Synthesis of carbon-14 labelled midazolam

Article abstract of DOI:10.1002/jlcr.1620

Cytochrome P450 (CYP) enzymes are responsible for much of the phase I oxidative metabolism observed in vivo. Many important pharmaceutical compounds are metabolized by CYP. Co-administration of a drug with another agent can alter the efficacy or the toxic

Full text of DOI:10.1002/jlcr.1620

Research Article
Received 21 January 2009,
Revised 7 May 2009,
Accepted 18 May 2009
Published online 3 July 2009 in Wiley Interscience
(www.interscience.wiley.com) DOI: 10.1002/jlcr.1620
Synthesis of carbon-14 labelled midazolam
Ã
John A. Easter,a Richard C. Burrell,a Samuel J. Bonacorsi Jr,b and
Balu Balasubramanianb
Cytochrome P450 (CYP) enzymes are responsible for much of the phase I oxidative metabolism observed in vivo. Many
important pharmaceutical compounds are metabolized by CYP. Co-administration of a drug with another agent can
alter the efficacy or the toxicity, especially in cases where drug clearance depends primarily on the CYP metabolism.
Compounds that induce or inhibit the CYP activity are often used in drug–drug interaction studies. Midazolam is one such
compound that is routinely used in drug–drug interaction studies because it is a known substrate for CYP3A enzymes. The
synthesis of this important tool molecule has been documented, but unfortunately a detailed preparation of carbon-14-
labelled midazolam has not been reported in the literature. This paper describes a two-step synthesis leading to
[14C]midazolam. A total of 4.5mCi of [14C]midazolam was obtained having a specific activity of 120.1lCi/mg (39.12 mCi/mmol).
The radiochemical purity as determined by HPLC was 99.8% and the overall radiochemical yield was 9%.
Keywords: midazolam; carbon-14; isotope labelling; drug–drug interaction
carbon-14-labelled midazolam was never reported. This paper
describes the synthesis of carbon-14-labelled midazolam ([14C]-1).
Introduction
As part of drug-candidate screening and the drug discovery and
N
development process, investigators often conduct in vitro drug
metabolism studies to assess the potential for Cytochrome
P450 (CYP)-based drug interactions.1 These studies characterize
the metabolic pathway of the drug and the potential of other
drugs to modify the metabolism.1 Drug–drug interactions are
thought to be one of the most important factors in severe
adverse drug responses. Knowing this important fact, most
pharmaceutical companies now consider it mandatory to
minimize, as much as possible, the potential for drug–drug
interactions.2,3 Most drug–drug interactions occur when two
drugs share a common clearance pathway involving a drug-
metabolizing enzyme or a transporter.4 Once the metabolic
pathways of a drug candidate are understood, there is a need to
also assess the potential impact of inhibition of these pathways
by other drugs.
The largest group of enzymes known for the phase I oxidative
metabolism is the CYP family of enzymes. CYP enzymes are
located in various tissues throughout the body, with the
liver being the largest source. Two intestinal CYP isoforms
that account for 70% of the total intestinal CYP activity
include CYP3A4 and CYP3A5. Since 450% of CYP-metabolized
drugs are substrates for CYP3A, this enzyme is the source of
numerous drug–drug interactions.5 Several CYP3A probe
drugs have been identified to detect and quantify poten-
tial CYP3A-mediated drug–drug interactions.6 Midazolam,
with its known CYP3A4- and CYP3A5-mediated metabolism,
has been an important tool molecule for predicting drug–drug
interactions.6
C
N
14
N
Cl
F
[
14C]-1
Discussion and results
The syntheses of unlabelled and stable isotope-labelled midazolam
have been detailed in the literature.7,8 These procedures were
adopted for use in the carbon-14 synthesis using labelled ethyl
acetimidate hydrochloride as shown in Scheme 1. Compound 2,
which was the starting material for our carbon-14-labelled synthesis,
was prepared in seven steps from (2-amino-5-chlorophenyl)(2-
fluorophenyl)methanone following a literature procedure.13,14
solution of the amine 2 in ethanol/tetrahydrofuran at 01C
A
aDepartment of Chemical Synthesis, Bristol-Myers Squibb Research and
Development, 5 Research Parkway, Mail Stop 4BC-313, Wallingford, CT 06492,
USA
bDepartment of Chemical Synthesis, Bristol-Myers Squibb Research and
Development, Route 206 and Province Line Road, Princeton, NJ 08540, USA
The syntheses of unlabelled, stable-labelled, and tritium-labelled
midazolam have been detailed in the literature.7–9 The use
of carbon-14-labelled midazolam in metabolism studies has
also been reported.10–12 Surprisingly, a detailed synthesis of
*Correspondence to: John A. Easter, Department of Chemical Synthesis, Bristol-
Myers Squibb Research and Development, 5 Research Parkway, Mail Stop 4BC-
313, Wallingford, CT 06492, USA.
E-mail: john.easter@bms.com
J. Label Compd. Radiopharm 2009, 52 419–421
Copyright r 2009 John Wiley & Sons, Ltd.

J. A. Easter et al.
H
N
N
C
HCl
OEt
N
C
14
H2N
14
C
N
14
N
H
MnO2
Toluene
N
N
reflux
EtOH/THF
0 oC
Cl
N
N
Cl
Cl
F
F
F
14C]-3
2
[
14C]-1
[
14
Scheme 1. Synthesis of carbon-14-labelled midazolam ([14C]-1).
was treated with 50 mCi (specific activity 56 mCi/mmol) of in ethanol (2 mL) and THF (1 mL) at 01C was added [14C-imino]ethyl
carbon-14-labelled ethyl acetimidate hydrochloride, followed by an acetimidate hydrochloride (110 mg, 0.890 mmol, 50 mCi), followed
equivalent portion of unlabelled ethyl acetimidate hydrochloride to by unlabelled ethyl acetimidate hydrochloride (110 mg,
give imidazoline [14C]-3 after 3 h at room temperature (RT). 0.890 mmol).7,8 The reaction was stirred for 30 min at 01C, then
Conversion of imidazoline into imidazole was carried out using warmed to RT and stirred for an additional 3 h. At that time the
MnO2 as the oxidizing agent. This conversion was very sensitive to reaction was concentrated under reduced pressure. The residue was
the water content of the MnO2 used. Sigma-Aldrich supplies several dissolved into 15 mL of dichloromethane and extracted with 20 mL
grades of manganese oxide. When high-purity, Reagent Plus of 5% NaHCO3. The layers were separated and the aqueous was
(499%), or activated (85%) grades were used, little to no conversion extracted with 15 mL of dichloromethane. The combined dichlor-
to the desired midazolam was observed. However, the rate of extent omethane extracts were washed with saturated brine, dried over
of conversion increased when the activated (85%) MnO2 was dried anhydrous sodium sulfate, and concentrated under reduced
under vacuum at 1251C for 24 h. These results indicate that the pressure to yield crude [14C]-3 (298 mg, 0.909 mmol, 102% yield).
water content of MnO2 used dramatically affects the reaction.
A solution of [14C]-3 in dry toluene was treated with dried
The radiochemical purity as determined by HPLC was 65%.
MnO2 and refluxed for 3 h. After workup and purification by (E)-8-chloro-6-(2-fluorophenyl)-1-methyl-4H-benzo[f][14C]i-
midio[1,5-a][1,4]diazepine, [14C]midazolam ([14C]-1)
reversed-phase HPLC, 4.5 mCi of [14C]-1 was obtained having a
specific activity of 120.1 mCi/mg (39.12mCi/mmol). The radio-
chemical purity as determined by HPLC was 99.8% and the overall
radiochemical yield from labelled acetimidate hydrochloride was
9%. Radiolabelled [14C]-1 was relatively unstable in the solid at
high specific activity. At 120 mCi/mg (39.12mCi/mmol), [14C]-1
decomposed at a rate of 6.8% per month into three primary
impurities. Owing to the limited quantity synthesized, stabiliza-
tion of radiolytic decomposition was not explored.
To a suspension of crude [14C]-3 (298 mg, 0.909 mmol) in toluene
(15 mL) was added activated manganese(IV) oxide (1500 mg,
17.25 mmol) (dried under vacuum at 1251C for 24 h). The mixture
was heated at 1171C for 3 h. After cooling to RT, the reaction
mixture was filtered, rinsed with toluene (5 mL) and DCM (5 mL). The
filtrate was concentrated under reduced pressure to afford crude
carbon-14-labelled midazolam. The crude material was dissolved in
5 mL of acetonitrile. The solution was purified by semi-preparative
HPLC in 16 Â 0.3 mL injections on a Pack Pro C18 column
(19 Â 150 mm, Solvent A = 95:5 v/v 0.01 M ammonium acetate:
acetonitrile; B = 5:95 v/v 0.01 M ammonium acetate:acetonitrile,
gradient: 100% A 0–5 min, 100–0% A 5–25 min, 0–100% A
25–30 min, flow: 15 mL/min, wavelength: 215 nm). The pooled
fractions of [14C]-1 were partially concentrated under reduced
pressure to remove the acetonitrile. The remaining aqueous was
extracted with 2 Â 20 mL of DCM. The combined DCM extracts
were washed with saturated brine, dried over anhydrous sodium
sulfate, and concentrated under reduced pressure. The residue was
dried under vacuum at RT to yield 4.5 mCi of [14C]-1, having a
specific activity of 120.1 mCi/mg (39.12 mCi/mmol). The radio-
chemical purity as determined by HPLC was 99.8% and the overall
radiochemical yield from labelled acetimidate hydrochloride was 9%.
NMR (CDCl3) d 2.95 ppm (S, 3), 4.08 ppm (dd, 1, J= 12.84 Hz),
5.18 ppm (dd, 1, J= 13.6 Hz), 7.0–7.69 (m, 8, Aromatic-H). NMR
spectrum is consistent with the published spectrum of unlabelled
midazolam.7
Experimental
All reagents were obtained from Aldrich Chemical Company and
used without further purification. Carbon-14-labelled ethyl acet-
imidate hydrochloride was obtained from ViTrax Company. All
experimental procedures were optimized using unlabelled materi-
als. All glassware was dried and purged with nitrogen or argon
before use. All reactions were monitored by HPLC using the
following conditions: YMC Pack Pro C18 column s-3 mm
(4.6 Â 150 mm). Solvent A = water with 0.05% TFA; B = acetonitrile
with 0.05% TFA. Gradient: 100% A 0–5 min, 100–0% A 5–20 min, 0%
A 20–25 min, 0–100% A 25–30 min. Flow 1 mL/min, wavelength
215 nm. Chemical and radiochemical purity was determined by
HPLC using a Varian ProStar system (Model 210) equipped with a
Varian ProStar PDA (Model 330) and a Beta-Ram Detector (IN/US
Systems, Inc.). Specific activity was determined by gravimetric
analysis using liquid scintillation counting (Wallac Model 1409).
(E)-8-chloro-6-(2-fluorophenyl)-1-methyl-3a,4-dihydro-3H-
benzo[f-14C]imidazo[1,5-a]diazepine ([14C]-3)
Conclusions
To a solution of (E)-(7-chloro-5-(2-fluorophenyl)-2,3-dihydro-1H- Carbon-14-labelled midazolam was synthesized in two steps
benzo[e][1,4]diazepin-2-yl)methanamine11,12 (2, 270 mg, 0.890 mmol) from 50 mCi of radiolabelled precursor to afford 4.5 mCi (9%) of
www.jlcr.org
Copyright r 2009 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2009, 52 419–421

J. A. Easter et al.
the product with a specific activity of 120.1 mCi/mg and good
radiochemical purity. The radiolabelled product was observed to
be relatively unstable at high specific activity. At 120 mCi/mg
(39.12 mCi/mmol), [14C]-labelled midazolam decomposed at a
rate of 6.8% per month. The importance of using activated and
dried manganese oxide for the oxidation step in the preparation
of [14C]-labelled midazolam should also be noted.
[3] D. G. Bailey, J. R. Bend, J. M. Arnold, L. T. Tran, J. D. Spence, Clin.
Pharmacol. Ther. 1994, 60, 301.
[4] A. D. Rodrigues, Drug– drug Interactions, Marcel Dekker, New York,
NY, 2002.
[5] S. R. Penzak, K. H. Busse, S. M. Robertson, E. Formentini,
R. M. Alfaro, R. T. Davey Jr., J. Clin. Pharmacol. 2008, 48, 671.
[6] D. S. Streetman, J. S. Bertino, A. N. Nafziger, Pharmacogenetics
2000, 10, 187.
[7] A. Walser, L. E. Benjamin Sr., T. Flynn, C. Mason, R. Schwartz,
R. I. Fryer, J. Org. Chem. 1978, 43, 936.
Acknowledgement
[8] Y. Zhang, P. W. K. Woo, J. Hartman, N. Colbry, Y. Huang,
C. C. Huang, Tetrahedron Lett. 2005, 46, 2087.
[9] J. Allen, D. M. Brasseur, B. DeBruim, M. Denoux, S. Perard,
N. Philippe, S. N. Roy, J. Labelled Compd. Radiopharm. 2007, 50,
342.
[10] G. Lappin, W. Kuhnz, R. Jochemsem, J. Kneer, A. Chaudhary,
B. Oosterhuis, W. J. Drijfhout, M. Rowland, R. C. Garner, Clin.
Pharmacol. Ther. 2006, 80, 203.
The authors would like to thank the Bristol-Myers Squibb
Discovery Chemical Synthesis group for the synthesis of the
starting material 2.
[11] G. K. Woo, T. H. Williams, S. J. Kolis, D. Warinsky, G. J. Sasso,
M. A. Schwartz, Xenobiotica 1981, 11, 373.
[12] P. Heizmann, W. H. Ziegler, Arzneim.-Forsch. 1981, 31, 2220.
References
[1] R. Yuan, S. Madani, X. X. Wei, K. Reynolds, S. M. Huang, Drug Metab. [13] L. H. Sternbach, R. I. Fryer, W. Metlesics, E. Reeder, G. Sach,
Dispos. 2002, 30, 1311.
[2] S. V. Mandlekar, A. V. Rose, G. Cornelius, B. Sleczka, C. Caporuscio,
J. Wang, P. H. Marathe, Xenobiotica 2007, 37, 923.
G. Saucy, A. Stempel, J. Org. Chem. 1963, 27, 3788.
[14] A. Walser, L. E. Benjamin, T. Flynn, C. Mason, R. Schwartz, R. I. Fryer,
J. Org. Chem. 1978, 43, 936.
J. Label Compd. Radiopharm 2009, 52 419–421
Copyright r 2009 John Wiley & Sons, Ltd.
www.jlcr.org