Enantioselective synthesis of (R)- and (S)-N-Boc-morpholine-2-carboxylic acids by enzyme-catalyzed kinetic resolution: application to the synthesis of reboxetine analogs

Article abstract of DOI:10.1016/j.tetlet.2008.11.025

The (R)- and (S)-N-Boc-morpholine-2-carboxylic acids 9 and 10 were prepared using an enantioselective synthesis employing a highly selective enzyme-catalyzed kinetic resolution of racemic n-butyl 4-benzylmorpholine-2-carboxylate (11) as the key step. Acids 9 and 10 were then converted efficiently and stereoselectively to reboxetine analogs 3 and 4.

Full text of DOI:10.1016/j.tetlet.2008.11.025

Tetrahedron Letters 50 (2009) 389–391
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Enantioselective synthesis of (R)- and (S)-N-Boc-morpholine-2-carboxylic
acids by enzyme-catalyzed kinetic resolution: application to the synthesis
of reboxetine analogs
*
*
Paul V. Fish , Malcolm Mackenny , Gerwyn Bish, Timothy Buxton, Russell Cave, David Drouard,
David Hoople, Alan Jessiman, Duncan Miller, Christelle Pasquinet, Bhairavi Patel, Keith Reeves,
Thomas Ryckmans, Melanie Skerten, Florian Wakenhut
Discovery Chemistry, Pfizer Global Research and Development, Sandwich Laboratories, Ramsgate Road, Sandwich, Kent CT13 9NJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The (R)- and (S)-N-Boc-morpholine-2-carboxylic acids 9 and 10 were prepared using an enantioselective
synthesis employing a highly selective enzyme-catalyzed kinetic resolution of racemic n-butyl 4-ben-
zylmorpholine-2-carboxylate (11) as the key step. Acids 9 and 10 were then converted efficiently and ste-
reoselectively to reboxetine analogs 3 and 4.
Received 29 September 2008
Revised 24 October 2008
Accepted 7 November 2008
Available online 12 November 2008
Ó 2008 Elsevier Ltd. All rights reserved.
O
O
( )-Reboxetine is an orally active, selective noradrenaline reup-
take inhibitor (NRI) developed and launched for the treatment of
depression.1,2 The (+)-(S,S)-enantiomer of reboxetine (1) is cur-
rently undergoing advanced clinical evaluation as a potential treat-
ment of fibromyalgia and neuropathic pain.3 Esreboxetine (1) is
more potent than the reboxetine racemate with respect to inhibit-
ing the uptake of noradrenaline and much weaker at inhibiting up-
take of serotonin; the combination of these two pharmacologies
makes esreboxetine highly selective.2
S
R
NH
NH
R
S
O
O
Cl
EtO
MeO
2: ( )-(RR,SS)
3: (-)-(R,R)
1: (+)-(S,S)-esreboxetine
Further investigation of the reboxetine template identified ( )-2
as a lead with potential biological activity for dual noradrenaline
and serotonin reuptake inhibition.4,5 Evaluation of the single enan-
tiomers 3 and 4 in pre-clinical disease models required the synthe-
sis of multigram quantities of suitably functionalized, homochiral
morpholine synthons. In this Letter, we disclose a new and highly
efficient synthesis of homochiral (R)- and (S)-N-Boc-morpholine-2-
carboxylic acids 9 and 10, and describe their conversion to reboxe-
tine analogs 3 and 4.
The challenge of creating the molecular architecture of esre-
boxetine (1) has led to a number of noteworthy syntheses in recent
years (Fig. 1). Henegar and Cebula reported a process development
for the synthesis of esreboxetine succinate employing a Sharpless
asymmetric epoxidation of cinnamyl alcohol to epoxy alcohol 5.6
Esreboxetine has also been prepared enantioselectively from cin-
namyl bromide using a Sharpless asymmetric dihydroxylation as
the key step to furnish diol 6.7 In a conceptually different approach,
esreboxetine has been prepared from commercially available chiral
4: (+)-(S,S)
pool reagents, (1R,2R)-2-amino-1-phenyl-1,3-propanediol (7)8 and
(S)-3-amino-1,2-propanediol (8).9
Our approach to the reboxetine template, as exemplified by
compounds 3 and 4, employed a new enantioselective synthesis
OH
O
Br
OH
OH
5
6
(S,S)-1
OH OH
HO
NH2
OH
NH2
* Corresponding authors. Tel.: +44 (0)1304 644589 (P.V.F.); tel.: +44 (0)1304
644697 (M.M.).
E-mail addresses: paul.fish@pfizer.com (P.V. Fish), malcom.mackenny@pfizer.
com (M. Mackenny).
8
7
Figure 1. Chiral synthons employed for the synthesis of esreboxetine 1.
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.11.025

390
P. V. Fish et al. / Tetrahedron Letters 50 (2009) 389–391
where the key (R)- and (S)-morpholine acids 9 and 10 were pre-
pared by an enzyme-catalyzed kinetic resolution of racemic ester
11 (Scheme 1).10,11 This route was attractive to us as it allowed
for the late-stage installation of both the aryl and aryloxy rings
providing flexibility in the synthesis of additional analogs. Further-
more, the resolution proved to be highly selective and operation-
ally simple to perform on large scale.
Racemic ester 11 was prepared in two steps and in high yield
from readily available starting materials. Condensation of N-ben-
zylethanolamine (12) with 2-chloroacrylonitrile followed by t-
BuOK-promoted cyclization gave morpholine nitrile 13,12 which
then underwent alcoholysis with n-BuOH (or other alcohols, e.g.,
EtOH) in the presence of concd H2SO4 to give 11.
nation of the n-Bu ester with lipase Candida rugosa13 was com-
pletely stereoselective, with the enzyme catalyzing the
hydrolysis of the (S)-ester to give (S)-acid 15 whilst leaving the
(R)-ester 14 untouched.14,15 On scale-up, resolution of 11 (10 g)
gave ester 14 (4.7 g, 94%; >99% ee) and acid 15 (3.9 g, 97%; >99%
ee). The isolation of the ester 14 was straightforward as the organic
layer was simply separated, dried, and concentrated in vacuo. In
these early experiments, the reaction was buffered to pH 7.2, and
so isolation of acid 15 from the aqueous phase required the use
of ion exchange chromatography (Dowex 50W8X200 eluting with
aqueous ammonia). The resolution was subsequently performed in
water to give 14 and 15 in identical yields as before and with no
loss in selectivity. The isolation of both ester 14 and acid 15 could
now be achieved simply by separating the two layers and evapo-
rating the solvent. This resolution has been performed on >100 g
scale.16
Racemic ester 11, along with other esters (e.g., Et), were
screened against a panel of 19 broadly specific lipases. The combi-
The N-benzyl-protecting group of 14 and 15 was then ex-
changed to the corresponding N-Boc amines,17 and base hydrolysis
of the n-butyl ester gave the acid 9 which was isolated as the lith-
ium salt 9a for convenience. Hence, (R)- and (S)-morpholine acids 9
and 10 were prepared efficiently in just five and four steps with
overall yields of 46% and 53%, respectively; both acids were ob-
tained with very high chiral purity (>99% ee).
The conversion of acid 9 to 3 is outlined in Scheme 2, where
the second stereocenter was introduced by a diastereoselective
reduction of ketone 17. Activation of 9 with 1-propanephos-
phonic anhydride (T3P) followed by reaction with HN(Me)OMe
gave Weinreb amide 16 in good yield (>97% ee), and then treat-
ment of 16 with the Grignard reagent PhMgBr gave the phenyl
ketone 17 with no detectable loss in ee. Reduction of 17 was
achieved most selectively with zinc borohydride18 to give
(S,R)-alcohol 18 with creation of the second stereocenter with
good diastereoselectivity (S,R:R,R 16:1). The high diastereoselec-
tivity was attributed to a Zn-chelated transition state (I); this
outcome is entirely consistent with the model of Frein and
Rovis.19
HO
O
a, b
HN
Ph
N
Ph
NC
12
13
c
O
Ph
n-BuO
N
O
11
{
{
Zn
H
d
O
H
O
B
H
N
H
Ph
Boc
O
R
O
S
H
Ph
n-BuO
N
N
Ph
HO
+
I
O
O
14: 94% yield
>99% e.e.
15: 97% yield
>99% e.e.
The introduction of the aryl ether was most efficiently achieved
from 18 with a two-step process of mesylation and displacement.
Reaction of 18 with MeSO2Cl gave mesylate 19, and displacement
of the MsO group of 19 with 4-chloro-2-methoxyphenol furnished
the corresponding (R,R)-aryloxy ether 20 in good yield, with com-
plete inversion of the benzylic stereocenter and no detectable loss
of ee. This two-step process proved to be superior to Mitsunobu
reactions with alcohol 18 as reaction rates were faster, yields were
higher, and product isolation was straightforward. Finally, depro-
tection of the N-Boc amine of 20 with HCl afforded the (R,R)-amine
3. The (S,S)-enantiomer 420 was prepared by an identical sequence,
but starting with the (S)-acid 10.
e, f
e
O
O
N
N
Boc
HO2C
Boc
HO2C
R
S
In summary, (R)- and (S)-N-Boc-morpholine-2-carboxylic
acids 9 and 10 were prepared using an enantioselective syn-
9
10
9a: Li+ salt
thesis employing
resolution of racemic n-butyl 4-benzylmorpholine-2-carboxyl-
ate (11) as the key step. Acids and 10 were then con-
a highly specific enzyme-catalyzed kinetic
Scheme 1. Reagents and conditions: (a) CH2@C(Cl)CN, Et2O, 40 °C, quant.; (b) t-
BuOK, DME, reflux 65%; (c) n-BuOH, concd H2SO4, reflux, 85%; (d) lipase Candida
rugosa, t-BuOMe–H2O, rt, 24–30 h; (e) 2,5-dihydrotoluene, 10%-Pd/C, Boc2O, EtOH,
reflux, quant.; (f) LiOH (0.95 equiv), THF–H2O, 0 °C, 90%.
9
verted efficiently and stereoselectively to reboxetine analogs
3 and 4.

P. V. Fish et al. / Tetrahedron Letters 50 (2009) 389–391
391
O
R
O
O
R
c
b
S
X
N
N
N
Boc
Boc
Boc
R
O
O
OH
9: X = OH
17
18
a
16: X = N(Me)OMe
d
O
O
R
O
R
e
R
O
f
R
O
S
NH
N
N
Boc
Cl
Boc
R
OMs
MeO
(-)-(R,R)-3
Cl
MeO
20
19
Scheme 2. Reagents and conditions: (a) T3P, HN(Me)OMe, NEt3, CH2Cl2, then aq K2CO3, 85%; (b) PhMgBr, THF, rt, 98%; (c) Zn(BH4)2, Et2O, rt, 87%, ratio (S,R):(R,R) 16:1; (d)
MeSO2Cl, NEt3, CH2Cl2, 0 °C, quant.; (e) 2-MeO–4-ClC6H3OH, Cs2CO3, microwave in THF or reflux in dioxane, 95%; (f) HCl in dioxane, CH2Cl2, 95%.
12. King, F. D.; Hadley, M. S.; Joiner, K. T.; Martin, R. T.; Sanger, G. J.; Smith, D. M.;
Smith, G. E.; Smith, P.; Turner, D. H.; Watts, E. A. J. Med. Chem. 1993, 36, 683.
13. Lipase Candida rugosa is commercially available from Sigma (Catalogue No. L-
Acknowledgments
We thank Brain Samas, Neil Feeder, and Cheryl Doherty for single
crystal X-ray structure determinations. We are grateful to members
of the Physical Sciences Group for spectroscopic and analytical
services. We also thank Mark Andrews, Christopher Ashcroft, David
L. Gray, David Hepworth, and Cory Stiff for helpful discussions.
1754; 1140 units/mg).
14. The absolute stereochemistry was determined by conversion of 10 to an
authentic sample of (S,S)-esreboxetine (1) and was confirmed by single crystal
X-ray structure determinations of the (R,R)-3 hemi-fumarate and (S,S)-4
PhSO3H salts. Crystallographic data (excluding structure factors) for the
structures in this Letter have been deposited with the Cambridge
Crystallographic Data Center as supplementary publication numbers CCDC
680330 and 680331. Copies of these data can be obtained free of charge on
application to CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK.
References and notes
15. Enantioselectivities were determined by chiral HPLC (Chiralcel OJ-H,
250 Â 4.6 mm; hexane/0.1% TFA in EtOH, 80:20 isocratic, 1 mL/min).
16. Typical procedure: A solution of 11 (50 g, 0.18 mol) in t-BuOMe (1 L) was added
to a solution of Candida rugosa (1 g) in H2O (1 L), and the mixture was stirred
thoroughly at room temperature for 24–30 h (monitored by HLPC). The phases
were separated, and the aqueous layer was washed with t-BuOMe (0.5 L). The
combined organic layers were washed with H2O (0.3 L), were dried (MgSO4),
and were evaporated in vacuo to give (R)-ester 14 (24.9 g) as a clear oil. The
aqueous layer was evaporated in vacuo to give (S)-acid 15 (19.9 g) as a white
1. Wong, E. H. F.; Sonders, M. S.; Amara, S. G.; Tinbolt, P. M.; Piercey, M. F. P.;
Hoffmann, W. P.; Hyslop, D. K.; Franklin, S.; Porsolt, R. D.; Bonsignori, A.;
Carfagna, N.; McArthur, R. A. Biol. Psychiatry 2000, 47, 818.
2. Hajos, M.; Fleishaker, J. C.; Filipiak-Reisner, J. K.; Brown, M. T.; Wong, E. H. F.
CNS Drug Rev. 2004, 10, 23.
3. Pfizer Pipeline as of 28 February, 2008; see: www.pfizer.com. [Search date: 19
September, 2008].
4. Fish, P. V.; Deur, C.; Gan, X.; Greene, K.; Hoople, D.; Mackenny, M.; Para, K. S.;
Reeves, K.; Ryckmans, T.; Stiff, C.; Stobie, A.; Wakenhut, F.; Whitlock, G. A.
Bioorg. Med. Chem. Lett. 2008, 18, 2562.
5. Caution must be exercised, when evaluating biological activities of racemic
compounds as the contribution to the observed effects may not be equal for
each enantiomer; this is further complicated, when seeking dual activities at
two biological targets.
6. Henegar, K. E.; Cebula, M. Org. Process Res. Dev. 2007, 11, 354.
7. Siddiqui, S. A.; Narkhede, U. C.; Lahoti, R. J.; Srinivasan, K. V. Synlett 2006, 1771.
8. Metro, T.-X.; Pardo, D. G.; Cossy, J. J. Org. Chem. 2008, 73, 707.
9. Brenner, E.; Baldwin, R. M.; Tamagnan, G. Org. Lett. 2005, 7, 937.
solid. Data for 14:
[
a
]
À31.5 (MeOH;
c 1.2); >99% ee by chiral HPLC,
D
tR = 6.5 min; 1H NMR (CD3OD, 400 MHz) d 0.93 (t, J = 7.2 Hz, 3H), 1.34 (m, 2H),
1.59 (m, 2H), 2.41–2.30 (m, 2H), 2.59 (m, 1H), 2.84 (m, 1H), 3.52 (dd, J = 15.5
Hz, J = 12.9 Hz, 2H), 3.67 (m, 1H), 3.98 (m, 1H), 4.14 (m, 2H), 4.20 (m, 1H), 7.25
(m, 1H), 7.30 (m, 4H); LRMS ES+ m/z 278 (MH+); Anal. Calcd for C16H23NO3: C,
69.26; H, 8.49; N, 5.02. Found C, 69.31; H, 8.50; N, 5.03. Data for 15: [a]D +35.9
(MeOH; c 1.26); >99% ee by chiral HPLC, tR = 10.2 min; 1H NMR (CD3OD,
400 MHz) d 2.80–2.66 (m, 2H), 3.00 (m, 1H), 3.35 (m, 1H), 3.70 (m, 1H), 4.00 (s,
2H), 4.04 (m, 1H), 4.13 (m, 1H), 7.47–7.36 (m, 5H); LRMS ES+ m/z 222 (MH+).
17. In the context of the synthesis of 3 and 4, swapping the N-protecting group
from Bn to Boc had a beneficial effect on the yields of subsequent reactions and
simplified deprotection at the final step.
10. Henegar has recently reported
a concise synthesis of (S)-morpholine-2-
carboxylic acid from (R)-epichlorohydrin. This Note summarizes current
methods for the synthesis of racemic and single enantiomers of morpholine-
2-carboxylic acids, and includes a comprehensive summary of their use in the
preparation of pharmaceutically active compounds; see: Henegar, K. E. J. Org.
Chem. 2008, 73, 3662.
18. (a) Ramu, B. C. Synlett 1993, 885; (b) Oishi, T.; Nakata, T.. In Encyclopedia of
Reagents forOrganic Synthesis; Paquette, L. A., Ed.; Wiley: Chichester, 1995; Vol. 8,
pp 5536–5539; (c) Narasimhan, S.; Balakumar, R. Aldrichim. Acta 1998, 31, 19.
19. Frein, J. D.; Rovis, T. Tetrahedron 2006, 62, 4573.
11. Racemic ethyl 4-benzylmorpholine-2-carboxylate has been resolved by chiral
HPLC; see: Campbell, G. I.; Cases-Thomas, M. J.; Man, T.; Masters, J. J.; Rudyk, H.
C. E.; Walter, M. W. WO Patent 047272, 2005; Chem. Abstr. 2005, 142,
482071.
20. Data for 4 HCl salt: mp 148 °C; [a]
+14.4 (MeOH; c 0.20); 1H NMR (CD3OD,
D
400 MHz) d 3.05–3.20 (m, 3H), 3.25 (d, 1H), 3.78–3.87 (m, 4H), 4.08–4.20 (m,
2H), 5.31 (d, 1H), 6.70 (m, 2H), 6.95 (s, 1H), 7.28–7.44 (m, 5H); LRMS APCI m/z
334 (MH
+); >99% ee by chiral HPLC.