Resolution of δ-lactams provides access to nonracemic benzoquinolinones: The synthesis of LY300502 and LY300503

Full text of DOI:10.1021/jo9601425

4450
J . Org. Chem. 1996, 61, 4450-4454
Notes
Syn th esis of LY191704. The benzoquinolinone nu-
cleus of 1 and related compounds were previously pre-
pared1 by conversion of the 2-tetralone derivative into
an enamine followed by aza-annulation5 with an acryl-
amide derivative. Kursanov-type reduction of the ene-
lactam (triethylsilane-trifluoroacetic acid, TES-TFA)6
gave mixtures of the cis- and trans-benzoquinolinones (eq
1).
Resolu tion of δ-La cta m s P r ovid es Access to
Non r a cem ic Ben zoqu in olin on es: Th e
Syn th esis of LY300502 a n d LY300503
Bret A. Astleford, J ames E. Audia, J ack Deeter,
Perry C. Heath, Samantha K. J anisse, Thomas J . Kress,
J ames P. Wepsiec, and Leland O. Weigel*
Chemical Process Research and Development,
Lilly Research Laboratories, A Division of Eli Lilly &
Company, Indianapolis, Indiana 46285-4813
Received J anuary 22, 1996
Recently, LY191704 (1) was synthesized1 and identified
as a selective, potent, noncompetitive type I inhibitor of
5-R-reductase in cell cultures derived from human fore-
skin fibroblasts.2 This activity may be useful in the
development of new drugs for the treatment of acne, male
pattern baldness, benign prostatic hyperplasia, or pros-
tatic cancer and may augment the utility of finasteride
(2).3 As part of the development portfolio4 for this new
drug, a reliable and efficient process for the preparation
of large amounts of 1 was required. In addition, the
enantiomers of 1, LY300502 (3) and LY300503 (4), were
required for preclinical evaluation and potential com-
mercial development.
When an aza-annulation reaction was conducted as
described by Ninomiya5c with 6-chloro-2-tetralone7 (5),
problems of regioselectivity were encountered which have
not been previously detailed.8 Fusion of enamine of 67a
with acrylamide5c (Ninomiya’s protocol) gave 8, the
isomer 10,9 and 3-5% of the nonconjugated isomer5d
9
(eq 2).
Upon isolation, a 60:40 mixture of 8:10 was obtained
in 75% yield.8a The in situ ratio of 8:10 was highly
(5) (a) Stork, G. Pure Appl. Chem. 1968, 17, 383. (b) Stork, G.;
Kretchmer, R. A.; Schlessinger, R. H. J . Am. Chem. Soc. 1968, 90, 1647.
(c) Ninomiya, I.; Higuchi, S.; Mori, T. J . Chem. Soc., Chem. Commun.
1971, 457. (d) El-Barbary, A. A.; Carlsson, S.; Lawesson, S. O.
Tetrahedron 1982, 405. (e) Review of enamine mediated aza-annula-
tion: Stille, J . R.; Barta, N. S. In Studies in Natural Products
Chemistry; Rahman, A., Ed.; Elsevier: New York, 1996; Vol. 17, in
press.
(6) (a) Cannon, J . G.; Chang, Y.; Amoo, V. E.; Walker, K. A. Synthesis
1986, 494. (b) Using isomer-free 8, we conducted a series of reductions
isothermally without solvent (to increase rate) and measured the % of
cis-13: temperature °C, (% cis): -15° (2-3%); 0° (7-9%); 34° (14%);
83° (18%).
We disclose herein the synthesis of 1 with a high
degree of regiochemical control and enhanced stereose-
lectivity. We also present a new and practical method
for the resolution of 1 into 3 and 4.
(1) (a) For synthesis and SAR of LY191704, see: J ones, C. D.; Audia,
J . E.; Hirsch, K. S.; Lawhorn, D. E.; McQuaid, L. A.; Neubauer, B. L.;
Pike, A. J .; Pennington, P. A.; Stamm, N. B.; Toomey, R. E. J . Med.
Chem. 1993, 36, 421. (b) Neubauer, B. L.; Goode, R. L.; Gray, H. M.;
Hanke, C. W.; Hirsch, K. S.; Hsiao, K. C.; J ones, C. D.; Lawhorn, D.
E.; McQuaid, L.; Toomey, R. E.; Valia, K.; Audia, J . E. Drugs Future,
1995, 20, 144. (c) Hirsch, K. S.; J ones, C. D. EP Appl. 531026, August
20, 1992.
(7) (a) Nixon, J . A.; Pioch, R. P.; Schaus, J . M.; Titus, R. D., European
Pat. Appl. EP 343830 A2, November 29, 1989. (b) For the preparation
of 6-chloro-2-tetralone, see Rosowsky, A.; Battiglia, J ; Chen, K. K. N.;
Modest, E. J . J . Org. Chem. 1968, 33, 4288. (c) Suryawanshi, S. N.;
Fuchs, P. L. J . Org. Chem. 1986, 51, 902.
(2) (a) Hirsch, K. S.; J ones, C. D.; Audia, J . E.; Anderson, S.;
McQuaid, L.; Stamm, N. B.; Pennington, P.; Russell, D. W.; Neubauer,
B. L.; Toomey, R. E. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5277. (b)
Holt, D. A.; Levy, M. A.; Ladd, D. L.; Erb, J . M.; Heaslip, J . I.; Brandt,
M.; Metcalf, B. W. J . Med. Chem. 1990, 33, 937.
(3) Rasmusson, G. H.; Reynolds, G. F.; Steinberg, N. G.; Walton,
E.; Patel, G. F.; Liang, T.; Casccieri, M. A.; Cheeung, A. H.; Brooks, J .
R.; Berman, C. J . Med. Chem. 1986, 29, 2298.
(8) (a) Astleford, B. A.; Deeter, J .; Kress, T. J .; J anisse, S. K.; Weigel,
L. O.; Wepsiec, J . P. 206th National American Chemical Society
Meeting, Chicago, IL ORGN075, August 23, 1993. (b) When the aza-
annulation with â-tetralone was performed in our labs, the yield
reported by Ninomiya was reproduced. The isomer distribution derived
from 5 is due, in part, to the inductive/resonance effects of the
6-halogen substituent.
(9) For similar observations of regioselectivity in the alkylation of
tetralone enamines due to 1-8 peri-effects, see: (a) Volpe, T.; Revial,
G.; Pfau, M.; d’Angelo, J . Tetrahedron Lett. 1987, 28, 2367. (b) d’Angelo,
J . Tetrahedron Lett. 1988, 29, 4427.
(4) Weigel, L. O.; Astleford, B. A. Chirality in Industry - II; Collins,
A. N., Crosby, J ., Sheldrake, G. N., Ed.; Wiley: New York, 1996, in
press.
S0022-3263(96)00142-9 CCC: $12.00 © 1996 American Chemical Society

Notes
J . Org. Chem., Vol. 61, No. 13, 1996 4451
Sch em e 1
variable and ranged from approximately 60:40 to 85:15.
Although we observed only the ∆1,2 enamine isomer 6 by
1H NMR, compound 10 arose9 from the reaction of the
2,3
∆
enamine 7. Necessarily, the alkylation of 7 was
substantially faster than reaction of 6. Interestingly, the
ratio of 8:10 varied10a during the course of the reaction,
which suggested that the equilibrium and/or relative rate
of alkylation of the enamine 6 and 7 changed with the
changing reaction medium.10b
The structure assignment of 10 was in contrast to
results recently reported by Cannon’s group in the aza-
annulation of 8-methoxy-2-tetralone.11 Therefore, the
structure of 10 was unambiguously established. The
purified mixture of 8 and 10 from above was subjected
to reduction (TFA-TES) and methylated (vide infra) to
give a mixture of 1 and 11 (equation 3). Structure
determination by single crystal X-ray analysis of a
purified sample of 1112 corroborated our assignment of
10. In addition, the reduction of 10 had proceeded
stereospecifically11b to give the trans isomer 11.11c
In previous work we found the product distribution of
the aza-annulation process with 2-tetralones could be
controlled.8 Reaction of the pyrrolidine enamine 6 pro-
ceeded well in solution with excess acrylamide. Under
optimized laboratory conditions (e.g., isopropyl acetate,
(10) (a) The ratio of 8 to 10 was measured throughout the reaction
(10 mmol, 5 equiv of acrylamide and 3 mol % of TsOH, 87 °C): time h,
(8:10); 2.5 (94:6); 6 (92:8); 144 (85:15). (b) The pyrrolidine released
underwent Michael addition to acrylamide (exothermic) at a faster rate
than aza-annulation. These results suggest that use of 1.75 equiv of
acrylamide as recommended by Ninomiya (see ref 5c) represents 88%
of theory for complete conversion. A patent reported that 2.4 equiv of
acrylamide was optimal: Tanabe Seiyaku KK J P-48023779, March 27,
1973.
(11) (a) Cannon reported three isomers from aza-annulation with
8-methoxy-2-tetralone and acrylamide (total yield 30%, ratio of 19:20:
21, 77:17:6). Reduction (TES-TFA) of 19 and 21 reportedly gave the
trans isomer but 20 gave the cis isomer. Cannon, J . G.; Kirschbaum,
K. S. Synthesis 1993, 1151. (b) Mellin reported the mixture from ref
11a gave 65:35 cis/ trans mixture upon reduction with TES-TFA:
Mellin, C.; Vallagrda, J .; Nelson, D. L.; Bjork, L.; Yu, H.; Anden, N.
E.; Csoregh, L.; Arvidsson, L. E.; Hacksell, U. J . J . Med. Chem. 1991,
34, 497. (c) The spectral data of 6 was similar to the data for structure
II in reference 11a. We are exploring the results of the aza-annulation
and reduction sequence reported by Cannon (ref 11a) and Mellin (ref
11b).
89 °C, 4 mol % TsOH, 4-5 equiv of acrylamide) the in
situ ratio of 8:10 was improved to 95:5, and 8 crystallized
from the reaction medium in 77% yield (Scheme 1).
Reduction of amide 8 to the trans lactam 12 with excess
TES-TFA1,6a in methylene chloride was slow due to the
poor solubility of 8. Unfortunately, the reduction also
yielded 13% of the cis isomer 13 (Scheme 1). This
amount of isomer proved difficult to remove and sub-
stantially reduced the recovery of 12 upon crystallization.
However, reduction of 8 at (-10 °C with slow warming to
35 °C) without solvent under triphasic conditions (solid
8 + TFA + TES)8 reduced the in situ amount of the cis
isomer 13 to 6-7%. Moreover, the isolation and purifica-
tion was facilitated by the direct isolation of the TFA
complex 12.13 The complex 12 (containing 4% of 13) was
isolated in 95% yield after trituration with hot heptane
(Scheme 1). Recrystallization(s) of 12 (or the free amide)
from various solvents afforded pure 12. However, re-
moval of low levels of 13 by selective methanolysis in a
subsequent step proved more efficient (vide infra).
Methylation of the free amide of 12 using sodium
hydride-methyl iodide has been reported.1 For larger
preparations, a novel water-based methylation of 12 was
(12) Compound 11 crystallized in the monoclinic space group P21/n
with a unit cell having the dimensions a ) 13.442(2) Å, b ) 6.851(1)
Å, c ) 13.700(3) Å, and â ) 94.25° (2) with a calculated density of
1.318 g/cm-3. The data was collected on an automated four-circle
diffractomer using monochromatic copper radiation. The structure was
solved with the use of direct methods and was refined by least-squares
with anisotropic temperature factors for all atoms except hydrogen.
All hydrogen atoms were included at calculated positions. The final
R-factor was 0.05. An ORTEP plot of trans-(()-11 is included in the
supporting information. The author has deposited atomic coordinates
for this structure with the Cambridge Crystallographic Data Centre.
The coordinates can be obtained, on request, from the Director,
Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge,
CB2 1EZ, UK.
(13) The formation of TFA complexes provided a useful method for
isolation and purification of a number of related benzoquinolinone
derivatives. Unpublished observations, L. Weigel.

4452 J . Org. Chem., Vol. 61, No. 13, 1996
Notes
developed.14 Reaction of 12 with methyl chloride in
THF-H2O in the presence of sodium iodide and 50%
sodium hydroxide gave 1 in 91% yield mixed with 2% of
the cis isomer 14 (Scheme 1). The methyl chloride used
for this step had been generated and trapped in the next
step of the synthesis sequence. By using methyl chloride
in this manner, some potential environmental concerns
regarding the transport and utilization of large amounts
of methyl iodide were obviated.
Resolu tion of LY191704. LY300502 (3) and LY-
300503 (4) have been resolved by the separation of the
diastereomeric derivatives (2-R-phenethyl auxiliary on
the nitrogen of the free amide of 12)15 or by direct
chromatographic separation of 1.16 Production of large
quantities of 3 and 4 by these methods was not deemed
feasible. Classical acid-base type resolutions of δ-lac-
tams similar to 1 are not generally possible. We envi-
sioned that one tenable strategy might be defined by a
sequence involving δ-lactam alcoholysis to the ester17 and
an appropriate resolution18 of the amine with a nonra-
cemic acid (HA) followed by lactamization (eq 4). This
simple approach for the resolution of lactams has not
been previously described.19
cis amide 13 proved inert to methanolysis and remained
soluble throughout the subsequent operations. Without
isolation, 15 was converted into a diastereomeric mixture
of salts 17 and 18 by reaction with the monosodium salt
of R,R-di-p-toluoyl-L-tartaric acid (DTTA, inexpensive
form)18 and enough NaHCO3 to neutralize any residual
HCl (Scheme 1). Solutions of monoprotic salts 17 and
18 remained acidic during crystallization. Therefore,
closure of the free amine to the lactam was less
problematic.18b Appropriate handling of this mixture (see
Experimental Section) and crystallization from methanol
gave pure 17 (41% yield, 82% of theory; 98% de). The
filtrates of 17 were evaporated, and 18 was purified by
crystallization from CH2Cl2-H2O (45.5% yield; 91% of
theory; 98% de9c). If required, the chemical and the
diastereomeric purity of both 17 and 18 were upgraded
by repeating the simple solvent slurries (high tempera-
tures promote lactamization). Neutralization of 18 (or
17) under mild biphasic conditions with toluene-aqueous
NaHCO3 produced 3 (or 4) in high yields (Scheme 1).18
Con clu sion . In summary, we have demonstrated
control of regioselectivity in the aza-annulation of 5 to
8, defined a protocol for enhanced stereoselectivity in the
reduction of 8 to 12, and presented a new resolution
protocol of benzoquinolinones (1 resolved into 3 and 4).
In addition, the synthesis route generated methyl chlo-
ride (utilized in the alkylation step), provided an effective
purification method of benzoquinolinones via TFA com-
plex formation, and allowed resolution of both antipodes
of LY191704 (1) with the use of only R,R-di-p-toluoyl-L-
tartaric acid. The overall process is straightforward and
yields (30% for 3 and 25% for 4 from 5) have allowed the
preparation of suitable quantities for clinical develop-
ment. These observations have alreday proved useful in
the synthesis of related systems.21,22
In the event, the tricyclic lactam 1 (containing 2% of
the cis isomer 14) underwent direct acid-catalyzed meth-
anolysis (3-6 d) to 15 in high yields.17a,b The metha-
nolysis was faster in the presence of water. When 1 was
heated to reflux with methanol and excess concentrated
hydrochloric acid for 9 h, complete consumption of the
lactam was observed (Scheme 1).20 Some amino acid (16)
was observed but this byproduct was simply esterified
by codistillation of methanol and water during workup.
(During this stage, methyl chloride was trapped in THF
and used in the methylation step above, Scheme 1). The
Exp er im en ta l Section 23
8-Ch lor o-1,4,5,6-tetr a h yd r oben zo[f]qu in olin -3-on e (8). A
solution of the 6-chloro-2-tetralone7b (5, 107 g, 591 mmol),
pyrrolidine (56 mL, 1.13 equiv), and p-TsOH (0.17 g, 0.15 mol
%) in toluene (900 mL) was heated to reflux under nitrogen for
2 h during which time water was collected with a Dean-Stark
trap. This solution was concentrated in vacuo and the enamine
67a was triturated with heptane (350 mL). The heptane was
removed by cannula and the resulting solid dried under vacuum
(yield, 125 g). A portion (2.88 g) of the air-sensitive enamine
was heated to reflux with acrylamide (3.51 g, 4 equiv; 85-87
°C, 72 h) in isopropyl acetate (50 mL) containing p-TsOH (0.11
g). The mixture was cooled in an ice bath and filtered to give
2.44 g of pure 8 (85% from 6, 77% from 5) after vacuum drying:
TLC Rf 0.36; mp 227-230 °C (lit.1c mp 229-230 °C); IR (KBr) ν
1695, 1665 cm-1; MS m/ z 233 (M+), 235 (34%, M+ + 2); HPLC
(system A) 10.9 min (>99%); (system B) 4.37 min (>99%); GC,
6.8 min; 1H NMR (DMSO-d6) δ 9.43 (1H, s), 7.2-7.3 (3H, m),
2.83 (2H, m), 2.6-2.4 (4H, m), 2.28 (2 H, m); 13C NMR (DMSO-
d6) δ 169.68, 134.89, 134.63, 133.69, 128.47, 126.75, 126.11,
(14) Kress, T. J . Unpublished material.
(15) Audia, J . E.; Lawhorn, D. E.; Deeter, J . B. Tetrahedron Lett.
1993 34, 7001.
(16) Abel, A. D.; Erhard, K. F.; Yen, H. K.; Yamashita, D. S.; Brandt,
M.; Mohammed, H.; Levey, M. A.; Holt, D. A. Bioorg. Med. Chem.
Lett.1994, 1365.
(17) (a) Audia, J . E.; J ones, C. D.; McQuaid, L.; Weigel, L. US 5 334
767 August 2, 1994. (b) For methanolysis of N-BOC-δ-lactams under
basic conditions, see: Flynn, D. L.; Zelle, R. E.; Grieco, P. A. J . Org.
Chem. 1983, 48, 2424. (c) For ring-opening of N-tosyl lactams, see: Bon,
E.; Bigg, D. C. H.; Betrand, G. J . Org. Chem. 1994, 59, 1904 and
references therein. (d) For acid-catalyzed methanolysis of N-acyl
lactams see Dixit, N. A.; Tandel, S. K.; Rajappa, S. Tetrahedron Lett.
1994 35, 6133. (e) For reversal of the selectivity in reference 17b, see:
Wei, Z. Y.; Knaus, E. E. Tetrahedron Lett. 1994 35, 847.
(18) (a) Astleford, B. A.; Audia, J . E.; Dunigan, J . M.; J anisse, S.
K.; Kennedy, J .; Kress, T. J .; Waggoner, R.; Wepsiec, J . P.; Weigel, L.
O. 206th National American Chemical Society Meeting, Chicago, IL
ORGN76, August 23, 1993. (b) Attempted resolution of the nonmethy-
lated 12 using the resolution sequence shown in eq 4 was problematic
due to facile lactamization.
(21) Kuo, F.; Wheeler, W. J . J . Labeled Compd. Radiopharm. 1994,
34, 915.
(22) Weigel, L. O. 12th Process Development Symposium, Robinson
College, Cambridge, England, December 12, 1994.
(23) Physical data was gathered as reported previously, see: Weigel,
L. O.; Dunigan, J . J . Org. Chem. 1991, 56, 6225. The de’s of 17 and 18
were determined by closure to 4 and 3 (Scheme 1, step 5) and HPLC
on a Chiralcel OD column with 10% i-PrOH-hexane at 2.0 mL min at
40 °C and 220 nm. (3, tr ) 16.1 min; 4, tr ) 14.0 min); RP HPLC,
System A: Waters Nova-Pak C-18 column at 2.0 mL/min with
acetonitrile-water (27:73 + 0.67% NH4OAc w/v); System B: Zorbax
RX C-8 with acetonitrile-water (1:1) + 0.50% NH4OAc w/v at 2.0 mL/
min. Capillary GC: DB1 column 30 m x 0.25 µm isothermal at 250
°C; TLC, silica gel 60 F254 (0.25 mm) using ethyl acetate-hexane-
methanol (50:50:1) with a Shimadzu CS-930 Scanner 255 nm or 520
nm after development with phosphomolybdic acid.
(19) For
a similar approach involving enzymatically mediated
closure of amino esters to lactams, see Gutman, A. L.; Meyer, E.; Yue,
X.; Abell, C. Tetrahedron Lett. 1992, 33, 3943.
(20) Other strong acids (H2SO4, TsOH, MeSO3H, etc.) catalyzed the
methanolysis but gave nonvolatile sulfonic acid esters (e.g. TsOMe)
which are toxic and problematic during purification of intermediates.

Notes
J . Org. Chem., Vol. 61, No. 13, 1996 4453
122.36, 105.73, 30.21, 26.91, 24.36, 20.14. Anal. Calcd for C13
-
crystallization ensued, more heptane (200 mL) was added with
stirring and the mixture was concentrated (over 3 h, 15-20 °C)
under vacuum to 75 mL, cooled, and filtered to provide 1 (21.6
g, 91%): TLC, Rf 0.16; mp24 90-93 °C (Lit1c mp 82 °C); GC, 5.6
min; HPLC (system B), 5.54 min.; IR ν 1630, 1487, 1397, 1260
cm-1; UV, λ (ꢀ): 279 (600), 272 (600); MS m/ z 249 (M+), 251
H
12ClNO: C, 66.81; H, 5.18, N, 5.99. Found: C, 66.54; H, 5.34;
N, 6.27.
7-Ch lor o-1,2,3,4,4a ,5-h e xa h yd r ob e n zo[g]q u in olin -2-
on e (10). Tetralone 5 (201 g, 1.11 mol), pyrrolidine (81.8 g, 1.03
equiv), and p-TsOH (0.44 g) were converted to the enamine as
described above. The enamine was immediately treated with
molten acrylamide (237 g, 3.33 mol) at 95 °C for 22 h. This
mixture was poured into water (70 °C, 3.5 L) with good
mechanical stirring and cooled to 0 °C after which the super-
natant was decanted. The solid was then stirred for 25 h with
methanol (350 mL) and filtered to provide 196 g (75%) of a 60:
40 mixture of 8:10 (HPLC and 1H NMR assay). A portion of
this mixture was slurried with three portions of diethyl ether
to give pure 10: TLC, Rf 0.35; mp 238-241 °C; IR ν 3388, 1676,
1646 cm-1; UV, λ (ꢀ): 297 (27000), 226 (16000), 215 (14000); MS
m/ z 233 (M+), 235 (33%, M+ + 2); HPLC (system A), 16.2 min
(97%); HPLC (system B), 4.4 min ; GC: 6.4 min; 1H NMR
(DMSO-d6) δ 9.53 (1H, s), 6.91-7.2 (3H, m), 5.86 (1H, s), 2.91
(1H, m), 2.7-2.2 (4H, m), 2.03-1.85 (2H, m), 1.57 (2H, m); 13C
NMR δ 169.45, 141.41, 134.05, 133.61, 128.32, 126.54, 126.26,
1
(M+ + 2); HPLC, min (97.5% 1 + 2% 9); H NMR δ 7.24-7.15
(3H, m), 3.21 (1H, m), 3.03 (3H, s), 3.12-2.9 (2H, m), 2.8-2.4
(5H, m), 1.68 (2H, m); 13C NMR δ 170.31, 137.30, 135.58, 132.44,
128.38, 126.31, 126.26, 60.77, 40.95, 32.32, 3047, 28.16, 27.77,
23.89. Anal. Calcd for C14H16ClNO: C, 67.33; H, 6.46; N, 5.61;
Cl, 14.20. Found: C, 67.36; H, 6.55; N, 5.67; Cl, 14.64.
7-Ch lor o-1-m eth yl-1,2,3,4,4a,5,10,10a-tr a n s-octah ydr oben -
zo[g]qu in olin -2-on e (11). A portion of the mixture of 8:10 from
the aza-annulation reaction above was reduced (TES-TFA) and
methylated as described above. The mother liquors from the
crystallization of 1 were concentrated under vacuum, digested
in refluxing hexane (20 mL/g), and decanted. The solid was
recrystallized three times from hexane to afford 11: TLC, Rf
0.15; HPLC (system B), 5.7 min; mp 134-136 °C; IR ν 1630 cm-1
;
UV λ (ꢀ): 204 (21,600), 271 (600), 274 (600); MS m/ z 249 (M+,
100%), 251 (M+ + 2, 32%); GC, 5.8 min; 1H NMR δ 7.15 (3H,
M), 3.34 (2H, M), 3.03 (3H, s), 2.75 (1H, m) 1.87 2.73-2.37 (4H,
m) (2H, m), 1.63 (1H, m); 13C NMR δ 137.2, 132.3, 131.7, 131.3,
128.2, 122.7, 60.6, 36.7, 36.2, 36.1, 32.6, 31.2, 26.8. Anal. Calcd
for C14H16ClNO: C, 67.33; H, 6.46; N, 5.61; Cl, 14.20. Found:
C, 67.37; H, 6.31; N, 6.03.
125.64, 101.09, 33.97, 31.40, 30.99, 26.10. Anal. Calcd for C13
12ClNO: C, 66.81; H, 5.18; N, 5.99. Found: C, 66.63; H, 5.13;
N, 5.91.
-
H
8-Ch lor o-1,2,3,4,4a ,5,6,10b-tr a n s-octa h yd r oben zo[f]qu in -
olin -3-on e-Tr iflu or oa cetic Acid (1:1) (12). The lactam 8
(43.7 g, 187 mmol) was charged into a three-neck RB flask with
dry triethylsilane (130 mL, 94.6 g, 814 mmol, 4.35 equiv). The
reaction flask was placed in a cooling bath (-10 °C) and with
good mechanical stirring, trifluoroacetic acid (130 mL, 192 g,
1.69 mol, 9.0 equiv) was added over 20 min. The mixture was
stirred (-10 °C, 7 h; 0 °C, 14 h; 25 °C, 20 h; 35-37 °C, 1 h) and
monitored by HPLC analysis (time h/% conversion/% cis 13; 7/38/
3.5; 24/76/4.8; 44/99/6). The mixture was then concentrated
under vacuum and the solid residue digested in heptane (200
mL, 55 °C, 1 h), cooled (ice bath), and filtered. The solid was
treated with heptane again (300 mL, 60 °C, 1 h; 25 °C) and
filtered to give 12 (62.2 g, 95% yield) mixed with 4.5% of 13.
This material was suitable for use in the next step: TLC , Rf
0.12; mp: 138-141 °C (lit.1c mp 230-231 °C for free amide of
12); HPLC (system B), 3.84 min; IR ν 1635 cm-1; UV λ (ꢀ): 204
(21,600); MS m/ z 235(M+, 100%), 237 (M + 2, 34%); 1H NMR δ
8.17 (1H, s), 7.27-7.05 (3H, m), 3.38 (1H, m), 2.93 (2H, m), 2.53-
2.8 (4H, m), 2.18 (1H, m), 1.95-1.4 (2H, m); 13C NMR δ 170.44,
138.31, 135.75, 130.76, 128.18, 127.54, 125.67, 54.05, 39.47,
31.14, 28.57, 27.46, 24.61. Anal. Calcd for C15H15ClF3NO3: C,
51.52; H, 4.32; N, 4.01, F, 16.30. Found: C, 51.82; H, 4.41; N,
4.04; F, 16.07.
8-Ch lor o-1,2,3,4,4a ,5,6,10b-cis-octa h yd r oben zo[f]qu in o-
lin -3-on e (fr ee a m id e of 13). The heptane washes from the
preparation of 12 (1:1 of 12:13) were neutralized with aqueous
sodium bicarbonate and evaporated. The solid was purified by
preparative HPLC on silica gel with 20% THF in CH2Cl2 to give
the free amide of 13: TLC Rf 0.10; mp 185-187 °C; IR ν 1635
cm-1; UV λ (ꢀ): 204 (21600); MS m/ z 235 (M+, 100%), 237 (M+
+ 2, 38%); HPLC (system B), 3.57 min; 1H NMR δ 7.3-7.1 (3H,
m), 6.47 (1H, s), 3.36 (1H, m), 2.95-2.83 (2H, m), 2.74-2.3 (4H,
m), 2.12-2.03 (1H, m), 1.93-1.60 (2H, m); 13C NMR δ 172.51,
137.67, 135.35, 132.02, 129.89, 128.72, 126.61, 51.11, 35.57,
30.04, 27.45, 26.81, 26.67. Anal. Calcd for C13H14ClNO: C,
66.24; H, 5.99; Cl, 15.04; N, 5.94. Found: C, 66.20; H, 6.10; Cl,
15.49; N, 5.98.
8-Ch lor o-4-m eth yl-1,2,3,4,4a,5,6,10b-tr a n s-octah ydr oben -
zo[f]qu in olin -3-on e (1, LY191704). The TFA complex 12 (33.2
g, 95 mmol) in warm THF (125 mL) was washed with aqueous
sodium hydroxide (2 × 25 mL, 2 N). The THF layer was treated
with sodium iodide (23.5 g, 158 mmol), aqueous sodium hydrox-
ide (50% w/w, 47 mL), and a solution of methyl chloride in THF
(15% w/v, 125 mL; 18.7 g, 3.9 equiv, obtained from the metha-
nolysis below) in a pressure reactor for 33 h at 26-27 °C. The
mixture was then cooled in an ice bath and neutralized with 12
N HCl (pH 2-3), and enough water was added to dissolve the
inorganic salts. The layers were separated, and the aqueous
layer was extracted with CH2Cl2 (200 mL) after which the
organic extracts were combined, washed twice with saturated
brine and dried (4 Å molecular sieves, 33.2 g), filtered, and
concentrated under vacuum to 50 mL. Heptane (100 mL) was
added to a cloud point, and the mixture was seeded with 1. After
Crystals of 11 for single crystal X-ray analysis12 were grown
by isothermal evaporation of solvent (270 mg 11 with 17 mL
hexane and 3 mL of EtOAc, 24 h).
6-Ch lor o-1,2,3,4-t e t r a h yd r o-2(S)-(m e t h yla m in o)-1(S)-
n a p h th a len ep r op a n oic Acid Meth yl Ester -(2R,3R)-Bis[(4-
m eth ylben zoyl)oxy]bu ta n ed ioic Acid (1:1) (17). Lactam 1
(LY191704, 24.9 g, 100 mmol, containing 2% 14) was refluxed
with methanol (12.5 mL) and hydrochloric acid (12 N, 16.2 mL)
for 4 h. Fresh HCl (4 mL) and methanol (3.5 mL) were added,
and the mixture was heated another 5 h. The MeCl evolved
during this portion of the process was entrained in THF at -78
°C and employed in the conversion of 12 into 1 above. The
mixture was cooled and concentrated to 45 g at 40-43 °C .
Analysis at this point by HPLC showed 97% conversion (based
upon formation 15 and the amino acid 16). Three portions of
methanol (150 mL) were added, and the reaction mixture was
concentrated under vacuum at 45 °C. The mixture was diluted
to 240 mL with methanol and heated at 50 °C for 1 h (HPLC
analysis showed >97% of 15). The mixture was diluted in a
500.0 mL volumetric flask with MeOH and titrated (0.224
mequiv acid/mL, 0.12 equiv of excess HCl). In a separate vessel,
(R,R)-di-p-toluoyl-L-tartaric acid monohydrate (DTTA, 40.4 g,
100 mmol) was dissolved in methanol (600 mL) at 45 °C and
treated with NaHCO3 (9.41 g dissolved in 44 mL of warm water,
112 mmol). The solution containing 15 was added at 15 °C over
10 min with good stirring to the sodium DTTA solution. The
mixture was cooled to 4 °C for 1 h and filtered to give 17 (92%
de; the methanol filtrate was set aside for the isolation of 18,
see below). In turn, the CH2Cl2 filtrate from the isolation of 18
from below was concentrated in vacuo, and the residue (mostly
17) was triturated with methanol (100 mL). The two solid
portions of 17 were combined and stirred with methanol (600
mL, 35 °C to -10 °C) affording 17 as a monohydrate (28.0 g,
82% of theory); mp 120-130 °C dec; IR ν 3800-2500, 1725, 1612
cm-1; UV, λ (ꢀ): 270 (2200), 239 (28000), 204 (46200); MS m/ z
282 (M+ for amine), 284 (M+ + 2 for amine), 386 (M+ for DTTA);
KF, 2.4% water; HPLC (system A), 0.50 min (DTTA, 49%), 1.3
min (amino ester of 17, 49%); HPLC (system B), 1.71 min (amino
ester of 17); Chiral HPLC; 14.0 min (98% de); [R]D -66.6°, [R]365
-343° (c ) 1.00, MeOH); 1H NMR (DMSO-d6) δ 11.0-8.0 (1H, b
s), 7.85 (4H, d), 7.37 (4H, d), 7.2-7.0 (3H, m), 5.63 (2H, s), 3.58
(3H, s), 3.38 (1H, m), 2.75 (1H, m), 2.88 (2H, m), 2.45 (3H, s),
2.37 (6H, s), 2.35 (2H, m), 1.8-1.53 (4 H, m); 13C NMR (DMSO-
d6, partial) δ 130.90, 129.41, 129.33, 128.30, 127.94, 127.07,
126.83, 126.83, 126.11, 125.75, 72.14, 59.93, 55.28, 51.41, 51.36,
40.24, 38.65, 32.09, 31.08, 30.70, 30.35, 27.72, 23.06, 21.23, 19.20.
Anal. Calcd for C15H20ClNO2-C20H18O8‚H2O. C, 61.22; H, 5.84;
(24) Compounds 1, 3, and 4 (although chemically pure) often exhibit
poor melting point and DSC behavior due to polymorphism. Unpub-
lished data of M. Kearns, C. Pasini, and L. Weigel.

4454 J . Org. Chem., Vol. 61, No. 13, 1996
Notes
N, 2.04; O, 25.61; Cl, 5.17. Found: C, 61.49; H, 6.14; N, 1.87;
O, 25.06; Cl, 5.47.
Rf 0.16; DSC24 76 °C (lit15,17b mp 71.5-73 °C); IR ν 3005, 1627,
1485 cm-1; UV λ (ꢀ): 279 (600), 271 (640), 205 (20700); MS m/ z
249 (M+), 251 (M+ + 2, 33%); HPLC (system A) 3.4 min (>99%);
HPLC (system B), 5.54 min (99%); enantiospecific HPLC: 32.3
min (99.8% ee); GC, 5.6 min; [R]D -86° [R]365 -289° (c ) 1.00,
MeOH); 1H NMR δ 7.29 (3H, m), 3.22 (1H, m), 3.03 (3H, s), 3.0-
2.4 (7H, m), 2.8-2.6 (2H, m); 1H NMR (benzene-d6) δ 2.47 (3H,
s); 13C NMR δ 185.34, 162.96, 137.28, 135.53, 132.53, 128.43,
126.32, 60.84, 40.96, 32.26, 30.57, 28.18, 27.78, 23.86. Anal.
Calcd for C14H16ClNO: C, 67.33; H, 6.46; N, 5.61; O, 6.41; Cl,
14.20. Found: C, 67.12; H, 6.54; N, 5.80; O, 6.21; Cl, 13.96.
6-Ch lor o-1,2,3,4-t et r a h yd r o-2(R)-(m et h yla m in o)-1(R)-
n a p h th a len ep r op a n oic Acid Meth yl Ester -(2R,3R)-Bis[(4-
m eth ylben zoyl)oxy]bu ta n d ioic Acid (1:1) (18). The meth-
anol filtrates from the isolation of 17 were immediately
concentrated under vacuum to 125 g (<30 °C) and diluted with
CH2Cl2-water (1:1, 550 mL). This mixture was concentrated
and fresh CH2Cl2 (275 mL) was added. The mixture was cooled
(-4 °C, 0.5 h) and filtered under slight positive pressure to give
18 (33.6 g, 92% de). The CH2Cl2 portion of the filtrate was
utilized in the isolation of 17 (see above). A portion of 18 (5.00
g of 33.6 g) was stirred with 5:1 CH2Cl2-water (60 mL) at 35
°C. The mixture was cooled to -10 °C for 1.5 h and filtered to
give 18 (4.63 g, 91% of theory): mp 140-145 °C dec; KF, 0.43%
water; IR (KBr) ν 3800-2500, 1725, 1612 cm-1; UV λ (ꢀ): 270
(2200), 239 (28000), 204 (46200); MS m/ z 282 (M+ for amine),
284 (M+ + 2 for amine), 386 (M+ for DTTA), 668 (M+); HPLC
(system A) 0.48 min (>49% DTTA), 1.28 min (amino ester of
18, >49%); HPLC (system B), 1.71 min (amino ester of 18);
enantiospecific HPLC: 16.1 min (98% de); [R]D -94.5°; [R]365
-445.5° (c ) 1.00, MeOH); 1H NMR (DMSO-d6), identical to the
spectrum of 17; 13C NMR (CDCl3-DMSO-d6) δ 170.57, 165.75,
144.03, 136.81, 133.32, 132.44, 131.10, 130.03, 129.12, 128.62,
128.38, 126.78, 12666, 126.28, 72.98, 60.74, 57.50, 51.70, 40.72,
39.33, 30.92, 30.87, 30.67, 28.10, 27.58, 23.77, 21.66, 20.54.
Anal. Calcd for C35H38ClNO10: C, 62.92; H, 5.73; N, 2.10; Cl,
5.31. Found: C, 62.89 H, 5.58; N, 1.94; Cl, 5.34.
(+)-(4a S,10bS)-4-Meth yl-8-ch lor o-1,2,3,4,4a ,5,6,10b-tr a n s-
octa h yd r oben zo[f]qu in olin -3-on e (4, LY300503). Conver-
sion of 17 (50.0 g, 98% de) as described above provided 4 (18.0
g, 97%). TLC, IR, UV, MS, GC and HPLC (systems A and B)
identical to 3 above; DSC24 73 °(lit.15,17b mp 71-73 °C); chiral
HPLC 14.0 min (98% ee); [R]D +84.5° [R]365 +300° (c ) 1.00,
MeOH). Anal. Calcd for C14H16ClNO: C, 67.33; H, 6.46; N, 5.61;
O, 6.41; Cl, 14.20. Found: C, 67.12; H, 6.54; N, 5.80; O, 6.21;
Cl, 13.96.
Ack n ow led gm en t. We thank Marvin Hansen, Bret
Huff, J ohn Stille, J oseph Kennedy, J ohn Bowers, and
Michael Martinelli for useful input, materials, or de-
velopment of analytical assays. We also thank the
Molecular Structure Division of Eli Lilly and Co.
(-)-(4aR,10bR)-4-Meth yl-8-ch lor o-1,2,3,4,4a,5,6,10b-tr a n s-
octa h yd r oben zo[f]qu in olin -3-on e (3, LY300502). A mixture
of 18 (40.0 g, 99.8% de, anhydrous), toluene (800 mL), and 5%
aqueous sodium bicarbonate (800 mL) was stirred (30 min, 25
°C). The water layer was separated and washed with toluene
(200 mL). The combined organic extracts were washed with
water (100 mL) and saturated sodium chloride (100 mL) and
warmed to 60-80 °C until analysis by HPLC indicated complete
lactamization. The mixture was filtered over 4 Å molecular
sieves and evaporated in vacuo to afford 3 (14.9 g, 100%): TLC,
Su p p or tin g In for m a tion Ava ila ble: 1H NMR for com-
pounds 1, 3, 4, 6, and 8-12, 13 (free amide), 17, 18, HETCOR
data for 11, and an ORTEP plot of 11 (15 pages). This material
is contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
J O9601425