The first synthetic application of a monooxygenase employing indirect electrochemical NADH regeneration

Full text of DOI:10.1002/1521-3773(20010105)40:1<169::AID-ANIE169>3.0.CO;2-T

COMMUNICATIONS
aza-RGD mimetics showed a clear preference for the avb3
receptor.
The First Synthetic Application of a
Monooxygenase Employing Indirect
Lead structure 5 was found to exhibit a relatively high
degree of polarity; an unfavorable pharmacokinetic profile
can thus be expected for this compound. However, it has been
shown in the past that a hydrophobic residue in the b-position
to the carboxy group is tolerated by the avb3 receptor.[4d] For
this reason, we replaced the terminal carboxamide group with
a phenyl residue. The less polar RGD mimetic 10 exhibited
similar selectivity and considerably increased activity on the
avb3 receptor compared to the polar mimetic 5.
The results obtained would appear to support our concept
for identifying new, low molecular weight integrin ligands
through application of combinatorial solid-phase synthesis,
biological on-bead evaluation, and mass spectrometry to the
selected compounds.
Electrochemical NADH Regeneration**
Frank Hollmann, Andreas Schmid,* and
²
Eberhard Steckhan
In memory of Eberhard Steckhan
One of the most important challenges in applying mono-
oxygenase reactions in vitro is to find an effective regener-
ation system for the necessary co-enzyme (mostly
NAD(P)H). The well-established methods for the regener-
ation of the nicotinamide co-enzyme mainly consist of an
enzyme-coupled approach utilizing formate dehydrogena-
se[1, 4c] (for NAD(P)H) or glucose-6-phosphate dehydrogenase
(for NADPH).[2] Additionally, non-enzymatic redox catalysts
have been developed and successfully applied to NAD(P)H-
dependent dehydrogenases.[3] Thus only the producing en-
zyme and a mediator together with the electrode, as a source
of reducing equivalents, are needed.
Here we report on the first application of an isolated
monooxygenase with an indirect electrochemical regenera-
tion of NADH. The enzyme employed is the 2-hydroxybi-
phenyl-3-monooxygenase (HbpA, E.C. 1.14.13.44), a member
the class of flavine-dependent monooxygenases, from P. aze-
laica.[4] The homotetramer with a total mass of 256 kDa
catalyzes the specific ortho-hydroxylation of several a-sub-
stituted phenol derivatives (Scheme 1). To the best of our
knowledge no chemical counterpart with comparable specif-
icity is known.
Received: June 30, 2000
Revised: October 9, 2000 [Z15368]
[1] a) F. Balkenhohl, C. von dem Bussche-Hünnefeld, A. Lansky, C.
Zechel, Angew. Chem. 1996, 108, 2436 ± 2488; Angew. Chem. Int. Ed.
Engl. 1996 35, 2289 ± 2337, and references therein; b) P. Wentworth,
K. D. Janda, Curr. Opin. Biotechnol. 1998, 9, 109 ± 115, and references
therein; c) R. S. Houghten, C. Pinilla, J. R. Appel, S. E. Blondelle,
C. T. Dooley, J. Eichler, A. Nefzi, J. M. Ostresh, J. Med. Chem. 1999,
42, 3743 ± 3778, and references therein.
[2] A. Furka, F. Sebestyen, M. Asgedom, G. Dibo, Int. J. Pept. Protein Res.
1991, 37, 487 ± 493.
[3] K. S. Lam, S. E. Salmon, E. M. Hersh, V. J. Hruby, W. M. Kazmierski,
R. J. Knapp, Nature 1991, 354, 82 ± 84.
[4] a) R. Haubner, D. Finsinger, H. Kessler, Angew. Chem. 1997, 109,
1440 ± 1456; Angew. Chem. Int. Ed. Engl. 1997, 36, 1374 ± 1389, and
references therein; b) J. Samanen, Z. Jonak, D. Rieman, T.-L. Yue,
Curr. Pharm. Des. 1997, 3, 545 ± 584, and references therein; c) M. S.
Goligorsky, H. Kessler, V. I. Romanov, Nephrol. Dial. Transplant.
1998, 13, 254 ± 263, and references therein; d) R. M. Scarborough,
Curr. Med. Chem. 1999, 6, 971 ± 981, and references therein.
[5] a) E. Ruoslahti, M. D. Pierschbacher, Cell 1986, 44, 517 ± 518; b) E.
Ruoslahti, M. D. Pierschbacher, Science 1987, 238, 491 ± 497; c) T. A.
Springer, Nature 1990, 346, 425 ± 434; d) M. Pfaff in Integrin ± Ligand
Interaction (Ed.: J. A. Eble), Springer, Heidelberg, 1997, pp. 101 ± 121.
[6] a) J. Wermuth, PhD Thesis, Technische Universität München (Ger-
many), 1996; b) J. S. Schmitt, PhD Thesis, Technische Universität
München (Germany), 1998.
Scheme 1. Specific ortho-hydroxylation of a-substituted phenols catalyzed
by 2-hydroxybiphenyl-3-monooxygenase. R ˆ alkyl (Et, Pr, iPr), aryl (Ph,
2-HOC6H4), Hal (F, Cl, Br).
[7] G. B. Fields, R. L. Noble, Int. J. Pept. Protein Res. 1990, 35, 161 ± 214,
and references therein.
For the regeneration of NADH we applied the
[Cp*Rh(bpy)Cl]Cl complex which had been developed in
our group (Cp* ˆ C5Me5; bpy ˆ 2,2'-bipyridine). The corre-
sponding hydridorhodium complexes, which can be generated
either electrochemically by cathodic reduction at ꢀ750 mV
(versus Ag/AgClsat.) or chemically with formate, transform
[8] a) C. P. Holmes, J. Org. Chem. 1997, 62, 2370 ± 2380; b) C. P. Holmes,
D. G. Jones, J. Org. Chem. 1995, 60, 2318 ± 2319.
[9] For the synthesis of building blocks A3±6
, see the Supporting
Information. Although the regioisomers of compound A5 could be
separated by HPLC, they were used as a mixture of isomers for the
library synthesis.
[10] C. Gibson, S. L. Goodman, D. Hahn, G. Hölzemann, H. Kessler, J.
Org. Chem. 1999, 64, 7388 ± 7394.
‡
NAD(P) efficiently into the enzymatically active 1,4-
NAD(P)H form[3, 5] (Scheme 2). The conversion rates
[11] Six different building blocks in group A were used for the library syn-
thesis, whereas building blocks A1 and A2 were used as a racemate and
A5 as a racemate and a mixture of regioisomers. In consideration of the
isomers in group A, 11 different building blocks have been employed.
[12] The guanylation was performed as per: Y. Wu, G. R. Matsueda, M.
Bernatowicz, Synth. Commun. 1993, 23, 2055 ± 3060; the pyrimidyla-
tion was performed as per: C. Gibson, H. Kessler, Tetrahedron Lett.
2000, 41, 1825 ± 1728.
[13] The substrate BCIP is usually used together with the oxidation agent
4-nitro-blue-tetrazolium (NBT). However, NBT proved to be incom-
patible with our libraries because it reacted with some of the resin-
bound compounds and led to false-positive results.
[*] Dr. A. Schmid, Dipl.-Chem. F. Hollmann
Institut für Biotechnologie
ETH Hönggerberg
8093 Zurich (Switzerland)
Fax : (‡ 41)1-633-1051
E-mail: andreas.schmid@biotech.biol.ethz.ch
²
Prof. Dr. E. Steckhan
Institut für Organische Chemie und Biochemie
der Universität Bonn (Germany)
[**] This work was supported by BASF AG. We thank DEGUSSA AG for
the gift of chemicals.
[14] I. Ojima, S. Chakravarty, Q. Dong, Bioorg. Med. Chem. 1995, 3, 337 ±
360, and references therein.
Angew. Chem. Int. Ed. 2001, 40, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4001-0169 $ 17.50+.50/0
169

COMMUNICATIONS
Scheme 2. Indirect electrochemical (A) and chemical regeneration of NADH (B).
(Table 1) of the enzymatic reaction with reduced coenzyme
and under indirect chemical regeneration of NADH (Sche-
me 2B) prove the applicability of the proposed regeneration
system. Hence we advanced the enzymatic reaction to the
indirect electrochemical regeneration of NADH (Scheme 2A).
0.27 mmolLꢀ1 hꢀ1 with an air input of 10 cm3 minꢀ1. No
generation of NADH is detectable at a flow rate of
15 cm3 minꢀ1. This inhibition however is reversible: after
ceasing the air input both the rate of NADH production
and the concentration of NADH recover to their maximum
values.
Table 1. Comparison of the productivity rates (dc/dt) of 2,3-dihydroxybi-
phenyl with reduced coenzyme (NADH), under indirect chemical and
indirect electrochemical NADH regeneration.[a]
Regeneration
HbpA [U] dc/dt [mmolminꢀ1
]
Rel. activity[b] [%]
none[c]
2.24
2.60
2.09
2.33
1.83
93.3
89.6
9.65[f]
chemically[d]
electrochemically[e] 19[f]
[a] General conditions: phosphate buffer (50 mm, pH 7.5, saturated with
air); Tˆ 258C; [Cp*Rh(bpy)] ˆ 0.5 mm; [2-hydroxybiphenyl] ˆ 0.8 mm.
[b] Productivity relative to the amount of HbpA applied. [c] [NADH] ˆ
‡
‡
1.0 mm. [d] [NAD ] ˆ 1.09 mm; [HCOONa] ˆ 150 mm. [e] [NAD ] ˆ
1.0 mm ; air intake ˆ 10 cm3 minꢀ1; [2-hydroxybiphenyl] ˆ 2 mm; [FAD] ˆ
20 mm; E ˆ ꢀ750 mV (versus Ag/AgClsat.). [f] On account of the fast HbpA
inactivation (see text) large amounts of enzyme had to be applied; the air
intake is rate limiting.
Figure 1. Catalytic effects of [Cp*Rh(bpy)Cl]Cl (5 Â 10ꢀ4 m in 0.1m TRIS/
HCl buffer pH 7.5; TRIS ˆ tris(hydroxymethyl)aminomethane) in the
This approach was examined by cyclic voltammetry. Even
at proton concentrations less than 10ꢀ8 m (pH 8) the uptake of
a proton into the ligand sphere of the reduced rhodium
species proceeds quite quickly so that no oxidation peak is
visible. Strong increases in the peak current (catalytic effects)
are detected, even at high sweep rates in the presence of
‡
presence of various concentrations of NAD (0 (ÐÐ), 5 Â 10ꢀ4 (±±±),
and 1  10ꢀ3 m (- - - -)); v ˆ 81 mVsꢀ1; degassed buffers.
‡
NAD (Figure 1). The reduction peak potential lies at
ꢀ750 mV (versus Ag/AgClsat.), which is more positive by far
‡
than the first reduction potential of NAD (ꢀ920 mV versus
the saturated calomel electrode (SCE)).[6] Thus, coenzyme
inactivation by direct electron transfer becomes negligible.
The role of molecular oxygen was examined in detail. We
found that it inhibits the transfer of hydride from the reduced
‡
rhodium complex to NAD both in the chemical and the
electrochemical regeneration. As can be clearly seen in
Figure 2 the regeneration rate of NADH strictly depends on
the rate of air input. For example, the rate of NADH
production decreases from 1.1 (without air input) to
Figure 2. Influence of O2 on the indirect electrochemical regeneration of
NADH; [Cp*Rh(bpy)Cl]Cl (0.1 mm in 50 mm potassium phosphate buffer
3
ꢀ1
&
*
pH 7.5) ( no air insertion; air input 10 cm min until t ˆ 25 min).
170
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4001-0170 $ 17.50+.50/0 Angew. Chem. Int. Ed. 2001, 40, No. 1

COMMUNICATIONS
We characterized the product from the reaction of the
hydridorhodium complex and molecular oxygen as hydrogen
peroxide. We postulate its formation as depicted in
Scheme 3.[7] In addition, molecular oxygen is reduced at the
applied cathode potential and also yields hydrogen peroxide.
In the course of the electro-enzymatic conversion it is
dismutated into O2 and H2O by catalase.
ation process) can be handled better. In preliminary studies
the actual production time could be enhanced by more than
2.5 fold to over 8 h.
In conclusion, we could for the first time couple a flavine-
dependent monooxygenase reaction to an indirect electro-
chemical regeneration of NADH. The productivity rates of
this not-optimized process (204 mgLꢀ1 hꢀ1) are already at
about 50% of the in vivo process (390 mgLꢀ1 hꢀ1)[8] and the
in vitro process with enzymatic regeneration of the coenzyme
(405 mgLꢀ1 hꢀ1).[4c]
.
Experimental Section
The rhodium catalyst was prepared according to the method of Kölle and
Grätzel.[9]
HbpAwas enriched from the recombinant E. coli JM101 [pHBP461].[8] The
raw extract was purified and enriched by expanded-bed chromatography.[4c]
Enzymatic activity was determined by a UV-spectrometric assay.[4a]
Cyclovoltammogramms were taken with a BAS-100B/W analyzer (Bio-
analytical Systems) in an undivided cell employing a glassy carbon working
electrode (i.d. ˆ 3 mm), a Pt-wire counter electrode, and an Ag/AgClsat.
reference electrode (Cypress Systems). Electroenzymatic conversions were
performed in a thermostated electrolysis cell. A cylindrical carbon-felt
electrode (Vˆ 27 cm3) served as the cathode and the potential was adjusted
versus a Ag/AgClsat. reference electrode. A divided cell was achieved by
placing the platinum counter electrode within a dialysis membrane. The
following components were dissolved in 100 mL of potassium phosphate
Scheme 3. Postulated mechanism of the reaction between the hydrido-
rhodium complex and molecular oxygen.
During the enzymatic reaction without external input of
oxygen the catechol production stops on account of a lack of
molecular oxygen in the reaction medium (typically at
product concentrations between 0.2 and 0.4 mm). With the
active input of air into the reaction medium new product is
formed. Thus productivity rates of up to 1.1 mmolLꢀ1 hꢀ1
(204 mgLꢀ1 hꢀ1) are achieved. The mediatorꢁs turnover fre-
quency can be calculated to be 11 hꢀ1.
‡
buffer (50 mm, pH 7.5): 2-hydroxybiphenyl (2 mm), NAD (0.2 mm),
[Cp*Rh(bpy)Cl]Cl (0.1 mm), FAD (20 mm), catalase (250000 U), and
HbpA (19 U). Oxygen was supplied by a heterogeneous input of air. The
input rate was adjusted before each enzymatic conversion in such a way
‡
that the NAD reduction rate reached its maximum value.
The reaction was followed by HPLC analysis on a RP-18 column eluated
with methanol/water(0.1% H3PO4) (60/40). No by-products were detected.
The product was sublimated and characterized by NMR and MS analysis.
The rate-determining step is the inhibition of the mediator
by molecular oxygen rather than the heterogeneous charge
transfer at the cathode. Control experiments with homoge-
neous chemical regeneration of the hydridorhodium complex
under otherwise comparable conditions show that the pro-
ductivity rates are in the same range (for example 0.673 and
0.668 mmolLꢀ1 hꢀ1 at an air intake of 8 cm3 minꢀ1 with total
conversions of 100 and 95% under electrochemical and
chemical regeneration, respectively).
The poor long-term stability of the previous electro-
enzymatic conversions is the result of the experimental setup:
the input of heterogeneous air results in enzyme inactivation.
In addition HbpA adsorbes onto the electrode surface and
thus is exposed to locally high H2O2 concentrations that
cannot be sufficiently dismutated by catalase. The half-time
values for HbpA activity in phosphate buffer under the
conditions of the indirect chemical regeneration are in the
range of 30 ± 40 h (depending on the speed of air input),
whereas during electrolysis (depending on the amount of
HbpA applied and the mass of the electrode) no enzymatic
activity is detectable after about 2 h.
Received: June 27, 1999 [Z15343]
[1] a) W. Hummel, E. Schmidt, C. Wandrey, M.-R. Kula, Appl. Microbiol.
Biotechnol. 1986, 25, 175 ± 185; b) R. Rissom, U. Scharz-Linek, M.
Vogel, V. I. Tishkov, U. Kragl, Tetrahedron: Asymmetry 1997, 15, 2523 ±
2526.
[2] C.-H. Wong, G. M. Whitesides, J. Am. Chem. Soc. 1981, 103, 4890 ±
4899.
[3] a) D. Westerhausen, S. Herrmann, W. Hummel, E. Steckhan, Angew.
Chem. 1992, 104, 1496 ± 1498; Angew. Chem. Int. Ed. Engl. 1992, 31,
1529 ± 1531; b) R. Ruppert, S. Herrmann, E. Steckhan, Tetrahedron
Lett. 1987, 52, 6583 ± 6586.
[4] a) W. A. Suske, M. Held, A. Schmid, T. Fleischmann, A. G. Wubbolts,
H.-P. E. Kohler, J. Biol. Chem. 1997, 272, 24257 ± 24265; b) W. A.
Suske, W. J. H. van Berkel, H.-P. E. Kohler, J. Biol. Chem. 1999, 274,
33355 ± 33365; c) A. Schmid, J. Mol. Cat. B 2000, in press; d) A.
Schmid, H.-P. Kohler, K.-H. Engesser, J. Mol. Cat. B 1998, 5, 311 ± 316.
[5] a) E. Steckhan, S. Herrmann, R. Ruppert, E. Dietz, M. Frede, E. Spika,
Organometallics 1991, 10, 1568 ± 1577; b) R. Ruppert, S. Herrmann, E.
Steckhan, J. Chem. Soc. Chem. Commun. 1988, 1150 ± 1151; c) H. C. Lo,
O. Buriez, J. B. Kerr, R. H. Fish, Angew. Chem. 1999, 111, 1524 ± 1527;
Angew. Chem. Int. Ed. 1999, 38, 1429 ± 1432.
[6] C. O. Schmackel, K. S. V. Santhanam, P. J. Elvin, J. Am. Chem. Soc.
1975, 97, 5083 ± 5092.
[7] N. N. Greenwood, A. Earnshaw, Chemie der Elemente, 1st ed., VCH,
Weinheim, 1988, pp. 798 ± 831.
[8] M. Held, A. Schmid, H.-P. E. Kohler, W. A. Suske, B. Witholt, M. G.
Wubbolts, Biotechnol. Bioeng. 1999, 62, 641 ± 648.
[9] U. Kölle, S.-S. Kang, P. Infelta, P. Comte, M. Grätzel, Chem. Ber. 1989,
122, 1869 ± 1880.
To overcome this stability problem we are evaluating
possibilities of immobilizing the enzyme onto a solid matrix
thus protecting it from the electrode. Using this division of the
whole process setup into an electrochemical and an enzymatic
part, the ambivalent role of molecular oxygen (substrate of
the enzymatic reaction and inhibitor of the NADH regener-
Angew. Chem. Int. Ed. 2001, 40, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4001-0171 $ 17.50+.50/0
171