Synthesis of (+/-) - oxohexahydrofuro[3,2-b]pyrroles (pyrrolidine-trans-lactones) using acyl-iminium chemistry

Article abstract of DOI:10.1055/s-1998-1965

In a new synthesis of pyrrolidine-trans-lactones, an acyliminium pyrrolidine reacts with silyl ketene acetals. This reaction selectively generates C2, C3 trans stereochemistry.

Full text of DOI:10.1055/s-1998-1965

December 1998
SYNLETT
1375
Synthesis of (+/-) - Oxohexahydrofuro[3,2-b]pyrroles (Pyrrolidine-trans-lactones) Using
Acyl-Iminium Chemistry
*
Simon J.F. Macdonald, Julie E. Spooner, and Michael D. Dowle
Enzyme Chemistry II, GlaxoWellcome Medicines Research Centre, Gunnels Wood Road, Stevenage, SG1 2NY, UK
Fax: +(0) 1438 763 616
Received 8 October 1998
Abstract: In a new synthesis of pyrrolidine-trans-lactones, an acyl-
iminium pyrrolidine reacts with silyl ketene acetals. This reaction
selectively generates C2, C3 trans stereochemistry.
Serine protease inhibition is a field of enormous importance to the
pharmaceutical industry in the search for new therapies for, amongst
others, respiratory and cardiovascular diseases. Therefore our
Scheme 2
1
2
3,4
introduction of a new scaffold known as pyrrolidine trans-lactones
1
4
(and lactams), which are strained 5,5-trans fused bicycles, is both
timely and significant. We have studied these systems during a
programme to develop inhibitors of human neutrophil elastase (HNE)
for respiratory diseases and which has led to development candidates.
We required a flexible approach allowing stereocontrol and facile
introduction of substituents α to the lactone carbonyl (eg n-propyl in 1).
cyclised to give the protected cyclic aminal 7. Heating a toluene solution
of 7 at reflux gave the 2-pyrroline 5 in 62% yield over the 3 steps.
We then examined the potential of the pyrroline 5 to act as an epoxide
precursor. Although oxidations of 2-pyrrolines are now known, at the
time of this work we were unable to find precedent for such
6,7
4a
Our previous synthesis featured the creation of the pyrrolidine ring
and a stereocentre via an intramolecular Michael reaction. However it
lacked stereoselectivity and as inhibition of HNE requires specific
relative stereochemistry (as in 1) it was inefficient. This letter describes
a new route which generates the stereocentres on a preformed
pyrrolidine. It addresses some of the stereochemical issues, uses
chemistry suitable for larger scale work and avoids, for example, low
4a
temperature alkylations.
In our retrosynthesis, concern for the potential lability of the highly
strained trans-lactone meant that its creation was deferred to the end of
the synthesis: hence the initial disconnection to the hydroxy-acid 2
(Scheme 1). This in turn was disconnected to either the epoxide 3 or the
diol 4 – a convenient acyl-iminium precursor. Both compounds may be
derived from the benzyl carbamate of 2-pyrroline 5.
Scheme 3
transformations. Whilst conventional peracid epoxidation in
6b
dichloromethane failed to provide identifiable products, treatment
with N-bromosuccinimide in aqueous dioxan gave the bromohydrin 8 as
1
8
a single diastereomer (Scheme 3). H-NMR studies confirmed the
regio- and stereo-selectivity across C2 and C3 which suggested that the
reactivity profile of the epoxide 3 might provide the required control to
access the products of interest (such as 2). If pure pyrroline 5 is used the
crude bromohydrin 8 is stable when stored below room temperature.
o
After treatment of the bromohydrin with sodium hydride in THF (-70 C
to RT), (which produced a precipitate - presumed to be NaBr), addition
of the sodium salt of diethyl malonate gave a low yield of the alcohol 9
9
as the only isolated product; the hydroxyl had been transposed from C2
to C3 suggestive of an epoxide-like intermediate. In a separate
experiment, after treatment of 8 with NaH, dimethylketene acetal and
Scheme 1
BF ·OEt were added. This gave the bromo-ester 10 (19%) and the
3
2
10
5
reduced bromide 11 (8%)
suggesting that an acyl-iminium
Although Thaning and Wistrand have described preparations of
intermediate rather than an epoxide intermediate was more likely.
Whilst unoptimised, the poor yields of these processes led us to focus on
the diol 4 (R = H) as an acyl-iminium precursor. There is precedent for
analogues of 4 an alternative procedure was developed from the
2-pyrroline 5. This would allow access to 3 or 4 and is also potentially a
precursor for asymmetric oxidations which have just recently been
11
6
such reactions on pyrrolidine derivatives.
described for these substrates. Our synthesis (Scheme 2) started from
Dihydroxylation of 5 under standard conditions gave a mixture of the
cis and trans diols 12 in 91% crude yield (Scheme 4). Acetylation
commercially available 4-aminobutanal diethylacetal 6 which was
protected as its benzyl carbamate and then treated with pyridinium
tosylate in aqueous acetone. The resultant aldehyde spontaneously
12
gave
a
mixture of di-acetates 13 in 32-49% yield after

1376
LETTERS
SYNLETT
Acknowledgements: We are grateful for analytical and spectroscopic
assistance from Trevor Cholerton, Steve Richards, Dave O’Brian, Ian
Davidson and Dave Paul. We also thank Professor Tim Gallagher and
Lee Harrison for assistance and advice in the preparation of the
manuscript.
References and Notes.
(1) Vender, R.L. J. Invest. Med. 1996, 44, 531
(2) Ripka, W.C. Curr. Opin. Chem. Biol. 1997, 1, 242.
(3) The IUPAC name for pyrrolidine trans-lactones is 5-oxo-
hexahydrofuro[3,2-b]pyrroles.
(4) (a) Macdonald, S.J.F., Belton, D.J., Buckley, D.M., Spooner, J.E.,
Anson, M.S., Harrison, L.A., Mills, K., Upton, R.J., Dowle, M.D.,
Smith, R.A., Risley, C., Molloy, C.J. J. Med. Chem., in press. (b)
Dowle, M.D.; Finch, H.; Harrison, L.A.; Inglis, G.G.A.; Johnson,
M.R.; Macdonald, S.J.F.; Shah, P.; Smith, R.A. WO 9736903 A1
971009.
(5) (a) Thaning, M.; Wistrand, L-G. J. Org. Chem. 1990, 55, 1406. (b)
Thaning, M.; Wistrand, L-G. Acta Chem. Scand. 1989, 43, 290. (c)
Thaning, M.; Wistrand, L-G. Acta Chem. Scand. 1992, 46, 194.
Scheme 4
(6) (a) Sunose, M.; Anderson, K.M.; Gallagher, T.; Macdonald, S.J.F.
Tetrahedron. Lett. in press. (b) Sugisaka, C.H.; Carroll, P.J.;
Correia, C.R.D. Tetrahedron Lett. 1998, 39, 3413.
13
chromatography which, as expected, proved to be ideal acyl-iminium
ion precursors.
Reaction of 13 with trimethylallylsilane and BF ·OEt gave the 2-
allylpyrrolidine 15 in 80% yield as a 3:1 mixture of trans:cis isomers as
previously described in the literature. Treatment of the di-acetate 13
with substituted ketene acetals and BF ·OEt gave good yields of all
3
2
(7) Burgess, L.E.; Gross, E.K.M.; Jurka, J. Tetrahedron Lett. 1996,
37, 3255.
5a
1
(8) Data for 8: H-NMR (250MHz, CDCl ) δ 7.40-7.25 (m, 5H,
3
3
2
C H ), 5.67 (d, J = 3Hz, 0.67H, major rotamer CHOH), 5.61 (d, J
trans products across C2-C3 of the pyrrolidine 16, 17 and 18 and β and
α mixtures of the n-propyl group for 17 and 18. No cis-products across
C2-C3 were detected. We believe the trans stereoselectivity is
attributable to the greater steric bulk of the ketene acetals and the
conformation of the cations 14 and 19 which preferentially allow
nucleophilic attack from the less hindered exo face. It is interesting to
note the improved α-product selectivity when the bulk of the silyl
group of the ketene acetal group is increased. Further variation of the
O-substituents and/or the geometry of the ketene acetal and the Lewis
6 5
= 3Hz, 0.33H, minor rotamer CHOH), 5.22-5.10 (m, 2H, PhCH ),
2
4.22 (m, 1H, CHBr), 3.73-3.62 (m, 2.67H, NCH + major rotamer
2
CHOH), 3.33 (d, J = 3Hz, 0.33H, minor rotamer CHOH), 2.74-
2.56 (m, 1H, NCH HCH), 2.24-2.13 (m, 1H, NCH HCH);
2
2
assignments were supported by decoupling experiments and
o
warming to 55 C coalesced the signals at δ 5.67 and 5.61, the
OH’s. The other signals simplify.
1
o
13
(9)
H-NMR at 55 C, decouplings and C-NMR studies support the
14
acid may affect the selectivity further. The stereochemical ratios of the
regiochemistry of 9. The key H2 proton is masked by the CH ’s of
2
15
β:α propyl groups in 17 and 18 were established by analytical HPLC.
the malonate. Spectroscopic correlation with 20, and the fact that
cis-hydroxyester affords spontaneously cis-lactone, supports the
trans assignment.
This was subsequently confirmed by conversion of 17 into the trans-
lactone 22 (Scheme 5).
1
o
(10) Data for 10: H-NMR (400MHz, CDCl , 50 C), δ 7.38-7.23 (m,
3
5H, C H ), 5.19 (s, 2H, PhCH ), 4.62 (s, 1H, NCHCMe ), 4.50
6
5
2
2
(broad d, J = 3Hz, 1H, CHBr), 3.96 (m, 1H, NHCH), 3.66 (s, 3H,
CO CH ), 3.45 (m, 1H, NHCH), 2.38 (m, 1H, NCH HCH), 2.26
2
3
2
(m, 1H, NCH HCH); assignments supported by decoupling and
2
HMQC experiments.
(11) On pyrrolidones: (a) Dai, W-M.; Nagao, Y.; Fujita, E.
Heterocycles 1990, 30, 1231 and references therein. (b) Bernadi,
A.; Micheli, F.; Potenza, D.; Scolastic, C.; Villa, R. Tetrahedron
Lett. 1990, 31, 4949. (c) Koot, W-J.; Ginkel, R.; Kranenburg, M.;
Hiemstra, H.; Louwrier, S.; Moolenaar, M.J.; Speckamp, W.N.
Tetrahedron Lett. 1991, 32, 401; (d) Pilli, R.A,; Russowsky, D. J.
Org.Chem. 1996, 61, 3187. On carbamates of pyrrolidines: (f)
Kraus, G.A.; Neueschwander, K. J. Chem. Soc., Chem. Commun.,
1982, 134. (g) Asada, S.; Kato, M.; Asai, K.; Ineyama, T.; Nishi,
S.; Izawa, K.; Shono, T. J. Chem. Soc., Chem. Commun. 1989,
486; (h) David, M.; Dhimane, H.; Vanucci-Bacque, C.; Llommet,
G. Synlett 1998, 206.
Scheme 5
16
Removal of the acetate from 17 and saponification under standard
conditions gave efficiently the hydroxy-acid 21. Yamaguchi
cyclisation at 0.025mM in toluene gave a 53% yield of 1:2 β:α propyl-
17
1
18,19
trans-lactones 22 (ratio determined by H-NMR).
Analysis of the
1
H-NMR spectra was aided by comparison with the spectra of pure
2
samples of α and β analogues prepared by other routes.
1
NB. All new compounds gave satisfactory H-NMR, IR and
microanalytical data.
(12) The mixture of cis and trans diols may indicate the facile
formation of an acyl-iminium ion from the cis diol which reacts

December 1998
SYNLETT
1377
with water from the less hindered face to give the trans diol, or
ring opening of the cyclic aminal and reclosing.
The HPLC data for 20 was obtained on an ABZ column at 215nm
with isochratic elution (30% MeCN in H O 0.12%
+
2
trifluoroacetic acid) with a flow rate of 1mL/min. The α-isomer of
20 retained at 32.31min and the β-isomer at 36.26min. Data for
(13) The yield was dependent on the mode of work-up and/or the
stability of the product to chromatography on silica gel.
1
o
20: H-NMR (250MHz, CDCl , 55 C) δ 7.40-7.28 (m, 5H), 5.15
3
(14) We are exploring such variation with the 3-aminopyrrolidine
(s, 2H), 4.46-4.37 (m, 1H), 4.18-4.05 (m, 2H), 3.99-3.90 (m, 1H),
3.83-3.63 (m, 1H), 3.43-3.30 (m, 1H), 3.15-2.97 (bm, 0.6H), 2.84-
2.62 (bm, 0.4H), 2.28-1.13 (m, 10H), 0.85 (t, J = 7Hz, 3H).
4
series to access trans-lactams.
The experimental procedure for the preparation of 18 is
representative. To 2,3-diacetoxy-1-benzyloxycarbonylpyrrolidine
13 (0.049g, 0.153mmol) and the ketene acetal - (E)-(1,1-
dimethylethyl)[(1-ethoxy-1-butenyl)oxy]dimethylsilane - (0.161g,
(17) Inanaga, J.; Katsuki, S.; Takimoto, S.; Ouchida, S.; Inone, K.;
Nakano, A.; Okukuda, N.; Yamaguchi, M. Chem. Lett. 1979,
1021.
o
0.659mmol) in dry dichloromethane (5mL) at –70 C was added
1
o
boron trifluoride etherate (37.6µL, 0.306mmol) dropwise over
(18) Data for 22: H-NMR ( 250MHz, CDCl , 55 C) δ 7.39-7.26 (m,
3
o
5min. After 2h at –70 C, tlc showed absence of starting material
5H), 5.23-5.03 (m, 2H), 4.26-4.13 (m, 0.6H), 3.98-3.47 (m, 3H),
3.29-3.18 (m, 0.4H), 3.03-2.88 (m, 0.6H), 2.83-2.73 (m, 0.4H),
2.41-2.28 (m, 1H), 2.13-1.15 (m, 5H), 0.91-0.78 (m, 3H). This
data correlates with material available by a different route
so water (10mL) was added and the aqueous layer separated. After
extraction with dichloromethane (2 x 7mL), the combined organic
layers were washed with brine (7mL), dried (MgSO ), and
4
concentrated in vacuo. Flash chromatography on silica (Merck
9385) eluting with 7:3 hexane:ethyl acetate afforded 18 (β:α
30:70) (0.041g, 68%).
(unpublished results); IR (KBr diffuse reflectance) ν
1715 cm .
1796,
C=O
-1
(19) The dimethyl analogue 16 was processed in an identical manner to
17 with yields of 93% for the hydrolyses and 61% for the trans-
Analytical HPLC data was obtained on an Inertsil ODS2 column
1
at 215nm with gradient elution, solvent A (H O + 0.1% H PO )
2
3
4
lactonisation. Data for the dimethyl trans-lactone: H-NMR
o
and solvent B (95% MeCN + 5% H O 0.1% H PO ) 0 to 100% B
2
3
4
(250MHz, CDCl , 55 C) δ 7.38-7.28 (m, 5H, C H ), 5.16 (d, J =
3
6 5
over 40min with a flow rate of 1mL/min. The α-isomer of 17 and
18 retained at 31.61min and the β-isomer at 31.94min.
12Hz, 1H, PhHCH), 5.10 (d, J = 12 Hz, 1H, PhHCH), 4.06 (ddd, J
= 10, 10, 5.5Hz, 1H, H3), 3.89 (t, J = 10Hz, 1H, H5α), 3.68 (ddd,
J= 10, 10, 5.5Hz, 1H, H5β), 3.23 (d, J = 10Hz, 1H, H2), 2.38-2.28
(m, 1H, H4β), 2.10-1.92 (m, 1H, H4α), 1.45 (s, 3H, Me), 1.17 (s,
(16) After removal of the acetate there was no evidence for the cis-
hydroxyester or the corresponding cis-lactone. The selective
hydrolysis of the acetate was carried out to correlate 20 with
material available from a different route (unpublished results).
3H, Me) assignments confirmed by decoupling experiments; IR
-1
(CHCl solution) ν
1791, 1705 cm .
3
C=O