Asymmetric synthesis of a prostaglandin D2 receptor antagonist

Article abstract of DOI:10.1021/jo048305+

(Chemical Equation Presented). An asymmetric synthesis was developed for the production of a prostaglandin D₂ receptor antagonist for the treatment of allergic rhinitis. The stereogenic center was set using asymmetric allylic alkylation chemist

Full text of DOI:10.1021/jo048305+

Asymmetric Synthesis of a Prostaglandin D2 Receptor Antagonist
Kevin R. Campos,* Michel Journet, Sandra Lee, Edward J. J. Grabowski, and
Richard D. Tillyer
Department of Process Research, Merck Research Laboratories,
P.O. Box 2000, RY800-B267, Rahway, New Jersey 07065-0900
kevin_campos@merck.com
Received September 23, 2004
An asymmetric synthesis was developed for the production of a prostaglandin D2 receptor antagonist
for the treatment of allergic rhinitis. The stereogenic center was set using asymmetric allylic
alkylation chemistry, and the core of the structure was constructed via Pd-catalyzed N-cyclization/
Heck methodology. The synthesis relies on a late stage indoline oxidation which does not racemize
the product.
Introduction
Results and Discussion
To determine the optimal method for the synthesis of
this highly functionalized indole, several classical meth-
ods for the synthesis of indoles were attempted.4 Fischer
indole cyclization of hydrazine 2 with optically pure
ketoester 3 provided the advanced intermediate 4 in good
yield albeit with complete racemization of the stereogenic
center (eq 1).5 Likewise, condensation of aniline 5 with
optically pure 3 followed by intramolecular Heck reaction
rapidly provided the desired indole 6, but again as a
racemic mixture (eq 2).
Prostaglandin D2 (PGD2) is the major cyclooxygenase
metabolite of arachidonic acid produced by mast cells in
response to antigen challenge.1 It has been proposed that
excess production of PGD2 causes the inflammation
commonly observed in allergic diseases such as allergic
rhinitis, asthma, and atopic dermatitis.2 Efforts to de-
velop a PGD2 receptor antagonist have identified 1 as a
promising lead in the alleviation of various allergic
disorders.3 In this paper, the full details of the asym-
metric synthesis of 1 are disclosed.
(1) Lewis, R, A.; Soter, N. A.; Diamond, P. T.; Austen, K. F.; Oates,
J. A.; Roberts, L. J. J. Immunol. 1982, 129, 1627-1631.
(2) (a) Matsuoka, T.; Hirata, M.; Tanaka, H.; Takahashi, Y.; Murata,
T.; Kabashima, K.; Sugimoto, Y.; Kobayashi, T.; Ushikubi, F.; Aze, Y.;
Eguchi, N.; Urade, Y.; Yoshida, N.; Kimura, K.; Mizoguchi, A.; Honda,
Y.; Nagai, H.; Narumiya, S.; Kato, M.; Watanabe, M.; Vogler, B.; Awen,
B.; Masuda, Y.; Tooyama, Y.; Yoshikoshi, A. Science 2000, 287, 2013-
2016. (b) Charlesworth, E. N.; Kagey-Sobotka, A.; Schliemer, R. P.;
Norman, P. S.; Lichtenstein, L. M. J. Immunol. 1991, 149, 671-676.
(c) Proud, D.; Sweet, J.; Stein, P.; Settipane, R. A.; Kagey-Sobotka, A.;
Freidlander, M.; Lichtenstein, L. M. J. Allergy Clin. Immunol. 1990,
85, 896-905. (d) Murray, J. J.; Tonnel, A. B.; Brash, A. R.; Roberts, L.
J.; Gosset, P.; Workman, R.; Capron, A.; Oates, J. N. Eng. J. Med. 1986,
315, 800-804.
The observed lability of the stereogenic center to
racemization forced us to seek alternative methods to
(3) Tsuri, T.; Honma, T.; Hiramatsu, Y.; Mitsumori, S.; Inagaki, M.;
Arimura, A.; Yasui, K.; Asanuma, F.; Kishino, J.; Ohtani, M. J. Med.
Chem. 1997, 40, 3504-3507.
10.1021/jo048305+ CCC: $30.25 © 2005 American Chemical Society
Published on Web 12/01/2004
268
J. Org. Chem. 2005, 70, 268-274

Asymmetric Synthesis of a Prostaglandin D2 Receptor Antagonist
SCHEME 1. Retrosynthetic Analysis
FIGURE 1. X-ray structure of ligand/Pd complex 9.
isolated yield, 98% LCAP, 96% ee). Decarboxylation of
10 was readily accomplished in HOAc at 120 °C to afford
chiral acid 8 in 85% yield. Coupling of 8 with 2,6-dibromo-
4-fluoroaniline (7) via the acid chloride produced amide
11, which was isolated via recrystallization in 82%
yield.
synthesize 1. We envisioned that indoline 23 would be a
suitable precursor to 1 that would unmask the indole
functionality at a late stage (Scheme 1). We believed that
23 could be produced from bicyclic lactam 14 via in-
tramolecular Heck reaction. We felt that 14 would be
readily accessible from 2,6-dibromo-4-fluoroaniline (7)
and (R)-2-cyclopentene-1-acetic acid (8). This strategy
was attractive because the stereogenic center was intro-
duced early in the synthesis, the design was amenable
to the indole substitution pattern in 1, and the starting
materials were commercially available in bulk.
SCHEME 3. Alternative Synthesis of 11
SCHEME 2. Synthesis of 11
Alternatively, Meldrum’s acid could be used as the
nucleophile in Trost’s asymmetric allylic alkylation to
afford the desired adduct in similar yield, but slightly
lower enantioselectivity (90% ee). The advantage of this
method lies in the ability to directly couple this inter-
mediate with 2,6-dibromo-4-fluoroaniline in refluxing
toluene in 90% isolated yield (Scheme 3). Although the
process was two steps shorter, the diminished enanti-
oselectivity in the allylic alkylation made this method less
attractive.
Preliminary studies toward intramolecular N-cycliza-
tion to form bicyclic lactam 14 were focused on an
iodolactamization/elimination sequence (Scheme 4). Sub-
jection of 11 to conditions precedented in the literature7
for regioselective N-cyclization afforded an 88:12 mixture
of regioisomers favoring the desired, N-cyclized product
12 over the undesired O-cyclized product 13.8 The crude
mixture of 12 and 13 was treated with DBU in toluene,
which cleanly formed the elimination products 14 and
15. The two regioisomers were readily separated by
column chromatography to afford 14 in 85% yield over
the two-step process.
In accordance with literature precedent, (R)-2-cyclopen-
tene-1-acetic acid (8) was produced via Trost asymmetric
allylic alkylation of cyclopentenyl acetate with dimethyl
malonate (Scheme 2).6 Replacement of NaH/Hex4NBr
with bistrimethylsilylacetamide (BSA) as the base re-
sulted in a more practical, reproducible procedure. The
reaction was run at 0 °C with 4 mol % Pd, a 1:1 ratio of
catalyst to ligand, and only 1.05 equivalents of dimeth-
ylmalonate to afford a 92% yield of the desired malonate
adduct in 96% ee. We discovered that subjection of the
crude allylic alkylation mixture to hydrolysis conditions
(KOH/MeOH) generated a precipitate, which was deter-
mined to be ligand/Pd complex 9 by X-ray crystal
structure analysis (Figure 1). Removal of 9 via filtration
of the resulting slurry afforded a solution of the potas-
sium salt of malonic acid adduct 10, which crystallized
upon acidification, extraction, and concentration (88%
Subjection of 14 to a variety of Heck reaction condi-
tions9 afforded product 16 very cleanly; however, most
(4) Gribble, G. W. J. Chem. Soc., Perkin Trans. 1 2000, 1045-1075
and references therein.
(7) Arunachalam, T.; Fan, H.; Pillai, K. M. R.; Ranganathan, R. S.
J. Org. Chem. 1995, 60, 4428-4438.
(5) (a) Gribble, G. W. Contemp. Org. Synth. 1994, 1, 1. (b) Hughes,
D. L. Org. Prep. Proc. Int. 1993, 25, 607-632. (c) Robinson, B. The
Fischer Indole Synthesis; Wiley-Interscience: New York, 1982.
(6) (a) Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921-
2944. (b) Trost, B. M.; Bunt, R. C. J. Am. Chem. Soc. 1994, 116, 4089-
4090.
(8) The O-cyclized product 13 was a 2:1 mixture of E/Z imidate
isomers.
(9) (a) Dounay, A. B.; Overman, L. E. Chem. Rev. 2003, 2945-2964.
(b) Link, J. Org. React. 2002, 157. (c) Beletskaya, I. P.; Cheprakov, A.
V. Chem. Rev. 2000, 100, 3009-3066.
J. Org. Chem, Vol. 70, No. 1, 2005 269

Campos et al.
SCHEME 5. Synthesis of Bromoester 18
SCHEME 4. Iodolactamization/Elimination
receptor antagonist 1. In fact, we believed that the
desired indole product could be obtained from 16 via
isomerization of the olefin to the desired position; how-
ever, all attempts to isomerize 16 were unsuccessful,
presumably due to steric constraints of the tricyclic
system.13 To reduce the strain of the tricyclic system, the
lactam was hydrolyzed (KOH, MeOH, 50 °C) to cleanly
afford the bicyclic amino acid intermediate 17 (Scheme
5). Subjection of 17 to isomerization conditions (RhCl3,
EtOH)14 afforded the desired indole 18 in 75% yield;
however, the enantiomeric excess of the product was only
57%. The remainder of the material had reverted to
lactam 16. The lactamization side reaction was a recur-
ring problem with substrate 17 under any Lewis acidic
conditions, presumably due to the nucleophilicity of the
amine.
conditions suffered from incomplete conversion (50% yield
at best) with the remainder being starting material.
Complete conversion was achieved when the reaction was
performed in DMF/water in the absence of phosphine
ligand, which delivered 16 in 80% isolated yield (eq 3).10
To deactivate the amine toward acylation, the sulfone
moiety was introduced at this stage of the synthesis.
Copper-mediated coupling of sodium sulfinate15 with
bromo acid 17, followed by esterification afforded sulfone/
ester intermediate 20 in 85% yield over the two steps
(Scheme 6). This substrate did not suffer the same
lactamization side reaction observed with 17, presumably
due to the attenuated nucleophilicity of the vinylogous
sulfonamide.
With a method for the conversion of 11 to 16 in hand
(three steps), we sought to streamline this process
through catalytic methods. It is precedented that in-
tramolecular N-cyclizations across an olefin can be
achieved using a stoichiometric amount of a PdII source.11
We envisioned that reaction conditions should exist
whereby PdII would mediate the N-cyclization while the
byproduct of this cyclization (Pd0) would catalyze the
Heck reaction.12 Indeed, when acyclic amide 11 was
subjected to stoichiometric Pd(OAc)2 and K2CO3 in DMF
at 50 °C, the Heck adduct 16 was isolated directly in 40%
yield (eq 4). Efforts to improve this process are ongoing.
SCHEME 6. Synthesis of Sulfone/Ester 20
Subjection of 20 to the previously developed isomer-
ization conditions (RhCl3/EtOH) afforded clean conver-
sion to indole 21; however, the enantiomeric excess was
only 55%, similar to that observed with 17 (Table 1).
The synthesis of Heck adduct 16 represented the
complete synthesis of the carbon skeleton of the PGD2
(12) For recent examples of tandem Pd-catalyzed processes, see:
Pinho, P.; Minnaard, A. J.; Feringa, B. L. Org. Lett. 2003, 259-261
and references therein.
(10) The benefits of Heck conditions performed in aqueous systems
have been reported. (a) Zhao, F.; Shirai, M.; Arai, M. J. Mol. Catal. A
2000, 154, 39. (b) Demik, N. N.; Kabachnik, M. M.; Novikova, Z. S.;
Beletskaya, I. P. Russ. J. Org. Chem. 1995, 31, 57. (c) Zhang, H. C.;
Daves, G. D. Organometallics 1993, 12, 1499. (d) Bumagin, N. A.; More,
P. G.; Beletskaya, I. P. J. Organomet. Chem. 1989, 371, 397. .
(11) (a) Shimada, T.; Nakamura, I.; Yamamoto, Y. J. Am. Chem.
Soc. 2004, 10546-10547. (b) Danishefsky, S.; Taniyama, E. Tetrahe-
dron Lett. 1983, 24, 15.
(13) It is quite possible that Bredt’s rule applies to this system which
would prohibit olefin isomerization to either “bridgehead” position.
Shea, K. J. Tetrahedron 1980, 1683-1715 and references therein.
(14) (a) Clive, D. L. J.; Daigneault, S. J. Org. Chem. 1991, 56,
5-5289. (b) Andrieux, J.; Barton, D. H. R.; Patin, H. J. Chem. Soc.,
Perkin Trans. 1 1977, 359. (c) Grieco, P. A.; Nishizawa, M.; Marinovic,
N.; Ehmann, W. J. J. Am. Chem. Soc. 1976, 98, 7102.
(15) (a) Baskin, J. M.; Wang, Z. Tetrahedron Lett. 2002, 8479-8483.
(b) Suzuki, H.; Abe, H. Tetrahedron Lett. 1995, 36, 6239-6242.
270 J. Org. Chem., Vol. 70, No. 1, 2005

Asymmetric Synthesis of a Prostaglandin D2 Receptor Antagonist
TABLE 1. Isomerization of 20 with RhCl3
time (h)
21 (% ee)/22 (% ee)
1
5
18
50 (96):50 (96)
80 (65):20(96)
100 (55):0
SCHEME 7. Proposed Mechanism for RhCl3 Isomerization of 20
When conversion and enantioselectivity were monitored
over time, a new intermediate, determined by NMR to
be 22, appeared within 1 h of the start of the reaction
and slowly converted to indole 21. Additionally, the
enantiomeric excess of indole product 21 after 1 h was
very high (96% ee) but slowly degraded over the course
of the reaction. No racemization was observed when
optically pure indole 21 was subjected to the reaction
conditions, so we presume that the loss of enantioselec-
tivity was the result of conversion of intermediate 22 to
21.
These data suggest that two reaction pathways were
operating in the isomerization of 20 to 21; one pathway
that rapidly produced the indole with no racemization
and another pathway that slowly produced the indole via
intermediate 22 with concomitant loss of enantiomeric
integrity (Scheme 7). In accord with the literature, the
isomerization of 20 with RhCl3 is a likely kinetic process
that indiscriminately isomerizes the olefin in a “clock-
wise” or “counterclockwise” fashion.14 Isomerization of the
olefin in a “counterclockwise” direction affords intermedi-
ate A, which was not detectable, but presumably under-
goes rapid isomerization to indole 21, according to
literature precedent.16 Molecular modeling analysis of
these intermediates at the PM3 level shows the following
relative energy levels: 20 > intermediate A > 21, which
supports our proposal.
Isomerization of the olefin in the “clockwise” direction
delivers intermediate 22, which was observed in the
reaction and appears to be on the pathway which results
in racemization. The buildup of 22 during the reaction
suggests that further isomerization in the “clockwise”
direction is impaired. According to molecular modeling,
intermediate B, the next species on the pathway of
“clockwise” isomerization, is much higher in energy than
20. This calculated activation barrier would explain the
buildup of 22 during the reaction. Under the acidic
reaction conditions, intermediate B would rapidly isomer-
ize via protonation/deprotonation of the enamine/iminium
species to the more stable indole 21. Despite the proximal
stereogenic center, the protonation sequence appears to
be nonselective, resulting in the production of racemic
indole 21.
An alternative to the isomerization of intermediate 20
was to reduce the substrate to the fully saturated indoline
23 and subsequently oxidize to the indole (Scheme 8).
Reduction was accomplished with Crabtree’s catalyst in
THF to afford the indoline 23 in 90% yield.17 Oxidation
of the indoline with a variety of typical reagents (DDQ,18
(16) (a) Forbes, I. T. Tetrahedron Lett. 1999, 9293-9295. (b) Tietze,
L. F.; Buhr, W. Angew. Chem., Int. Ed. Engl. 1995, 1366-1368. (c)
Kaufman, M. D.; Greico, P. A. J. Org. Chem. 1994, 59, 7-7198. (d)
Simoji, Y.; Saito, F.; Tomita, K.; Morisawa, Y. Heterocycles 1991, 2389-
2397.
(17) Alternatively, Wilkinson’s catalyst was also found to be effec-
tive; however hydrogenation with Pd/C was found to produce isomer-
ization products with some degradation of enantiomeric excess.
J. Org. Chem, Vol. 70, No. 1, 2005 271

Campos et al.
MHz, CDCl3) 7.36 (d, 2H, J ) 7.6 Hz), 6.99 (br s, 1 H), 5.82
(m, 2H), 3.26 (m, 1H), 2.51 (dd, 1H, J ) 14.4, 6.8 Hz), 2.40 (m,
3H), 2.32 (m, 1H), 1.60 (m, 1H); 13C NMR (400 MHz, CDCl3)
δ 29.7, 31.9, 42.4, 119.4, 119.7, 124.1, 124.2, 131.4, 131.6, 133.7,
159.1, 161.7, 170.6; TLC Rf ) 0.25 (60% hexane, 40% EtOAc).
Anal. Calcd for C13H12Br2FNO: C, 41.41; H, 3.21; N, 3.71.
Found: C, 41.42; H, 3.04; N, 3.68. Separation of enantiomers
was accomplished by LC analysis with supercritical CO2
(Chiralcel OJ, 1.5 mL/min, 200 bar, 4-40% methanol, at 2%/
min, tR ) 10.2, 10.5 min).
SCHEME 8. Endgame Chemistry
(3aR,6S,6aS)-1-(2,6-Dibromo-4-fluorophenyl)-6-iodo-
hexahydrocyclopenta[b]pyrrol-2(1H)-one (12). To a solu-
tion of amide 11 (33.4 g, 0.089 mol) in dioxane (1.8 L) were
added MeOH (500 mL) and 5 M NaOH(aq). The mixture was
stirred for 10 min, and then N-iodosuccinimide (54 g, 0.24 mol)
was added in one portion and stirred for 3 h. The reaction was
quenched with 1 M Na2S2O3 (1 L) and extracted from the
organic with MTBE (500 mL). The aqueous layer was back-
extracted twice with MTBE (500 mL), and the combined
organics were dried over Na2SO4 and concentrated. The crude
reaction mixture was used directly in the next step. The assay
yield of 12, based on a purified standard of 12, was 39.3 g
(88%). A portion of the material was purified by flash chro-
matography (85:15 hexane/EtOAc) for characterization pur-
NBS,19 Pd/C,20 salcomine21) proved unsuccessful; how-
ever, oxidation with MnO2 in benzene at 50 °C produced
the indole 21 in 80% yield. Benzylation of 21 with
4-chlorobenzyl chloride and Cs2CO3 followed by hydroly-
sis in the same pot afforded the PGD2 receptor antagonist
1 in 90% isolated yield.
This synthetic route provided PGD2 receptor antagonist
1 in 12 steps and 23% overall yield (96% ee) with no loss
of enantiomeric purity over the entire sequence of reac-
tions. Key points in the synthesis are early stage intro-
duction of the stereogenic center via Trost’s asymmetric
allylic alkylation, Pd-mediated cyclization/Heck reaction
to afford the entire carbon skeleton in one step, and late-
stage oxidation of a fully functionalized indoline to the
indole. Efforts to further improve the synthesis of 1 are
ongoing.
poses: [R]23 -29.2 (c ) 0.0082, CH2Cl2, 96% ee); IR (film)
D
3074, 2959, 1714, 1585, 1566, 1461 cm-1; 1H NMR (400 MHz,
CDCl3) 7.44 (dd, 1H, J ) 7.6, 2.8 Hz), 7.42 (dd, 1H, J ) 7.6,
2.8 Hz), 5.00 (dd, 1 H, J ) 8.0, 2.0 Hz), 4.36 (dt, 1H, J ) 4.4,
2.0 Hz), 3.14 (m, 1H), 2.91 (dd, 1H, J ) 18.0, 10.8 Hz), 2.44
(m, 2H), 2.37 (dd, 1 H, J ) 18.0, 4.4 Hz), 2.09 (m, 1H), 1.72
(m, 1H); 13C NMR (400 MHz, CDCl3) δ 28.6, 33.3, 34.3, 36.7,
37.4, 75.1, 120.3, 120.5, 120.6, 120.8, 123.8, 123.9, 126.2, 160.0,
162.6, 173.8; TLC Rf ) 0.25 (70% hexane, 30% EtOAc). Anal.
Calcd for C13H11Br2FINO: C, 31.05; H, 2.20; N, 2.78. Found:
C, 31.19; H, 2.02; N, 2.70.
(3aR,6aS)-1-(2,6-Dibromo-4-fluorophenyl)-3,3a,4,6a-tet-
rahydrocyclopenta[b]pyrrol-2(1H)-one (14). To a solution
of crude iodolactamization product containing 12 (39.3 g assay,
0.078 mol) in toluene (400 mL) was added DBU (36 mL, 0.234
mol), and the mixture was heated to reflux. The reaction
mixture was stirred for 18 h, cooled to rt, diluted with toluene
(200 mL), and washed with 1 M HCl (300 mL). The layers were
separated, and the organic layer was dried over Na2SO4. The
crude oil was purified by flash chromatography to yield 14
Experimental Section
2-[(1R)-Cyclopent-2-en-1-yl]-N-(2,6-dibromo-4-fluorophe-
nyl)acetamide (11). To a solution of (R)-2-cyclopentene-1-
acetic acid (8, 52.4 g, 0.415 mol) and DMF (1 mL) in CH2Cl2
(500 mL) at 4 °C was added oxalyl chloride (188 mL, 1.76 mol)
over 45 min such that gas evolution from the reaction was
controlled. Upon complete addition, the reaction mixture was
aged for 30 min and then slowly warmed to rt. The reaction
mixture was aged at rt for 30 min and checked for conversion
(GC). The reaction mixture was concentrated by distillation
to remove CH2Cl2 and excess oxalyl chloride. The resulting
crude acid chloride was added via addition funnel to a solution
of 2,6-dibromo-4-fluoroaniline (7, 112 g, 0.415 mol) in N,N-
dimethylacetamide (500 mL) at 4 °C. The reaction mixture was
aged for 4 h at which time the reaction was complete by HPLC
analysis. The reaction mixture was diluted with MTBE (2 L)
and washed with 1.7% NH4Cl(aq) solution (1.7 L) and then with
water (1.7 L). The organic layer was dried over Na2SO4 and
concentrated. The resulting crude oil was recrystallized from
70:30 hexane/acetone to yield 101 g of 11. A second crop of 11
(18.9 g) was recovered from the mother liquors to afford a total
(28.6 g, 98%) as a solid: mp 91.5-93.0 °C; [R]23 -88.0 (c )
D
0.0084, CH2Cl2, 96% ee); IR (film) 3064, 2915, 2848, 1707,
1585, 1565, 1464, 1386, 1230 cm-1; 1H NMR (400 MHz, CDCl3)
7.40 (dd, 1H, J ) 5.2, 2.8 Hz), 7.39 (dd, 1H, J ) 5.2, 2.8 Hz),
5.98 (m, 1H), 5.72 (dq, 1H, J ) 5.6, 2.0 Hz), 5.04 (m, 1H), 3.20
(m, 1H), 2.85 (dd, 1H, J ) 17.5, 10 Hz), 2.79 (m, 1H), 2.43 (dd,
1H, J ) 17.6, 6.4 Hz), 2.38 (m, 1H); 13C NMR (400 MHz, CDCl3)
δ 33.9, 38.0, 39.2, 70.4, 120.0, 120.2, 120.5, 124.3, 125.9, 126.0,
128.7, 134.3, 162.3, 173.5. TLC Rf ) 0.25 (60% hexane, 40%
EtOAc). Anal. Calcd for C13H10Br2FNO: C, 41.63; H, 2.69; N,
3.73. Found: C, 41.44; H, 2.47; N, 3.70. Separation of enan-
tiomers was accomplished by LC analysis with supercritical
CO2 (Chiralpak AS, 1.5 mL/min, 200 bar, 4-40% methanol,
at 2%/min, tR ) 9.3, 9.9 min).
(2aS,9bS,9cS)-6-Bromo-8-fluoro-2a,3,9b,9c-tetrahydro-
4H-benzo[b]cyclopenta[gh]pyrrolizin-4-one (16). To a
solution of amide 14 (3.64 g, 9.71 mmol) in DMF (49 mL) was
added water (5.5 mL) followed by Pd(OAc)2 (433 mg, mmol)
and K2CO3 (3.4 g, mmol). The reaction mixture was heated to
100 °C and stirred for 12 h. The reaction mixture was diluted
with EtOAc (150 mL) and washed with 1 M HCl (150 mL).
The resulting slurry was filtered over Celite, and then the
layers were separated. The organic layer was washed twice
with water (50 mL) and subsequently dried over Na2SO4 and
concentrated. The resulting crude oil was purified by flash to
of 120 g, 76% isolated yield of 11: mp ) 175.6-176.4 °C; [R]23
D
-26.6 (c ) 0.0086, 96% ee, CH2Cl2); IR (film) 3213, 3163, 3002,
2918, 2850, 1653, 1589, 1569, 1522, 1460 cm-1; 1H NMR (400
(18) No reaction was observed when 20 was treated with DDQ. Lee,
S.; Lim, H.-J.; Cha, K.; Sulikowski, G. A. Tetrahedron 1997, 53, 16521-
16532.
(19) Treatment with NBS afforded decomposition, which was the
result of overoxidation of 20. Ko, C.-W.; Chou, T. J. Org. Chem. 1998,
63, 4645-4653.
(20) No oxidation was observed using Pd/C in a variety of solvents.
Pelcman, B.; Gribble, G. W. Tetrahedron Lett. 1990, 31, 2381-2384.
(21) No oxidation was observed. Inada, A.; Nakamura, Y.; Morita,
Y. Chem. Lett. 1980, 12.
afford 16 (2.28 g, 80%) as a solid: mp 110.7-112.4 °C; [R]23
D
+59.6 (c ) 0.0072, CH2Cl2, 96% ee); IR (film) 3076, 3050, 2946,
1721, 1606, 1587, 1468, 1423, 1247 cm-1; 1H NMR (400 MHz,
272 J. Org. Chem., Vol. 70, No. 1, 2005

Asymmetric Synthesis of a Prostaglandin D2 Receptor Antagonist
CDCl3) 7.12 (dd, 1H, J ) 8.8, 2.4 Hz), 6.91 (ddd, 1H, J ) 7.6,
2.4, 0.4 Hz), 5.69 (dt, 1H, J ) 5.6, 1.6 Hz), 5.64 (dt, 1H, J )
5.6, 2.0 Hz), 5.25 (t, 1H, J ) 6.0 Hz), 4.08 (m, 1H), 3.51 (m,
1H), 3.12 (dd, 1H, J ) 16.0, 7.6 Hz), 2.44 (d, 1H, J ) 16.0 Hz);
13C NMR (400 MHz, CDCl3) 39.7, 44.2, 54.3, 72.2, 110.9, 111.1,
111.6, 118.7, 118.9, 130.3, 133.3, 133.9, 137.4, 140.5, 140.6,
159.0, 161.5, 175.8; TLC Rf ) 0.25 (65% hexane, 35% EtOAc).
Anal. Calcd for C13H9BrFNO: C, 53.09; H, 3.08; N, 4.76.
Found: C, 54.06; H, 2.85; N, 4.77.
crude product was assayed by HPLC for 20 using a sample
purified by recrstallization from 70:30 hexane/EtOAc (1.42 g,
96%). The crude material was used directly in the next
reaction; however, a small sample was purified by recrystal-
lization (70:30 hexane/EtOAc) for characterization purposes:
mp 118.0-119.4 °C; [R]23D -55.2 (c ) 0.0075, CH2Cl2, 96% ee);
1
IR (film) 3403, 2909, 2844, 1731, 1476, 1295 cm-1; H NMR
(400 MHz, CDCl3) 7.12 (dd, 1H, J ) 8.4, 2.0 Hz), 7.04 (d, 1H,
J ) 7.6 Hz), 5.70 (m, 2H), 5.53 (br s, 1 H), 4.85 (m, 1H), 4.45
(d, 1H, J ) 8.4 Hz), 3.76 (s, 3H), 3.36 (m, 1H), 2.99 (s, 3H),
2.60 (dd, 1H, J ) 16.0, 5.6 Hz), 2.53 (dd, 1H, J ) 16.0, 10.4
Hz); 13C NMR (400 MHz, CDCl3) 35.0, 42.5, 46.7, 51.9, 53.8,
64.2, 111.4, 111.7, 117.5, 117.8, 130.1, 133.9, 135.3, 145.8,
173.2; TLC Rf ) 0.30 (70% hexane, 30% EtOAc). Anal. Calcd
for C15H16FNO4S: C, 55.37; H, 4.96; N, 4.31. Found: C, 55.63;
H, 4.75; N, 4.25. Separation of enantiomers was accomplished
by HPLC analysis (Chiralcel AD, 1.0 mL/min, 90% hexane,
10% 2-propanol, tR ) 12.4, 13.2 min).
[(3S,3aS,8bS)-5-Bromo-7-fluoro-3,3a,4,8b-tetrahydro-
cyclopenta[b]indol-3-yl]acetic Acid (17). To a solution of
amide 16 (1.73 g, 5.88 mmol) in MeOH (19 mL) was added 3
M KOH(aq) (8.5 mL), and the mixture was heated to 50 °C
overnight. The next day, the solution was cooled to room
temperature, concentrated, and then diluted with MTBE (100
mL). The resulting solution was quenched with 3 M HCl (30
mL) and diluted with water (50 mL). The layers were sepa-
rated, and the organic solution was washed once with water
(50 mL). The organic layer was dried over Na2SO4 and
concentrated to yield 17 as a solid which was suitable for
further reaction (1.77 g, 92%). A small sample was purified
by flash chromatography for characterization purposes: mp
Methyl [(3R,3aS,8bS)-7-Fluoro-5-(methylsulfonyl)-1,2,3,-
3a,4,8b-hexahydrocyclopenta[b]indol-3-yl]acetate (23).
To a solution of 20 (937 mg assay, 2.88 mmol) in THF (10 mL)
was added Crabtree’s catalyst (220 mg, 10 mol %), and then
the mixture was evacuated and purged with H2 five times. The
reaction mixture was stirred overnight, concentrated to dry-
ness, and purified by flash chromatography (70:30 hexane/
EtOAc to 60:40 hexane/EtOAc) to yield 23 as a solid (890 mg,
95%): mp 95.3-96.4 °C; [R]23D -59.8 (c ) 0.0101, CH2Cl2, 96%
172.4-173.5 °C; [R]23 -102.0 (c ) 0.0049, CH2Cl2, 96% ee);
D
1
IR (film) 3383, 3056, 1696, 1588, 1479, 1211 cm-1; H NMR
(400 MHz, CDCl3) 6.93 (ddd, 1H, J ) 8.4, 2.4, 0.4 Hz), 6.82
(ddd, 1H, J ) 7.6, 2.4, 0.8 Hz), 5.69 (m, 1H), 4.79 (dd, 1H, J )
8.0, 6.4 Hz), 4.49 (m, 1H), 3.31 (m, 1H), 2.81 (dd, 1H, J ) 16.0,
10.4 Hz), 2.69 (dd, 1H, J ) 16.0, 4.8 Hz); 13C NMR (400 MHz,
CDCl3) 34.5, 46.5, 56.0, 63.7, 110.3, 110.6, 116.4, 116.6, 130.6,
131.6, 131.7, 133.1, 178.4; TLC Rf ) 0.25 (50% hexane, 50%
EtOAc). Anal. Calcd for C13H11BrFNO2: C, 50.02; H, 3.55; N,
4.49. Found: C, 49.69; H, 3.33; N, 4.15. Separation of enan-
tiomers was accomplished by LC analysis with supercritical
CO2 (Chiralpak AD 1.5 mL/min, 200 bar, 4-40% (25 mM
iBuNH2 in MeOH), tR ) 13.5, 13.8 min).
ee); IR (film) 3408, 2951, 2848, 1733, 1479, 1298, 1133 cm-1
;
1H NMR (400 MHz, CDCl3) 7.10 (dd, 1H, J ) 8.4, 2.4 Hz), 6.95
(dd, 1H, J ) 8.0, 2.0 Hz), 5.40 (s, 1H), 4.53 (sept, 1H, J ) 3.2
Hz), 3.89 (t, 1H, J ) 8.8 Hz), 3.72 (s, 3H), 3.01 (s, 3H), 2.53
(dd, 1H, J ) 15.6, 8.8 Hz), 2.47 (dd, 1H, J ) 15.6, 4.4 Hz),
2.30 (dsept, 1H, J ) 6.0, 2.8 Hz), 2.04 (dsept, 1H, J ) 6.4, 2.8
Hz), 1.79 (m, 2H), 1.22 (dq, 1H, J ) 12.4, 6.8 Hz); 13C NMR
(400 MHz, CDCl3) 29.9, 33.9, 34.4, 42.4, 42.6, 46.2, 51.7, 65.8,
111.2, 111.4, 116.1, 117.9, 118.2, 138.9, 139.0, 146.9, 153.7,
156.1, 173.2; TLC Rf ) 0.25 (60% hexane, 40% EtOAc). Anal.
Calcd for C15H18FNO4S: C, 55.03; H, 5.54; N, 4.28. Found: C,
54.99; H, 5.42; N, 4.23. Separation of enantiomers was
accomplished by HPLC analysis (Chiralcel AD, 1.0 mL/min,
90% hexane, 10% 2-propanol, tR ) 11.7, 12.2 min).
[(3S,3aS,8bS)-7-Fluoro-5-(methylsulfonyl)-3,3a,4,8b-
tetrahydrocyclopenta[b]indol-3-yl]acetic Acid (19). To a
solution of bromo acid 17 (1.61 g, 5.16 mmol) in DMSO (120
mL) were added CuI (3.91 g, 20.6 mmol), 5 M NaOH (1.2 mL),
and then NaSO2Me (2.10 g, 20.6 mmol). The reaction mixture
was evacuated and purged three times and then heated to 120
°C. The reaction mixture was stirred at 120 °C overnight,
cooled to rt, diluted with EtOAc (250 mL) and water (250 mL),
and quenched with 3 N HCl (8 mL). The resulting mixture
was filtered over Celite, and then the layers were separated.
The organic layer was washed twice with water (100 mL), dried
over Na2SO4, and concentrated. The crude product was assayed
by HPLC for 19 based on a purified standard of 19 (1.42 g,
88%). The crude material was used directly in the next
reaction; however, a small sample was purified for character-
ization purposes: mp 173.5-175.2 °C; [R]23D -54.5 (c ) 0.0051,
THF, 96% ee); IR (film) 3399, 3015, 2952, 1701, 1476, 1434,
1293, 1133 cm-1; 1H NMR (400 MHz, THF-d8) 7.14 (dd, 1H, J
) 8.0, 2.0 Hz), 7.06 (dd, 1H, J ) 8.8, 2.4 Hz), 5.77 (m, 1H),
5.71 (m, 2H), 4.84 (m, 1H), 4.45 (d, 1H, J ) 8.4 Hz), 3.30 (q,
1H, J ) 7.2 Hz), 2.92 (s, 3H), 2.52 (m, 2H); 13C NMR (400 MHz,
THF-d8) 34.5, 41.4, 47.0, 53.7, 64.3, 110.8, 111.1, 116.7, 117.0,
117.6, 129.7, 134.0, 135.7, 135.8, 145.9, 153.4, 155.7, 175.0;
TLC Rf ) 0.25 (95% CH2Cl2, 5% MeOH). Anal. Calcd for C14H14-
FNO4S: C, 54.01; H, 4.53; N, 4.50. Found: C, 53.94; H, 4.33;
N, 4.37. Separation of enantiomers was accomplished by LC
analysis with supercritical CO2 (Chiralcel OBH, 1.5 mL/min,
200 bar, 4-40% methanol, at 2%/min, tR ) 11.9, 14.4 min).
Methyl [(3R)-7-Fluoro-5-(methylsulfonyl)-1,2,3,4-tet-
rahydrocyclopenta[b]indol-3-yl]acetate (21). To a solution
of 23 (738 mg assay, 2.24 mmol) in benzene (10 mL) was added
MnO2 (4.9 g, 0.056 mol), then the mixture was heated to 50
°C overnight. The next day, the reaction mixture was cooled
to rt, filtered over Celite, and then washed with EtOAc. The
resulting solution was concentrated to dryness, and the
product was purified by flash chromatography (50:50 hexane/
EtOAc) to yield 21 as a solid (582 mg, 80%): mp 141.2-143.4
°C; [R]23 -96.2 (c ) 0.0073, CH2Cl2, 96% ee); IR (film) 3414,
D
2953, 2849, 1718, 1575, 1476, 1305, 1129 cm-1; 1H NMR (400
MHz, CDCl3) 9.79 (br s, 1 H), 7.36 (m, 2H), 3.79 (s, 3H), 3.62
(m, 1H), 3.14 (s, 3H), 2.95-2.72 (m, 4H), 2.57 (dd, 1H, J )
16.8, 10.8 Hz), 2.17 (m, 1H); 13C NMR (400 MHz, CDCl3) 23.1,
35.2, 35.7, 39.1, 44.7, 52.0, 108.1, 108.4, 110.2, 110.5, 119.7,
121.9, 122.0, 126.8, 126.9, 133.2, 150.0, 155.1, 157.4, 173.4;
TLC Rf ) 0.25 (50% hexane, 50% EtOAc). Anal. Calcd for
C15H16FNO4S: C, 55.37; H, 4.96; N, 4.31. Found: C, 55.37; H,
4.76; N, 4.25. Separation of enantiomers was accomplished by
HPLC analysis (Chiralcel AD, 1.0 mL/min, 90% hexane, 10%
2-propanol, tR ) 10.7, 12.5 min).
[(3R)-4-(4-Chlorobenzyl)-7-fluoro-5-(methylsulfonyl)-
1,2,3,4-tetrahydrocyclopenta[b]indol-3-yl]acetic Acid (1).
To a solution of 21 (500 mg assay, 1.54 mmol) in DMAc (5
mL) at rt were added Cs2CO3 (1.5 g, 4.61 mmol) and 4-chlo-
robenzyl bromide (379 mg, 1.85 mmol). The reaction mixture
was stirred for 3 h, 5 N NaOH (1.23 mL) was added to the
reaction mixture, and the mixture was heated to 50 °C for 2
h. The reaction mixture was cooled to 0 °C, and 5 N HCl (3.0
mL) was added slowly, keeping the temperature below 15 °C.
The reaction mixture was diluted with water (5 mL) and
Methyl [(3S,3aS,8bS)-7-Fluoro-5-(methylsulfonyl)-3,-
3a,4,8b-tetrahydrocyclopenta[b]indol-3-yl]acetate (20).
To a solution of acid 19 (1.42 g assay, 4.57 mmol) in MeOH
(20 mL) was added concd H2SO4 (15 drops), and the mixture
was heated to 50 °C. The reaction mixture was stirred for 2 h,
cooled to rt, and concentrated. The crude reaction mixture was
diluted with EtOAc (50 mL), washed with 1 M NaHCO3 (50
mL), and then washed twice with water (50 mL × 2). The
organic layer was dried over Na2SO4 and concentrated. The
J. Org. Chem, Vol. 70, No. 1, 2005 273

Campos et al.
extracted with isopropyl acetate (10 mL). The organic layer
was dried over Na2SO4 and concentrated to dryness. The
product was purified by flash chromatography (50:50 hexane/
177.0. Anal. Calcd for C21H19ClFNO4S: C, 57.86; H, 4.39; N,
3.21. Found: C, 57.81; H, 4.33; N, 3.24. Separation of enan-
tiomers was accomplished by HPLC analysis (Chiralcel OJR,
1.5 mL/min, 99.9% MeOH, 0.1% trifluoroacetic acid, tR 2.5, 3.3
min).
EtOAc) to yield 1 as a solid (604 mg, 90%): mp 175 °C; [R]23
D
1
-29.3 (c ) 1, MeOH, 99.8% ee); H NMR (400 MHz, CDCl3)
7.65 (dd, 1H, J ) 9.2, 2.4 Hz), 7.42 (dd, 1H, J ) 8.0, 2.8 Hz),
7.23 (m, 2H), 6.67 (m, 2H), 6.14 (d, 1H, J ) 17.6 Hz), 5.65 (d,
1H, J ) 17.6 Hz), 3.47 (m, 1H), 2.95 (m, 1H), 2.82 (m, 2H),
2.71 (s, 3H), 2.61 (dd, 1H, J ) 16.0, 2.0 Hz), 2.45 (dd, 1H, J )
16.0, 10.4 Hz), 2.31 (m, 1H); 13C NMR (400 MHz, CDCl3) 22.8,
35.4, 35.8, 38.4, 44.4, 50.1, 110.7, 111.0, 112.6, 112.9, 121.4,
125.9, 126.6, 129.2, 132.5, 133.4, 137.3, 151.8, 154.6, 157.1,
Supporting Information Available: Experimental pro-
cedures and characterization data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
JO048305+
274 J. Org. Chem., Vol. 70, No. 1, 2005