Pictet-Spengler based synthesis of a bisarylmaleimide glycogen synthase kinase-3 inhibitor

Article abstract of DOI:10.1021/ol101405g

A practical synthesis of the glycogen synthase kinase-3 (GSK3) inhibitor bisarylmaleimide 1 has been accomplished employing Pictet-Spengler methodology to access the indole 7-position in preparing the benzodiazepine tricyclic fragment. A seven-step linear sequence that starts with commercially available 5-fluoroindole 7 affords the bisarylmaleimide 1 in 33% overall yield.

Full text of DOI:10.1021/ol101405g

ORGANIC
LETTERS
2010
Vol. 12, No. 16
3700-3703
Pictet-Spengler Based Synthesis of a
Bisarylmaleimide Glycogen Synthase
Kinase-3 Inhibitor
Nicholas A. Magnus,* Christopher P. Ley, Patrick M. Pollock, and
James P. Wepsiec
Eli Lilly and Company, Chemical Product Research and DeVelopment DiVision,
Indianapolis, Indiana 46285
magnus_nicholas@lilly.com
Received July 7, 2010
ABSTRACT
A practical synthesis of the glycogen synthase kinase-3 (GSK3) inhibitor bisarylmaleimide 1 has been accomplished employing Pictet-Spengler
methodology to access the indole 7-position in preparing the benzodiazepine tricyclic fragment. A seven-step linear sequence that starts with
commercially available 5-fluoroindole 7 affords the bisarylmaleimide 1 in 33% overall yield.
Type 2, or noninsulin-dependent, diabetes mellitus ac-
counts for approximately 90% of all cases of diabetes and
is characterized by an inability of the body to effectively
respond to insulin secreted by the pancreas, or insulin
resistance.1 Glycogen synthase kinase-3 inhibitor (GSK3)
is involved in signaling from the insulin receptor. Inhibi-
tors of GSK3 are expected to effect lowering of plasma
glucose similar to insulin, making GSK3 an attractive
target for the treatment of type 2 diabetes. The GSK3
inhibitor bisarylmaleimide 1, illustrated in Figure 1, has
Figure 1. Bisarylmaleimide 1.
been studied at Lilly for the treatment of type 2 dia-
betes.2
The published synthesis of bisarylmaleimide 1 by Engler
and Henry et al. describes the preparation of the tricyclic
ring system of the indole fragment via the Bartoli indole
synthesis.2,3 The Bartoli reaction gives access to an indole
with an aldehyde at the 7-position for subsequent preparation
of the required seven-membered benzodiazepine type ring
system.
(1) (a) Cohen, P.; Frame, S. Nat. ReV. Mol. Cell. Biol. 2001, 2, 769. (b)
Cohen, P.; Goedert, M. Nat. ReV. Drug. Disc. 2004, 3, 479.
(2) Engler, T. A.; Henry, J. R.; Malhotra, S.; Cunningham, B.; Furness,
K.; Brozinick, J.; Burkholder, T. P.; Clay, M. P.; Clayton, J.; Diefenbacher,
C.; Hawkins, E.; Iversen, P. W.; Li, Y.; Lindstrom, T. D.; Marquart, A. L.;
McLean, J.; Mendel, D.; Misener, E.; Briere, D.; O’Toole, J. C.; Porter,
W. J.; Queener, S.; Reel, J. K.; Owens, R. A.; Brier, R. A.; Eessalu, T. E.;
Wagner, J. R.; Campbell, R. M.; Vaughn, R. J. Med. Chem. 2004, 47, 3934.
10.1021/ol101405g  2010 American Chemical Society
Published on Web 07/19/2010

The synthesis starts with 5-fluoro-2-nitrobenzaldehyde 2,
which is prepared via reduction and oxidation of the
corresponding acid (Scheme 1).2,4 5-Fluoro-2-nitrobenzal-
Scheme 2. Retrosynthetic Plan for a Tricyclic Indole
Scheme 1. Intermediates for Bartoli-Based Synthesis of 6a
Fischer indole synthesis with 4-fluorophenylhydrazine and
pyruvic acid followed by decarboxylation,7 and metalation
methods by Schlosser.8 Alkylation of 7 with 2-chloroethy-
lamine to give intermediate 8 has precedence utilizing a
quaternary ammonium salt to promote phase-transfer chem-
istry.9 Next, we speculated that indole 8 with the ethylamine
tether could be taken directly into a Pictet-Spengler cy-
clization to give tricycle 11 based on literature precedent
employing paraformaldehyde and TFA at reflux.10 If the
conversion of 8 to 11 failed, the alternative path of reducing
the 5-member indole ring of 8 to the indoline 9 to induce
aniline-type direction and promote the Pictet-Spengler
cyclization to give tricycle 10 is also precedented.9b,11 Lastly,
the oxidation/dehydrogenation of N-alkylated indolines simi-
lar to 10 to afford indoles can be accomplished by a variety
of methods.12
The N-alkylation of indole 7 with 2-chloroethylamine
hydrochloride 12 performed most reliably in the presence
of excess 85% KOH pellets at about 23 °C (Scheme 3).
Lesser amounts of KOH, or use of the powdered form, often
gave stalled reactions and over alkylation. Research into this
matter led to the conclusion that the desired selective mono-
alkylation was taking place on the surface of the base (85%
KOH pellet) in the omega phase.13 Therefore, the use of
excess 85% KOH pellets ensured that an omega phase was
dehyde 2 is converted to the di-n-butyl acetal 3, which gives
the necessary steric bulk to favor the Bartoli reaction.3c The
Bartoli reaction and subsequent hydrolysis affords 5-fluoro-
7-formylindole 4. The 7-formylindole 4 is elaborated via
reductive amination with ethanolamine, BOC protection, and
activation as a mesylate 5 to promote closure of the
7-membered ring to afford 6a. The intermediate mesylate 5
has stability issues due to intramolecular attack of the BOC
group on the mesylate, which gives appreciable amounts of
oxazolidinone 6b at temperatures above 0 °C.5 In addition,
protection group chemistry is necessary since attempts to
replace the BOC group with the required piperidine urea of
1 failed to give the desired mesylate activation for 7-mem-
bered ring closure.
To avoid the low temperature Bartoli reaction conditions
(e-40 °C) that require a large excess of vinyl Grignard
reagent (4 equiv) and afford modest yields (∼50%), we first
focused on alternative methods for preparing the tricyclic
indole component. Secondary goals were to maximize
efficiency by avoiding protection group chemistry and to
delay construction of the bisarylmaleimide until late in the
synthesis. Bisarylmaleimides are insoluble in most solvents,
making them difficult to manipulate and purify. Therefore,
a strategy involving late construction of the bisarylmaleimide
ring system made sense to increase flexibility of solvent and
reagent choices during the synthesis and enhance the purity
of the final product.
(6) (a) Clark, R. D.; Repke, D. B. J. Heterocycl. Chem. 1985, 22, 121.
(b) Madge, D. J.; Hazelwood, R.; Iyer, R.; Jones, H. T.; Salter, M. Bioorg.
Med. Chem. Lett. 1996, 6, 857.
(7) (a) Bergmann, E. D.; Pelchowicz, Z. J. Chem. Soc. 1959, 1913. (b)
Bratulescu, G. Tetrahedron Lett. 2008, 49, 984.
(8) Schlosser, M.; Ginanneschi, A.; Leroux, F. Eur. J. Org. Chem. 2006,
13, 2956.
(9) (a) Cuadro, A. M.; Matia, M. P.; Garcia, J. L.; Vaquero, J. J.; Alvarez-
Builla, J. Synth. Commun. 1991, 21, 535. (b) Welmaker, G. S.; Sabalski,
J. N. Tetrahedron Lett. 2004, 45, 4851.
With these goals in mind, Pictet-Spengler methodology
to access the indole 7-position in preparing the benzodiaz-
epine portion of 1 was proposed (Scheme 2). In Scheme 2,
the starting 5-fluoroindole 7 is commercially available and
has been prepared via the Leimgruber-Batcho method,6
(10) Zhang, L.-H.; Meier, W.; Wats, E.; Costello, T. D.; Ma, P.;
Ensinger, C. L.; Rodgers, J. M.; Jacobsen, I. C.; Rajagopalan, P. Tetrahedron
Lett. 1995, 36, 8387.
(11) (a) Archer, S.; Lewis, T. R.; Unser, M. J.; Hoppe, J. O.; Lape, H.
J. Am. Chem. Soc. 1957, 79, 5783
.
(12) (a) Naruto, S.; Yonemitsu, O. Chem. Pharm. Bull. 1980, 28, 900.
(b) Gribble, G. W.; Pelcman, B. J. Org. Chem. 1992, 57, 3636. (c) Hara,
T.; Mori, K.; Mizugaki, T.; Ebitani, K.; Kaneda, K. Tetrahedron Lett. 2003,
44, 6207. (d) Hara, T.; Mori, K.; Mizugaki, T.; Ebitani, K.; Kaneda, K.
Tetrahedron Lett. 2003, 44, 6207. (e) Kamata, K.; Kasai, J.; Yamaguchi,
K.; Mizuno, N. Org. Lett. 2004, 6, 3577. (f) Lakatosh, S. A.; Luzikov,
Y. N.; Preobrazhenskaya, M. N. Tetrahedron 2005, 61, 8241. (g) Haider,
N.; Kaeferboeck, J. Tetrahedron 2004, 60, 6495.
(3) (a) Bartoli, G.; Palmieri, G.; Bosco, M.; Dalpozzo, R. Tetrahedron
Lett. 1989, 30, 2129. (b) Bosco, M.; Dalpozzo, R.; Bartoli, G.; Palmieri,
G.; Petrini, M. J. Chem. Soc., Perkin Trans. 2 1991, 657. (c) Dobson, D. R.;
Gilmore, J.; Long, D. A. Synlett 1992, 79
.
(4) Lautens, M.; Fang, Y.-Q. J. Org. Chem. 2008, 73, 538
.
(5) Curran, T. P.; Pollastri, M. P.; Abelleira, S. M. R. J.; McCollum,
T. A.; Rowe, C. G. Tetrahedron Lett. 1994, 35, 5409.
Org. Lett., Vol. 12, No. 16, 2010
3701

Scheme 3. Pictet-Spengler Based Synthesis of Bisarylmaleimide 1
maintained which afforded high conversions to the desired
mono-alkylated indole 13. Maintaining the reaction temper-
ature at about 23 °C was also important for high conversions
to 13 due to competitive degradation of 12 to aziridine, which
was more pronouced at higher temperatures. The N-akylated
indole 13 was purified via isolation as a hydrochloride salt
in 80% yield.
The reduction of indole 13 to indoline 14 was attempted
via hydrogenolysis with palladium and platinum which
typically gave unreliable conversions and over reduction,
respectively.9b,14 The use of triethylsilane in TFA was useful
for the transformation of 13 to 14, but working up the
corrosive TFA reaction media to isolate the product was not
desirable.15 Reduction of 13 to 14 was most productive with
borane pyridine complex in AcOH at 23 °C.16 The indoline
14 was first isolated as a hydrochloride salt, but chloride
was found to suppress the subsequent Pictet-Spengler
chemistry presumably via the formation of stable chlorohy-
drins. However, 14 could be purified and isolated in 87%
yield as an acetate salt which was compatible with the
reaction media for the next step.
indoline 10 was isolated as an oil and was found to perform
best in the necessary oxidation/dehydrogenation to give an
indole if the piperidine urea was installed first. With that,
10 was dissolved in xylenes and treated with TEA and
1-piperidinecarbonyl chloride 15 at 20 °C to afford indoline
urea 16 as an oil. After workup, the xylenes solution of 16
was treated with palladium on carbon, and cyclohexene was
introduced to remove hydrogen from the system and suppress
reversion of 17 to 16. Heating this mixture to 140 °C
produced the indole 17 as a crystalline solid in 72% yield
from 14.
The tricyclic indole urea 17 was suspended in MTBE at
about 3 °C, and oxalyl chloride was added to the mixture. It
was necessary to warm this suspension to 20 °C to induce
acylation of the indole 3-position. The suspension was cooled
to 10 °C and a large excess of MeOH added to the reaction
to generate the methyl ester 18 still as a suspension.17 At
that point, the ester 18 could be isolated by simple filtration
and washing with MeOH in 82% yield.
The ester 18 was combined with the known amide 192 in
DMF, and the resulting slurry was cooled to about 5 °C. A
solution of tBuOK in THF was added to the ester 18/amide
19 mixture at 5-7 °C to give a homogeneous solution. It
was important to keep the temperature at or below 5-7 °C
during the base addition as the intermediate imide 20 formed
and cyclized via an intramolecular Perkin-type condensation
to give 21, which is slow to dehydrate to the biarylmaleimide
1 (Scheme 4).17 If the base addition was conducted at higher
temperatures, the dehydration event that liberates KOH
started to occur prior to completion of the initial imide
formation resulting in considerable hydrolysis of the starting
ester 18 and loss of desired product yield. After the
maleimide formation was complete, the resulting slurry was
diluted with water and DMF, and the THF was removed from
Attempts to induce Pictet-Spengler cyclizations at the
indole 13 oxidation state failed in our hands probably due
to the electron deficiency of the fluorinated benzenoid ring.
However, the indoline 14 underwent smooth transformation
to the tricyclic indoline 10 in the presense of 1 equiv of
formalin, AcOH, and catalytic H2SO4 at 70 °C. The tricyclic
(13) (a) Halpern, M. E. In Phase-Transfer Catalysis: Mechanisms and
Syntheses; Liotta, C. L., Berkner, J., Wright, J., Fair, B., Eds.; ACS
Symposium Series 659; American Chemical Society: Washington, DC,
1997; pp 29-40. (b) Yadav, G. D.; Jadhav, Y. B. Org. Process Res. DeV.
2003, 7, 588
.
(14) Shaw, J. E.; Stapp, P. R. J. Heterocycl. Chem. 1987, 24, 1477
.
(15) (a) Lanzilotti, A. E.; Littell, R.; Fanshawe, W. J.; McKenzie, T. C.;
Lovell, F. M. J. Org. Chem. 1979, 44, 4809. (b) Magnus, P.; Gallagher, T.;
Schultz, J.; Orr, Y.-N.; Ananthanarayan, T. P. J. Am. Chem. Soc. 1987,
109, 2706. (c) Hobson, L. A.; Nugent, W. A.; Anderson, S. R.; Deshmukh,
S. S.; Haley, J. J.; Liu, P.; Magnus, N. A.; Sheeran, P.; Sherbine, J. P.;
Stone, B. R. P.; Zhu, J. Org. Process Res. DeV. 2007, 11, 985.
(17) (a) Faul, M. M.; Winneroski, L. L.; Krumrich, C. A. J. Org. Chem.
1998, 63, 6053. (b) Faul, M. M.; Winneroski, L. L.; Krumrich, C. A.
Tetrahedron Lett. 1999, 40, 1109.
(16) Yasuo, K.; Kunio, S.; Schun-ichi, Y. Synthesis 1978, 447.
3702
Org. Lett., Vol. 12, No. 16, 2010

Scheme 4. Mechanism for Maleimide Formation
the mixture via distillation (failure to remove THF from the
system resulted in complex mixed solvates of the crystals
of 1 during the isolation/purification). The resulting mixture
was quenched with aqueous AcOH and seeded with 1 during
a further water addition to induce crystallization of 1 in 81%
isolated yield with >99% purity.
The seven-step linear sequence detailed in Scheme 3 that
starts with commercially available 5-fluoroindole 7 affords
the bisarylmaleimide 1 in 33% overall yield. This linear
sequence can be run in standard reactors with no need for
temperatures below about 0 °C. There is no protection group
chemistry, and the assembly of the insoluble maleimide
structure is reserved until the last step which gives flexibility
with regards to reagent and solvent choices and ensures high
purity final product. The overall result is an efficient and
reliable synthesis that provides access to large quantities of
the bisarylmaleimide 1.
Acknowledgment. We are grateful to Mr. Eric D. Crockett
(Eli Lilly and Company) for performing all of the high-
resolution mass spectrometer experiments and Mr. Chad E.
Hadden (Eli Lilly and Company) for assistance with inter-
pretation of the NMR spectra.
Supporting Information Available: Experimental pro-
cedures and spectral data for compounds 13, 14, 10, 16-18,
and 1. This material is available free of charge via the Internet
at http://pubs.acs.org.
OL101405G
Org. Lett., Vol. 12, No. 16, 2010
3703