Practical asymmetric synthesis of a selective endothelin A receptor (ETA) antagonist

Article abstract of DOI:10.1021/ol016601s

Equation presented A practical, chromotography-free asymmetric synthesis was developed for the large scale preparation of an endothelin receptor antagonist 2. This synthesis includes a new efficient process for the preparation of 6-bromo-2,3-dihydrobenzof

Full text of DOI:10.1021/ol016601s

ORGANIC
LETTERS
2001
Vol. 3, No. 21
3357-3360
Practical Asymmetric Synthesis of a
Selective Endothelin A Receptor (ETA)
Antagonist
Zhiguo J. Song,* Matthew Zhao, Lisa Frey, Jing Li, Lushi Tan, Cheng Y. Chen,
David M. Tschaen, Richard Tillyer, Edward J. J. Grabowski, Ralph Volante, and
Paul J. Reider
Department of Process Research, Merck Research Laboratories, P.O. Box 2000,
Rahway, New Jersey 07065
Yoshiaki Kato, Shigemitsu Okada, Takayuki Nemoto, Hiroki Sato,
Atsushi Akao, and Toshiaki Mase
Process R & D, Banyu Pharmaceutical Co. Ltd., Okazaki, Aichi 4440858, Japan
zhiguo_song@merck.com
Received August 16, 2001
ABSTRACT
A practical, chromotography-free asymmetric synthesis was developed for the large scale preparation of an endothelin receptor antagonist 2.
This synthesis includes a new efficient process for the preparation of 6-bromo-2,3-dihydrobenzofuran, a stereoselective conjugate addition of
an aryllithium followed by stereospecific addition of the Grignard reagent of the top aryl bromide, and an aminophosphate-mediated sterospecific
intramolecular enolate alkylation, which led to the formation of the five-membered ring bearing three contiguous asymmetric centers.
Endothelin receptor antagonists are currently being evaluated
as useful therapeutic agents for the treatment of hypertension,
congestive heart failure, and renal diseases.1,2 We recently
reported an efficient large scale synthesis of an early drug
candidate 1.3 In this paper, we report the development of a
large scale, economical synthesis of a similar compound 2,
which is under development as an endothelin A receptor
selective antagonist. Although structurally similar to 1,
compound 2 offers some unique and significant synthetic
challenges, which are described herein. Our approach to the
synthesis of the core ring system of compound 2 is similar
to the previously reported synthesis for 1. As outlined in
Scheme 1, the key steps include a Michael addition of the
bottom arylmetal to the Michael acceptor and addition of
top arylmetal to the Michael addition product, followed by
cyclization to form the five-membered ring with correct
stereochemistry. To accomplish this goal, we needed to
develop an efficient synthesis for the top aryl fragment as
6-bromo-2,3-dihydrobenzofuran and an efficient method to
introduce the isopropyl amino group to the pyridine ring. In
the process of realizing this strategy, we found that substan-
tial modifications to the known chemistry were required for
the strategy to be successful.3
(1) (a) Astles, P. C.; Brown, T. J.; Halley, F.; Handscombe, C. M.; Harris,
N. V.; Majid, T. N.; McCarthy, C.; McLay, I. M.; Morley, A.; Porter, B.;
Roach, A. G.; Sargent, C.; Smith, C.; Walsh, R. J. A. J. Med. Chem. 2000,
43, 90 and references therein. (b) Ishikawa, K.; Nagase, T.; Mase, T.,
Hayama, T.; Ihara, M.; Nishikibe, M.; Yano, M. PCT Int. Appl. WO
9505374, 1995.
(2) Clark, W. M. Curr. Opin. Drug DiscoVery DeV. 1999, 2, 565.
(3) (a) Song, Z. J.; Zhao, M.; Desmond, R.; Devine, P.; Tschaen, D. M.;
Tillyer, R.; Frey, L.; Heid, R.; Xu, F.; Foster, B.; Li, J.; Reamer, R.; Volante,
R.; Grabowski, E. J.; Dolling, U.-H.; Reider, P. J.; Okada, S.; Kato, Y.;
Mano, E. J. Org. Chem. 1999, 64, 9658-9667. (b). Devine, Paul N.;
Desmond, R.; Frey, L.; F.; Heid, R. M.; Song, Z.; Tillyer, R. D.; Tschaen,
D. M.; Zhao, M.; Kato, Y.; Mano, E.; Okada, S.; Kato, S.; Mase, T. Yuki
Gosei Kagaku Kyokaishi 1999, 57, 1016-1025
10.1021/ol016601s CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/27/2001

butoxide and ethylene glycol, compound 7 was isolated
directly from the reaction mixture as a solid in 87% yield
upon addition of water. Conversion of the alcohol 7 to the
corresponding bromide 8 was carried out in toluene with
PBr3, and the bromide was used without purification.5b
Treatment of the bromide with n-butyllithium led to lithium-
bromide exchange followed by intramolecular alkylation of
the resulting aryllithium to give the desired 6-bromo-2,3-
dihydrobenzofuran 5 in 80% yield from the alcohol 7.5c This
compound was isolated from methanol-water as a crystalline
solid.5d
Scheme 1
Preparation of the Michael Reaction Acceptor 12. The
key Michael acceptor 12 was synthesized according to
Scheme 3. Starting with the readily available 2,6-dibromo-
pyridine, direct displacement of one of the bromides with
lithium benzylisopropyl amide in toluene went smoothly to
give the aminopyridine 10.6a,7 Interestingly, the desired
product was not formed in THF as reaction solvent. Formy-
lation with Vilsmeier’s reagent generated from DMF/POCl3
gave the aldehyde 11 with the desired regiochemistry. The
side reaction in this step was the displacement of the bromide
by the chloride, which was minimized by using excess (4
equiv) Vilsmer’s reagent and lower temperature (35-40 °C).
Bromoaldehyde 11 was converted into the tert-butyl ester
12 in 85% yield via a Heck reaction with 1.05 equiv of tert-
butyl acrylate in N,N-dimethyl acetamide with NaOAc‚3H2O
and PdCl2(dppf) (CH2Cl2 complex) as the catalyst at 80 °C.6b
Compounds 10-12 were oils and therefore were not purified.
Synthesis of 6-Bromo-2,3-dihydrobenzofuran (5). As
shown in Scheme 2, the commonly used synthetic route to
Scheme 2
Conjugate Addition. To introduce the bottom chiral
center in 2, we envisioned the strategy of using chiral amino
alcohol or diamine auxiliary to form the intermediate like
13, which could undergo diastereoselective conjugate addi-
tion to the unsaturated ester by the bottom arylmetal species.
On the basis of our earlier experience with conjugate
additions of this type, we expected the addition of the
aryllithium to intermediate such as 13 would be straightfor-
ward.3,4 However, we found that extensive screening was
necessary to find the right protecting groups on the substrates,
the right chiral auxiliary, and conditions for the reaction to
proceed with high yield and diastereoselectivity. As shown
in Scheme 3, the best combination we found was N-benzyl
protection for the amino substituent of 12 along with the
trityl protecting group on the alcohol of the bottom aryl
compound (Ar2)8 and (S,S)-pseudoephedrine as the auxiliary.
The bottom aryllithium (Ar2Li) was prepared at low tem-
perature (typically e50 °C), and the enoate 13 (prepared by
reacting 12 with (S,S)-pseudoephedrine) was added to the
aryllithium solution at temperatures typically lower than -50
°C. After the auxiliary was removed with citric acid, the
this compound is from 3-bromophenol, relying on a non-
regioselective Friedel-Crafts cyclization of 3 to 4 followed
by low yielding reduction (due to debromination).5a We
desired a practical synthesis of this compound that obviated
both of these problems. Thus, when readily available 1,4-
dibromo-2-fluorobenzene (6) was treated with potassium tert-
(4) (a) Frey, L. F.; Tillyer, R. D.; Caille, A.-S.; Tschaen, D. D.; Dolling,
U.-H.; Grabowski, E. J. J.; Reider, P. J. J. Org. Chem. 1998, 63, 3120. (b)
Alexakis, A.; Sedrani, R.; Mangeney, P.; Normant, J. F. Tetrahedron Lett.
1988, 29, 411.
(5) (a) Paper from Banyu Medicinal Chemistry, manuscript in preparation.
(b) The reaction was carried out at 90 °C with 0.45 equiv of PBr3. After 2
h, an additional 0.1 equiv of PBr3 and 0.3 equiv of water were added to get
complete conversion. (c) Bradsher, C. K.; Reames, D. C. J. Org. Chem.
1981, 46,1384. (d) Data for 5: white crystals, mp 42-45 °C; 1H NMR
(250 MHz, CDCl3, ppm) 3.15 (t, 2H, J ) 8.8 Hz), 4.58 (t, 2H, J ) 8.8
Hz), 6.9-7.1 (m, 3H); 13C NMR (100 MHz, CDCl3, ppm) 161.1, 126.2,
125.8, 123.3, 120.8, 112.9, 71.9, 29.3; HPLC 95% pure by area.
(6) (a) Data for 10: colorless oil; 1H NMR (250 MHz, CDCl3, ppm)
1.21 (d, J ) 6.7 Hz, 6 H), 4.52 (s, 2H), 5.11 (hep, J ) 6.7 Hz, 1H), 6.18
(d, J ) 8.4 Hz, 1H), 6.0.68 (d, H)7.4 Hz, 1H), 7.14 (dd, J ) 8.4, 7.4 Hz,
1H), 7.2-7.4 (m, 5H). (b) Data for 12: 1H NMR (250 MHz, CDCl3, ppm)
1.25 (d, 6H, J ) 6.7 Hz), 1.54 (s, 9H), 4.68 (s, 2H), 5.35 (br, m, 1H), 6.41
(d, J ) 10 Hz, 1H), 6.89 (d, J ) 15 Hz, 1H), 7.15-7.36 (m, 5H), 7.83 (d,
J ) 10 Hz, 1H), 8.29 (d, J ) 15 Hz, 1H), 10.20 (s 1H).
(7) Satisfactory 13C NMR and elemental analysis results were obtained.
(8) The trityl ether (Ar2Br) was made from 2-bromo-5-methoxybenzyl
chloride and Ph3COH in the presence of t-BuOK.
3358
Org. Lett., Vol. 3, No. 21, 2001

Scheme 3
desired product 14 was produced in 81% yield and 92%
lower (entries 1 and 5). In addition, it was found that as
protecting group in the bottom aryl fragment (Ar2 in Scheme
3), the trityl group (Tr ) trityl) gave much better yield for
the Michael addition than p-methoxybenzyl.
ee.7-9 Several other auxiliaries and protecting group com-
binations gave less satisfactory results. As shown in Scheme
4 and Table 1, as chiral auxiliary, (S,S)-pseudoephedrine (18)
Grignard Addition. Adding the Grignard reagent of the
top aryl bromide (Ar1MgBr) solution to aldehyde 14 in THF
at -70 to -78 °C gave the product 15 in high yield (>90%)
but with only modest diastereoisomeric ratio (dr) of 5/1.10
A high stereoselectivity is important for controlling this chiral
center in the final product. To improve the selectivity, a
variety of solvents and additives were investigated; the key
results are summarized in Table 2. Different arylmetals gave
very different reactivity and selectivity. The aryl Grignard
gave the best selectivity (entries 1-3). The aryllithium
reacted readily but with much lower diastereoselectivity,
while the arylzinc (prepared from aryllithium and zinc
chloride) was unreactive. It was also found that more polar
solvents improved the diastereoselectivity, e.g., the reaction
in NMP gave product with higher dr (15/1), while toluene
gave lower dr (5.6/1). The ratio was further improved to 25/1
by running the reaction at low temperature with NMP/THF
mixed solvent.
Scheme 4
gave better diastereoselectivity than N-mehyl-(S)-phenyl-
glycinol (19) or N-methyl (1S,2R)-1-amino-2-indanol (20)
(entries 1-3). The sterically more hindered tert-butyl esters
gave better yields than the isopropyl ester of the enoate
(entries 1 and 4). The protecting group on the isopropylamino
group (PG) also affects the diastereoselectivity of the
conjugate addition, e.g., when the benzyl protecting group
was replaced with benzoyl, the ee of the product was much
Cyclization. In our reported synthesis of 1, we described
a novel phosphate-mediated stereospecific intramolecular
ester enolate alkylation of an intermediate similar to 15.3
Table 1. Stereoselectivity of the Conjugate Addition
entry
substrate
auxiliary
% ee
% yield
1
2
3
4
5
12
12
12
12a
12b
18
19
20
18
18
92
89
80
91
57
81
58
70
65
78
1
(9) Data for 14: solid; H NMR (400 MHz, CDCl3, ppm) 9.64 (s, 1H),
7.77 (d, J ) 8.9 Hz, 1H), 7.5-7.6 (m, 6H), 7.2-7.4 (m, 15H), 6.94 (d, J
) 2.4 Hz, 1H), 6.3-6.8 (br, m, 3H), 5.09 (dd, J ) 4.0, 11 Hz, 1H), 4.85
(br, s, 1H), 4.55 (d, J ) 7.1 Hz, 1H), 4.42 (Abq, J ) 11.2, 23.9 Hz, 2H),
3.78 (s, 3H), 3.15 (br, m, 1H), 2.7-2.8 (m, 1H), 1.26-1.31 (m, 6H), 1.19
(s, 9H).
Org. Lett., Vol. 3, No. 21, 2001
3359

equiv of (Me2N)2POCl and 5.2 equiv of NaHMDS at -20
to 0 °C. After aqueous acid workup the reaction mixture
was treated with concentrated HCl in acetonitrile to remove
both the trityl and the tert-butyl groups. The resulting acid
was crystallized as the benzylamine salt 17 in 67% overall
yield from the Grignard addition product 15.11,7 This is the
first crystallized intermediate from the beginning of the
synthesis, and all of the reactions were performed using crude
intermediates, which is a demonstration of the robustness
of the process. The ee of the acid in 17 was upgraded by
this crystallization from 90% to >99%.
Table 2. Diastereoselectivity of the Addition to Aldehyde 14
by Ar1MX
entry
solvent
THF
THF
THF
PhMe
NMP
NMP/THF(1/1)
MX
temp
dr
1
2
3
4
5
6
MgBr
Li
ZnCl
MgBr
MgBr
MgBr
-78 °C
-78 °C
rt
-78 °C
-20 °C
-50 °C
5.0/1
nr
5.6/1
15/1
25/1
To finish the synthesis, the free acid from the benzylamine
salt 17 was hydrogenated in MeOH at 40 °C under 40 psi
hydrogen. The product was dissolved in THF to remove
catalyst, and final product 2 was crystallized from methanol
in 90-95% yield and ∼99% pure by HPLC.12,7
Following the same protocol, we treated compound 15 with
1.5 equiv of diethyl chlorophosphate [(EtO)2POCl] and 5
equiv of lithium hexamethyldisilazide (LiHMDS), expecting
an stereospecific SN2 displacement of the alcohol phosphate
by the tert-butyl ester enolate to give 16. Instead, a complex
product mixture was produced. We reasoned that the amino
substituent on the pyridine ring of 15 is much more electron-
donating (than the butyl group in compound 1) so that the
phosphate intermediate from the alcohol decomposes readily.
Therefore, other less active leaving groups such as phosphite,
carbonate, carbamate, and phosphoroamide were investigated.
Only the phosphoroamide offered promising results. Thus,
treatment of 15 with (Me2N)2POCl and LiHMDS or NaH-
MDS in THF or a THF/toluene mixture afforded the
cyclization product 16. The reaction was significantly faster
and cleaner with NaHMDS. It was also faster in THF than
in a toluene/THF mixture. Under optimized conditions, the
crude Grignard addition product 15 was treated with 1.6
In summary, this work has expanded the scope and defined
some of the limitations of this strategy for synthesizing this
class of endothelin antagonists. Careful evaluation of the
chemistry allowed for modification of the reactivity of the
substrates and reagents, which led to significant improve-
ments of the key transformations
Supporting Information Available: Procedure for the
preparation of compounds 2, 5, 7, 8, 10, 11, 12, 14, 15, and
17. 13C NMR data for compounds 2, 14, and 17. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL016601S
(11) Data for 17: mp 135-137 °C; 1H NMR (400 MHz, CD3OD, ppm)
7.4 (m, 5H), 6.6-7.25 (m, 11H), 6.21(d, J ) 87. Hz, 1H), 4.4-4.9 (m,
5H), 4.35-4.55 (m, 5H), 3.97 (s, 2H), 3.76 (s, 3H), 3.18 (t, J ) 8.9 Hz,
1H), 3.11 (t, J ) 8.6 Hz, 2H), 1.10 (d, J ) 6.8 Hz, 3H), 1.04 (d, J ) 6.6
Hz, 3H).
(10) Data for 15: (solid); 1H NMR (C6D6) of the major isomer has a
peak at 5.11 ppm (ABq, J ) 10.5, 4.3 Hz)) and the minor isomer at 5.30
ppm (Ab q, J ) 11.5, 3.5 Hz). The diastereoisomeric ratio was determined
by HPLC on an YMC PVA column. The stereochemistry of the alcohol is
assigned on the basis of the subsequent chemical transformations.
(12) Data for 2: mp >200 °C; 1H NMR (400 MHz, DMSO-d6, ppm)
6.6-7.2 (m, 7H), 6.2-6.3 (m, 2H), 4.4-4.7 (m, 5H), 4.34 (d, J ) 8.4 Hz,
1H), 3.6-3.8 (m, 1H), 3.1 (m, 1H), 3.14 (t, J ) 8.6 Hz, 2H), 3.01 (t, J )
8.7 Hz, 1H), 1.04 (d, J ) 6.4 Hz, 3H), 0.98 (d, J ) 6.4 Hz, 3H).
3360
Org. Lett., Vol. 3, No. 21, 2001