An efficient synthesis of a potent PPARpan agonist

Article abstract of DOI:10.1021/jo061295n

An efficient synthesis of 2-{4-[({4-{[4-(4-methoxyphenyl)-piperazin-1-yl] methyl}-2-[4-(trifluoromethyl)phenyl]-1,3-thiazol-5-yl}methyl)thio]phenoxy} -2-methylpropanoic acid (1), a potent PPARpan agonist, is described. The seven-step synthesis, which affo

Full text of DOI:10.1021/jo061295n

PPARpan agonist.7 The original synthesis of 1 (Scheme 1), not
designed for atom efficiency or waste minimization, involved
14 linear steps and was not amenable for large-scale preparation
due to many safety and scalability issues within the synthesis.7
To provide sufficient quantities of 1 for preclinical and clinical
studies, an efficient, practical, and scalable route was required.
As shown in the retrosynthetic analysis depicted in Scheme 2,
we envisioned that a regioselective carbon-sulfur bond forma-
tion could be achieved directly from the diol 3 and the
thiophenol 2 under acidic conditions.8 We predicted that a stable
carbocation would be generated preferably at the C-5 hy-
droxymethyl of the thiazole ring in diol 3.9 Thus, an SN1
displacement should be more favored at the C-5 hydroxymethyl
group,10 leading to a very efficient synthesis of 1.
An Efficient Synthesis of a Potent PPARpan
Agonist
Jiasheng Guo,* Greg A. Erickson, Russ N. Fitzgerald,
Richard T. Matsuoka, Stephen W. Rafferty,†
Matthew J. Sharp, Barry R. Sickles, and James C. Wisowaty
Synthetic Chemistry, Chemical DeVelopment, GlaxoSmithKline,
Research Triangle Park, North Carolina 27709
jiasheng.x.guo@gsk.com
ReceiVed June 22, 2006
In this paper, we report an efficient and practical synthesis
of 1 in 7 steps with 30% overall yield.
The synthesis of 1 began with condensation of commercially
available diethyl 2-chloro-3-oxosuccinate (4) with 4-(trifluo-
romethyl)thiobenzamide (5) in ethanol, which afforded the
diester 6 in 83% yield (Scheme 3).
Reduction of 6 with LAH yielded diol 3. To avoid undesired
mono- and bisdesfluoro byproducts (7 and 8), which carried
through to 1 as major impurities, the reaction temperature was
maintained below -10 °C. Thus the reduction was carried out
at -10 to -15 °C in THF followed by acid workup with
aqueous sulfuric acid, affording 3 in 83% yield without the
formation of the desfluoro side products.
We next turned our attention to the regioselective carbon-
sulfur bond formation at the C-5 hydroxymethyl. We explored
two approaches to form the carbon-sulfur bond: synthesis of
the thiophenol 2 and in situ coupling with 3; direct coupling of
3 with commercially available 4-hydroxythiophenol. To pursue
An efficient synthesis of 2-{4-[({4-{[4-(4-methoxyphenyl)-
piperazin-1-yl]methyl}-2-[4-(trifluoromethyl)phenyl]-1,3-
thiazol-5-yl}methyl)thio]phenoxy}-2-methylpropanoic acid
(1), a potent PPARpan agonist, is described. The seven-step
synthesis, which afforded 1 in 30% overall yield, includes a
highly regioselective carbon-sulfur bond formation via
coupling of a bishydroxymethylthiazole (3) with 4-hydrox-
ythiophenol, displacement of the remaining alcohol through
a three-step telescoped sequence involving an efficient
cleavage of an aryl mesylate, and an efficient and practical
method of introducing an isobutyric acid fragment.
(7) Banker, P.; Cadilla, R.; Lambert, M. H., III; Rafferty, S. W.;
Sternbach, D. D.; Sznaidman, M. L. PCT Int. Appl. WO2002059098.
(8) For examples of acid-promoted displacement of activated alcohols,
such as benzylic alcohols, by thiols, see: (a) Micha-Screttas, M.; Screttas,
C. G. J. Org. Chem. 1977, 42, 1462. (b) Guindon, Y.; Frenette, R.; Fortin,
R.; Rokach, J. J. Org. Chem. 1983, 48, 1357. (c) Manabe, K.; Limura, S.;
Sun, X.-M.; Kobayashi, S. J. Am. Chem. Soc. 2002, 124, 11971. (d)
Breitschuh, R.; Seebach, D. Synthesis 1992, 83. (e) Stewart, A. S. J.; Drey,
C. N. C. J. Chem. Soc., Perkin Trans. 1 1990, 1753. (f) Bouwman, E.;
Driessen, W. L. Synth. Commun. 1988, 18, 1581.
(9) Under acidic conditions, 3 may form carbocations at two positions,
C-5 hydroxymethyl carbon (carbocation A) or C-4 hydroxymethyl carbon
(carbocation B). The carbocation A has greater stability because of the
increased delocalization due to resonance (7 canonical forms) within the
thiazole and the benzene rings, while the carbocation B (3 canonical forms)
is stabilized only by delocalization with a 4,5-double bond and the mercapto
atom.
Peroxisome proliferator-activated receptors (PPARs) are
ligand-sensitive nuclear factors with three subtypes: PPARR,
PPARγ, and PPARδ. PPARs play an important role in many
cellular functions and have emerged as therapeutic targets for
the treatment of human metabolic diseases.1-6 To address
multiple metabolic disorders more effectively, the simultaneous
activation of PPARR, PPARγ, and PPARδ by a single com-
pound (i.e., a PPARpan agonist) is being pursued. Furthermore,
the simultaneous activation of individual PPARs with a single
pan molecule may result in synergistic effects, affording
enhanced biological activity and increased potency.6 On the basis
of preclinical data, compound 1 was identified as a potent
† Department of Medicinal Chemistry.
(1) Wilson, T. M.; Brown, P. J.; Sternbach, D. D.; Henke, B. R. J. Med.
Chem. 2000, 43, 527.
(2) Sternbach, D. D. Ann. Rep. Med. Chem. 2003, 38, 71.
(3) Kersten, S.; Desvergne, B.; Wahli, W. Nature 2000, 405, 421.
(4) Berger, J.; Moller, D. E. Annu. ReV. Med. 2002, 53, 409.
(5) Evans, R. M.; Barish, G. D.; Wang, Y.-X. Nat. Med. 2004, 10, 1.
(6) Evans, J. L.; Lin, J. J.; Goldfine, I. D. Curr. Diabetes ReV. 2005, 1,
299.
(10) On the basis of MOPAC/PM3 calculations, it was found that the
carbocation A is 11 kcal/mol lower in energy than that of the carbocation
B. This result indicates that the primary product should be the expected
one.
10.1021/jo061295n CCC: $33.50 © 2006 American Chemical Society
Published on Web 09/12/2006
8302
J. Org. Chem. 2006, 71, 8302-8305

SCHEME 1
SCHEME 2
SCHEME 3
SCHEME 4
in a single pot, giving thioether products while minimizing
potential disulfide formation (Scheme 4).12
Although these conditions provided a convenient single-pot
process, the reduction was found to be very exothermic under
anhydrous conditions, and the exotherm was not addition
controlled. An induction period followed by an extremely large
and sudden exotherm was observed even when only small
amounts of Me2SiCl2 had been added.13 However, the presence
of a small amount of residual water in the sulfonyl chloride
moderated the exotherm, and by slow addition of Me2SiCl2 to
a mixture of the sulfonyl chloride and zinc dust in isopropyl
acetate, the reaction proceeded smoothly under addition control
with only a mild, addition-controlled exotherm. The reduction
of the sulfonyl chloride to the thiophenol was typically complete
the former approach, it was necessary to prepare the required
O-substituted thiophenol 2. It had been reported11 that aromatic
sulfonyl chlorides are reduced to aromatic thiophenols with Zn/
Me2SiCl2 in dichloroethane, in the presence of DMA (N,N-
dimethylacetamide) or DMF. DMA and DMF are reported to
serve as activating agents, converting sulfonyl chlorides to
Vilsmeyer-type reactive iminium salts. Further, it was reported
that, in the absence of DMA or DMF, no reaction had taken
place. In our efforts to replace 1,2-dichloroethane with a more
desirable solvent for large-scale preparation, we identified
isopropyl acetate. Further, we found that reduction of the
sulfonyl chlorides in isopropyl acetate did not require DMA or
DMF as additives and allowed in situ coupling to be achieved
(12) Martin, M. T.; Thomas, A. M.; York, D. G. Tetrahedron Lett. 2002,
43, 2145.
(11) (a) Uchiro, H.; Kobayashi, S. Tetrahedron Lett. 1999, 40, 3179.
(b) Uchiro, H.; Kobayashi, S. Jpn. Kokai Tokkyo Koho JP2000256306.
(13) The sulfonyl chloride was reduced to a sulfinic acid after the initial
vigorous exotherm.
J. Org. Chem, Vol. 71, No. 21, 2006 8303

SCHEME 5
SCHEME 6
SCHEME 7
on conclusion of the Me2SiCl2 addition. The exotherm could
be further reduced by adding a mixture of a sulfonyl chloride
and Me2SiCl2 (0.25-0.75 equiv) in isopropyl acetate to a
suspension of zinc dust in isopropyl acetate containing small
amounts of water (0.5 equiv). In these reactions, Me2SiCl2 first
reacts with water to form HCl and polysiloxane. The sulfonyl
chloride is reduced by Zn/HCl to a sulfinic acid, which
subsequently is reduced by Zn/Me2SiCl2 to thiophenol under
anhydrous conditions. Alternatively, a mild reduction can be
achieved via a single-pot process by reducing sulfonyl chlorides
to sulfinic acids with Zn/HOAc/H2O followed by reduction of
12
the sulfinic acids to thiols with Zn/Me2SiCl2.
Using the reaction conditions described above, the sulfonyl
chloride 97 was reduced with Zn/Me2SiCl2 in isopropyl acetate
to the thiophenol 2, which was then coupled in situ with 3,
affording the desired coupling product 10 with excellent
regioselectivity (Scheme 5).
Of the two potential side products, the bis-coupling product
11 and C-4 regioisomer, only the bis-coupling product, 11, was
produced, and in small amounts. This impurity was removed
by crystallization from EtOH/isooctane (1:1), giving the desired
product 10 in 55% isolated yield with 98% purity. The mono-
coupling regioisomer at the C-4 hydroxymethyl carbon was not
detected at all.
methanesulfonic acid was achieved with excellent regioselec-
tivity, affording 12 in 63% isolated yield with 98% purity
(Scheme 6).
Mesylation of the hydroxymethyl in 10, followed by dis-
placement of the primary mesylate with 1-(4-methoxyphenyl)-
piperazine and saponification of the ester with aqueous sodium
hydroxide, afforded 1. The efficiency of the process was
successfully demonstrated on a 50 L scale, producing several
batches of 1 (up to 2 kg per batch).
Again, the undesired mono-coupling product at C-4 hy-
droxymethyl was not detected, while the bis-coupling side
product 13 (approximately 1:13 vs the desired product) was
removed during crystallization. To assemble the piperazine
fragment, the hydroxymethyl group in 12 needed to be converted
to an active leaving group. Treatment of 12 with methane-
sulfonic anhydride and Hunig’s base in dichloromethane thus
gave a bismesylate (Scheme 7).
Having established regioselective control and crystallization
methods for rejection of bisalkylated products, an alternative
and more efficient preparation was accomplished via direct
coupling of 3 with 4-hydroxythiophenol. The coupling of 3 with
4-hydroxythiophenol in MeCN/toluene in the presence of
Use of methanesulfonic anhydride as the mesylating agent
was found to be essential since treatment of 12 with methane-
sulfonyl chloride resulted in the corresponding alkyl chloride.
8304 J. Org. Chem., Vol. 71, No. 21, 2006

The chloride was found to be a poor substrate for subsequent
displacement with the piperazine. The alkyl mesylate, on the
other hand, was readily displaced with 1-(4-methoxyphenyl)-
piperazine, while the aryl mesylate served as a protecting group
for the phenol. Although the use of sulfonates as protecting
groups for phenols has recently attracted more attention,14,15
removal of an aryl mesylate or aryl tosylate often requires
vigorous conditions.15,16 However, our research indicated that
aryl mesylate 14 could be readily cleaved by treatment with
20-40 mesh NaOH beads in acetone or THF/acetone containing
small amounts (1-2 equiv) of water at room temperature. The
reaction is heterogeneous. The added water probably increases
the solubility of NaOH in the reaction medium thus increasing
the nucleophilicity of the hydroxide anion and consequently the
reaction rate. The reaction proceeded very slowly under
anhydrous conditions. Use of 20-40 mesh NaOH beads was
found to be necessary to achieve the mild and efficient cleavage
of the aryl mesylate. The demesylation proceeded very slowly
when other forms of NaOH, such as flakes, pellets, or aqueous
solution, were used. Using the conditions described, demesy-
lation of 14 was complete within 1 h and afforded phenol 15 in
87% isolated yield.
duced via alkylation of 15 with ethyl 2-bromoisobutyrate and
K2CO3 in ethanol followed by hydrolysis. However, ethyl
methacrylate, which tends to polymerize under the reaction
conditions, was obtained as a major side product. Polymethacry-
late, although not an issue for small-scale synthesis, posed
quality issues on scale. Additionally, polymer-coated vessels
posed cleaning issues. Alternatively, the 2-isobutyric acid
fragment was assembled via the Bargellini chemistry17 (1,1,1-
trichloro-2-methyl-2-propanol with NaOH). However, the exo-
thermic nature of the reaction posed a safety issue18 for large-
scale preparation. To address these issues, we developed an
efficient, controlled, and practical method using 2-bromoisobu-
tyric acid as an alkylating agent to introduce the isobutyric acid
moiety.18 Thus, alkylation of 15 with 2-bromoisobutyric acid
in 2-butanone furnished 1 in excellent yield, while avoiding the
polymerization and safety problems. The purity of the final
product 1 was consistently over 99%.
In summary, we have developed a highly efficient, practical,
and scalable route for the synthesis of the potent PPARpan
agonist 1, with 30% overall yield, shortening the original
synthesis from 14 linear steps to 7 steps. Acid-catalyzed
coupling of the diol 3 with 4-hydroxythiophenol provides an
efficient and highly regioselective means of forming the
carbon-sulfur bond. To the best of our knowledge, there is no
precedent in the literature for this kind of differentiation on the
thiazole nucleus. A multistep sequence involving a mild
deprotection of a phenol mesylate introduces the piperazine
fragment efficiently. Finally, using 2-bromoisobutyric acid as
the alkylating agent assembles the isobutyric acid moiety
effectively and efficiently. The route is free of column chro-
matography and is amenable for large-scale preparation of 1.
The final step of the route required assemblage of the
2-isobutyric acid fragment, which had previously been intro-
(14) (a) Briot, A.; Baehr, C.; Brouillard, R.; Wagner, A.; Mioskowski,
C. Tetrahedron Lett. 2003, 44, 965. (b) Ritter, T.; Stanek, K.; Larrosa, I.;
Carreira, E. M. Org. Lett. 2004, 6, 1513.
(15) (a) Carr, J. A.; Bisht, K. S. Org. Lett. 2004, 6, 3297. (b) Bensel,
N.; Pevere, V.; Desmurs, J. R.; Wagner, A.; Mioskowski, C. Tetrahedron
Lett. 2002, 43, 4281. (c) Birkbeck, A. A.; Brkic, Z.; Giles, R. G. F.
Tetrahedron Lett. 2004, 45, 6147.
(16) Greene, T. W.; Wuts, P. G. M. ProtectiVe Groups in Organic
Synthesis, 3rd ed.; John Wiley & Sons: New York, 1999.
Acknowledgment. We thank Dr. Alan Millar for helpful
comments and discussions.
(17) (a) Bargellini, G. Gazz. Chim. Ital. 1906, 36, 337. (b) Bargellini,
G. Atti Accad. Naz. Lincei Cl. Sci. Fis. Mater. Nat. Rend. 1906, 15I, 582.
(c) Weizmann, C.; Sulzbacher, M.; Bergmann, E. J. Am. Chem. Soc. 1948,
70, 1153. (d) Gilman, H.; Wilder, G. R. J. Am. Chem. Soc. 1955, 77, 6644.
(e) Clewley, R. G.; Fischer, A.; Henderson, G. N. Can. J. Chem. 1989, 67,
1472. (f) Cvetovich, R. J.; Chung, J. Y. L.; Kress, M. H.; Amato, J. S.;
Matty, L.; Weingarten, M. D.; Tsay, F.-R.; Li, Z.; Zhou, G. J. Org. Chem.
2005, 70, 8560.
Supporting Information Available: Experimental procedures
1
and H and 13C NMR spectra for compounds 1, 3, 6, 10, 12, and
15. This material is available free of charge via the Internet at
http://pubs.acs.org.
(18) Davis R.; Fitzgerald R.; Guo, J. Synthesis 2004, 1959.
JO061295N
J. Org. Chem, Vol. 71, No. 21, 2006 8305