Mechanism of enzymatic Birch reduction: Stereochemical course and exchange reactions of benzoyl-CoA reductase

Article abstract of DOI:10.1021/ja805091w

Dearomatizing benzoyl-coenzyme A reductases (BCR) from facultatively anaerobic bacteria are key enzymes in the anaerobic degradation of aromatic compounds. They catalyze the ATP-dependent reduction of benzoyl-CoA (BCoA) to cyclohexa-1,5-diene-1-carboxyl-CoA (dienoyl-CoA). A Birch reduction mechanism involving alternate electron transfer and protonation steps has been proposed for BCR. In this work we reacted BCoA in H₂O and D₂O, and d₅-BCoA in H₂O with BCR and the second enzyme of the pathway, dienoyl-CoA hydratase (DCH). The 1,4 hydration product formed from the dienoyl-CoA, 6-hydroxycyclohex-1-ene-1-carbonyl-CoA, was analyzed by several NMR techniques. The results obtained indicate that BCR stereoselectively forms the trans-dienoyl-CoA product, and DCH stereoselectively catalyzes a trans-1,4 water addition. Moreover, unexpected proton exchanges at C-2 and C-6 were observed. They indicate that a free radical intermediate with an unusual low pK_a is formed during BCR catalysis. This finding provides evidence for the proposed Birch reduction mechanism of BCR and is in agreement with the establishedradical mechanism of homologous α-hydroxyacyl-CoA dehydratases. Copyright

Full text of DOI:10.1021/ja805091w

Published on Web 10/01/2008
Mechanism of Enzymatic Birch Reduction: Stereochemical Course and
Exchange Reactions of Benzoyl-CoA Reductase
Ba¨rbel Thiele,† Oliver Rieder,‡ Bernard T. Golding,§ Michael Mu¨ller,‡ and Matthias Boll*,†
Institute of Biochemistry, UniVersity of Leipzig, 04103 Leipzig, Germany, Institute of Pharmaceutical and Medical
Chemistry, UniVersity of Freiburg, 79104 Freiburg, Germany, and School of Chemistry, Newcastle UniVersity, NE1
7RU Newcastle, U.K.
Received July 2, 2008; E-mail: boll@uni-leipzig.de
Scheme 1. Reaction Catalyzed by BCR and DCH in D2O
The Birch reduction is a widely used synthetic tool in organic
chemistry that achieves 1,4-dihydro additions to benzenoid and other
aromatic compounds. The reaction proceeds by alternate electron
transfer and protonation steps to the aromatic ring and requires
solvated electrons, which are usually generated by dissolving an
alkali metal in liquid ammonia.1 Considering these nonphysiological
conditions it is remarkable that a similar reaction exists in biology:
the dearomatizing benzoyl-coenzyme A reductase (BCR) plays a
key role in the anaerobic degradation of aromatic compounds.2-4
Table 1. Signal Couplings in H-H COSY and NOESY NMR
Experiments; the Relative Integrals of 6-OH-enoyl-CoA 3 Formed
from BCoA in H2O (a), d5-BCoA in H2O (b), and BCoA in D2O (c)
Are Shown
In anaerobic bacteria many low-molecular aromatic growth sub-
strates are converted to the central intermediate benzoyl-CoA
(BCoA), which serves as substrate for BCR.5 The enzyme catalyzes
the reduction of BCoA (1) to cyclohexa-1,5-diene-1-carboxyl-CoA
(dienoyl-CoA, 2) rather than the kinetically favored 2,5-dienoyl-
CoA isomer (Scheme 1).6 A mechanism similar to the classical
Birch reduction has been suggested in which the rate limiting first
electron transfer yields a radical anion.7,8 The CoA ester moiety is
considered to stabilize this intermediate by formation of a relatively
stable thioester ketyl radical. Remarkably, BCR couples electron
transfer to the aromatic ring from the donor reduced ferredoxin
(Fd) to a stoichiometric ATP hydrolysis, a reaction that has long
been considered as an exclusive feature of nitrogenase.2,9 The
oxygen sensitive BCR has so far only been isolated from the
facultatively anaerobic bacterium Thauera aromatica.2
signal (ppm) assignment
H-H-COSY
NOESY
a
b
c
1.5
H-4;5
1.7; 2.03;
2.18; 4.51
1.5
1.5; 2.18; 7.1 1.5; 2.18; 7.1
1.5; 2.03; 7.1 1.5; 2.03; 7.1
1.5
2.03; 2.18
1.7; 2.03;
3
<2
>2
2.18; 4.51
1.5
1.7
H-5
H-3
H-3
H-6
H-2
1
1
1
1
1
1
<0.1
1
0.26
1
1
2.03
2.18
4.51
7.1
0.7
0.9
0.65
0.28
1.5
2.03; 2.18
Table 2. Assignment of the Proton Signals of 6S-6-OH-enoyl-CoA (3)
Previous attempts to elucidate the stereochemistry of the BCR reac-
tion (syn or anti dihydro addition) by NMR analysis of the dienoyl-
CoA product were not successful due to ambiguous data obtained
from the protons at C-4. To overcome these problems we aimed to
determine the stereochemistry at the level of the 1,4-hydration
product of dienoyl-CoA, 6-hydroxycyclohex-1-ene-1-carboxyl-CoA
(6-OH-enoyl-CoA, 3), which is formed by the subsequent enzyme
in the pathway, dienoyl-CoA hydratase (DCH)10 (Scheme 1). We
present results obtained from reacting BCoA in H2O/D2O, and d5-
BCoA in H2O with BCR and DCH.
label
proton signal (ppm)
calculated
label
proton signal (ppm)
calculated
2a
3a
3b
4a
7.10
2.18
2.03
1.50
6.95
2.37
2.17
1.81
4b
5a
5b
6a
1.50
1.70
1.50
4.51
1.81
2.05
1.89
4.89
additions to 1, further NMR experiments were conducted. The
results of H-H COSY and NOESY experiments are summarized
in Table 1.
The copurification of BCR/DCH was conducted as described with
the expected specific activities (for BCR 0.25 µmol min mg-1);
reversed phase HPLC analysis confirmed that BCoA was converted
to dienoyl-CoA and 6-OH-enoyl-CoA in a 1:1 ratio.6 The CoA
esters were synthesized as described.6 Enzymatic conversion of 1
mM CoA esters was carried out under a N2-atmosphere in 10 mL
of the established assay buffer using 7.5 mM Ti(III)-citrate as
electron donor.2 The reaction products were isolated by preparative
reversed phase HPLC and freeze-dried.
The signal at 1.5 ppm showed a relative integral of three protons,
which derive from the protons on more than one carbon atom. The
downfield signal at 7.1 ppm is assigned to the alkene proton (H-
2), whereas the signal at 4.51 belongs to the carbinol proton (H-6).
The signals at 2.03 and 2.18 showed COSY crosspeak signals with
the alkene proton and are therefore assigned to the H-3 protons.
The signal at 1.5 ppm showed COSY crosspeaks to all other protons
except to that assigned to the 7.1 ppm signal. Thus the protons
accounting for this signal are at C-4 and C-5. The 1.7 ppm signal
showed only crosspeaks with the 1.5 ppm signal, which points to
a proton at C-5. The NOESY experiments support these findings.
The data were additionally refined by calculating the 1H NMR shifts
using the Perch-NMR-Predictor 1.3 software (Table 2).
Assignment of NMR Data Obtained from 6-OH-enoyl-CoA.
The 1H NMR spectrum of nondeuterated 3 exhibited signals at 1.5,
1.7, 2.03, 2.18, 4.51 and 7.1 ppm accounting for eight protons as
described.6 To elucidate the relative stereochemistry for the
† University of Leipzig.
‡ University of Freiburg.
§ Newcastle University.
If formed in H2O, compound 3 should carry three additional
hydrogens from the solvent at C-3, C-4, and C-5 (Scheme 1 and Table
9
14050 J. AM. CHEM. SOC. 2008, 130, 14050–14051
10.1021/ja805091w CCC: $40.75
2008 American Chemical Society

C O M M U N I C A T I O N S
Scheme 2. Differentially Labelled 6-OH-enoyl-CoA Species
Scheme 4. (A) Proposal for Birch Reduction Mechanism of BCR
and (B) Comparison with the ꢀ Deprotonation of the Methyl
Thiopropionate Radical 13 to the Radical Anion 14
Formed from BCoA in D2Oa
a Relative abundance: 4, 25-30%; 5, 55-60%; 6 and 7, 5-10% each;
calculations were derived from 1H-NMR peak integrals, an overalying signal
from a CoA-derived proton was subtracted.
Scheme 3
Mechanism of BCR. The unusual proton exchanges at C-2 and
C-6 during the two-step formation of 6-OH-enoyl-CoA cannot be
explained by DCH reaction. Instead, exchanges on the level of a radical
intermediate during BCR catalysis appear to be most plausible (Scheme
4). After the first electron transfer, the highest electron densities of
the resulting radical anion are at the C-2, C-4, and C-6 positions with
10a and 10b representing resonance structures.7 Protonation and further
conversion of 10a yield 11a and 12, protonation of 10b yields 11b.
The alternative protonations are both plausible but require a sterically
flexible donor (e.g., H2O). The pKa of enoyl thioester radicals, such
as 11b, are remarkably low (pKa ) 14 for the ꢀ deprotonation of 13
to 14).12 The resonance-based stability of 10b and the proposed partial
protonation of the carbonyl oxygen by the enzyme should shift the
pKa to a value below 10, which could explain the observed exchanges.
As the free radical 11b is more stable than 11a, the latter should be
reduced more readily. Moreover BCR may govern selective electron
transfer to 11a but not to 11b.
The formation of a ketyl radical intermediate has recently been
demonstrated during the mechanistically sophisticated dehydration
of an R-hydroxyacyl-CoA compound.13 The corresponding dehy-
dratases are the only enzymes that are homologous to BCR, and
similar mechanisms via radical intermediates have been proposed
many years ago.8
Conclusions. The results obtained provide the first evidence that
BCR reaction stereoselectively yields the trans-dienoyl-CoA prod-
uct. Moreover, the unexpected exchanges at C-2 and C-6 provide
evidence for the proposed Birch reduction mechanism. They can
be explained by the low pKa of a radical intermediate, and are in
perfect agreement with the established radical mechanism of
homologous R-hydroxyacyl-CoA dehydratases.
2). However, NMR analysis of 3 in D2O or from d5-BCoA in H2O
revealed four different products with additional deuterium/hydrogen
incorporations at C-2 and C-6. Thus, in addition to the expected
trideuterated (4), tetradeuterated (5, 6) and pentadeuterated (7)
compounds were also identified, indicating unexpected exchange
reactions at C-2 and C-6 (Scheme 2). HSQC-NMR analysis revealed
that no carbon carries more than one incorporated H or D from the
solvent; each of the carbons in position 3-5 carry one hydrogen and
one deuterium in all compounds.
Stereochemistry of DCH Reaction. For easier presentation of the
data obtained R and S are henceforth used. Here they are arbitrarily
chosen as only the relative stereochemistry could be solved (e.g., 6S-
or 6R-6-OH-enoyl-CoA). Notably, the ring nomenclature inverts during
the two-step formation of 6-OH-enoyl-CoA by BCR and DCH (Scheme
1
1). H NMR of 3 shows no signal at 1.7 ppm (Table 1). Thus, the
position 5a carries a deuterium and position 5b is hydrogenated after
the two-step reaction process (Table 2). The integral of the signal at
2.18 ppm was much lower than that of the signal at 2.03 ppm, which
suggests that the deuteration mainly occurred at position 3a. These
findings yield the structure 8a. The second step introduces a deuterium
at C-3 (previously C-5) and a deuteroxy group at C-6 (C-2). C-3 and
C-6 are S configured. Consequently the hydration step occurred as a
1,4-anti addition, which is consistent with the mechanism of 1,2-syn-
additions to 2-monoenoyl-CoA compounds such as crotonyl-CoA.11
Stereochemistry of BCR Reaction. The stereocenters created
by the BCR reaction are located at C-4 and C-5 in 8. From the
considerations made above the C-3-S and the C-5-R configurations
are established; these hydrogen atoms/deuterons are cis configured.
However, it was not possible to deduce the configuration of C-4
directly via correlations with C-3 or C-5.
Supporting Information Available: Full NMR spectra and results
from MOPAC calculations. This material is available free of charge
via the Internet at http://pubs.acs.org.
References
Calculations using MOPAC and Austin Model 1 theory revealed
that the distance between H-3b and H-4b (cis) is 2.41 Å. The
distance between H-3b and H-2 is 2.62 Å, whereas the distance
between H-3b and H-4a (trans) is 3.09 Å (Supporting Information).
We therefore applied quantitative NOESY techniques and monitored
the reflexes at 1.5 ppm (H-4a and H-4b) and at 7.1 ppm (H-2) when
irradiating the proton at 2.03 ppm (H-3b). If the proton at C-4 is
located at 4b, a stronger quantitative response than observed at 7.1
ppm is expected; if located at 4a, a much smaller response should
be observed. Only a clear response at 7.1 ppm was detected but
none at 1.5 ppm, while in the unlabeled compound a response at
1.5 ppm is visible (Table 1). Therefore we assigned the proton to
the 4a position completing the stereochemistry of both reactions
(8b). Accordingly, the BCR catalyzed reduction is trans selective.
(1) Birch, A. J.; Hinde, A. L.; Radom, L. J. Am. Chem. Soc. 1981, 103, 284–9.
(2) Boll, M.; Fuchs, G. Eur. J. Biochem. 1995, 234, 921–33.
(3) Boll, M. J. Mol. Microbiol. Biotechnol. 2005, 10, 132–42.
(4) Boll, M. Biochim. Biophys. Acta 2005, 1707, 34–50.
(5) Gibson, J.; Harwood, C. S. Annu. ReV. Microbiol. 2002, 56, 345–69.
(6) Boll, M.; Laempe, D.; Eisenreich, W.; Bacher, A.; Mittelberger, T.; Heinze,
J.; Fuchs, G. J. Biol. Chem. 2000, 275, 21889–95.
(7) Mo¨bitz, H.; Boll, M. Biochemistry 2002, 41, 1752–8.
(8) Buckel, W.; Keese, R. Angew. Chem., Int. Ed. 1995, 34, 1502–6.
(9) Unciuleac, M.; Boll, M. Proc. Natl. Acad. Sci.U.S.A. 2001, 98, 13619–24.
(10) Laempe, D.; Eisenreich, W.; Bacher, A.; Fuchs, G. Eur. J. Biochem. 1998,
255, 618–27.
(11) Willadsen, P.; Eggerer, H. Eur. J. Biochem. 1975, 54, 247–52.
(12) Smith, D. M.; Buckel, W.; Zipse, H. Angew. Chem., Int. Ed. 2003, 42,
1867–1870.
(13) Kim, J.; Darley, D. J.; Buckel, W.; Pierik, A. J. Nature 2008, 452, 239–42.
JA805091W
9
J. AM. CHEM. SOC. VOL. 130, NO. 43, 2008 14051