4-Phenylpyridine glucagon receptor antagonists: Synthetic approaches to the sterically hindered chiral hydroxy group

Article abstract of DOI:10.1016/S0040-4039(02)00840-7

Systematic evaluation of the structure-activity relationships of a new class of 4-aryl-pyridine glucagon antagonists led to the discovery of potent analogues bearing a key secondary hydroxy moiety as seen in compound 1a. Due to the importance of this new class of compounds, it became necessary to establish an efficient synthesis of the pure enantiomer. A resolution and two chiral syntheses of alcohol 1a were discovered and herein presented.

Full text of DOI:10.1016/S0040-4039(02)00840-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 4455–4458
4-Phenylpyridine glucagon receptor antagonists: synthetic
approaches to the sterically hindered chiral hydroxy group
Gaetan H. Ladouceur,* James H. Cook, Elizabeth M. Doherty, Dierk Giebel and
William R. Schoen
Department of Chemistry Research, Bayer Research Center, 400 Morgan Lane, West Haven, CT 06516, USA
Received 16 January 2002; revised 29 April 2002; accepted 30 April 2002
Abstract—Systematic evaluation of the structure–activity relationships of a new class of 4-aryl-pyridine glucagon antagonists led
to the discovery of potent analogues bearing a key secondary hydroxy moiety as seen in compound 1a. Due to the importance
of this new class of compounds, it became necessary to establish an efficient synthesis of the pure enantiomer. A resolution and
two chiral syntheses of alcohol 1a were discovered and herein presented. © 2002 Elsevier Science Ltd. All rights reserved.
Recently, we reported our research efforts in the dia-
betes area which led us to the discovery of a new series
of potent non-peptidic human glucagon receptor antag-
onists.1 Compound 1a, which was a potent antagonist
(IC50=110 nM, Scheme 1), was chosen as a representa-
tive analog for in vivo efficacy and toxicological stud-
ies. Amongst the modifications investigated throughout
the SAR optimization of our class of antagonists, the
secondary hydroxy moiety was found to be the key
pharmacophoric element affording high affinity to the
glucagon receptor.
hol group revealed itself as a poor nucleophile, reacting
only partially with bulky chiral auxiliaries such as
Mosher’s reagent.3 Thus, we decided to convert the
alcohol into a salt-forming derivative with an elec-
trophile lacking steric bulk. Accordingly, alcohol 1 was
treated with a base, such as methyllithium, followed by
phthalic anhydride to produce the ester 3 (Scheme 2).
Resolution of these derivatives can be accomplished by
F
F
HO
1) MeLi / THF
OHC
C3H7
C3H7
Compound 1a was originally synthesized by direct alkyl-
ation of its aldehyde precursor (2) with methyllithium
(Scheme 1). This procedure delivered a racemic mixture
of alcohols (1) and the more active enantiomer (1a) was
obtained by chiral chromatography.2 Although this
synthesis of the racemate was straightforward, it was
difficult to obtain large quantities of the pure desired
enantiomer (1a). The need for bulk quantities of the
pure enantiomer to support several biological studies
prompted us to explore alternative synthetic routes. We
discovered that the high level of steric hindrance caused
by the neighboring groups of the secondary alcohol
dictated its unique reactivity and the approach of
reagents to the sp2 center.
2) chiral chromatography
N
N
2
1a
Scheme 1. Achiral synthesis of compound 1.
F
F
HO2C
O
HO
O
1. MeLi / THF
C3H7
C3H7
O
2.
N
N
O
3
1
O
NH2
1.
Our initial attempts focused on a resolution of the
racemic alcohol 1 by conventional methods. The alco-
2. crystallization
3. NaOH / EtOH
1a
* Corresponding author. Tel.: +1-203-812-2055; fax: +1-203-812-2650;
e-mail: gaetan.ladouceur.b@bayer.com
Scheme 2. Chiral resolution of compound 1.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00840-7

4456
G. H. Ladouceur et al. / Tetrahedron Letters 43 (2002) 4455–4458
O
S
F
F
fractional crystallization in the presence of a chiral
amine.4 In our case, quick success was achieved by
fractional crystallization of salts with R-(+)-a-methyl-
benzylamine.5,6 Compound 1a was then easily produced
by a nearly quantitative cleavage of the ester group,
giving an overall yield of about 30%.
CH2Li
O
S
HO
4
OHC
C3H7
C3H7
p-Tol
THF
N
N
5a (S,S)
5b (S,R)
2
The starting material remaining in the mother liquors
(ꢀ70%), enriched with the undesired enantiomer, could
be submitted to a recycling process. The starting mate-
rial recovered was oxidized to the corresponding ketone
(MnO2/CH2Cl2), then reduced to the racemic alcohol 1
(LAH/THF) in a nearly quantitative yield. Alcohol 1
was submitted once more to the resolution procedure as
described in Scheme 2.
Chromatography
H2, Raney Ni
EtOH, rt
5a (S,S)
1a
Scheme 4. Chiral synthesis compound 1.
F
F
We later improved this process by synthesizing com-
pound 3 in a one-pot procedure directly from the
aldehyde 2 (Scheme 3), instead of a two-step sequence.
O
O
HO
O
MnO2 / CH2Cl2
S
S
C3H7
C3H7
p-Tol
p-Tol
reflux
We also investigated enantioselective syntheses through
chiral alkylation, since several research groups have
been successful with this approach. We were particu-
larly interested in the work of Braun and Hild,7 who
have shown that the zinc derivative of methyl-p-tolyl
sulfoxide can be added to benzaldehyde with high
diastereoselectivity. The resulting b-hydroxysulfoxide
can undergo reductive cleavage by Raney Ni to give a
secondary alcohol as in compound 1a. However, no
reaction occurred when we treated aldehyde 2 with the
same reaction conditions, or with harsher conditions,
such as refluxing tetrahydrofuran. On the other hand,
aldehyde 2 reacted cleanly with lithiated (S)-methyl
sulfoxide 4, but gave only a 1.3:1 ratio of desired (5a)
to undesired (5b) diastereomers (Scheme 4). The
diastereomers were separated by standard chromato-
graphy and the desired isomer was treated with Raney
nickel in ethanol to give 1a (96% yield, 99% e.e). This
reduction proceeded without racemization, in contrast
to a published report.7
N
N
6
5
-78oC
LAH / THF
F
O
HO
H2, Raney Ni
EtOH, rt
1a
S
C3H7
p-Tol
N
5a
Scheme 5. Inversion of the chirality of the secondary alcohol.
sition. The resulting large amounts of aluminum salts
generated by this excess might account for the relatively
modest yield (72%).
The high diastereoselectivity achieved during the reduc-
tion of the ketone 6 offered the option to eliminate the
chromatographic separation of the two diastereomers
(5a and 5b) obtained in Scheme 4. Both diastereomers
were oxidized together by manganese dioxide to give
ketone 6 (87%), then submitted to the same desulfuriza-
tion process as previously described.
In an alternative process, 1a was obtained through a
stereoselective reduction of a b-ketosulfoxide (Scheme
5). Oxidation of the alcohol 5 with manganese dioxide
provided the corresponding ketone 6 in 87% yield.
Reduction of 6 with lithium aluminum hydride8–11 at
−78°C gave the desired diastereomer in >98%
This sequence was convenient for producing a few
grams of 1a. However, several problems were encoun-
tered when a larger scale synthesis was attempted. In
particular, both methylsulfoxide intermediates (5 and 6)
were poorly soluble and required very large amounts of
solvent, making the process impractical for scale-up.
Furthermore, the (S)-methyl p-tolyl sulfoxide precursor
was available only in small quantities and at a high
cost. Therefore, we continued to explore other synthetic
approaches. We were also intrigued by the unique
features of the molecule and were anxious to learn
more about its reactivity towards other enantioselective
reagents.
1
diastereomeric excess as determined by H NMR. The
rate of this reaction was extremely slow, unless a large
excess of lithium aluminum hydride was used (14
equiv.). Allowing the reaction mixture to warm to 0°C
did not hasten the reaction but instead led to decompo-
F
F
HO2C
O
O
1. MeMgBr / THF
OHC
C3H7
C3H7
O
2.
N
N
O
3
2
O
We then decided to investigate enzymatic resolutions as
they offer several potential advantages: low cost (cata-
lytic), simplicity and high selectivity.12 Enantioselective
Scheme 3. Alternative synthesis of compound 3.

G. H. Ladouceur et al. / Tetrahedron Letters 43 (2002) 4455–4458
4457
esterification of the racemic alcohol of 1 appeared to be
the most interesting procedure because the reactions are
usually performed in organic solvents. Unfortunately,
alcohol 1 did not esterify when treated with a wide
variety of enzymes; PPL,13 Bacillus subtilis,14 Candida
cylindracea,15 Pseudomonas fluorescens,16 lipoprotein
lipase,17 lipase PS-3018 and chymotrypsin.19 The lack of
reactivity of the compound is probably related to the
steric effects caused by the bulky groups adjacent to the
alcohol. Similarly, the reverse reaction, the enantiose-
lective saponification of the corresponding acetate of 1,
was unsuccessful. However, this type of reaction must
be performed in a buffer solution, which gives inade-
quate solvation for the highly lipophilic phenylpyridine
acetate.
Compound 1a was predominantly obtained when (+)-
N-methylephedrine was used while its enantiomer is
obtained with (−)-N-methylephedrine. The next objec-
tive was to improve the enantiomeric excess (e.e.) of the
reduction. Enantioselective reactions are often tempera-
ture and solvent dependent. Accordingly, we attempted
this reaction at 0°C and a ratio of 7:3 was then
obtained with a comparable yield (ꢀ90%). This inter-
esting result clearly demonstrated the direct effect of the
temperature on the enantioselectivity of the reduction.
It was then obvious to try even lower temperatures. At
−78°C, the desired product was produced with a ratio
of 97:3 (94% e.e.) with an excellent yield (90–95%). This
highly enriched alcohol can then be recrystallized from
a mixture of ethanol and water to give 1a (>99% e.e.).
This new procedure became the most efficient and
simple sequence thus far and has been successfully
applied on several other analogs bearing the same
sterically hindered secondary alcohol.
We then focused our efforts on exploring enantioselec-
tive reduction of ketone 7 (Scheme 6). This field has
been extensively studied and, in general, has been very
successful with rigid chiral hydride reagents, such as the
efficient CBS reduction.20 However, when we tried these
conditions on the ketone 7, no reaction occurred. In
fact, sodium borohydride itself was reacting only very
slowly with ketone 7 (<50% conversion after 3 days).
Likewise, DIP Chloride21 and Binal22 did not react with
ketone 7. In any case, we always faced the same prob-
lem, the chiral reagents used were too sterically hin-
dered to react efficiently with the bulky ketone. Limited
success was finally obtained with a procedure developed
by Terashima (LAH, ethylaniline and N-methyl
ephedrine).23,24 Although the conversion was poor
(<10%), we could demonstrate that some enantioselec-
tivity was achieved during the course of the reaction
(ꢀ2:1 mixture of enantiomers) and we decided to inves-
tigate this reaction further.
This new chiral reducing agent could potentially have
other applications on similar sterically hindered
ketones. A detailed procedure is included as additional
information.25
Acknowledgements
We are grateful to A. Paiva and L. Huang for efficient
LC/MS analyses.
References
1. Ladouceur, G. L.; Cook, J. H.; Doherty, E. M.; Schoen,
W. R.; MacDougall, M. L.; Livingston, J. N. Bioorg.
Med. Chem. Lett. 2002, 12, 191.
2. Chiralpak AD column; isocratic elution 1.5 ml/min. (99%
hexane/methyl t-butyl ether), retention times (4a, 7.3 min;
4b, 9.9 min).
3. Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95,
512.
4. Ingersoll, A. W. In Organic Reactions; Adams, R., Ed.;
Wiley: New York, 1944; Vol. 2, Chapter 9.
As we tried to explain the very low yield of the reac-
tion, a key discrepancy was noted: the alcohol pro-
duced was not the result of a reduction of the ketone
(7) by Terashima’s reagent but rather by its less hin-
dered precursor, the complex LAH/N-methylephedrine.
Indeed, during the preparation of the chiral reducing
agent, a small portion of the reagent did not react with
ethylaniline (<10%) which readily reduces the ketone
where Terashima’s reagent does not, even at higher
temperature. In order to take advantage of this obser-
vation, the reduction was subsequently performed with
the complex LAH/N-methylephedrine, which yielded
approximately 90% of the corresponding alcohol with
the same enantiomeric ratio as previously observed. To
our knowledge, no one else has reported an enantiose-
lective reduction with this complex and the mechanistic
approach of the reagent toward the substrate remains
unclear.
5. Gough, G. A. C.; Hunter, H.; Kenyon, J. J. Chem. Soc.
1926, 2053.
6. Methanol (5.25 mL) was added to compound 3 (23.1 g,
47 mmol) in hot hexane (350 mL), then the suspension
was heated to reflux. The suspension dissolved after
addition of (R)-(+)-a-methylbenzylamine (6.06 mL, 47
mmol) and the salt precipitated after a few minutes. After
45 min stirring, the precipitate was filtered and washed
with cold hexane (2×10 mL) to afford 11.36 g (40%, 91%
ee) of the salt. The salt was dissolved in hot methanol
(150 mL) then hot hexane (300 mL) was added. The
methanol was removed with a cooled Dean–Stark trap.
The heat bath was removed as soon as the solution
became turbid. After 1 h of stirring at room temperature
the precipitate was filtered and washed with cold hexane
(2×10 mL) to afford 7.95 g (70%, 98.5% ee) of the salt.
The salt was cleaved with a mixture of 20% aqueous
sodium hydroxide solution (50 mL) and ethanol (50 mL)
F
F
F
HO
HO
O
MnO2
Chiral
C3H7
C3H7
C3H7
CH2Cl2
reduction
N
N
N
1a
1
7
Scheme 6. Enantioselective synthesis of alcohol 1a.

4458
G. H. Ladouceur et al. / Tetrahedron Letters 43 (2002) 4455–4458
24. Kawasaki, M.; Susuki, Y.; Terashima, S. Chem. Lett.
at reflux for 30 min. The reaction mixture was extracted
1984, 239.
with diethylether (3×100 mL). The combined organic
phases were washed with brine and dried over MgSO4
and the solvent removed in vacuo. The residue was
purified by filtration through silica gel (¥ 1.5¦×5¦, 200
mL 30% CH2Cl2/hexane, 300 mL CH2Cl2) to afford 4.23
g (95%, 98.5% ee) as a white solid: mp 105–6.5°C;
25. To a solution of (1S,2R)-(+)-N-methylephedrine (31.1 g,
0.174 mol) in ether (208 mL) was added lithium alu-
minum hydride (1 M/Et2O, 1.5 equiv., 174 mL) dropwise
at 0°C under argon (gas evolution, the first 50 mL of
lithium aluminum hydride reacted strongly). The reaction
was refluxed for 1.5 h turning from a clear solution to a
white milky solution. The reaction was cooled to −78°C
for a dropwise addition of 7 (39.53 g, 0.116 mol) in
diethyl ether (300 mL) cooled to 0°C, (ꢀ2 mL/min; the
reaction temperature was monitored, not allowing the
temperature to rise above −60°C). After the addition was
completed, the addition funnel was washed with another
50 mL of dry ether. The reaction was kept at −78°C for
4.0 h and then allowed to stir and warm to room
temperature overnight. The reaction was quenched at 0°C
with isopropanol (70 mL) and diluted with ether (700
mL), washed with water (4×500 mL), 10% HCl (2×500
mL), brine (2×500 mL), and dried with MgSO4. Filtra-
tion and concentration afforded a white solid. The
product was filtered through a pad of silica (600 g, 10%
ether/hexane) to afford the desired alcohol (36.67 g,
92%). HPLC (chiralcel OD-H c066-013-50418 (4.6×150
mm)) showed 97% ee favoring the first alcohol enan-
tiomer. Rf=0.4 (CH2Cl2); The alcohol (32.1 g, 93.5
mmol) was dissolved in ethanol (820 mL) and water (480
mL) was slowly added at 60°C. The solution was heated
to 75°C for 5 min and the heating and the stirrer bar were
removed. The solution was seeded with pure 1a every 5
min while it was allowed to cool to room temperature.
When the seeds did not dissolve any longer (after 1.5 h),
more seeds were added. The crystals are filtered after 16
h, dried in the vacuum oven and dissolved in
dichloromethane (50 mL). The solvent is removed in
vacuo and the residue dried to afford 20.4 g (63%, 99.4%
ee) as a white solid. mp=106–108°C; 1H NMR (300
MHz, CDCl3): l 7.11 (m, 3H), 7.04 (m, 1H), 4.64 (dq,
J=3.7, 6.6 Hz, 1H), 3.73 (sept, J=6.6 Hz, 1H), 3.18
(sept, J=6.6 Hz, 1H), 2.15 (m, 2H), 1.56 (d, J=3.7 Hz,
1H), 1.39 (d, J=6.6 Hz, 3H), 1.28 (m, 14H), 0.73 (t,
J=7.35 Hz, 3H). FAB-MS Calcd for (C22H30FNO) 343,
found 344 (M+H). Anal. Calcd for C22H30FNO: C, 76.93;
H, 8.80; N, 4.08; F, 5.53. Found: C, 77.02; H, 8.75; N,
4.02; F, 5.62. [h]D +40.7 (c 1.22×10−2 M; CH2Cl2).
1
Rf=0.39 (CH2Cl2); H NMR (CDCl3, 300 MHz) l 7.11
(m, 3H), 7.04 (m, 1H), 4.64 (dq, J=3.7, 6.6 Hz, 1H), 3.73
(sept, J=6.6 Hz, 1H), 3.18 (sept, J=6.6 Hz, 1H), 2.15
(m, 2H), 1.56 (d, J=3.7 Hz, 1H), 1.39 (d, J=6.6 Hz, 3H),
1.28 (m, 14H), 0.73 (t, J=7.35 Hz, 3H); MS-FAB m/e
Calcd for (C22H31FNO, M+H) 344, found 344; Anal.
Calcd for C22H30FNO: C, 76.93; H, 8.80; N, 4.08; F,
5.53; Found: C, 77.20; H, 8.97; N, 4.01; F, 5.60; [h]D
+38.5 (c 1.35×10−2 M; CH2Cl2).
7. Braun, M.; Hild, W. Chem. Ber. 1984, 117, 413.
8. Solladie, G.; et al. Tetrahedron Lett. 1982, 23, 5047.
9. Solladie, G.; et al. Tetrahedron Lett. 1985, 26, 435.
10. Kosugi, H.; Konta, H.; Uda, H. J. Chem. Soc., Chem.
Commun. 1985, 211.
11. Carreno, M. C.; et al. J. Org. Chem. 1990, 55, 2120.
12. Wong, C.-H.; Whitesides, G. M. Enzymes in Organic
Chemistry; Pergamon, 1994.
13. Babiak, K. A.; Ng, J. S.; Dygas, J. H.; Weyker, C. L.;
Wang, Y. F.; Wong, C. H. J. Org. Chem. 1990, 55, 3377.
14. Mori, K.; Akao, H. Tetrahedron 1980, 36, 91.
15. Iriuchijima, S.; Kojima, N. J. Chem. Soc., Chem. Com-
mun. 1981, 185.
16. Schneider, M.; Laumen, K. J. Chem. Soc., Chem. Com-
mun. 1988, 1459.
17. Andersch, P.; Schneider, M. P. Tetrahedron: Asymmetry
1993, 4, 2135.
18. Berkowitz, D. B.; Danishefsky, S. J. Tetrahedron Lett.
1991, 32, 5497.
19. Roy, R.; Rey, A. W. Tetrahedron Lett. 1987, 28, 4935.
20. Corey, E. J.; Bakshi, R. K.; Shibata, S. J. Org. Chem.
1988, 53, 2861 and references cited therein.
21. Chandrasekharan, J.; Ramachadran, P. V.; Brown, H. C.
J. Org. Chem. 1985, 50, 5446.
22. Noyori, R.; Tomino, I.; Yamada, M.; Nishisawa, M. J.
Am. Chem. Soc. 1984, 106, 6717.
23. Terashima, S.; Tanno, N.; Koga, K. J. Chem. Soc.,
Chem. Commun. 1980, 1026.