Biooxidation of methyl group: Part 2. Evidences for the involvement of cytochromes P450 in microbial multistep oxidation of terfenadine

Article abstract of DOI:10.1016/j.molcatb.2010.07.017

The actinomycete Streptomyces platensis grown in culture medium containing soybean peptones can transform terfenadine, an antihistamine drug, into its active metabolite fexofenadine. The microbial oxidation of methyl group of terfenadine into carboxylic acid could be an alternative to chemical ways to produce fexofenadine. This bioconversion requires three oxidation steps: a hydroxylation of one methyl group followed by the oxidation of the corresponding alcohol into the aldehyde and finally its oxidation into the carboxylic acid. The oxidation reaction of each step has been studied. Terfenadine and intermediates incubated with whole cells were not oxidized under argon whereas their biotransformation under ¹⁸O₂-enriched atmosphere gave labeled fexofenadine. P450 inhibitors, such as clotrimazole or fluconazole, inhibited oxidation activity of each step. While the two last steps could be catalyzed by dehydrogenases or oxidases, this study strongly demonstrates the role of at least one, or possibly several cytochromes P450, in the oxidation of terfenadine into fexofenadine by S. platensis cells. To our knowledge, this is one of the few examples of involvement of P450s in such three steps oxidation of a xenobiotic.

Full text of DOI:10.1016/j.molcatb.2010.07.017

Journal of Molecular Catalysis B: Enzymatic 67 (2010) 172–178
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Biooxidation of methyl group: Part 2. Evidences for the involvement of
cytochromes P450 in microbial multistep oxidation of terfenadine
Amane El Ouarradi, Murielle Lombard, Didier Buisson∗
Laboratoire de Chimie et Biochimie Pharmacologiques et Toxicologiques, UMR 8601 CNRS, Université Paris Descartes 5, 45, rue des Saints-Pères, 75270 Paris Cedex 06, France
a r t i c l e i n f o
a b s t r a c t
Article history:
The actinomycete Streptomyces platensis grown in culture medium containing soybean peptones can
transform terfenadine, an antihistamine drug, into its active metabolite fexofenadine. The microbial oxi-
dation of methyl group of terfenadine into carboxylic acid could be an alternative to chemical ways to
produce fexofenadine. This bioconversion requires three oxidation steps: a hydroxylation of one methyl
group followed by the oxidation of the corresponding alcohol into the aldehyde and finally its oxidation
into the carboxylic acid. The oxidation reaction of each step has been studied. Terfenadine and inter-
mediates incubated with whole cells were not oxidized under argon whereas their biotransformation
under 18O2-enriched atmosphere gave labeled fexofenadine. P450 inhibitors, such as clotrimazole or
fluconazole, inhibited oxidation activity of each step. While the two last steps could be catalyzed by
dehydrogenases or oxidases, this study strongly demonstrates the role of at least one, or possibly several
cytochromes P450, in the oxidation of terfenadine into fexofenadine by S. platensis cells. To our knowl-
edge, this is one of the few examples of involvement of P450s in such three steps oxidation of a xenobiotic.
Received 24 March 2010
Received in revised form 2 July 2010
Accepted 27 July 2010
Available online 6 August 2010
Keywords:
Streptomyces platensis
Biotransformation
Hydroxylation
Aldehyde
Labeling study
P450 inhibitors
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
is the main mammalian metabolite of terfenadine. Because fexofe-
nadine chemical synthesis is laborious, microbial hydroxylation of
Biocatalysis and biotransformation have become a commonly
used tool in organic chemical synthesis, in particular to produce
synthons and metabolites of xenobiotics. They are used for the
synthesis of chiral molecules but also of achiral products where
a chemical process would not be possible. For example, chem-
ical oxidations of heteroaromatic molecules and non-activated
carbon-atom give side-products and are generally unspecific. On
the contrary, some examples describe the specific hydroxyla-
tion of methyl and methylene groups by microorganisms [1]. The
most appropriate activity could be obtained by screening isolated
enzymes or microorganisms. While some enzymes are commer-
cially available (hydrolases, reductases), oxygenases are generally
provided by microorganisms and used in whole cells [2]. However,
for preparative scale, an improvement of the biocatalyst is often
necessary [3].
the easily accessible terfenadine has been considered as the first
step. Moreover, according to Microbial Model of Mammalian Drug
Metabolism it was expected to biotransform terfenadine into fex-
ofenadine.
It was found that several microorganisms are able to catalyze
the multistep oxidation of terfenadine (Fig. 1) [6,7] and analogs [8].
Among them, we showed [9] that the fungus Absidia corymbifera
and the bacterium Streptomyces platensis are the most efficient in
fexofenadine synthesis. However, the activity per gram of cells is
too low for a scale-up process and fexofenadine production must
be optimized by proteins engineering. This scale-up requires the
precise knowledge of the enzymatic activity(ies) involved in this
multistep oxidation and the cloning of the proteins corresponding
genes [10].
While whole microbial cells contain a wealth of enzymes
holding different redox activities, various enzymatic systems can
catalyze these three reactions: the first step, hydroxylation of a
methyl group can only be catalyzed by monooxygenases, whereas
the two following oxidation steps could be catalyzed by either
dehydrogenases, oxidases or monooxygenases (a hem and non-
hem enzymes) [11–13]. Dehydrogenases are the most often cited in
biotechnological applications. For example, a recent work describes
[14] an efficient oxidation of alcohol into the corresponding acid
using whole cells of Brevibacterium sp. or a three enzymes system,
We are interested in the chemoenzymatic synthesis of fexofe-
nadine, an antihistamine drug devoid of cardiotoxicity [4,5] which
∗
Corresponding author. Present address: Muséum National d’Histoire Naturelle,
Centre National de la Recherche Scientifique, Unité Molécules de Communication
et Adaptation des Microorganismes, CP54, 57 rue Cuvier, 75005 Paris, France.
Tel.: +33 0 1 40 79 81 24; fax: +33 0 1 40 79 31 35.
E-mail address: buisson@mnhn.fr (D. Buisson).
1381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.07.017

A. El Ouarradi et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 172–178
173
Fig. 1. Biotransformation of terfenadine.
2-phenylethanol dehydrogenase, phenyl acetaldehyde dehydroge-
nase and NADH oxidase to regenerate the NAD+. In an industrial
process, the oxidation of methyl groups on aromatic heterocycles
to the corresponding carboxylic acids were achieved by enzymes
from wild type Pseudomonas putida, xylene monooxygenase (XMO),
benzyl alcohol dehydrogenase and benzaldehyde dehydrogenase
[15,16]. However, Schmid and coworkers showed [17] that XMO
has an alcohol and an aldehyde oxidizing activities, and used
it in synthesis of aryl carboxylic acids [18]. There are only few
examples of involvement of microbial cytochrome P450 in the for-
mation of carboxylic acid, including oxidation of 2-ethylhexanol
by CYP101 from P. putida [19], and formation of dicarboxylic acid
from alkanes in Candida tropicalis [20]. However, three fatty alco-
hol oxidases from the strain C. tropicalis, which could be involved,
were also characterized [21]. Some aldehyde oxidases are charac-
terized including one from a Streptomyces strain [22] and one from
a Brevibacillus sp. that is used to remove glutaraldehyde, a poten-
tial environmental pollutant [23]. Finally, some fungal extracellular
hem thiolate peroxidases are capable of performing oxidation of
methyl groups leading to the corresponding acids [24].
In some cases, there is a lack of knowledge of the enzymes
involved in the preparation of acids by microbial oxidation [25–30].
Schwartz et al. reported the oxidation of ebastine to carebastine by
C. blabesleena and suggested that the two alcohol oxidation steps
were catalyzed by oxido-reductases [31].
Contrary to Schwartz suggestion, we obtained some results in
agreement with a monooxygenase-dependent mechanism for the
three steps oxidation involved in the fexofenadine formation by
S. platensis. Firstly, oxidation of hydroxyterfenadine to fexofena-
dine occurs with cells grown in culture medium containing soybean
peptone, which are known to induce cytochrome P450 in Strepto-
myces griseus [32]. Secondly, oxidation of hydroxyterfenadine in
18O2-enriched atmosphere lead to labeled fexofenadine, resulting
to incorporation of one atom of dioxygen [33].
Then, prior to endeavor to purify the enzymes involved in the
multistep oxidation of terfenadine by S. platensis cells, it was nec-
essary to identify the precise class of these oxidizing enzymes. We
describe here the investigations performed to inquire the type of
enzyme implicated in each oxidation step. Alcohol dehydrogenase
and monooxygenase-dependent oxidative activities were seeked
in vitro with cell-free extracts. However, as P450-monooxygenases
dependent activities that result from a multi-component systems
with a relatively low natural level of expression [34], are rarely pre-
served during preparation of cell-free extract, indirect approaches
are often used to decipher the precise type of enzymes involved in
oxidation activities [35–39]. Using this strategy, we performed the
biotransformation of terfenadine, and of the corresponding alco-
hol and aldehyde intermediates under 18O2 atmosphere or in the
presence of P450 inhibitors. Our results suggest that the overall
biotransformation arise from three cytochrome P450 dependent
oxidation steps.
2. Materials and methods
2.1. Chemicals
Terfenadine was purchased from Sigma, hydroxyterfenadine
was prepared as previously reported [33] and the correspond-
ing aldehyde was obtained as described in this work. Dioxygen
18O2 (99% atom 18O) and H218O was purchased from Cortecnet
(Paris) and Euriso-top, respectively. Clotrimazole is obtained from
Sigma chemicals (Lyon, France) and fluconazole from Pfizer (Orsay,
France).
2.2. Preparation of aldehyde 3
Hydroxyterfenadine 2 (96.6 mg, 0.2 mmol), TEMPO (31.2 mg,
0.2 mmol) and tetrabutylammonium chloride (22.2 mg, 0.08 mmol)
were dissolved in CH2Cl2 and buffer (NaHCO3/K2CO3 pH 8.3,
2 ml). The mixture was stirred at room temperature and N-
chlorosuccinimide (53.4 mg, 0.4 mmol) was added in four portions
(2 h). After 1 h, aqueous layer was extracted with CH2Cl2 (two
times). The combined organic layers were washed with brine,
dried over anhydrous sodium sulfate and evaporated. The residue
was chromatographed on silica gel (dichloromethane/methanol:
97/3) to afford 3 in 45% yield. 1H RMN (CDCl3, 500 MHz): ı = 9.50
(s, 1 H, CHO), 7.53–7.20 (m, 14 H, Har), 4.67 (m, 1H, CHOH),
3.26–2.97 (m, 2H, CHNCHꢀ), 2.72 (m, 2H, CH2(CH2)2CHOH), 2.56
(t, H, J = 5.00 Hz, CH), 2.51–2.34 (m, 2H, CHNCHꢀ), 2.03–1.99 (m, 2H,
CHCH2NCH2CHꢀ), 1.91–1.76 (m, 4H, CH2CH2CHOH), 1.60–1.49 (m,
2H, CHCH2NCH2CHꢀ), 1.47 (s, 6H, 2 CH3). RMN (CDCl3, 500 MHz)
ı = 203.7 (CHO), 146.9, 145.2, 141.4, 129.8, 128.2, 127.7, 127.0, 80.5,
74.1, 58.2, 54.5, 51.6, 44.2, 38.3, 25.4, 23.9, 23.1. HRMS m/z calcd for
C
32H40NO3 [M + H]+ 486.3008, Found. 486.3010.
2.3. Bacteria and culture conditions
S. platensis NRRL 2364 cultures were maintained on agar slants
(ISP medium 2) and stored at 4 ◦C. Liquid culture media containing
(per l) glucose 16 g, yeast extract (DIFCO) 4 g, malt extract (DIFCO)
10 g (YM medium), and glucose 16 g, yeast extract 4 g, malt extract
10 g and soybean peptones (Organotechnie) 5 g (YMS medium)
were sterilized without glucose at 120 ◦C for 20 min. Microorgan-
ism was cultivated at 30 ◦C for 48 h in an orbital shaker (200 rpm).
Cell dry weight (CDW) was obtained by centrifugation of 100 ml
culture medium and biomass was dried at 100 ◦C for 24 h.

174
A. El Ouarradi et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 172–178
2.4. Incubation with cell-free extracts
20 ␮l, fluconazole 40 ␮l) to obtain final concentration (clotrima-
zole: 0.2, 0.3, 0.4, 0.5 and 0.7 mM; fluconazole: 10 and 15 mM). DMF
was added in controls without inhibitor. Substrates were added
in solution in DMF (20 ␮l) to obtain final concentration (0.2 g l−1).
Incubations were monitored by HPLC throughout 48 h. Experi-
ments in presence of both inhibitors were conducted with final
concentrations clotrimazole (0.5 mM) and fluconazole (15 mM) and
hydroxyterfenadine (0.2 g l−1) or 2-phenylpropionaldehyde.
S. platensis cells (500 ml) were cultivated in YMS medium in a 2-l
Erlenmeyer flask. After 48 h, cells were collected by centrifugation.
The bacterial pellets were suspended in TEG buffer (50 mM Tris–HCl
pH 7.6, EDTA 1 mM, Glycerol 10%) containing chicken egg lysozyme
and Dnase I. After 2 h of incubation at 30 ◦C, cells were subsequently
disrupted by addition of detergent-based Bug Buster reagent, in the
presence of protease inhibitors. The cells were then sonicated and
centrifuged (7000 × g) at 4 ◦C to remove cells debris. The cloudy
supernatant, which contained both cytosolic and membrane mate-
rial, was fractionated by a centrifugation at 100 000 × g for 1 h, at
4 ◦C. The supernatant fraction, which contained the soluble pro-
teins, was decanted. The pellet containing membrane proteins was
resuspended in TEG buffer. Protein concentration of both cytosolic
and membrane fractions was estimated using Bradfod method.
Substrate hydroxyterfenadine (10–50 ␮M) and soluble or mem-
brane fractions (100–400 ␮g) were incubated in the presence of
NAD+ or NADP+ (1 mM) in TEG buffer, in a total reaction volume
of 400 ␮l. Reaction mixtures were incubated at 30 ◦C for up to
2 h, and methanol (1 volume) was added to quench the reaction.
The precipitated proteins were removed by centrifugation (10 min,
10 000 × g) and the supernatants were analyzed by HPLC.
2.6. Analyses of metabolites
Supernatants were analyzed by HPLC performed on a Gilson
HPLC interfaced to computer using the Gilson Unipoint software.
The system involved pump 305 and 306, gradient dynamic mixer
811B and autoinjector 234 and the detector was Shimadzu-SDP6A
model.
The column (Agilent, C18, 5 ␮m (250× 4.6)) was in oven (Shi-
madzu CTO-10A model) at 40 ◦C and eluted (flow rate 1 ml/min)
with a gradient solvent system: isocratic 70% (NH4OAc 0.1 M/30%
CH3CN) for 7 min followed by gradient to 100% (40% NH4OAc
0.1 M/60% CH3CN) in 5 min and held 18 min. The detection was at
UV 230 nm and sample volumes were 20 ␮l.
LC–MS data were performed with a Surveyor-LCQ Advantage
mass spectrometer, using the same conditions except a flow rate
of 0.2 ml/min and UV. The mass spectrometer was in ESI-positive
mode, using a 4 kV capillary tube voltage and an inlet temperature
of 275 ◦C.
2.5. Incubations with whole cells
The biotransformations were performed in culture medium
or in citrate buffer pH 5 at 30 ◦C in an orbital shaker (200 rpm)
and substrate was added in DMF. Biotransformation was moni-
tored as followed. Samples (800 ␮l) were diluted with methanol
(700 ␮l), mixed vigorously and centrifuged at 10 000 × g for 5 min.
The resulting supernatants were micro-filtered (0.45 ␮m) and ana-
lyzed by HPLC.
3. Results and discussion
The main goal of our study is to obtain an effective method of
fexofenadine production from terfenadine. This objective needs to
confirm the chemical nature of the intermediates and to determine
the type of enzyme catalyzing each reaction step.
2.5.1. Incubations of hydroxyterfenadine in presence of
terfenadine
S. platensis was grown in YM medium (500 ml), harvested, and
resuspended in citrate buffer pH 5 (125 ml). Cells suspension was
separate in 5 flasks and terfenadine was added in DMF (25 ␮l) to
3.1. Incubation with intermediate as substrate
The putative intermediates, alcohol 2 and aldehyde 3, respec-
tively were obtained by microbial hydroxylation as described [33]
and oxidation of alcohol 2 as depicted in Section 2 (Section 2.2).
We showed that both intermediates were oxidized into fex-
ofenadine by S. platensis NRRL 2364 cells cultured either in YM
or YMS culture medium, and this result confirmed the oxidation
pathway of terfenadine into fexofenadine (Fig. 1). The hydrox-
yterfenadine oxidation activity of YM-cells was surprising because
incubations of terfenadine with cells obtained in these condi-
tions afforded hydroxyterfenadine as the main product and a low
amount of fexofenadine. To explain these results, we supposed
an influence of terfenadine in the hydroxyterfenadine oxida-
tion. To verify this hypothesis, cells produced in YM culture
medium were incubated with hydroxyterfenadine (0.15 g l−1) in
the presence of terfenadine at various concentrations (0–0.2 g l−1).
Incubations were performed with 14 g of CDW l−1 in citrate buffer
pH 5 at 30 ◦C for 20 h (Fig. 2). At low concentrations of terfe-
nadine (0.01–0.025 g l−1), fexofenadine formation increased with
terfenadine concentration, while concentration of hydroxyterfe-
nadine was lower than initial concentration. This is the result of
hydroxylation of terfenadine in concentration-dependant manner
and of oxidation of hydroxyterfenadine in concomitant reactions.
For higher concentration of terfenadine (0.025–0.2 g l−1), hydrox-
yterfenadine formation increased with terfenadine concentration,
whereas fexofenadine formation decreased under these conditions.
Therefore, at 0.2 g l−1 of terfenadine, no formation of fexofenadine
was observed and the concentration of hydroxyterfenadine was
higher than its initial concentration.
obtain final concentrations: 0, 0.01, 0.025, 0.05, 0.1 and 0.2 g l−1
.
After 30 min, hydroxyterfenadine was added in DMF (final concen-
tration 0.15 g l−1). Biotransformation were monitoring as described
above. After 20 h, samples (5 ml) of cells suspensions were taken
and 1 ml of methanol was added. Mixtures were vigorously stirred,
centrifuged and analyzed by HPLC.
2.5.2. Biotransformation under 18O2 atmosphere
S. platensis was grown in YM or YMS medium (50 ml), harvested,
and resuspended in citrate buffer pH 5 (50 ml) in which nitrogen
was bubbled for 5 min. 2 ml of cell suspension were transferred in a
10-mlvial, which wassealed byturnoffflangestopped. Theremain-
ing air was removed by vacuum using needle and replaced by
nitrogen three times. Finally, the head space of vials were evacuated
and replaced with pure oxygen enriched with 18O2 (99 atom% 18O),
substrates were added by injection of 4 ␮l of solution in DMF (final
concentration 0.2 g l−1, 0.42 mM) and vials were placed in rotary
shaker at 200 rpm and 30 ◦C. After 48 h (hydroxyterfenadine and
aldehyde) or 120 h (terfenadine) of incubation, 400 ␮l of reaction
mixtures were withdrawn, 350 ␮l of methanol was added. After
vigorous agitation and centrifugation, supernatants were analyzed
by LC–MS.
2.5.3. Effects of cytochrome P450 inhibitors on oxidation
S. platensis was grown in flask (500 ml) containing YM or YMS
culture medium (250 ml). Culture was separated in 24× 10 ml in
25-ml flask. Inhibitors were added in solution in DMF (clotrimazole

A. El Ouarradi et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 172–178
175
Table 1
Comparison of oxidative activity catalyzed by whole cells grown in YMS and YM culture medium.
Substrate
Product
Ra
Terfenadine 1a
Hydroxyterfenadine 1b
1,12
Hydroxyterfenadine 1b
Fexofenadine 1c
1,46
Aldehyde 1d
Fexofenadine 1c
1,02
a
Ratio of specific activity (YMS/YM).
Thus, cells grown in YM culture medium were able to oxi-
dize hydroxyterfenadine to produce fexofenadine in two conditions
either when hydroxyterfenadine is used as substrate or when ter-
fenadine has disappeared more or less completely. We strongly
suggest that this phenomenon is explained by terfenadine inhi-
bition phenomenon. The same observation was described in the
oxidation of toluene and pseudocumene by xylene monooxygenase
[18].
ence of the oxidized form of the nicotinamide cofactor, NAD+ or
NADP+. No oxidation of hydroxyterfenadine was observed in these
conditions. On the contrary to flavin or P450-dependent monooxy-
genases which are known to be multi-protein systems, alcohol
dehydrogenases do not require any other protein cofactor. Thus
this result suggested that no alcohol dehydrogenase was involved
in the transformation of the alcohol intermediate.
The best oxidizing activities were observed with cells grown
in YMS culture medium. On the one hand, these cells are able to
produce fexofenadine in the presence of terfenadine. On the other
hand, hydroxyterfenadine oxidation was approx. 1.5-fold higher
than oxidation by cells grown in YM culture medium, while no sig-
nificant difference between YM and YMS-cells was observed for
terfenadine hydroxylation and aldehyde oxidation (Table 1). These
results are in agreement with the soybean peptone-induction of
alcohol-oxidative activity not inhibited by terfenadine.
3.3. Biotransformations under controlled atmosphere
These experiments were performed with S. platensis grown in
YM or YMS culture media. Under argon, terfenadine was not oxi-
dized into hydroxyterfenadine as expected when the reaction was
catalyzed by a monooxygenase. In these conditions, when hydrox-
yterfenadine and aldehyde were added as substrates, there are
no formations of aldehyde and/or acid, but the reduction of alde-
hyde was observed. These results showed survival of cells or active
enzymes in these conditions and suggested that the oxidations of
alcohol and aldehyde were dioxygen-dependant.
3.2. Incubation with cell-free extracts
To investigate the dioxygen function in these reactions, ter-
fenadine, hydroxyterfenadine and aldehyde were incubated with
whole cells of S. platensis under 18O2-enriched atmosphere
(99 atom% 18O). The reaction mixtures were analyzed by liquid
chromatography–mass spectrometry in ESI-positive mode and the
m/z values at 488, 486 and 502 Da for standard hydroxyterfenadine,
aldehyde and fexofenadine correspond to [M + H]+ molecular ions,
respectively. The results are summarized in Table 2.
In both culture conditions, mass spectrum of hydroxyterfena-
dine formed from terfenadine showed incorporation of one atom
of oxygen, with the normalized ratios of [M + H]+ to [M + H]+ + 2
molecular ion peaks of 10:90 and 20:80 for YM and YMS culture
conditions, respectively (entries 1 and 2).
Further oxidations of hydroxyterfenadine afford fexofenadine,
which is formed by incorporation of a second oxygen-atom.
In monooxygenase-dependent reaction, this oxygen-atom is
incorporated from molecular dioxygen whereas in dehydrogenase-
dependent reaction, fexofenadine is formed from aldehyde by
incorporation of one oxygen atom from water. Then, when terfe-
nadine was added as a substrate in an 18O2-enriched atmosphere,
fexofenadine formed via monooxygenase catalyzed oxidation
would give an [M + H]+ ion at 506 Da, 4 Da greater than that of
the standard, whereas fexofenadine formed via dehydrogenases
(alcohol and aldehyde dehydrogenase) catalyzed oxidation would
give an [M + H]+ ion at 504 Da, 2 Da greater than that of the
standard. In this study, fexofenadine (49%) was obtained after 120-
h incubation of terfenadine with S. platensis whole cells grown
in YMS-culture medium and mass spectrum showed a ratio of
[M + H]+/[M + H]+ + 2/[M + H]+ + 4 molecular ion peaks of 7:64:29
(entry 2). Thus, we observed an incorporation of two 18O atoms
into the carboxylic acid with a [M + H]+ + 4 enrichment of 29%.
Mass spectra of fexofenadine formed when hydroxyterfena-
dine was added as substrate in incubations with S. platensis whole
cells grown in YM and YMS-culture medium showed a ratio of
[M + H]+/[M + H]+ + 2/[M + H]+ + 4 molecular ion peaks of 5:84:11
and 6:80:14, respectively (entries 3 and 4). Fexofenadine obtained
in incubations performed with aldehyde as substrate, was entirely
mono-labeled (entries 5 and 6). Obviously, oxidation of aldehyde
was catalyzed by monooxygenase(s).
Experiments were conducted with cell-free extracts on ter-
fenadine metabolism, in order to determine if flavin or P450
dependent monooxygenases were involved in the oxidation of the
methyl group of the t-butyl moiety. S. platensis was cultivated
in YMS culture medium and cells were disrupted as described
in Section 2 (Section 2.6). Terfenadine was incubated with solu-
ble or membrane fractions or a mixture of them, in the presence
of the reduced form of the nicotinamide cofactor, NADH, H+
or NADPH, H+. No oxidation of terfenadine was observed with
either compartment cells, in these conditions. Indeed, such a
loss of enzymatic activity is often observed in the case of multi-
protein complexes, as P450-dependent monooxygenases usually
are. Therefore, this result indicated that P450s could be responsible
for the first step of terfenadine oxidation, i.e. its hydroxyla-
tion.
In the same manner, experiments were conducted on hydrox-
yterfenadine metabolism, in order to determine if an alcohol
dehydrogenase was involved in the oxidation of the alcohol into the
corresponding aldehyde. Hydroxyterfenadine was incubated with
soluble or membrane fractions or a mixture of them, in the pres-
Fig. 2. Concentration of fexofenadine and hydroxyterfenadine after 20 h of incu-
bation of hydroxyterfenadine (0.15 g l−1) in presence of different concentrations
of terfenadine. Biotransformations by S. platensis-whole cells grown in YM culture
medium (65 h). Symbols: square, hydroxyterfenadine; triangle, fexofenadine.

176
A. El Ouarradi et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 172–178
Table 2
LC/MS analysis of hydroxyterfenadine 2 and Fexofenadine 4 formed from incubation of terfenadine 1 (120 h), hydroxyterfenadine 2 (48 h) and corresponding aldehyde 3
under 18O2-atmosphere.
Substrate
Culture media
Incubation products
2
4
a
a
%
[M + H]+/[M + H]+ + 2b
%
[M + H]+/[M + H]+ + 2/[M + H]+ + 4b
1
1
2
2
3
3
YM
YMS
YM
YMS
YM
YMS
80
33
15
10
0
10/90
20/80
90/10
86/14
0
49
85
90
100
100
7/64/29
5/84/11
6/80/14
3/97/0
0
3/97/0
a
Relative amounts of oxidation products (difference with 100% corresponding to terfenadine).
Normalized ratios of isotopic compounds, representative of 3 experiments, except for aldehyde.
b
Since the incorporation of the second oxygen-atom occurs dur-
involved in the secondary metabolism [42]. According to this fact,
the effect of cytochromes P450 selective inhibitors on terfenadine
oxidative pathway was investigated.
ing the oxidation of aldehyde, the loss of labelling observed in
fexofenadine from incubation of terfenadine could result from an
exchange with water through equilibrium between aldehyde and
hydrate form. We demonstrated this rapid exchange of oxygen
atom by incubation of aldehyde in 18OH2 and monitoring by mass
spectrometry. After 30 min aldehyde was entirely labeled.
Several mechanisms have been proposed for aldehyde oxidation
into carboxylic acid by monooxygenase, including hydroxylation
of the aldehyde or of its hydrated form as described for acetalde-
hyde oxidation by cytochrome P450 2E1 [40]. However, study of
oxidation of 11-oxo-ꢀ8-tetrahydrocannabinol [41] rather showed
the monooxygenation of the aldehyde. Results of this study were
in agreement with oxidation of aldehyde form because the acid
formed from the aldehyde was entirely labeled. Indeed, in mech-
anism involving the hydrated form, the mono-labeled ortho acid
intermediate shall give unlabeled acid in large amount.
Fig. 3 shows the proposed pathway to explain the formation
of doubly and mono-labeled fexofenadine from the oxidation of
terfenadine, hydroxyterfenadine and aldehyde. The formation of a
doubly labeled fexofenadine demonstrated that labeled aldehyde
was stable enough to be oxygenated before exchange of oxygen
atom. To explain this result it could be suggested that one enzyme
catalyze the oxidation of alcohol and of aldehyde without its releas-
ing from the active site.
Cytochrome P450 selective inhibitors are extensively used to
determine which enzyme isoform catalyzes the oxidation of drugs
[43]. Involvement of CYP3A4 [44] and 2D6 [45] in terfenadine
metabolism was demonstrated by this method. The approach is
also used to establish the participation of cytochrome P450 in
microbial biotransformation because the direct demonstration of
their contribution is difficult [36,39,46–48]. We tested five com-
pounds (clotrimazole, fluconazole, methyrapone, menadione and
SKF525A) from different classes of P450 inhibitors on the biotrans-
formation of terfenadine, hydroxyterfenadine and aldehyde by S.
platensis cells grown in culture medium containing or not soybean
peptones. Inhibitors were added in solution in DMF to incubations
(final concentration 0.5 mM except for fluconazole 15 mM) 30 min
before substrate. The formation of oxidation product was moni-
toring by HPLC analysis during 24 h. All compounds showed an
inhibition effect, but because problems of co-elution with substrate
or oxidation products, only clotrimazole and fluconazole were used
in further investigation.
Kinetic studies of the inhibition of S. platensis cells catalyzed
oxidation of terfenadine 1, hydroxyterfenadine 2 and aldehyde 3
by clotrimazole and fluconazole were performed at various clotri-
mazole (0–0.7 mM), fluconazole (0–15 mM) concentrations. Fig. 4
shows their inhibitory activity against S. platensis cells grown in
YM (Fig. 4A) or in YMS (Fig. 4B) culture medium oxidation of
three substrates. The inhibition effect was inhibitor concentration
dependent. Clotrimazole caused the almost complete inhibition
of terfenadine and aldehyde oxidation, and did not inhibit com-
pletely hydroxyterfenadine oxidation. The inhibitory activity of
fluconazole depended on cells culture condition and substrate.
Higher level of inhibition of terfenadine oxidation was observed
3.4. Effects of cytochrome P450 inhibitors
The insertion of one atom of oxygen into a non-activated CH
bond can be catalyzed by several heme or non-heme enzymes, as
methane monooxygenases (MMO), flavin monooxygenases (FMO)
or cytochromes P450-monooxygenases. Streptomyces species are
well known to contain a wealth of cytochromes P450, most of them
Fig. 3. Proposed pathways for oxidation of terfenadine 1, hydroxyterfenadine 2 and aldehyde 3 under 18O2-atmosphere.

A. El Ouarradi et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 172–178
177
Fig. 4. Effect of clotrimazole and fluconazole on oxidative activities of S. platensis. (A) Cells grown in YM culture medium and (B) cells grown in YMS culture medium. Data
are presented as the remaining relative activity as a function of inhibitor concentration. Control activities of oxidation were determined for each substrate. Calculations based
on peak areas HPLC analysis of the main metabolite after 24 h incubation of three different substrates: terfenadine, RCH3 (᭹); hydroxyterfenadine, RCH2OH (ꢀ); aldehyde
RCHO (ꢁ).
against cells grown in YMS-culture medium than against cells
grown in YM-culture medium while opposite effect was observed
in hydroxyterfenadine oxidation. These results suggested that
several monooxygenase activities were involved in hydroxyterfe-
nadine oxidation, some of them being inhibited by clotrimazole and
other by fluconazole. This hypothesis was confirmed by incubation
of hydroxyterfenadine with S. platensis cells in presence of mix-
ture of inhibitors, clotrimazole (0.5 mM) and fluconazole (15 mM).
Incubations were performed with cells grown in YM and YMS cul-
ture medium, and biotransformation was widely (20% remaining
activity) and almost completely inhibited, respectively (data no
showed). The control of an absence of effect of inhibitors on other
enzymatic activities and cells survival was accomplished. For that,
we studied the biotransformation of 2-phenylpropionaldehyde by
S. platensis in presence and in the absence of both inhibitors.
Reduction was observed in both conditions in similar manner
(data no showed) and we can therefore think reasonably that the
observed decrease of oxidizing activities results from inhibition of
cytochromes P450.
4. Conclusion
This study shows that S. platensis transforms terfenadine into
fexofenadine via hydroxyterfenadine and corresponding alde-
hyde, which has been synthesized by a chemoenzymatic way.
This metabolism involves cytochromes P450 in three oxidative
reactions: fexofenadine obtained from terfenadine, hydroxyterfe-
nadine and from aldehyde was labeled when incubations were
performed in 18O2-enriched atmosphere, and the reactions were
inhibited by cytochromes P450 inhibitors. The best hydroxyter-
fenadine oxidation activity was obtained when S. platensis was
grown in presence of soybean peptones. This is the first example
in biocatalytic application using microbial wild type strain where
involvement of monooxygenases was established in three-step oxi-
dation of a methyl group into carboxylic acid. Work is in progress
concerning the isolation, purification and cloning of one of the
cytochrome P450 catalyzing these oxidations.
Acknowledgments
This study shows that S. platensis metabolism of terfenadine
into fexofenadine involves cytochromes P450 in three oxidative
reactions and could be compared to mammalian metabolism. It
has been shown that CYP3A4 and CYP2J2 are responsible for the
formation of fexofenadine [49] and carebastine [50].
We thank Mrs. A. Triclin for technical assistance, and Drs. Pierre
Lafite, Patrick Dansette and Dr Chantal Blandin for fruitful discus-
sions. Part of this work was financially supported by RDR2 “Aller
vers une Chimie écocompatible”.
Genomic analysis of different Streptomyces species revealed the
presence of a large number of CYP genes: 18 CYP genes were found
in S. coelicolor [51], 15 in S. peucetius [52] and 33 in S. avermiti-
cilis [53]. In the vast majority of cases the physiological functions
of these CYP are unknown. However, because of the various types
of secondary metabolites with important medical application pro-
duced by Streptomyces species, it is envisaged that CYP enzymes
may have roles in their biosynthesis [54]. For example, it was
recently established that a CYP of S. avermiticilis (PtII) can cat-
alyze the two-step oxidation of methyl group of pentalenene to
pentalen-3al, but the enzyme involved in the oxidation of aldehyde
was not characterized yet [55]. Then, it would not be surprising
that several cytochromes P450 are involved in the three-step oxi-
dation of terfenadine, each reaction could be catalyzed by one
or more enzymes. Consequently, the multiplicity of the necessary
multi-component systems to produce fexofenadine from terfena-
dine makes difficult the improvement of an efficient biocatalytic
system [56].
References
[1] B. Buhler, A. Schmid, J. Biotechnol. 113 (2004) 183–210.
[2] A.J.J. Straathof, S. Panke, A. Schmid, Curr. Opin. Biotechnol. 13 (2002) 548.
[3] A.S. Bommarus, K.M. Polizzi, Chem. Eng. Sci. 61 (2006) 1004–1016.
[4] A. Markham, A. Wagstaff, Drugs 55 (1998) 269–274.
[5] K. Simpson, B. Jarvis, Drugs 59 (2000) 301–321.
[6] J. Meiwes, M. Worm, US Patent 5,990,127, Hoechst Marion Roussel, Deutchland
(1999).
[7] R. Azerad, J. Biton, I. Lacroix, WO Patent 47693, Hoechst Marion Roussel,
Puteaux, Fr (1999).
[8] A. El Ouarradi, I. Salar-Arnaud, D. Buisson, Tetrahedron 64 (2008) 11738–11744.
[9] C. Mazier, M. Jaouen, M.-A. Sari, D. Buisson, Bioorg. Med. Chem. Lett. 14 (2004)
5423–5426.
[10] S.G. Burton, D.A. Cowan, J.M. Woodley, Nat. Biotechnol. 20 (2002) 37–45.
[11] W. Kroutil, H. Mang, K. Edegger, K. Faber, Adv. Synth. Catal. 346 (2004) 125–142.
[12] F. Molinari, Curr. Org. Chem. 10 (2006) 1247–1263.
[13] M. Sono, M.P. Roach, E.D. Coulter, J.H. Dawson, Chem. Rev. 96 (1996)
2841–2887.
[14] J. Hirano, K. Miyamoto, H. Ohta, Tetrahedron Lett. 49 (2008) 1217–1219.
[15] a. Kiener, Angew. Chem. Int. Ed. Eng. 31 (1992) 774–775.

178
A. El Ouarradi et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 172–178
[16] N.M. Shaw, K.T. Robins, A. Kiener, Adv. Synth. Catal. 345 (2003) 425–435.
[17] B. Buhler, A. Schmid, B. Hauer, B. Witholt, J. Biol. Chem. 275 (2000)
10085–10092.
[18] B. Buhler, B. Witholt, B. Hauer, A. Schmid, Appl. Environ. Microbiol. 68 (2002)
560–568.
[19] K.J. French, M.D. Strickler, D.A. Rock, D.A. Rock, G.A. Bennett, J.L. Wahlstrom,
B.M. Goldstein, J.P. Jones, Biochemistry 40 (2001) 9532–9538.
[20] W.H. Eschenfeldt, Y.Y. Zhang, H. Samaha, L. Stols, L.D. Eirich, C.R. Wilson, M.I.
Donnelly, Appl. Environ. Microbiol. 69 (2003) 5992–5999.
[21] L.D. Eirich, D.L. Craft, L. Steinberg, A. Asif, W.H. Eschenfeldt, L. Stols, M.I. Don-
nelly, C.R. Wilson, Appl. Environ. Microbiol. 70 (2004) 4872–4879.
[22] H. Uchida, Y. Okamura, H. Yamanaka, T. Fukuda, S. Haneda, K. Aisaka, Y. Fujii,
World J. Microbiol. Biotechnol. 22 (2006) 469–474.
[23] Y. Maeda, A. Yagyu, A. Sakurai, Y. Fujii, H. Uchida, World J. Microbiol. Biotechnol.
24 (2008) 797–804.
[24] M. Hofrichter, R. Ullrich, Appl. Microbiol. Biotechnol. 71 (2006) 276–288.
[25] H. Schwartz, R.E. Licht, H.E. Radunz, Appl. Microbiol. Biotechnol. 40 (1993)
382–385.
[26] H. Pérez, H. Luna, N. Manjarerrez, A. Solis, M. Nunez, Biotechnol. Lett. 21 (1999)
855–858.
[27] N. Tarui, Y. Ikeura, H. Natsugari, K. Nakahama, J. Biosci. Bioeng. 92 (2001)
285–287.
[36] K. Mitsukura, Y. Kondo, T. Yoshida, T. Nagasawa, Appl. Microbiol. Biotechnol.
71 (2006) 502–504.
[37] B. Kolvenbach, N. Schlaich, Z. Raoui, J. Prell, S. Zuhlke, A. Schaffer, F.P. Guen-
gerich, P.F.X. Corvini, Appl. Environ. Microbiol. 73 (2007) 4776–4784.
[38] R. Seshadri, A.S. Lamm, A. Khare, J.P.N. Rosazza, Enzyme Microb. Technol. 43
(2008) 486–494.
[39] M. Sanchez-Gonzalez, J.P.N. Rosazza, J. Nat. Prod. 67 (2004) 553–558.
[40] L.C. Bell-Parikh, F.P. Guengerich, J. Biol. Chem. 274 (1999) 23833–23840.
[41] K. Watanabe, S. Narimatsu, I. Yamamoto, H. Yoshimura, J. Biol. Chem. 26 (1991)
2709–2711.
[42] D.C. Lamb, F.P. Guengerich, S.L. Kelly, M.R. Waterman, Expert Opin. Drug Metab.
Toxicol. 2 (2006) 27–40.
[43] O. Pelkonen, M. Turpeinen, J. Hakkola, P. Honkakoski, J. Hukkanen, H. Raunio,
Arch. Toxicol. 82 (2008) 667–715.
[44] C.H. Yun, R.A. Okerholm, F.P. Guengerich, Drug Metab. Disp. 21 (1993) 403–409.
[45] B.C. Jones, R. Hyland, M. Ackland, C.H. Tyman, D.A. Smith, Drug Metab. Disp. 29
(1998) 875–882.
[46] H.L. George, H.D. VanEtten, Fungal Genet. Biol. 33 (2001) 37–48.
[47] D. Buisson, J. Quintin, G. Lewin, J. Nat. Prod. 70 (2007) 1035–1038.
[48] X.L. Xin, D.H. Su, X.J. Wang, Q.P. Yuan, J. Mol. Catal. B: Enzyme 59 (2009)
201–205.
[49] K.H.J. Ling, G.A. Leeson, S.D. Burmaster, R.H. Hook, M.K. Reith, L.K. Cheng, Drug
Metab. Disp. 23 (1995) 631–636.
[28] J.W. Wong, H.A. Watson, J.F. Bouressa, M.P. Burns, J.J. Cawley, A.E. Doro, D.B.
Guzek, M.A. Hintz, E.L. McCormick, D.A. Scully, J.M. Siderewicz, W.J. Taylor, S.J.
Truesdell, R.G. Wax, Org. Proc. Res. Dev. 6 (2002) 477–481.
[29] M.S. Pacey, Enzymatic Oxidation, Pfizer Products Inc., NY, USA, 2003.
[30] D. Abanda-Nkpwatt, W. Schwab, J. Agric. Food Chem. 52 (2004) 5939–5942.
[31] H. Schwartz, A. Liebig-Weber, H. Hochstätter, H. Böttcher, Appl. Microbiol.
Biotechnol. 44 (1996) 731–735.
[50] K.H. Liu, M.G. Kim, D.J. Lee, Y.J. Yoon, M.J. Kim, J.H. Shon, C.S. Choi, Y.K. Choi, Z.
Desta, J.G. Shin, Drug Metab. Disp. 34 (2006) 1793–1797.
[51] D.C. Lamb, T. Skaug, H.L. Song, C.J. Jackson, L.M. Podust, M.R. Waterman, D.B.
Kell, D.E. Kelly, S.L. Kelly, J. Biol. Chem. 277 (2002) 24000–24005.
[52] N. Parajuli, D.B. Basnet, H.C. Lee, J.K. Sohng, K. Liou, Arch. Biochem. Biophys.
425 (2004) 233–241.
[32] F.S. Sariaslani, D.A. Kunz, Biochem. Biophys. Res. Commun. 141 (1986) 405–410.
[33] C. Mazier, M. Lombard, M.-A. Sari, D. Buisson, Biocatal. Biotransf. 25 (2007)
401–407.
[34] Y.J. Chun, T. Shimada, R. Sanchez-Ponce, M.V. Martin, L. Lei, B. Zhao, S.L.
Kelly, M.R. Waterman, D.C. Lamb, F.P. Guengerich, J. Biol. Chem. 282 (2007)
17486–17500.
[53] D.C. Lamb, H. Ikeda, D.R. Nelson, J. Ishikawa, T. Skaug, C. Jackson, S. Omura, M.R.
Waterman, S.L. Kelly, Biochem. Biophys. Res. Commun. 307 (2003) 610–619.
[54] B. Zhao, M.R. Waterman, Drug. Metab. Rev. 39 (2007) 343–352.
[55] R. Quaderer, S. Omura, H. Ikeda, D.E. Cane, J. Am. Chem. Soc. 128 (2006)
13036–13037.
[56] Y.J. Chun, T. Shimada, M.R. Waterman, F.P. Guengerich, Biochem. Soc. Trans. 34
(2006) 1183–1185.
[35] H. Ichinose, H. Wariishi, H. Tanaka, Biotechnol. Prog. 15 (1999) 706–714.