A novel and selective approach to enantiomerically pure bicyclic-trans-lactams via a titanium enolate of a thiopyridyl ester

Full text of DOI:10.1021/jo001364c

334
J . Org. Chem. 2001, 66, 334-336
viscous oil in 99% yield. Alternatively, on a larger scale
A Novel a n d Selective Ap p r oa ch to
En a n tiom er ica lly P u r e
Bicyclic-Tr a n s-La cta m s via a Tita n iu m
En ola te of a Th iop yr id yl Ester
a solvent exchange into dichloromethane (DCM) provided
a solution of the methyl ethers suitable for the next step.
NOE experiments were conducted on the enolate
formed when the known 2-pyridylthioester2 6 was treated
with titanium tetrachloride and di-isopropylethylamine
to generate the titanium enolate. The enolate was
prepared at -20 °C and the sample was then warmed to
the appropriate temperature and spectra recorded. The
ratio of the Z (major) to the E (minor) isomer was
monitored from -10 to +10 °C over 160 min and is
summarized in Table 1. It is apparent that the ratio of
Z/E enolates (initially 88.5:11.5) is slightly lower in this
study than that found by previous authors (95:5).3
Furthermore, it is interesting to note that the enolate
appears to degrade over time and especially at higher
temperatures, regenerating 6 and leading to a lower Z/E
ratio.
The titanium enolate of 6 was treated with 5 at 0-5
°C and led, after workup, to a mixture of the major
diastereoisomer 7 and minor diastereoisomer4 in a ratio
of ca. 12:1. This compares favorably with the 3:1 diaste-
reoisomer ratio seen with silylketene acetals under BF3‚
OEt2 catalysis.1 The ratio of 6:5 required to achieve rapid
conversion and optimum yield was found to be 2:1. No
reaction was seen at -10 °C, and hence, the reaction
temperature was determined by a compromise between
the instability of the enolate and its reactivity.
J ason W. B. Cooke,* Malcolm B. Berry,
Darren M. Caine, Kevin S. Cardwell, J ohn S. Cook, and
Anne Hodgson
Chemical Development, Glaxo Wellcome
Medicines Research Centre, Gunnels Wood Road,
Stevenage, Hertfordshire SG1 2NY, U.K.
jwbc16317@glaxowellcome.co.uk
Received September 14, 2000
The bicyclic-trans-lactam 1 (GW311616A) is a potent
inhibitor of human neutrophil elastase and is of interest
due to its potential as a treatment for respiratory disease
such as chronic bronchitis.1
Due to the requirements of a clinical program, we
needed to develop efficient and robust chemistry that was
suitable for scale-up. In an earlier paper,1 workers from
Glaxo Wellcome revealed several approaches to the
bicyclic-trans-lactam core which introduced the isopropyl
group with 3:1 selectivity in the N-acyliminium ion
coupling chemistry. Furthermore, a trifluoroacetamide
group was used as a protecting group and subsequently
had to be removed. The current work has focused on
improving the selectivity of the coupling and avoiding the
use of additional protecting groups. We have now im-
proved on these earlier routes by introducing a highly
selective titanium enolate N-acyliminium ion coupling
and using the methanesulfonamide group as both a
protecting group and required substituent. A further
advantage of the new route is that the methanesulfona-
mide group acts as an activating group, and hence, the
coupling product 7 can be cyclized directly to bicyclic-
trans-lactam 8 under mild conditions (see Scheme 1).
The dynamic kinetic resolution of racemic aminolactam
2, via the (+)-di-p-toluoyltartaric acid (DPTT) salt 3, has
already been described.1 Conversion of this salt into the
sulfonamide lactam 4 was carried out by dissolution in
hydrochloric acid, filtration of the DPTT, basification with
NaHCO3, and treatment with mesyl chloride to give 4 in
84% yield. The enantiomeric purity of 4 was determined
by HPLC and found to be 70% ee. Reduction of 4 was
carried out with LiBH4 in THF and the resultant “lactol”
converted in situ to the methyl ethers 5 (as an ap-
proximate 1:1 mixture of epimers) with methanolic HCl.
Evaporation of the solvent gave the methyl ethers 5 as a
A possible transition-state model to explain the ob-
served selectivity is presented in Figure 1. The structure
of the titanium enolate of 6 is the same as that proposed
by Cinquini and Cozzi5 but the N-acyliminium ion is not
linked to titanium through nitrogen as with an imine
condensation. It is conceivable that the sulfonamide
coordinates the titanium and this may provide an expla-
nation for the increased selectivity of the NHSO2Me
versus NHCOCF3 analogue (12:1 vs 2.4:1).
Cyclization of the crude 7 was readily effected, due to
the leaving group ability of 2-mercaptopyridine, using
cesium carbonate in DCM to give the CBZ bicyclic-trans-
lactam 8 directly in 33-39% yield for the two steps. The
enantiomeric purity of this material was also determined
by HPLC and found to be 99.4% ee after crystallization.
It was demonstrated, from analyzing crude material, that
the crystallization was crucial to enhance the modest
enantiomeric excess seen in the sulfonamide lactam 4 to
the high levels seen in purified 8.
Deprotection of 8 was carried out under transfer
hydrogenation conditions with Pd/C formic acid. The
resultant lactam was acylated with the known1 piperi-
dinocrotonic acid 9 via a mixed anhydride with isobutyl
chloroformate. The hydrochloride salt of 1 was crystal-
lized from ethyl acetate/ethanol in 76% yield for the two
steps. A single-crystal X-ray structure of 1 (see the
(2) Cinquini, M.; Cozzi, F.; Cozzi, P. G.; Consolandi, E. Tetrahedron
1991, 47, 8767-8774.
(3) Annunziata, R.; Cinquini, M.; Cozzi, F.; Cozzi, P. G. J . Org.
Chem. 1992, 57, 4155-4162. Enolisation was carried out with triethy-
lamine.
(1) Macdonald, S. J . F.; Clarke, G. D. E.; Dowle, M. A.; Harrison, L.
A.; Hodgson, S. T.; Inglis, G. G. A.; J ohnson, M. R.; Shah, P.; Upton,
R. J .; Walls, S. B. J . Org. Chem. 1999, 64, 5166-5175 and references
therein.
(4) The minor diastereoisomer, the isopropyl group epimer, was not
isolated and fully characterized but identified by LCMS.
(5) Benaglia, M.; Cinquini, M.; Cozzi, F. Eur. J . Org. Chem. 2000,
563-572.
10.1021/jo001364c CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/06/2000

Notes
J . Org. Chem., Vol. 66, No. 1, 2001 335
a
Sch em e 1
a
Key: (a) ref 1; (b) HCl then MsCl, NaHCO3, DCM; (c) LiBH4, THF then HCl, MeOH; (d) 6, TiCl4, i-Pr2NEt, DCM then 5; (e) Cs2CO3,
DCM; (f) HCO2H, Pd/C, DMF then 9, i-BuOCOCl, Et3N.
combined organic layer was washed with water and concentrated
to give 4 as a pale pink solid in 84% yield (2.1 g, 6.7 mmol).
Recrystallization from IMS gave colorless needles (1.4 g): mp
116-120 °C; [R]20 ) +44 (c 0.5, MeOH); HPLC tR 3.61min,
D
99.3% area. Chiral HPLC (Chiralcel AD, heptane/ethanol 40:
60, oven ambient, typical retention times: 4, 8.5min; enantiomer,
10.7min) 70% ee; 1H NMR (DMSO) δ 7.71 (1H, d, J ) 9.0), 7.46-
7.34 (5H, m), 5.25 (2H, s), 4.33 (1H, dt, J ) 11.6, 8.8), 3.75 (1H,
t, J ) 9.4), 3.58 (1H, td, J ) 10.5, 6.6), 3.05 (3H, s), 2.36 (1H,
F igu r e 1. Possible transition-state model.
m), 1.83 (1H, m); 13C NMR (DMSO) δ 172.1, 151.2, 136.0, 128.8,
128.5, 128.2, 67.6, 55.1, 42.8, 42.7, 26.2; HRMS calcd for
Ta ble 1. NOE Stu d y of Tita n iu m En ola te of 6
C13H16N2O5S‚H m/z 313.0861, found 313.0858. Anal. Calcd for
time (min)
T (°C)
ratio Z/E
% 6
C
13H16N2O5S: C, 50.0; H, 5.2; N, 9.0; S, 10.3. Found: C, 49.8;
H, 4.9; N, 8.6; S, 10.3.
0
30
60
100
130
160
-10
-10
-10
0
0
10
88.5:11.5
86.6:13.4
85.2:14.8
82.9:17.1
82.5:17.5
82.1:17.9
3.5
6.1
8.6
17.2
31
65.8
Meth yl Eth er s 5. A suspension of sulfonamide lactam 4 (5.0
g, 16.0 mmol) in THF (40 mL) was cooled to -15 °C and treated
with a 2 M solution of LiBH4 in THF (8.2 mL, 16.4 mmol) over
20 min. The solution was allowed to warm to 20 °C over 1 h and
then added dropwise to a solution of HCl in MeOH (from AcCl
2 mL and MeOH 40 mL) at 0 °C over 45 min. The solution was
allowed to warm to 20 °C over 45 min and then added dropwise
to a solution of saturated NaHCO3 (40 mL) over 1 h at 20 °C.
The solution was concentrated to 40 mL and then extracted with
DCM (2 × 40 mL). The extracts were concentrated to an oil,
redissolved in DCM (80 mL), and concentrated to give 5 (as a
mixture of diastereoisomers) as a viscous, pale yellow oil (5.2 g,
15.8 mmol, 99%): IR 1696; HPLC tR 3.99 min, 39.6% area, tR
Supporting Information) was determined which con-
firmed both the relative and absolute configuration.
In conclusion, this N-acyliminium ion condensation
with a titanium enolate of a 2-pyridylthioester represents
a novel and selective method for the synthesis of these
types of molecules. Furthermore, 2-mercaptopyridine is
readily displaced offering a mild cyclization method to
the sensitive bicyclic-trans-lactam core.
1
4.03 min, 48.2% area; H NMR (DMSO at 100 °C) δ 7.38-7.31
(5H, m), 7.07 and 6.62 (1H, 2 × br d), 5.18-4.99 (3H, m), 3.82-
3.76 (1H, m), 3.63-3.25 (2H, m), 3.38 and 3.30 (3H, 2 × s), 2.94
and 2.93 (3H, 2 × s), 1.25-0.93 (2H, m); 13C NMR (DMSO at
100 °C) δ 137.5, 137.4, 129.0, 128.5, 128.4, 128.2, 128.1, 94.0,
87.8, 67.1, 67.0, 57.8, 57.2, 55.9, 55.3, 44.6, 43.4, 41.7, 29.0, 28.1;
HRMS calcd for C14H20N2O5S‚NH4 m/z 346.1437, found 346.1432.
Exp er im en ta l Section
Melting points are uncorrected. 1H NMR spectra were re-
corded at 400 MHz, all coupling constants are in Hz and 13C
NMR spectra at 100 MHz. FTIR spectra were recorded in Nujol
mull. Elemental analyses were performed by Butterworth
Laboratories Limited. HPLC was carried out using the following
method: a gradient from 100% A to 95% B over 8min (A ) water/
TFA 1000:0.5. B ) acetonitrile/TFA 1000:0.5); column 50 mm
× 2.0 mm Luna C18(2), 3µm; flow rate 1.0 mL/min; temperature
40 °C; UV detection at 220 nm. HRMS were run on a quadrupole
time-of-flight mass spectrometer using +ve ion electrospray and
erythromycin as lock mass. Chiral HPLC was carried out using
the following method: isocratic with heptane/ethanol mixtures,
column Chiralcel AD or OD-H 250 × 4.6 mm, 5 µm; flow rate
1.0 mL/min; temperature ambient - 50 °C; UV detection at 215
nm.
Su lfon a m id e La cta m 4. A solution of 3 (3.4 g, 4.0 mmol) in
2 M hydrochloric acid (30 mL, 60 mmol) was stirred at 20 °C for
1 h. The solid precipitate (DPTT) was filtered off and washed
with water (10 mL). To the filtrate was added a solution of mesyl
chloride (0.93 mL, 11.9 mmol) in DCM (30 mL) followed by
aqueous NaHCO3 (50 mL) and the mixture stirred vigorously
for 30 min. Aqueous NaHCO3 (5 mL) was added and stirring
continued for 18 h with further additions of aqueous NaHCO3
to maintain the mixture at pH 8. The organic layer was
separated and the aqueous layer extracted with DCM. The
P yr r olid in e Th ioester 7. A solution of 2-pyridylthioester 6
(5.4 g, 27.6 mmol) in DCM (80 mL) was cooled to 0 °C and treated
with titanium tetrachloride (3.1 mL, 28.2 mmol) dropwise over
10 min. To the mixture was added diisopropylethylamine (4.9
mL, 28.2 mmol) over 5 min to generate the titanium enolate.
To the black enolate solution was added a solution of 5 (4.5 g,
13.7 mmol) in DCM (10 mL). The solution was stirred at 0-4
°C for 1 h and then quenched with 5% w/v aqueous citric acid
(30 mL). The phases were separated, and the organic layer was
washed with 5% w/v aqueous citric acid (30 mL) and brine (30
mL). The solution was concentrated to an orange oil (containing
mainly 6, 7, and minor amounts of the isopropyl group epimer:
HPLC-MS tR 4.99min, 2.7% area, tR 5.19min, 32.1% area, both
peaks give m/z 492 MH+) which can either be used directly in
the cyclization step or purified for characterization. The major
diastereoisomer 7 was isolated by crystallization from ethyl
acetate and cyclohexane (1:1) in 25% yield. Data for the major
isomer 7: mp 142-143 °C; [R]20 ) +8 (c 0.5, MeOH); IR 1675;
D
1H NMR (DMSO, mixture of rotamers, ratio 5:3) δ 8.63 (1H, br
d, J ) 3.0), 7.91 (1H, td, J ) 7.7, 1.8), 7.61-7.27 (8H, m), 5.24-
5.01 (2H, m), 4.12-4.00 (2H, m), 3.66-3.56 (1H, m), 3.30-3.24
(1H, m), 3.13 (1H, dd, J ) 9.0, 5.4), 2.92, 2.90 (3H, 2s), 2.17-
1.75 (3H, m), 1.06, 0.96, 0.92, 0.83 (6h, 4d, J ) 6.5); 13C NMR

336 J . Org. Chem., Vol. 66, No. 1, 2001
Notes
(DMSO) δ 199.3, 199.2, 154.6, 151.0, 150.8, 138.2, 137.4, 130.44,
128.8, 128.7, 128.5, 128.4, 128.1, 127.7, 124.6, 66.9, 66.4, 64.9,
64.4, 63.0, 61.8, 54.9, 54.0, 45.3, 44.9, 41.3, 41.2, 31.5, 30.7, 28.1,
28.0, 21.2, 21.1, 21.0, 20.9; HRMS calcd for C23H29N3O5S2‚H m/z
492.1627, found 492.1621. Anal. calcd for C23H29N3O5S2: C, 56.2;
H, 6.0; N, 8.6; S, 13.0. Found: C, 56.1; H, 5.8; N, 8.3; S, 12.9.
CBZ Bicyclic-Tr a n s-La cta m 8. A solution of the crude
pyrrolidine thioester 7 in DCM (70 mL) was treated with cesium
carbonate (7.5 g, 23.0 mmol) and the mixture stirred at 20 °C
overnight. Water (30 mL) was added and stirred to dissolve the
solids before the phases were separated. The DCM solution was
washed with 2 M HCl (30 mL) and then concentrated to 25 mL.
Ethyl acetate (25 mL) was added and the solution concentrated
to 25 mL, which initiated crystallization. The suspension was
treated with cyclohexane (50 mL) over 30 min at 55 °C and then
cooled to 20 °C and stirred for 5 h. The solid was collected by
filtration, washed with cyclohexane (2 × 40 mL), and dried in
vacuo overnight to give 8 in 39% yield over two steps (2.05 g,
mL, 32.3 mmol) and stirred for 15 min before the addition of
isobutyl chloroformate (3.8 mL, 29.3 mmol). The suspension was
stirred for 30 min at 5 °C before the solution of deprotected
sulfonamide was added. The mixture was stirred at 23 °C for
16 h and then the solids were filtered off and the filtrate
concentrated to 45 mL. The concentrate was partitioned between
EtOAc (100 mL) and 10% aqueous KHCO3 (100 mL). The
aqueous layer was separated and extracted with EtOAc (100
mL). The organic layers were washed with brine and then
combined and stirred for 15 min with Norit charcoal (2 g). The
mixture was filtered through Harborlite and concentrated to 32
mL. More EtOAc (40 mL) and EtOH (20 mL) were added, and
the solution was warmed to 30 °C and treated with acetyl
chloride (3 mL, 42 mmol). The resulting suspension was cooled
in ice, and the solid was collected by filtration, washed with a
mixture of EtOAc (52 mL) and EtOH (8 mL), and dried in vacuo
at 45 °C to give 1 as a white solid in 76% yield (8.7 g, 20 mmol):
mp 235 °C dec; [R]20 ) + 52 (c 0.5, MeOH); HPLC tR 3.00min,
D
5.4 mmol): mp 182 °C; [R]20 ) +59 (c 0.5, DCM); HPLC tR
99.7% area. Chiral HPLC (Chiralcel OD-H, heptane/ethanol 90:
10, oven 50 °C, typical retention times: 1 at 30.4 min, enanti-
omer at 22.7 min) 99.9% ee; IR 1743, 1667, 1626; 1H NMR
(CD3OD + DCl) δ 6.76 (2H, m), 4.02 (2H, m), 3.95 (2H, m), 3.73
(1H, td, J ) 11.7, 5.2), 3.49 (3H, m), 3.30 (1H, pent, J ) 1.6),
3.26 (3H, s), 3.03 (4H, m), 2.52 (1H, m), 2.15 (1H, m), 1.97-1.79
(5H, m), 1.53 (1H, m), 1.25, 1.00 (6H, 2d, J ) 7.1).
D
5.13min, 0.2% area (isopropyl group epimer), tR 5.28min, 98.8%
area (8); chiral HPLC (Chiralcel OD-H, heptane/ethanol 90:10,
oven 30°, typical retention times: 8 at 25 min, enantiomer at
24 min) 99.4% ee; IR 1749, 1707; 1H NMR (CDCl3) δ 7.39-7.32
(5H, m), 5.12, 5.10 (2H, AB system, J ) 12.0), 3.87-3.75 (2H,
m), 3.47 (1H, m), 3.29 (1H, d, J ) 11.7), 3.24 (3H, s), 2.77 (1H,
d, J ) 11.8), 2.70 (1H, br s), 2.55 (1H, m), 2.02 (1H, m), 1.16,
0.95 (6H, 2d, J ) 6.4).
Bicyclic-Tr a n s-La cta m 1. To a mixture of 8 (10 g, 26.3
mmol) and 10% palladium on charcoal (4 g, dry powder) in DMF
(200 mL) at 40 °C was added 98-100% formic acid (1.5 mL, 40
mmol) and the mixture stirred for 2.5 h. The reaction was cooled
to 20 °C before cyclohexene (3 mL, 30 mol) was added. After
being stirred for 1 h, the reaction mixture was filtered through
Harborlite and concentrated to 70 mL. A suspension of 9 (6.6 g,
32 mmol) in DMF (150 mL) at 5 °C was treated with Et3N (4.5
Ackn owledgm en t. We are grateful to Athina Chatzi,
Chris J ones, Sean Lynn, Nisha Mistry, Andrew Ray,
Alec Simpson, and Dave Sugden for analytical support.
Su p p or tin g In for m a tion Ava ila ble: Crystal structure
data for bicyclic-trans-lactam 1. This material is available free
of charge via the Internet at http://pubs.acs.org.
J O001364C