Regioselective preparation of 5-hydroxypropranolol and 4′-hydroxydiclofenac with a fungal peroxygenase

Article abstract of DOI:10.1016/j.bmcl.2009.04.015

An extracellular peroxygenase of Agrocybe aegerita catalyzed the H₂O₂-dependent hydroxylation of the multi-function beta-adrenergic blocker propranolol (1-naphthalen-1-yloxy-3-(propan-2-ylamino)propan-2-ol) and the non-steroidal anti-inflammatory drug diclofenac (2-[2-[(2,6-dichlorophenyl)amino]phenyl]acetic acid) to give the human drug metabolites 5-hydroxypropranolol (5-OHP) and 4′-hydroxydiclofenac (4′-OHD). The reactions proceeded regioselectively with high isomeric purity and gave the desired 5-OHP and 4′-OHD in yields up to 20% and 65%, respectively. ¹⁸O-labeling experiments showed that the phenolic hydroxyl groups in 5-OHP and 4′-OHD originated from H₂O₂, which establishes that the reaction is mechanistically a peroxygenation. Our results raise the possibility that fungal peroxygenases may be useful for versatile, cost-effective, and scalable syntheses of drug metabolites.

Full text of DOI:10.1016/j.bmcl.2009.04.015

Bioorganic & Medicinal Chemistry Letters 19 (2009) 3085–3087
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Regioselective preparation of 5-hydroxypropranolol and 40-hydroxydiclofenac
with a fungal peroxygenase
a
a
a
b
Matthias Kinne a, , Marzena Poraj-Kobielska , Elisabet Aranda , René Ullrich , Kenneth E. Hammel ,
*
Katrin Scheibner c, Martin Hofrichter a
a Unit of Environmental Biotechnology, International Graduate School of Zittau, Markt 23, 02763 Zittau, Germany
b USDA Forest Products Laboratory, Madison, WI 53726, USA
c University of Applied Sciences Lausitz, 01968 Senftenberg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
An extracellular peroxygenase of Agrocybe aegerita catalyzed the H2O2-dependent hydroxylation of
the multi-function beta-adrenergic blocker propranolol (1-naphthalen-1-yloxy-3-(propan-2-ylamino)-
propan-2-ol) and the non-steroidal anti-inflammatory drug diclofenac (2-[2-[(2,6-dichloro-
phenyl)amino]phenyl]acetic acid) to give the human drug metabolites 5-hydroxypropranolol (5-OHP)
and 40-hydroxydiclofenac (40-OHD). The reactions proceeded regioselectively with high isomeric purity
and gave the desired 5-OHP and 40-OHD in yields up to 20% and 65%, respectively. 18O-labeling experi-
ments showed that the phenolic hydroxyl groups in 5-OHP and 40-OHD originated from H2O2, which
establishes that the reaction is mechanistically a peroxygenation. Our results raise the possibility that
fungal peroxygenases may be useful for versatile, cost-effective, and scalable syntheses of drug
metabolites.
Received 30 January 2009
Revised 1 April 2009
Accepted 2 April 2009
Available online 9 April 2009
Keywords:
Peroxidase
Peroxygenase
Oxygenase
Cytochrome P450
Hydroxylation
5-Hydroxypropranolol
Propranolol
40-Hydroxydiclofenac
Diclofenac
Ó 2009 Elsevier Ltd. All rights reserved.
Ascorbic acid
Selective hydroxylations of aromatic compounds are among the
most challenging reactions in synthetic chemistry and have gained
steadily increasing attention during the last decade because hydrox-
ylated aromatics are important precursors and products in the phar-
maceutical industry.1 For example, 5-hydroxypropranolol (5-OHP),
a human metabolite of the beta blocker propranolol (1-naphtha-
len-1-yloxy-3-(propan-2-ylamino)propan-2-ol), is of pharmacolog-
ical interest as it is frequently used in metabolic studies and has been
demonstrated to be equipotent to propranolol as a b-receptor antag-
onist.2 Another important human drug metabolite is 40-hydrox-
ydiclofenac (40-OHD), a major metabolite of the anti-inflammatory
drug diclofenac (2-[2-[(2,6-dichlorophenyl)amino]phenyl]acetic
acid) in humans.3
cost-effective use of this approach. The chemical synthesis of 40-
OHD has likewise proved difficult.3 A simpler approach would
be to use an enzyme to hydroxylate these drugs in one step.
One possibility would be the use of monooxygenases such as
cytochrome P450s (cytP450s). This approach has been shown to
work for 40-OHD synthesis,3,5,6 but is currently restricted to
whole-cell biotransformations because cytP450s are poorly stable
and catalytically slow. Moreover, their intracellular location
makes them difficult to produce in quantity.7,8 Another approach,
the use of laboratory-evolved, engineered cytP450s for H2O2-
dependent hydroxylation of propranolol to 5-OHP via the so-
called ‘peroxide shunt’, has been demonstrated but needs further
optimization.9
Although a four-step chemical synthesis of 5-OHP from 1,5-
naphthalenediol is available,
problems with the removal of byproducts have prevented the
We used a recently discovered heme-thiolate enzyme from the
basidiomycete Agrocybe aegerita to hydroxylate propranolol and
diclofenac. This stable, secreted peroxygenase (AaP) oxidizes a
wide range of aromatic substrates and appears to be a versatile oxi-
doreductase for biotechnological applications.10–15 We treated
each of the compounds (0.5 mM) with purified AaP (0.4 lm) and
four equivalents of H2O2 (2 mM) in the presence of ascorbic acid
(4 mM). The last ingredient was added to prevent polymerization
a
low overall yield (<5%)4 and
Abbreviations: AaP, Agrocybe aegerita peroxygenase; cytP450, cytochrome P450;
40-OHD, 40-hydroxydiclofenac; 5-OHD, 5-hydroxydiclofenac; 4-OHP, 4-hydroxy-
propranolol; 5-OHP, 5-hydroxypropranolol.
* Corresponding author. Tel.: +49 3583 612723; fax: +49 3583 612734.
E-mail address: kinne@ihi-zittau.de (M. Kinne).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.04.015

3086
M. Kinne et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3085–3087
Figure 1. HPLC elution profiles showing products formed by AaP (2 U mlꢀ1, 0.4
lM) after conversion of propranolol (500 lM, left) and diclofenac (500 lM, right) in the
presence of ascorbic acid (4 mM). Controls without enzyme (I). Complete reactions (II). The reactions (pH 7) were started by addition of H2O2 (2 mM). Insets show UV/visible
absorption spectra of the reactant (right) and major product (left) in each chromatogram.18
of 5-OHP or 40-OHD, an undesirable side reaction attributable to
the general peroxidase activity of AaP.16 The mixtures were stirred
at room temperature for 3 min and stopped with 0.1 ml of 50% wt/
vol trichloroacetic acid, after which products were identified and
quantified against authentic standards by HPLC.17
tions conducted with more AaP did not give significantly higher
yields (data not shown). It appears more likely that the phenolic
products 5-OHP and 40-OHD prevented further oxidation of the
parent compounds because they also are AaP substrates. Under
our reaction conditions, these phenols probably consumed some
of the H2O2 by undergoing continuous, competitive AaP-catalyzed
oxidation to the 5-OHP and 40-OHD phenoxy radicals, which in
turn were continuously re-reduced to 5-OHP and 40-OHD by the
excess ascorbate we included. This conclusion is supported by
our observation that propranolol was rapidly polymerized when
the reactions were conducted in the absence of ascorbate (data
not shown).
When we conducted the AaP-catalyzed oxidation of proprano-
lol with H218O2 in place of H2O2, mass spectral analysis18 of the
resulting 5-OHP (Fig. 3) showed that the principal [MꢀH]ꢀ ion
had shifted from the natural abundance m/z of 274 to m/z 276
for propranolol. Similarly, the analogous experiment with diclofe-
nac showed that the principal ion of 40-OHD shifted from an m/z
of 310 to an m/z of 312 (data not shown). Experiments using
H218O gave no detectable 18O incorporation (data not shown),
as expected because phenolic oxygens are not readily exchange-
able with water under our reaction conditions.19 An additional
experiment with natural abundance H2O2 in an N2-purged reac-
tion mixture showed that propranolol production was not inhib-
ited by depletion of O2 (data not shown), and therefore O2 did not
contribute significantly as an electron acceptor. These results
show that the new phenolic oxygens in 5-OHP and 40-OHD orig-
inated from H2O2.
The reactions proceeded rapidly and regioselectively, convert-
ing about 20% of racemic propranolol to 5-OHP and about 65% of
diclofenac to 40-OHD (Figs. 1 and 2). For propranolol oxidation,
the formation of byproducts previously reported for the
cytP450-catalyzed reaction9 was insignificant: 4-hydroxypropra-
nolol and 1-naphthol occurred only in trace quantities, and
N-desisopropylpropranolol was not found. The enantiomeric
excess of S-5-OHP during AaP-catalyzed hydroxylation of
propranolol was less than 2% (data not shown), that is, the reac-
tion was not enantioselective. For diclofenac oxidation, the
human drug metabolite 5-OHD3 was not formed, but traces of
several other unidentified byproducts were detected. Control
reactions without AaP or with heat-inactivated enzyme gave no
conversion of propranolol or diclofenac.
The failure of the reactions to proceed to completion was
probably not a consequence of enzyme inactivation, because reac-
Since cytP450s and AaP are both heme-thiolate enzymes, the
catalytic cycle of AaP probably resembles the ‘peroxide shunt’
that is responsible for the hydroxylation of propranolol by some
engineered cytP450s.9 However, AaP appears to be the better
choice as a biocatalyst because it is easier to produce, is more
efficient, is more stable to H2O2, and in the case of propranolol
exhibits higher regioselectivity (Table 1). On the other hand,
although AaP regioselectively hydroxylates other precursors as
well,14 it has some limitations. For example, we have observed
that, although AaP efficiently hydroxylates N-phenylacetamide
to acetoaminophen (yields up to 80%, unpublished results), it
very poorly adds the second 3-hydroxyl needed to produce the
human drug metabolite 3-hydroxyacetaminophen. We are cur-
rently exploring the possibility that AaP may catalyze useful
one-step monooxygenations of other pharmaceutically relevant
aromatics.
Figure 2. Time course of AaP-catalyzed 5-hydroxylation of propranolol (reaction
conditions are the same as in Figure 1, standard deviation <5%).

M. Kinne et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3085–3087
3087
Figure 3. Mass spectra showing molecular ions of 5-OHP obtained from the oxidation of propranolol with AaP (1 U mlꢀ1, 0.2
lM) in the presence of natural abundance H2O2
(top) or H218O2 (bottom).19
6. Rushmore, T. H.; Reider, P. J.; Slaughter, D.; Assang, C.; Shou, M. Metab. Eng.
2000, 2, 115.
7. Urlacher, V. B.; Lutz-Wahl, S.; Schmid, R. D. Appl. Microb. Biotechnol. 2004, 64,
317.
8. Eiben, S.; Kaysser, L.; Maurer, S.; Kuhnel, K.; Urlacher, V. B.; Schmid, R. D. J.
Biotechnol. 2006, 124, 662.
Table 1
Conversion of propranolol (5 mM) to 5-OHP by AaP and cytP450.c
Enzyme in
reaction (
Conversion of
Reaction
Products
formed
lM)
propranolol to 5-OHP (%)
time (min)
9. Otey, C. R.; Bandara, G.; Lalonde, J.; Takahashi, K.; Arnold, F. H. Biotechnol.
Bioeng. 2006, 93, 494.
10. Ullrich, R.; Nuske, J.; Scheibner, K.; Spantzel, J.; Hofrichter, M. Appl. Environ.
Microbiol. 2004, 70, 4575.
11. Ullrich, R.; Hofrichter, M. FEBS Lett. 2005, 579, 6247.
12. Aranda, E.; Kinne, M.; Kluge, M.; Ullrich, R.; Hofrichter, M. Appl. Microbiol.
Biotechnol. 2009, 82, 1057.
AaPa (0.6)
13.6
0.5
2
180
1
4
CytP450b (5.0)
a
H2O2 concentration was 5 mM.
b
c
H2O2 concentration was 1 mM. Data are for mutant D6H10 in Ref. 9.
Conversion of diclofenac to 40-OHD under these reaction conditions was 30%.
13. Kluge, M.; Ullrich, R.; Dolge, C.; Scheibner, K.; Hofrichter, M. Appl. Microbiol.
Biotechnol. 2009, 81, 1071.
14. Kinne, M.; Ullrich, R.; Hammel, K. E.; Scheibner, K.; Hofrichter, M. Tetrahedron
Lett. 2008, 49, 5950.
Acknowledgments
15. Ullrich, R.; Dolge, C.; Kluge, M.; Hofrichter, M. FEBS Lett. 2008, 582, 4100.
16. Hofrichter, M.; Ullrich, R. Appl. Microbiol. Biotechnol. 2006, 71, 276.
17. HPLC was done on a Phenomenex Synergi Fusion RP-80A reverse phase column
We thank Ulrike Schneider and Monika Brandt for technical
assistance, and Martin Kluge (Inge) and Christiane Liers for useful
discussions. Thisworkwas supportedby theKonradAdenauerFoun-
dation (M.K.), the Fulbright Foundation (K.E.H.) and the Deutsche
Bundesstiftung Umwelt (M.H. and K.S., Project Number 13225-32).
(4.6 by 150 mm, 4 lm particle size). The column was eluted at 40 °C and a flow
rate of 1 ml minꢀ1 with a mixture of aqueous phosphoric acid solution (15 mM,
pH 3) and acetonitrile, 95:5, for 5 min, followed by a 25-min linear gradient to
100% acetonitrile. Elution times of propranolol, 5-OHP, 4-OHP, 1-naphthol, N-
desisopropylpropranolol, diclofenac, 5-OHD and 40-OHD were checked against
authentic standards.
18. Liquid chromatography/mass spectroscopic (LC/MS) analyses were performed
References and notes
using
150 mm, 5
a
reversed phase Synergi Gemini C6-Phenyl 110A column (4.6 by
m particle size). The isocratic mobile phase consisted of 5% vol/vol
l
1. Ullrich, R.; Hofrichter, M. Cell. Mol. Life. Sci. 2007, 64, 271.
acetonitrile and 95% aqueous 0.1% vol/vol ammonium formate that had been
adjusted to pH 10 beforehand with NaOH. The column was operated at 40 °C
and 1 ml minꢀ1 for 5 min. Electrospray ionization was performed in the
negative ionization mode.
2. Greenslade, F. C.; Newquist, K. L. Arch. Int. Pharmacodyn. Ther. 1978, 233, 270.
3. Webster, R.; Pacey, M.; Winchester, T.; Johnson, P.; Jezequel, S. Appl. Microbiol.
Biotechnol. 1998, 49, 371.
4. Oatis, J. E., Jr.; Russell, M. P.; Knapp, D. R.; Walle, T. J. Med. Chem. 1981, 24, 309.
5. Vail, R. B.;Homann, M. J.;Hanna,I.; Zaks, A. J. Ind.Microbiol. Biotechnol. 2005, 32, 67.
19. Koizumi, M.; Titani, T. Bull. Chem. Soc. Jpn. 1938, 13, 607.