In vitro biotransformations of the prostaglandin D2 (DP) antagonist MK-0524 and synthesis of metabolites

Article abstract of DOI:10.1016/j.bmcl.2006.10.055

Metabolites of the potent DP antagonist, MK-0524, were generated using in vitro systems including hepatic microsomes and hepatocytes. Four metabolites (two hydroxylated diastereomers, a ketone and an acyl glucuronide) were characterized by LC-MS/MS and

Full text of DOI:10.1016/j.bmcl.2006.10.055

Bioorganic & Medicinal Chemistry Letters 17 (2007) 301–304
In vitro biotransformations of the prostaglandin D2 (DP)
antagonist MK-0524 and synthesis of metabolites
a
a
a
Deborah A. Nicoll-Griffith,a, Carmai Seto, Yves Aubin, Jean Franc¸ois Levesque,
*
´
a
a
a
a
´
Nathalie Chauret, Stephen Day, Jose M. Silva, Laird A. Trimble,
Jean-Franc¸ois Truchon,a Carl Berthelette,a Nicolas Lachance,a Zhaoyin Wang,a
Claudio Sturino,a Matt Braun,b Robert Zambonia and Robert N. Younga
aMerck Frosst Centre for Therapeutic Research, PO Box 1005, Pointe Claire—Dorval, Que., Canada H9R 4P8
bMerck Research Laboratories, 126 Lincoln Avenue, Rahway, NJ 07065, USA
Received 7 September 2006; revised 15 October 2006; accepted 23 October 2006
Available online 25 October 2006
Abstract—Metabolites of the potent DP antagonist, MK-0524, were generated using in vitro systems including hepatic microsomes
and hepatocytes. Four metabolites (two hydroxylated diastereomers, a ketone and an acyl glucuronide) were characterized by LC–
1
MS/MS and H NMR. Larger quantities of these metabolites were prepared by either organic synthesis or biosynthetically to be
used as standards in other studies. The propensity for covalent binding was assessed and was found to be acceptable (<50 pmol-
equiv/mg protein).
ꢀ 2006 Elsevier Ltd. All rights reserved.
Prostaglandin D2, a major cyclooxygenase-derived
prostanoid is implicated in congestion, inflammatory
processes, and nicotinic acid-induced flushing.1–3 MK-
0524, [(3R)-4-(4-chlorobenzyl)-7-fluoro-5-methylsulfo-
nyl)-1,2,3,4-tetrahydrocyclopenta[b]indol-3-yl)acetic acid)],
is a potent and selective prostaglandin D2 receptor (DP)
antagonist that has been shown to inhibit antigen-induced
increases in nasal resistance in animals4 and nicotinic acid-
induced flushing in mice and humans.3
Intermediates in the metabolic reactions, or the products
themselves, may be reactive and give rise to covalent
protein binding. Evaluation of this is important since
covalent protein modification may lead to unwanted
toxicities or idiosyncratic reactions in humans.5
In order to generate metabolites for characterization,
incubations of MK-0524 were conducted under standard
conditions using suspended hepatocytes or hepatic
microsomal incubations (supplemented with NADPH
or uridinediphospho-glucuronic acid (UDPGA)) from
several species including human.6,7 All incubates were
analyzed by reverse-phase liquid chromatography-mass
spectrometry on a 2790 HT liquid chromatograph cou-
pled to a Waters 996 photodiode array detector and a
Micromass Quattro LC triple quadrupole mass spec-
trometer operated in negative ion mode. Figure 1 shows
the UV-chromatogram of a rat hepatocyte incubation.
As a part of drug discovery and development, the meta-
bolic profiles of new drug candidates must be character-
ized. Drug metabolites are the products of enzymatic
modifications such as oxidation or conjugate formation.
O
F
OH
Cl
N
O S O
HPLC–MS analysis indicated that two major metabo-
lites, M1 and M4, were a result of oxidation of MK-
0524 because the observed [MÀH]À ions were 16 Da
higher than MK-0524. Another oxidative metabolite
(M3) was also observed with a mass 14 Da higher than
MK-0524. MS fragmentation patterns for M1, M3,
and M4 were not diagnostic enough to pinpoint the sites
of oxidation. In the hepatocyte incubations, conjugated
MK-0524
Keywords: DP receptor antagonist; MK-0524; Biotransformations;
Metabolites; Reactive intermediates.
*
Corresponding author. Tel.: +1 514 428 8619; fax: +1 514 428
4900; e-mail: deborah_nicoll-griffith@merck.com
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.10.055

302
D. A. Nicoll-Griffith et al. / Bioorg. Med. Chem. Lett. 17 (2007) 301–304
MK-0524
Specifically, for preparative HPLC, a YMC ODS-AM
column (50 · 20 mm) with 20 mL/min mobile phase flow
rate was used. Electrospray mass spectrometric detec-
tion on a Micromass ZMD mass spectrometer was used
to automate peak detection and collection. Prior to
evaporation of the purified fractions, the pH was raised
to ꢀ9 to prevent metabolite degradation. Specifically,
200 lL of 8% aqueous ammonium hydroxide was added
per 20 mL of HPLC eluant. Once the volume had been
reduced to half by rotary evaporation, the remaining
solvent was removed by lyophilization.
100
M4
M1
%
0
M2 M8
M3
12.0
Tim
6.0
8.0
10.0
14.0
16.0
Figure 1. HPLC of a rat hepatocyte incubation of MK-0524. The
incubation was conducted using 50 lM MK-0524 and 106 cells for 2 h
at 37 ꢁC. HPLC analysis was performed on a YMC ODS-A S5 column
(4.6 · 150 mm) with UV detection at 305 nm. The mobile phase flow
rate was 1 mL/min with a gradient for which solvent A was 20 mM
ammonium acetate, 5% methanol in water and B was acetonitrile. The
acetonitrile was increased from 5% to 65% over 17.5 min. The peak at
8.2 min was present in blank incubations and is not related to MK-
0524.
1
The H NMR spectra of the resulting solids were ob-
tained on a Varian 600 MHz NMR. Initial attempts
indicated that the metabolites were not stable in
DMSO-d6, however, 2 mM Tris-d11 in D2O was used
successfully (pH ꢀ9.5). MK-0524 was fully assigned
1
and compared to the H NMR spectrum of M1 and
M4. All aromatic and benzylic protons were present in
both metabolites (Table 1). The two protons at position
1 were replaced by a single proton shifted down field by
more than 2 ppm, consistent with hydroxylation at car-
bon 1. The structures of M1 and M4 were proposed to
be hydroxylated materials, as shown in Figure 2, based
on these NMR data and the literature precedent of
3-methylindole oxidation.9 The absolute stereochemistry
of the hydroxylation products M1 and M4 could be
determined by comparing vicinal proton coupling con-
stants and molecular models of the metabolites. The vic-
inal coupling constants between the remaining proton at
position 1 and the two geminal protons at position 2
were 6.4 and 3.4 Hz for M1 and 6.9 and 1.3 Hz for
M4. The calculated coupling constants from the molec-
ular models allowed the identification of absolute stereo-
chemistry of the new hydroxyl groups on M1 and M4 to
metabolites were also observed (M2 and M8). HPLC–
MS analysis indicated a 176 Da mass shift increment
compared to MK-0524, indicating that glucuro-conju-
gates were formed. The mass spectrometric fragmenta-
tion pattern was not diagnostic enough to identify the
site of conjugation. Dog and human hepatocyte incuba-
tions gave similar profiles. Hepatic microsomal incuba-
tions in various species (rat, rabbit, dog, sheep,
squirrel monkey, rhesus monkey, and human) produced
the oxidative metabolites in the presence of NADPH
and the glucuronide in the presence of UDPGA, as as-
sessed by HPLC–MS.
Incubations of MK-0524 with microsomes expressing
individual recombinant cytochromes P450 (rCYPs,
GENTEST Corp., Woburn, USA) were performed un-
der standard conditions.7 CYP3A was by far the most
active in producing M1, M3, and M4, while CYP1A
and 2C9/19 also contributed to the formation of M1
and M4, respectively. A CYP3A4 bioreactor8 was there-
fore utilized to produce milligram quantities of M1 and
M4 for 1H NMR characterization and biological testing.
The bioreactor contained human CYP3A4, oxidoreduc-
tase (OR), and cytochrome b5 (b5) expressed with the
baculovirus-SF-9 cell expression system. Typically, a
2 L culture of SF-9 cells was incubated at 27 ꢁC and
infected with three viruses expressing the CYP3A4,
OR, and b5 genes and 24 h post-infection, hemin was
added. At 24 h post-addition of hemin, MK-0524 was
added and the culture was incubated at 27 ꢁC for anoth-
er 24 h or until the cell viability decreased below 50%. In
a 2 L culture, which consisted of a 25 lM solution of
MK-0524 (22 mg), a turnover of 80% was achieved,
which consisted of a mixture of 35% M1, 2% M3, and
43% M4, as indicated by reverse-phase HPLC–UV. Be-
cause the metabolites were partitioned between media
and cells, the latter were harvested by centrifugation.
The supernatant was set aside and the cell pellet was
extracted with a 1:1 mixture of acetonitrile and water.
Table 1. 1H NMR chemical shift and multiplicity data for MK-0524
and oxidative metabolites
1
8
2
O
F
13
3
OH
Cl
6
N
MK-0524
15
O S O
20
17
18
Protons
MK-0524
M1
5.49 (br s)
—
M2
M3
5.28 (dd)
—
1a
1b
2a
2b
3
2.91 (dd)
2.73 (m)
2.69 (m)
2.23 (m)
3.42 (m)
7.50 (d)
7.48 (d)
2.51 (dd)
2.23 (m)
5.97 (d)
5.68 (d)
6.70 (d)
7.20 (d)
2.98 (s)
—
—
2.57 (m)
2.42 (m)
3.54 (m)
7.65 (dd)
7.57 (dd)
2.48 (dd)
2.03 (dd)
5.91 (d)
5.60 (d)
6.63 (d)
7.20 (d)
2.99 (s)
3.21 (dd)
3.03 (m)
2.64 (d)
3.60 (m)
7.59 (dd)
7.71 (dd)
2.58 (dd)
2.17 (dd)
5.84 (m)
5.79 (m)
6.74 (d)
7.2 (d)
1.97 (d)
3.25 (m)
7.66 (dd)
7.57 (dd)
2.58 (dd)
2.24 (dd)
5.88 (d)
5.63 (d)
6.61 (d)
7.19 (d)
2.99 (s)
6
8
13a
13b
15a
15b
17
18
20
2.76 (s)
The metabolites were then isolated by using a combina-
tion of solid-phase extraction and preparative HPLC.
All samples were dissolved in 2 mM Tris-d11 in D2O. Numbering of
carbons is indicated below.

D. A. Nicoll-Griffith et al. / Bioorg. Med. Chem. Lett. 17 (2007) 301–304
303
OH
The small amount of M3 generated in the CYP 3A4 bio-
reactor precluded isolation for NMR characterization.
However, based on MS data, it was speculated that
M3 was the ketone resulting from oxidation of M1
and/or M4.
R2
R1
N
OH
OH
OH
O
O
F
F
O
O
OH
Cl
N
O
O
S O
O
S O
Cl
A standard of M3 was prepared by the synthesis as
shown in Scheme 1. Briefly, a DDQ oxidation of the es-
ter 611 afforded the ketone 7. Copper coupling of the
bromine with methanesulfinic acid sodium salt yielded
compound 8. Saponification of 8 afforded the synthetic
metabolite, M3. The structure of M3 was confirmed
by an HPLC co-injection of the authentic standard with
a rat hepatocyte incubation.
M1 R1 : H R2 : OH
M3 R1, R2 : =O
M2
M4 R1 : OH R2 : H
Figure 2. Structures of oxidative and conjugative metabolites of MK-
0524.
be syn and anti, to the carboxyl-containing side chain,
respectively.
Attempts to produce the alcohol metabolites, M1 and
M4, by reduction of M3 were unsuccessful, presumably
due to the electronic nature of the vinylogous indole
amide nitrogen. Also, attempts to oxidize MK-0524
using a variety of synthetic protocols failed to yield
the hydroxylated metabolites, thus confirming the neces-
sity of the CYP3A4 bioreactor synthesis, described
above.
M1 and M4 were found to interconvert under even
slightly acidic conditions such as a mixture of 14 mM
aqueous ammonium acetate and acetonitrile (65:35) at
ambient temperature (Fig. 3). An equilibrium was
reached after approximately 48 h, yielding an M1 to
M4 ratio of 2:3. However, if M1 and M4 were dissolved
in an acetonitrile solution containing ammonium
hydroxide (0.01% v/v) minimal interconversion was ob-
served after 24 h (<5%). Similar stability was observed
in rat plasma. This is consistent with the fact that plas-
ma becomes basic under ambient conditions, reaching
pH values greater than 8 within a few hours of
collection.10
Maximal production of the glucuronide conjugate, M2,
was achieved using dog hepatic microsomes fortified
with UDPGA.12 Other species gave lower turnover. A
25 mL incubation containing 2 mg/mL of dog hepatic
microsomal protein and 1.1 mg MK-0524 (100 lM)
afforded a 91% yield of M2 after 2 h, as indicated by re-
verse-phase HPLC–UV. The product was isolated using
a combination of solid-phase extraction and preparative
HPLC conditions similar to those described above, ex-
1
100
A
cept that the pH was not adjusted. H NMR of the iso-
% M1
%M4
lated material at 600 MHz (DMSO-d6) showed
characteristic resonance for the anomeric protons at
ꢀ5.3 ppm. The coupling constant was 8.1 Hz, consistent
with the trans-diaxial coupling expected for the forma-
tion of a b-linked O-acyl glucuronide as shown in Figure
2.12 Synthetic acyl glucuronide, M2, was produced
according to previously described methodology.13
80
60
40
20
0
Based on the known phenomena of acyl migration ob-
served with b-linked O-acyl glucuronides,14 M8 is pro-
posed to be the 2-O-linked acyl glucuronide. LC–MS
0
24
48
72
96
120
144
Time (h)
100
B
% M1
%M4
O
O
F
O
F
80
60
40
20
0
DDQ
O
O
N
N
Cl THF, H2O
35-50%
Br
Cl
Br
6
7
MeSO2Na, CuI
NMP 27%
O
O
O
O
F
F
NaOH
OH
0
24
48
72
96
120
144
O
N
N
Time (h)
THF, H2O
Cl
Cl
O
S O
50%
O S
O
Figure 3. Stability study of M1 (A) and M4 (B) in 14 mM aqueous
ammonium acetate and acetonitrile (65:35) at ambient temperature.
Analysis was conducted by HPLC–UV.
M2
8
Scheme 1. Synthesis of the ketone metabolite M2.

304
D. A. Nicoll-Griffith et al. / Bioorg. Med. Chem. Lett. 17 (2007) 301–304
Table 2. Covalent protein modification in vitro and in vivo
The biosynthetic and synthetic metabolites prepared as
described herein were used as standards for additional
MK-0524 metabolism studies.20–22
Test system
Rat
Human
Microsomes (with NADPH)a
Hepatocytesb
In vivoc:
33
38
16
14
Acknowledgments
Liver (24 h)
Plasma (24 h)
5
ND
ˆ
The authors thank Dr. Marc Bilodeau of Hopital St.
ND, not detected (limit of detection 5 pmol-equiv/mg protein).
a,bValues reported as pmol-equiv binding/mg protein at 1 h.18
cValues reported as pmol-equiv binding/mg protein.18
´
´
Luc (Montreal, Quebec) for providing human liver tis-
sue from which hepatocytes were isolated. Helpful dis-
cussions with Dr. Gary O’Neill during manuscript
preparation were especially welcomed.
spectra of this metabolite gave mass spectral data indi-
cating a molecular weight of M+176 and microsomal
incubations to form the glucuronide initially yielded
only M2. M8 was produced upon standing, with con-
comitant loss of M2, consistent with it being an M2
degradant. A stability study was performed at 37 ꢁC un-
der slightly basic conditions (1 mM M2 in 125 mM
phosphate buffer (pH 7.4) and DMSO-d6 (70:30). M2
converted to M8 and MK-0524 with a degradation
half-life of 95 min. A similar half-life has been reported
for the acyl glucuronide of indomethacin in phosphate
buffer at pH 7.4.14 It is known, however, that b-linked
O-acyl glucuronides are stable under acidic conditions15
and this was found to be the case with M2 (data not
shown).
References and notes
1. Lewis, R. A.; Soter, N. A.; Diamond, P. T.; Austen, K. F.;
Oates, J. A.; Roberts, L. J., II J. Immunol. 1982, 129, 1627.
2. Matsuoka, T.; Hirata, M.; Tanaka, H.; Takahashi, Y.;
Murata, T., et al. Science 2000, 287, 2013.
3. Cheng, K.; Wu, T.-J.; Wu, K. K.; Sturino, C.; Metters, K.,
et al. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 6682.
4. Sturino, C. F.; Lachance, N.; Boyd, M.; Berthelette, C.;
Labelle, M., et al. J. Med. Chem., in press.
5. Evans, D. C.; Watt, A. P.; Nicoll-Griffith, D. A.; Baillie,
T. A. Chem. Res. Toxicol. 2004, 17, 3.
6. Nicoll-Griffith, D. A.; Falgueyret, J. P.; Silva, J. M.;
Morin, P. E.; Trimble, L., et al. Drug Metab. Dispos.
1999, 27, 403.
7. Chauret, N.; Nicoll-Griffith, D. A.; Friesen, R.; Li, C.;
Trimble, L.; Dube, D.; Fortin, R.; Girard, Y.; Yergey, J.
Drug Metab. Dispos. 1995, 23, 1325.
8. Rushmore, T. H.; Reider, P. J.; Slaughter, D.; Assang, C.;
Shou, M. Metab. Eng. 2000, 2, 115.
9. Skiles, G. L.; Adams, J. D., Jr.; Yost, G. S. Chem. Res.
Toxicol. 1989, 2, 254.
10. Fura, A.; Harper, T. W.; Zhang, H.; Fung, L.; Shyu, W.
C. J. Pharm. Biomed. Anal. 2003, 32, 513.
11. Sturino, C. F.; Lachance, N.; Boyd, M.; Berthelette, C.;
Labelle, M., et al. Bioorg. Med. Chem. Lett. 2006, 16,
3043.
12. Nicoll-Griffith, D.; Yergey, J.; Trimble, L.; Williams, H.;
Rasori, R.; Zamboni, R. Drug Metab. Dispos. 1992, 20,
383.
13. Juteau, H.; Gareau, Y.; Labelle, M. Tetrahedron Lett.
1997, 38, 1481.
14. Bailey, M. J.; Dickinson, R. G. Chem. Biol. Interact. 2003,
145, 117.
The formation of CYP3A4-mediated oxidative metabo-
lites on the 3-methylene indole core9,16 and formation of
the acyl glucuronide and acyl migration products14,17
raised concerns that reactive species may be formed as
a result of metabolism. A tritiated analog of MK-0524
was synthesized4 and tested for covalent protein modifi-
cation according to standard procedures in liver micro-
somes fortified with NADPH and in hepatocytes using a
semi-automated membrane harvester to trap precipitat-
ed proteins.5,18 Despite the fact that significant amounts
of 3-methylene indole hydroxylation metabolites were
formed in microsomal incubations and acyl glucuronide
and rearranged glucuronides were formed in hepato-
cytes (cf. Fig. 1), the levels of binding were below the
target threshold of <50 pmol-equiv/mg protein at 1 h
(Table 2). This could be due to substitution of the indole
core with electron-withdrawing fluorine and methyl sul-
fone moieties.4 No tritiated water was detected by
HPLC in the incubates indicating that the amount of
covalent binding was not underestimated due to loss
of the tritium label.
15. van Breeman, R. B.; Feneslau, C. C.; Dulik, D. M. Adv.
Med. Biol. 1986, 197, 423.
16. Kassahun, K.; Skordos, K.; McIntosh, I.; Slaughter, D.;
Doss, G. A., et al. Chem. Res. Toxicol. 2005, 18, 1427.
17. Spahn-Langguth, H.; Benet, L. Z. Drug Metab. Rev. 1992,
24, 5.
In rats, after a 10 mg/kg po dose of tritiated MK-0524,
covalent binding was measured in liver homogenate
and plasma at 24 h post-dose. The procedure involved
exhaustive extraction of the protein pellet formed by
acetonitrile precipitation of the plasma or liver homog-
enates followed by exhaustive extraction with metha-
nol-ether (3:1).19 Covalent binding was shown to be
low or not detectable (Table 2). According to these re-
sults, there is no reason to believe that reactive metabo-
lite intermediates should mediate adverse events via
covalent modification of proteins.
18. Day, S. H.; Mao, A.; White, R.; Schultz-Utermoehl, T.;
Miller, R.; Beconi, M. G. J. Pharmacol. Toxicol. Methods
2005, 52, 278.
19. Pohl, L. R.; Branchflower, R. V. Methods Enzymol. 1981,
77, 43.
20. Schwartz, M. S.; Desai, R. B.; Bi, S.; Miller, A. R.;
Matuszewski, B. K. J. Chromatogr., B 2006, 837, 116.
21. Dean, B. J.; Chang, S.; Xia, Y.; Karanam, B.; Franklin R.
B. In preparation.
22. Chang, S. W.; Reddy, V., Pereira, Dean, B. J.; Karanam,
B. V.; Franklin R. B. In preparation.