A modern approach to the synthesis of 2-(4-chlorophenyl)[2- 14C]thiazol-4-ylacetic acid ([14C] fenclozic acid) and its acyl glucuronide metabolite

Article abstract of DOI:10.1002/jlcr.2985

An updated approach to the 1960s synthesis of [¹⁴C] fenclozic acid from labelled potassium cyanide is presented. By employing modern synthetic methodology and purification techniques, many of the inherent hazards in the original synthesis are avoided or significantly reduced. The concomitant labelled stereoselective synthesis of the key acyl glucuronide metabolite (the 1-β-O-acyl glucuronide) is also described. [¹⁴C]Fenclozic acid and its key acyl glucuronide metabolite, the 1-β-O-acyl glucuronide were prepared from labelled potassium cyanide. [¹⁴C]Fenclozic acid had previously been prepared in the 1960s, however in this paper we describe a shorter synthesis that avoids or reduces some of the inherent hazards present in the original synthesis. Copyright

Full text of DOI:10.1002/jlcr.2985

Research Article
Received 15 June 2012,
Revised 23 August 2012,
Accepted 17 October 2012
Published online 11 January 2013 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.2985
A modern approach to the synthesis of
14
2-(4-chlorophenyl)[2- C]thiazol-4-ylacetic
acid ([14C] fenclozic acid) and its acyl
glucuronide metabolite
*
David A. Killick and Nick Bushby
An updated approach to the 1960s synthesis of [14C] fenclozic acid from labelled potassium cyanide is presented. By employing
modern synthetic methodology and purification techniques, many of the inherent hazards in the original synthesis are
avoided or significantly reduced. The concomitant labelled stereoselective synthesis of the key acyl glucuronide metabolite
(the 1-b-O-acyl glucuronide) is also described.
Keywords: fenclozic acid; acyl glucuronide; cyanide
by filtration from the reaction mixture and used without further
purification in the next step.
Coupling of the cyanide 4 to 4-bromochlorobenzene was
achieved via a classical Rosenmund–von Braun reaction, heating
Introduction
Fenclozic acid (ICI 54,450) (Figure 1) was one of a series of substi-
tuted phenylthiazoylacetic acids synthesised by ICI in the 1960s
the two components in dimethyl formamide at 150–155 ꢀC for
7 h in a sealed tube. In addition to the harsh conditions used,
the reaction was quite labour intensive – it was noted that higher
yields were obtained if the reaction tube was removed every
20 min and thoroughly shaken.
for evaluation in the non-steroidal treatment of rheumatoid
arthritis. Early preclinical work showed the toxicology profile to
be benign, with safety findings in multiple species indicating
no evidence of overt liver injury. It was only when the compound
progressed to man, where a high incidence of jaundice was
recorded, that development was stopped.1
The thioamide 6 was then prepared by passing dry hydrogen
sulfide gas through a stirred solution of the chlorobenzonitrile 5
and triethylamine in absolute alcohol at 70 ꢀC for 2 h before
purifying by column chromatography. The Hantzsch thiazole
synthesis was used to prepare the heterocycle 7, reacting the
thioamide 6 with 1,3-dichloro-2-propanone. The final steps
involved displacement of the aliphatic chlorine with unlabelled
potassium cyanide in dimethyl sulfoxide to afford the nitrile 8
that was hydrolysed, without further purification, in refluxing
hydrochloric acid to give the carboxylic acid [14C]-1. Purification
by column chromatography afforded [14C] fenclozic acid [14C]-1
with a radiochemical purity of 99.7% and specific activity of
12.6 mCi/mg (3.19 mCi/mmol), constituting an overall radioche-
mical yield of 22% from potassium [14C] cyanide 3.
Drugs containing carboxylic acid moieties can be metabolised
in vivo to form 1-b-O-acyl glucuronides following enzyme
mediated reactions with uridine diphosphate-glucuronic acid.
Because acyl glucuronides have the ability to undergo reactions
resulting in covalent binding to endogenous proteins with the
potential to form immunogenic, and therefore toxic haptens,
they are of potential concern when developing a new drug, as
highlighted by recent Food and Drug Administration guidelines.2,3
A key question from a current perspective is whether modern
Drug Metabolism and Pharmacokinetics (DMPK) and Molecular
Toxicology screens would have flagged up fenclozic acid 1 as a risk
and stopped development earlier. This afforded the opportunity to
review the original synthesis4 and apply modern synthetic metho-
dology and techniques to resynthesise [14C] fenclozic acid [14C]-1
(Scheme 2). The subsequent synthesis of the [14C] acyl glucuronide
metabolite [14C]-2 (Figure 2) that has not been reported previously
is also described (Scheme 3).
A positive aspect to this synthetic approach was the minimal
purification needed for many of the steps; many of the crude
intermediates could be used directly in subsequent steps
Results and discussion
The first step of the original synthetic approach (Scheme 1) to
radiolabelled fenclozic acid [14C]-1 involved inorganic manipula-
tion of commercially sourced potassium [14C] cyanide 3 with
copper sulfate and sodium metabisulfite to prepare labelled
DMPK IM, AstraZeneca UK Ltd, Mereside, Alderley Park, Macclesfield, Cheshire
SK10 4TG, UK
*Correspondence to: Nick Bushby, DMPK IM, AstraZeneca UK Ltd., Mereside,
Alderley Park, Macclesfield, Cheshire SK10 4TG, UK.
copper(I) cyanide 4. The solid product could be readily isolated E-mail: Nick.Bushby@astrazeneca.com
J. Label Compd. Radiopharm 2013, 56 17–21
Copyright © 2013 John Wiley & Sons, Ltd.

D. A. Killick and N. Bushby
without further purification. Although recognising that the work
was conducted using the best available methods at the time, a
number of areas for potential improvement were quickly identi-
fied. The manipulation of inorganic potassium cyanide 3 to give
copper cyanide 4 could be avoided by employing a suitable
direct coupling method. In addition, some of the reaction condi-
tions could be considered harsh, with high temperatures and
strong acids and bases required. This was an important consid-
eration as we were proposing to prepare fenclozic acid [14C]-1
at a much higher specific activity, which would exacerbate any
stability issues. It was also hoped that the use of gaseous hydro-
gen sulfide could be avoided, and that the recurring use of
potassium cyanide could also be avoided. In addition, many of
the workup and purification procedures could be updated to
reflect modern working practices, avoiding, for example, the use
of benzene/chloroform mixtures for column chromatography as
described in some of the original experimental procedures.
Having reviewed the original synthesis, it was possible to
simplify the reaction sequence to just four steps (Scheme 2),
although still beginning from potassium [14C] cyanide 3 and retain-
ing the same site of labelling.
Figure 1. Fenclozic acid.
By employing suitable palladium coupling conditions5 (Tris
(dibenzylideneacetone) dipalladium(0)/1,10-bis(diphenylphosphino)
Figure 2. Fenclozic acid glucuronide.
Scheme 1. Original synthetic route to [14C] fenclozic acid [14C]-1.
Scheme 2. Revised synthetic route to [14C] fenclozic acid [14C]-1.
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 17–21

D. A. Killick and N. Bushby
ferrocene), the labelled potassium cyanide 3 could be directly and N-methylmorpholine in acetonitrile. After stirring at ambient
coupled to 4-chloroiodobenzene, exploiting the greater reactivity temperature for 24 h under nitrogen, the reaction mixture was
of the iodo substituent over the chloro to give the radiolabelled neutralised with acid resin and the crude product 10 purified by
chlorobenzonitrile 5. The modest temperature (40 ꢀC) required silica cartridge chromatography.
for the reaction contrasts sharply to the elevated temperatures
needed for the Rosenmund–von Braun reaction used in the challenging. After several days of recharging the Pd/C catalyst,
original synthesis. the reaction appeared to have stalled at 80% conversion. The
Removal of the benzyl protecting group from 10 proved more
The thioamide 6 was then accessed in a microwave mediated mixture was worked up and the crude product purified by prepara-
reaction,6 avoiding the use of gaseous hydrogen sulfide directly, tive HPLC to provide the acyl glucuronide metabolite 2 with a
by using solid ammonium sulfide. After heating for 30 min at radiochemical purity >98% at a specific activity of 12.9 mCi/mg.
80 ꢀC, the product was isolated as a solid with no further purifica- NMR spectroscopy (Figure 3) confirmed the excellent selectivity
tion necessary.
of the synthetic method towards the required b-anomer, seen at
The remaining stages were telescoped into a single pot proce- d 5.7 (3J = 7.92 Hz). There was no evidence of the presence of the
dure, again using a microwave approach, which allowed for some a-anomer (expected at approx. d 6.1 with a coupling constant of
extremely quick reaction times. In a change to the previously 3J = 3.7Hz).10
reported method,7 the thioamide 6 and methylchloroacetoacetate
were mixed with methanol and heated in the microwave for
30 min at 100 ꢀC to form the methyl ester of the parent compound.
2M Sodium hydroxide was added and the mixture reheated in the
microwave for 15 min further at 50ꢀC to hydrolyse the ester. After
acidification, the crude product was collected by filtration and
washed with water. The material was purified by preparative HPLC
to afford [14C] fenclozic acid [14C]-1 in 24% overall radiochemical
yield, with a radiochemical purity >99%, at a specific activity of
240 mCi/mg (61.8mCi/mmol). Although this yield was comparable
with that obtained through the original route, we believe that it
could be improved further by using freshly prepared potassium
[14C] cyanide 3 rather than the aged batch we had available
for this synthesis.
Experimental
1H-NMR spectra were recorded on either a Bruker (500MHz) or a Bruker
(600 MHz) in the deuterated solvent stated. Chemical shifts (d) in ppm are
quoted relative to DMSO-d6 (d 2.50) or acetone-d6 (d 2.09). Liquid chroma-
tography-mass spectrometry (LC-MS) data was obtained on a Waters Alli-
ance LC (Waters Corporation, MA, USA) with Waters ZQ mass detector. Ana-
lytical TLC was carried out on Merck 5785 Kieselgel 60F254 fluorescent plates
(Merck KGaA, Darmstadt, Germany). Microwave reactions were carried out
in a CEM Explorer (CEM Corporation, NC, USA), heating at 100W. Column
chromatography was performed using ISCO (Teledyne Isco, NE, USA) Com-
panions with pre-packed silica gel cartridges. The target compounds were
purified by preparative HPLC (Waters XBridge Prep C18 OBD column,
5 mm silica, 19mm diameter, 100 mm length), using decreasingly polar mix-
tures of water (containing 0.1% formic acid) and acetonitrile as eluents. Spe-
cific activities were measured gravimetrically with a Packard TriCarb 1900CA
Liquid Scintillation Analyser (Packard Instrument Company Inc., IL, USA).
Radiochemical purities were determined on an Agilent series 1100 HPLC
system coupled to a b-Ram Flow Scintillation Analyser.
With the successful completion of the radiolabelled drug
[
14C]-1, we turned our attention to the synthesis of a key meta-
bolite 2 (Scheme 3). Recent AstraZeneca collaborations with
academia have provided a viable approach to the routine synth-
esis of acyl glucuronide metabolites. Bowkett et al.8 have shown
that the synthesis of 1-b-O-acyl glucuronides can be achieved in
high yield and excellent b-selectivity through the selective acyla-
tion of allyl or benzyl D-glucuronate, followed by deprotection.
Benzyl protection was preferred for our work, as it avoided
possible Pd residue contamination noted in the literature9 when
removing allyl protecting groups.
4-chlorobenzo[14C]nitrile (5)
Tris(dibenzylideneacetone)dipalladium(0) (15.0 mg, 0.02 mmol) and 1,10-
bis(diphenylphosphino)ferrocene (30.4 mg, 0.06 mmol) were added to
1-chloro-4-iodobenzene (665 mg, 2.79 mmol) and potassium [14C] cya-
nide (164 mCi, 875 mCi/mg, 187 mg, 2.79 mmol) in acetone (1 mL) under
[14C]-1 was coupled to benzyl D-glucuronate 9 using O-(7-azaben-
zotriazol-1-yl)-N,N,N0,N0-tetramethyluronium hexafluorophosphate nitrogen. The resulting mixture was stirred at 40 ꢀC for 4.5 h.
Scheme 3. Synthetic route to [14C] fenclozic acid glucuronide [14C]-2.
J. Label Compd. Radiopharm 2013, 56 17–21
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

D. A. Killick and N. Bushby
Figure 3. 1H-NMR Spectra of [14C] fenclozic acid glucuronide (acetone-d6), showing the anomeric proton signal at d 5.7 (3J = 7.92 Hz).
3
The reaction mixture was evaporated to dryness and redissolved in (500 MHz, DMSO-d6) d 7.92 (2H, d, J = 8.22 Hz), 7.54 (3H, m), 3.78 (2H, s)
ethyl acetate (30 mL) and washed with water (30 mL). The organic layer
was dried over MgSO4, filtered and the filtrate was evaporated.
LCMS m/z (ESI+) (M + H)+ = 256.
The crude product was purified by flash silica chromatography, elution
gradient 0–30% EtOAc in isohexane. Pure fractions were evaporated to
dryness to afford 4-chlorobenzo [14C]nitrile (261 mg, 67%) as a yellow solid.
(2S,3S,4S,5R,6S)-benzyl 6-(2-(2-(4-chlorophenyl)[2-14C]
thiazol-4-yl)acetoxy)-3,4,5-trihydroxytetrahydro-2H-pyran-
2-carboxylate (10)
4-chlorobenzothio[14C]amide (6)
N-Methylmorpholine (0.173mL, 1.58 mmol) was added dropwise over a
period of 10 s to a stirring suspension of unlabelled fenclozic acid
(180 mg, 0.71 mmol), [14C] fenclozic acid (4.8mCi 20mg, 0.08mmol)
(2S,3S,4S,5R,6R)-benzyl 3,4,5,6-tetrahydroxytetrahydro-2H-pyran-2-carboxy-
late (224mg, 0.79mmol) and O-(7-azabenzotriazol-1-yl)-N,N,N0,N0-tetra-
methyluronium hexafluorophosphate (300mg, 0.79mmol) in acetonitrile
(10 mL) at 22 ꢀC. The resulting pale yellow solution was stirred under nitro-
gen for 24 h at room temperature.
4-Chlorobenzo[14C]nitrile (201 mg, 1.44 mmol) and ammonium sulfide
(0.983 mL, 7.21 mmol) were suspended in methanol (4 mL) and sealed
into a microwave tube. The reaction was heated to 80 ꢀC for 30 min in
the microwave reactor and cooled at room temperature. The reaction
mixture was diluted with water (8 mL) and the precipitate was collected
by filtration, washed with water (10 mL) and dried under vacuum to
afford 4-chlorobenzothio[14C]amide (179 mg) as a yellow solid, which
was used without further purification. LCMS m/z (ESIꢁ) (M ꢁ H)ꢁ = 172.
The reaction mixture was acidified by the addition of Amberlyst A-15
(H+, 2 equiv., 1.53mmol) and filtered. The filtrate was concentrated under
reduced pressure to leave a pale yellow solid that was dissolved in dichlor-
omethane:propan-2-ol (6:1.4, v:v) and purified by flash silica chromatogra-
phy, elution gradient 5–10% propan-2-ol in dichloromethane. Fractions
containing product were combined and concentrated under reduced
pressure to give a white solid. The material was further dried at 40ꢀC under
reduced pressure to leave (2S,3S,4S,5R,6S)-benzyl 6-(2-(2-(4-chlorophenyl)
[2-14C]thiazol-4-yl)acetoxy)-3,4,5-trihydroxy tetrahydro-2H-pyran-2-carbox-
2-(2-(4-chlorophenyl)[2-14C]thiazol-4-yl)acetic acid (Fenclozic
acid, [14C]-1)
4-Chlorobenzothio[14C]amide (179 mg, 1.03 mmol) and methyl 4-chloroa-
cetoacetate (121 ml, 1.03 mmol) were mixed with methanol (38 ml) and
sealed into a microwave tube. The reaction was heated to 100 ꢀC for
30 min in the microwave reactor and cooled at room temperature.
Sodium hydroxide (2M, aqueous) (1902 ml, 3.8 mmol) was added and
the mixture was heated in the microwave reactor to 50 ꢀC for 15 min.
The reaction mixture was acidified to pH 5–6 with 2M hydrochloric
acid. The precipitate was collected by filtration, washed with water
(2 mL) and dried under vacuum. The crude product was then purified
by preparative HPLC; fractions containing the desired compound were
ylate (2.70mCi, 231mg, 56%) as
a
white solid. LCMS m/z (ESI+)
(M+ H)+ = 520.
(2S,3S,4S,5R,6S)-6-(2-(2-(4-chlorophenyl)[2-14C]thiazol-4-yl)
acetoxy)-3,4,5-trihydroxy tetrahydro-2H-pyran-2-carboxylic
acid (Fenclozic acid acyl glucuronide, [14C]-2)
evaporated to dryness before drying further in a desiccator at 40 ꢀC to (2S,3S,4S,5R,6S)-benzyl 6-(2-(2-(4-chlorophenyl)[2-14C]thiazol-4-yl)acetoxy)-
afford 2-(2-(4-chlorophenyl)[2-14C]thiazol-4-yl)acetic acid (30 mCi, 3,4,5-trihydroxy tetrahydro-2H-pyran-2-carboxylate (2.70 mCi, 231 mg,
125 mg) as a white solid in an overall radiochemical yield of 24% from 0.44mmol) was dissolved in propan-2-ol (10mL) and tetrahydrofuran
potassium [14C]cyanide. Radiochemical purity >99% by HPLC. 1H NMR (5mL). To this solution, palladium on carbon 10% (47.1mg, 0.04 mmol)
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 17–21

D. A. Killick and N. Bushby
was added and the mixture was stirred at room temperature under
an atmosphere of hydrogen gas for 92h, with regular recharging of
the catalyst.
Acknowledgements
The authors wish to thank Eva Lenz and Howard Beeley for
performing the NMR analyses. Special thanks are extended to
Chris Page, Dr John Harding and Dr Ian Wilson.
After filtration and evaporation of the filtrate, the crude product was
purified by preparative HPLC; fractions containing the desired compound
were evaporated to dryness. Dichloromethane (1–2mL) was added to
the product in a vial and the mixture was evaporated to dryness. This
was repeated three times to give (2S,3S,4S,5R,6S)-6-(2-(2-(4-chlorophe-
nyl)[2-14C]thiazol-4-yl)acetoxy)-3,4,5-trihydroxytetra hydro-2H-pyran-2-
carboxylic acid (0.66 mCi, 51.3 mg, 27%) as an off white solid with a
radiochemical purity >98% by HPLC. 1H-NMR (600 MHz, acetone-d6)
d 8.03 (d 2H 3J = 8.22 Hz), 7.59 (s 1H), 7.56 (d 2H 3J = 8.22 Hz), 5.7 (d 1H
3J = 7.92 Hz), 4.07 (d 1H 3J = 9.39 Hz), 4.03 (s 2H), 3.73 (t 1H 3J = 9.39 Hz),
Conflict of Interest
The authors did not report any conflict of interest.
3.64 (t 1H 3J = 8.80 Hz), 3.51 (t 1H 3J = 8.51 Hz). LCMS m/z (ESI+) References
(M + H)+ = 430.
[1] S. Alcock, Proceedings of the European society for the study of drug
toxicity, 1970, 12.
[2] FDA. Guidance for industry. Safety testing of drug metabolites. FDA.
Rockville, MD, 2008.
Conclusion
[3] S. L. Regan, J. L. Maggs, T. G. Hammond, C. Lambert, D. P. Williams,
B. K. Park, Biopharm. Drug Dispos. 2010, 31, 367–395.
[4] D. F. White, J. Burns, J. Label. Compd. Radiopharm. 1976, 13, 393–400.
[5] K. Takagi, K. Sasaki, Y. Sakakibara, Bull. Chem. Soc. Jpn. 1991, 64, 1118–1121.
[6] M. C. Bagley, K. Chapaneri, C. Glover, E. A. Merritt, Synlett 2004, 2615–2617.
[7] C. S. Rooney, W. C. Randall, K. B. Streeter, C. Ziegler, E. J. Cragoe, H.
Schwam, S. R. Michelson, H. W. Williams, E. Eichler, D. E. Duggan,
E. H. Ulm, R. M. Noll, J. Med. Chem. 1983, 26, 700–714.
[8] E. R. Bowkett, J. R. Harding, J. L. Maggs, B. K. Park, J. A. Perrie, A. V.
Stachulski, Tetrahedron 2007, 63, 7596–7605.
[9] J. A. Perrie, J. R. Harding, D. W. Holt, A. Johnston, P. Meath, A. V. Stachulski,
Org. Lett. 2005, 7, 2591–2594.
[10] O. Corcoran, R. W. Mortensen, S. H. Hansen, J. Troke, J. K. Nicholson,
Chem. Res. Toxicol. 2001, 14, 1363–1370.
[14C] Fenclozic acid [14C]-1 has been successfully prepared using
modern synthetic methodology, reducing the risks associated
with the original synthesis. The radiolabelled acyl glucuronide
[
14C]-2 has also been chemically prepared for the first time, with
excellent b-selectivity.
In vitro metabolism studies using the labelled fenclozic acid
[
14C]-1 have shown extensive covalent binding, whereas studies
with the unlabelled acyl glucuronide (2) have demonstrated
rapid transacylation kinetics.11
Future work will involve more detailed metabolism studies
and the use of the radiolabelled acyl glucuronide [14C]-2 for
further covalent binding studies.
[11] E. Karlsson, et al. In preparation
J. Label Compd. Radiopharm 2013, 56 17–21
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org