ) 8.1, 1.5 Hz,
p-Ph), 7.04 (1H, t, J ) 8.1 Hz, m-Ph), 6.43 (1H, d, J ) 16.2 Hz, CH-
CO2Et), 5.20 (2H, s, CH2OCH3), 5.16 (2H, s, CH2OCH3), 4.25 (2H, q,
J ) 7.3 Hz, CH2CH3), 3.62 (3H, s, CH2OCH3), 3.59 (3H, s, CH2OCH3),
1.32 (3H, t, J ) 7.4 Hz, CH2CH3) ppm; 13C NMR (75.5 MHz, CDCl3)
167.01, 150.23, 145.92, 139.63, 129.59, 124.55, 120.23, 119.33, 118.03,
99.41, 95.12, 60.41, 57.82, 56.28, 14.30 ppm; m/z 296 (M+, 21%),
251 (M - CH2OCH3, 10%), 206 (30%), 45 (CH2OCH3+, 100%);
HRMS M+ 296.1271, C15H20O6 requires 296.1260.
Experimental Methods
General. 1H and 13C NMR spectra were recorded on a Joel 300-
MHz spectrometer in deuteriated solvents as described below. Mass
spectra and GCMS were recorded on a VG 70-250 mass spectrometer.
IR spectra were recorded on a 1600 series Perkin-Elmer FTIR infrared
spectrometer in NaCl discs. UV/visible spectra were recorded on a
Cary 1 UV/visible spectrophotometer using 1 mL quartz cells. ESR
spectra were recorded on a Bruker ECS106 ESR spectrometer by S.
Broadbridge (University of Southampton). GC was carried out using
a 25 m × 0.22 mm BP1 column with 25 µM film thickness and a
temperature gradient of 50 °C for 4 min followed by a 5 °C/min to
150 °C. HPLC was carried out on a Waters Associates chromatograph
using a Bio-Rad HPX-87H Organic Acids column (eluent 0.005 M
H2SO4 at 0.6 mL/min).
Chemicals. 2,3-Dihydroxyphenoxyacetic acid was synthesized from
pyrogallol according to the method of Christiansen.27 cis- and trans-
dimethylcyclopropane-1,2-dicarboxylates were synthesized according
to the method of van der Saal.18
trans-2,3-Dihydroxycinnamic Acid and 8-Hydroxycoumarin (8).
A solution of BBr3 (20 g, 80 mmol) in CH2Cl2 (50 mL) was maintained
at -78 °C under an atmosphere of N2. 2,3-Dimethoxycinnamic acid
(7.55 g, 34 mmol) in CH2Cl2 (50 mL) was added over 1 h with stirring.
The resultant mixture was allowed to warm to room temperature and
was stirred for a further 24 h. Careful addition of water (100 mL) to
the mixture was followed by extraction into ether (6 × 40 mL). The
Ethyl trans-2-(2,3-Bis(methoxymethyleneoxy)phenyl)cyclopropane-
1-carboxylate (10). NaH (60% mineral oil dispersion, 0.18 g, 4.5
mmol) was placed in a flask and washed with petroleum ether at 40-
60 °C (3 × 5 mL) by swirling and decanting. The system was
evacuated to remove all traces of petroleum ether. Powdered tri-
methoxysulfonium iodide (0.99 g, 4.5 mmol) was then added to the

8342 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Spence et al.
flask and the vessel filled with nitrogen. DMSO was added carefully
until the evolution of hydrogen ceased (∼20 min) to produced a milky
solution of the ylide. A solution of 9 (1.0 g, 3.38 mmol) in DMSO
(30 mL) was added dropwise to the ylide. The resulting mixture was
heated at 50 °C for 2 h and allowed to stir at room temperature for a
further 12 h. The resulting mixture was concentrated on a rotary
evaporator and diluted with water (100 mL). The product was extracted
into ether (4 × 50 mL), washed with water (5 × 50 mL), and dried
over MgSO4 and the ether was removed in Vacuo to yield a brown oil.
The crude material was chromatographed (silica gel, 50% EtOAc in
light petroleum) to give 10 as a pale yellow oil (0.55 g, 52%). IR
iodide (1.15 g, 5.35 mmol) was then added to the flask and the vessel
filled with nitrogen. DMSO was added carefully until the evolution
of hydrogen ceased (∼20 min) to produce a milky solution of the ylide.
A solution of 11 (0.63 g, 3.06 mmol) in DMSO (15 mL) was added
dropwise to the ylide. The resulting mixture was heated at 65 °C for
12 h and allowed to stir at room temperature for a further 24 h. The
resulting mixture was concentrated on a rotary evaporator and diluted
with water (100 mL). The product was extracted into ether (4 × 50
mL), washed with water (5 × 30 mL), and dried over MgSO4, and the
ether was removed under Vacuo to yield a brown oil. The crude
material was chromatographed (silica gel, 50% EtOAc in light
petroleum) to give a pale brown oil containing a 2:3 mixture of 12:11
(overall yield 0.55 g, 52%) which was not further separable.29 1H NMR
(300 MHz, CDCl3) 7.07-6.98 (3H, m, Ph), 5.19 (2H, s, CH2OCH3),
3.49 (3H, s, CH2OCH3), 2.57 (1H, td, J ) 8.8, 5.1 Hz, CH-Ar), 2.32
(1H, ddd, J ) 9.6, 7.4, 5.1 Hz, CHCO2R), 1.73 (1H, ddd, J ) 9.6, 8.1,
4.4 Hz, CH2), 0.98 (1H, dt, J ) 10.3, 5.1 Hz, CH2) ppm; 13C NMR
(75.5 MHz, CDCl3) 166.69, 144.45, 139.93, 124.11, 123.26, 120.82,
115.23, 95.48, 56.35, 20.13, 17.53, 14.48 ppm; m/z 220 (M+, 13%),
190 (M+ - CH2O, 17%), 45 (CH2OCH3+, 100%); HRMS M+ 220.0733,
C12H12O4 requires 220.0736.
(liquid film) 3040, 2956, 1722, 1602, 1583, 1475, 1439, 1271 cm-1
;
1H NMR (300 MHz, CDCl3) 6.97-6.89 (2H, m, o-, m-Ph), 6.46 (1H,
dd, J ) 7.4, 1.5 Hz, p-Ph), 5.14 (2H, s, CH2OCH3), 5.11 (2H, s,
CH2OCH3), 4.12 (2H, q, J ) 7.4 Hz, CH2CH3), 3.54 (3H, s, CH2-
OCH3), 3.45 (3H, s, CH2OCH3), 2.85 (1H, ddd, J ) 9.6, 6.6, 4.4 Hz,
CH-Ar), 1.81 (1H, ddd, J ) 7.3, 4.8, 4.8 Hz, CHCO2Et), 1.56 (1H,
ddd, J ) 8.3, 4.4, 4.4 Hz, CH2), 1.28 (1H, m, CH2), 1.23 (3H, t, J )
7.4 Hz, CH2CH3) ppm; 13C NMR (75.5 MHz, CDCl3) 173.36, 149.83,
145.92, 134.42, 124.34, 118.15, 114.63, 99.07, 95.03, 60.56, 57.42,
56.12, 23.77, 20.87, 16.25, 14.22 ppm; m/z 310 (M+, 11%), 234 (34%),
45 (CH2OCH3+, 100%); HRMS M+ 310.1410, C16H22O6 requires
310.1416.
cis-2-(2,3-Dihydroxyphenyl)cyclopropane-1-carboxylic Acid (7b).
To a 3:2 mixture of 11 and 12 (0.30 g), dissolved in methanol (1 mL),
was added a solution of NaOH (0.5 M, 20 mL). The reaction was
stirred at room temperature for 3 h. On completion (by TLC) the
mixture was acidified to pH 1 and stirred at room temperature for a
further 4 h. After this time the products were extracted into ether (3
× 50 mL), washed with brine (2 × 30 mL), dried over MgSO4, and
concentrated in Vacuo to give a brown solid containing both 7b and 8
(overall recovery 0.25 g). The crude mixture was dissolved in EtOAc
(20 mL) and extracted in portions of 50 mM potassium phosphate buffer
at pH 7.0 (5 × 10 mL). The combined aqueous layers were acidified
to pH 2.0, re-extracted into EtOAc, and dried over MgSO4 and the
solvent was removed in Vacuo to afford 70 mg of an orange solid,
shown by 1H NMR spectroscopy to contain a 4:1 mixture of 7b and 8,
and small amounts of the lactone form of 7b.29 IR (solution) 3422,
trans-2-(2,3-Dihydroxyphenyl)cyclopropane-1-carboxylic Acid
(7a). To a solution of 10 (0.90 g, 2.90 mmol) in methanol (1 mL) was
added a solution of NaOH (0.5 M, 20 mL). The reaction was stirred
at room temperature for 3 h. On completion the mixture was acidified
(2 M, HCl), extracted into ether (3 × 50 mL), washed with brine (2 ×
30 mL), dried over MgSO4, and concentrated in Vacuo to give a brown
carboxylic acid product (0.69 g, 85%). The carboxylic acid (0.6 g,
2.13 mmol) was taken up in methanol (1 mL) and placed under an
atmosphere of N2. To this solution was added 2 M HCl (30 mL) and
the resulting mixture was stirred at room temperature for 12 h. The
product was then extracted into ether (3 × 30 mL), washed with brine
(30 mL), dried over MgSO4, and evaporated to dryness to furnish an
orange solid. Recrystallization (ether/light petroleum) afforded 7a as
white crystals (0.30 g, 73%). A small amount of this material (∼1.0
mg) was further purified by HPLC on an organic acid column eluted
with 0.005 M H2SO4 which gave rise to a peak at 160 min. Mp 142-
1
2400, 1696, 1590 cm-1; H NMR (300 MHz, d6-acetone) 6.64 (1H,
dd, J ) 7.4, 1.5 Hz, o-Ph), 6.62 (1H, dd, J ) 7.4, 1.5 Hz, p-Ph), 6.55
(1H, t, J ) 7.4 Hz, m-Ph), 2.55 (1H, q, J ) 8.4 Hz, CH-Ar), 2.10 (1H,
ddd, J ) 8.4, 7.2, 5.1 Hz, CHCO2H), 1.53 (1H, ddd, J ) 8.1, 5.1, 4.5
Hz, CH2), 1.32 (1H, ddd, J ) 8.0, 7.2, 4.5 Hz, CH2) ppm; m/z 194
(M+, 43%), 176 (M - H2O, 97%), 147 (M - HCO2H - H, 100%);
HRMS M+ 194.0585, C10H10O4 requires 194.0579.
1
144 °C; IR (solution) 3399, 3169, 1690, 1592, 1476, 1283 cm-1; H
NMR (300 MHz, d6-acetone) 6.72 (1H, dd, J ) 7.4, 1.5 Hz, o-Ph),
6.62 (1H, t, J ) 7.4 Hz, m-Ph), 6.42 (1H, dd, J ) 7.4, 1.5 Hz, p-Ph),
2.71 (1H, ddd, J ) 8.8, 6.6, 4.4 Hz, CH-Ar), 1.82 (1H, ddd, J ) 8.1,
4.4, 4.4 Hz, CH2), 1.45 (1H, ddd, J ) 8.1, 4.4, 4.4 Hz, CH2), 1.36
(1H, ddd, J ) 8.8, 7.4, 4.4 Hz, CH2) ppm; 13C NMR (75.5 MHz, d6-
acetone) 174.75, 145.11, 144.82, 127.30, 120.98, 117.30, 113.83, 22.85,
21.31, 15.86 ppm; m/z 194 (M+, 61%), 176 (M - H2O, 100%), 147
(M - HCO2H - H, 99%), 77 (C6H5+, 52%), 51 (C4H3+, 30%); HRMS
M+ 194.0574, C10H10O4 requires 194.0579.
Enzyme Assays. Escherichia coli (2,3-dihydroxyphenyl)propionate
1,2-dioxygenase (MhpB) was purified as previously described.14
Alcaligenes eutrophus dioxygenase MpcI was purified to 50% homo-
geneity from an overexpressing strain of E. coli JM109/pAE166, as
described elsewhere.15 Escherichia coli 2-hydroxy-2-ketonona-2,4-
diene-1,9-dioic acid 5,6-hydrolase (MhpC) was purified from an
overexpressing strain of W3110/pTB9 as described elsewhere.17
MhpB and MpcI were activated prior to use by addition of 100 mM
ammonium iron(II) sulfate (5 µl) and 100 mM sodium ascorbate (5
µL) to apoenzyme (100 µL) at 0 °C, as previously described.14 MhpB
and MpcI were assayed by addition of re-activated enzyme to a solution
of substrate in 50 mM Tris at pH 8.0 and 20 °C. Appearance of ring-
fission products were observed for 1 at 394 nm (ꢀ ) 15 600 M-1
cm-1),14 for 7a at 405 nm, and for 2,3-dihydroxycinnamic acid at 460
nm. For 2,3-dihydroxyphenoxyacetic acid, where no strong absorbance
change was observed, O2 consumption was measured electrochemically
using a Clark-type oxygen electrode (Rank Bros.). MhpC was assayed
by monitoring the disappearance of the ring-fission products by UV
spectroscopy.17
Substrate analogues were assayed in triplicate in the concentration
range 0.5-5 Km. Apparent Km values were calculated from Lineweaver/
Burk and Eadie/Hofstee plots. kcat values were obtained using the
oxygen electrode: maximal rate values were compared with the
maximal rate of the natural substrate 1, whose kcat is previously
established.14 Large-scale conversions using MhpB were carried out
in 50 mM Tris buffer (pH 8.0), followed by acidification to pH 1 and
extraction of products into ethyl acetate.
8-(Methoxymethyleneoxy)coumarin (11). To an ice-cooled solu-
tion of 8 (0.30 g, 1.85 mmol) in CH2Cl2 (20 mL) and N,N-
diisopropylethylamine (dried over CaH2) (0.47 g, 0.64 mL; 3.70 mmol)
was added, dropwise, MOMCl (0.30 g, 0.28 mL, 3.70 mmol). The
reaction mixture was stirred on ice for 1 h before it was warmed to
room temperature and stirred for a further 20 h. The resulting mixture
was diluted with water (30 mL), extracted into ether (3 × 30 mL),
washed (10% NaHCO3, saturated NaCl), dried over MgSO4, and
concentrated in Vacuo to afford the desired product as a yellow oil
(0.37 g, 97%). IR (liquid film) 2962, 1718, 1609, 1567, 1463, 1153,
1
1048 cm-1; H NMR (300 MHz, CDCl3) 7.68 (1H, d, J ) 9.6 Hz,
CH-Ar), 7.31 (1H, dd, J ) 7.3, 1.5 Hz, o-Ph), 7.15 (1H, t, J ) 8.1 Hz,
m-Ph), 7.10 (1H, dd, J ) 8.1, 1.5 Hz, p-Ph), 6.39 (1H, d, J ) 9.6 Hz,
CH-CO2R), 5.26 (2H, s, CH2OCH3), 3.50 (3H, s, CH2OCH3) ppm;
13C NMR (75.5 MHz, CDCl3) 160.22, 144.59, 144.43, 143.72, 124.34,
120.94, 119.74, 118.60, 116.74, 95.51, 56.48 ppm; m/z 206 (M+, 48%),
176 (M - CH2O, 82%), 77 (C6H5+, 19%), 45 (CH2OCH3+, 100%);
HRMS M+ 206.0580, C11H10O4 requires 206.0579.
cis-2-(2-Hydroxy-3-(methoxymethylene)phenyl)cyclopropane-1-
carboxylic Acid Lactone (12). NaH (60% mineral oil dispersion, 0.21
g, 5.35 mmol) was placed in a flask and washed with light petroleum
(3 × 5 mL) by swirling and decanting. The system was evacuated to
remove all traces of light petroleum. Powdered trimethyloxosulfonium
GC-MS Experiments. To a solution of 7a (∼2 mg) in H2O (30
mL) was added Tris buffer (pH 8.0, 1 mL, 1.0 M). An aliquot of

Cis-Trans Isomerization of a Cyclopropyl Radical Trap
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8343
reactivated MhpB (200 µL, 70 units/mL) was added and the resulting
solution was stirred slowly at room temperature. Subsequent aliquots
of reactivated MhpB (100 µL) were added at 2, 4, and 6 min. After
10 min, aliquots of MhpC (20 × 30 µL, 25 units/mL) were added over
a further 10 min. The solution was acidified (concentrated HCl) to
pH 2 and the organic products were extracted into EtOAc (6 × 25
mL) and dried over MgSO4 and the solvent was removed in Vacuo.
The resulting residue was dissolved in ether (∼3 mL) and treated with
excess diazomethane. The excess diazomethane was removed by
bubbling nitrogen through the ethereal solution and the resulting solution
was concentrated in Vacuo to generate the derivatized enzyme products
as an orange oil. The oil was submitted for GC-MS along with a sample
of chemically synthesized standards. 7b was converted enzymatically
exactly as described for 7a. Conversion of 7a by MpcI was carried
out using reactivated MpcI (total volume 3 mL, ∼5 units/mL).
Conversion of 7a in D2O was carried out as above, except that the
Tris buffer had been lyophilized and re-suspended in D2O (30 mL). 7a
(2 mg) was stirred in the deuteriated buffer for 30 min prior to addition
of re-activated MhpB (3 × 100 µL in H2O, ∼70 units/mL) over 10
min. MhpC (20 × 10 µL in H2O, ∼80 units/mL) was added over 10
min, and the products isolated and derivatized as described above. The
final proportion of D2O was approximately 98%.
Enzyme-catalyzed formation of catechol semiquinone was carried
out by addition of mushroom tyrosinase (Sigma, 10 µg) to a solution
of catechol (0.2 mg) in 50 mM potassium phosphate buffer (pH 8.0),
total volume 1.0 mL. Semiquinone formation was observed by UV
spectroscopy (λmax 410 nm) and ESR spectroscopy.20 Semiquinone
formed from 7a was also observed by UV spectroscopy (λmax 425 nm).
To determine if isomerization of 7a occurs as a result of semiquinone
formation a sample of 7a (15 mg) was dissolved in 50 mM deuteriated
potassium phosphate buffer (pH 8.0) and examined by 1H NMR
spectroscopy. This sample was treated with tyrosinase (20 µg) of stirred
for 20 min, forming a yellow solution. The sample was examined by
1H NMR spectroscopy after 3 and 24 h. Only signals corresponding
to 7a were observed.
Acknowledgment. We would like to thank Dr. P. Fortnagel
(University of Hamburg) for the gift of JM109/pAE166 contain-
ing the Alcaligenes eutrophus mpcI gene, and Dr. Jonathan
Sanvoisin and John Pollard (University of Southampton) for
assistance with enzyme purification. This work was supported
by the Biomolecular Sciences panel of the Biotechnology and
Biological Sciences Research Council (grant GR/H55406).
Semiquinone Generation. Generation of the catechol semiquinone
under alkaline conditions was carried out by the method of Beck and
Nibler.19 In order to determine whether any isomerization of 7a occurs
as a result of semiquinone formation a 0.5 M solution of 7a in methanol
(1 mL) was mixed with 1 M sodium hydroxide (100 µL) until the
solution became yellow. The mixture was stirred for a further 10 min,
diluted with water, and acidified to pH 4. The product was extracted
into ethyl acetate, dried (MgSO4), and evaporated to dryness. Analysis
Supporting Information Available: 1H NMR spectra of
7a and 7b and GCMS data for 7a and 7b with MhpB/MhpC (8
pages). See any current masthead page for ordering and Internet
access instructions.
1
by H NMR spectroscopy revealed the presence of only 7a.
JA9607704