Synthesis of C-3 alkyl analogs of cocaine

Article abstract of DOI:10.1016/S0040-4039(01)01254-0

The first general stereocontrolled approach to C-3 alkyl analogs of cocaine is described. The route utilized versatile intermediates 2β-hydroxymethyl-3α-cyanotropane and 2β-hydroxymethyl-3β-cyanotropane.

Full text of DOI:10.1016/S0040-4039(01)01254-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 6259–6261
Synthesis of C-3 alkyl analogs of cocaine
Shi Xian Deng,* Dan Wen Huang and Donald W. Landry
Division of Clinical Pharmacology and Experimental Therapeutics, Department of Medicine, Columbia University,
630 West 168 Street, New York, NY 10032, USA
Received 9 May 2001; revised 10 July 2001; accepted 12 July 2001
Abstract—The first general stereocontrolled approach to C-3 alkyl analogs of cocaine is described. The route utilized versatile
intermediates 2b-hydroxymethyl-3a-cyanotropane and 2b-hydroxymethyl-3b-cyanotropane. © 2001 Elsevier Science Ltd. All
rights reserved.
No specific treatment exists for cocaine addiction or
overdose.1 Cocaine acts by blocking a neurotransmitter
reuptake transporter that normally removes dopamine
from a synapse in the reward pathway of the central
nervous system.2 Cocaine amplifies neurotransmission
in this pathway resulting in reinforcement of antecedent
behavior (i.e. cocaine self-administration). Small
molecule antagonists of cocaine have proven elusive,
perhaps because of the difficulties inherent in blocking
a blocker. In the course of exploring novel transition-
state analogs through which to elicit catalytic antibod-
ies that hydrolyze cocaine at its benzoyl ester,3 we
required access to the C-3 analog of cocaine 1. Herein,
we report the first general approach to the synthesis
of a C-3 alkyl analog of cocaine, achieved through
preparation of 2b-hydroxymethyl-3b-cyanotropane 2
(Scheme 1).
Myriad cocaine analogs have been synthesized for
structure–activity studies,5 but no C-3 alkyl analogs are
among them. Carbonꢀcarbon bond formation at C-3 is
sterically hindered and the relative stereochemistry at
C-2 and C-3 is difficult to control. The only reported
cocaine analogs with a carbon substituent at C-3 are
the ‘WIN’ series5 with a phenyl group or other aro-
matic moiety at this position. Most WIN compounds
are synthesized through Michael addition of arylmag-
nesium bromides to dehydroecgonine methyl ester,6 but
our attempts to expand this method to alkyl nucle-
ophiles were unsuccessful.
As a simple, alternative approach to C-3 analogs we
considered SN2 substitution of the mesylate of a previ-
ously reported alloecgonine methyl ester 3.7 Literature
precedents4 and our own initial trials indicated an
unavoidable elimination of the mesylate of 3 if the
methyl ester was unprotected. Thus, alloecgonine
methyl ester 3 was reduced to diol 4 by LiAlH4 in 87%
yield, and the primary alcohol of 4 selectively protected
with t-butyldimethylsilyl chloride to provide 5 in 91%
yield. Mesylation of alcohol 5 proceeded in 88% yield
and reflux of 6 in EtOH/H2O in the presence of KCN
provided the bisaxial a-nitrile 8 in 74% yield. Retention
of configuration at C-3 of 8 was unexpected, but could
be attributed to the participation of the tropane nitro-
gen (intermediate 7). Fortunately, the a-nitrile epimer-
ized with NaNH2 to the desired b-isomer 2 in 83%
yield. The configuration of the two nitriles was clearly
H3C
O
H3C
O
N
N
OCH3
1
OCH3
2
3
O
O
R
WIN series
cocaine
H3C
H3C
O
N
N
OCH3
OTBDMS
CN
O
1
established through the comparison of their H NMR
3β-cyano intermediate 2
C-3 alkyl analog 1
spectra with the corresponding isomers of cocaine. Spe-
cifically, the C-3 proton in 8 showed the same coupling
pattern and constants as allococaine (doublet at 2.90
ppm with a coupling constant J=5.3 Hz), whereas
the C-3 proton in 2 showed a pattern and coupling
* Corresponding author. Tel.: 212-305-1152; fax: 212-305-3475;
e-mail: sd184@columbia.edu
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01254-0

6260
S. X. Deng et al. / Tetrahedron Letters 42 (2001) 6259–6261
H3C
H3C
H3C
O
N
N
N
OR
OCH3
OR
iii
a) i
b) ii
OMs
OH
OH
4 R=H
5 R=TBDMS
3
6 R=TBDMS
H3C
N
H3C
N
OR
OR
CN
+
N
H3C
v
iv
OR
CN
8 R=TBDMS
2 R=TBDMS
7 R=TBDMS
Scheme 1. Reagents and conditions: (i) LiAlH4/THF, 3 h; (ii) TBDMSCl/imidazole, CH2Cl2; (iii) MsCl/Et3N, CH2Cl2; (iv)
KCN/EtOH:H2O (10:3), reflux, 4 h; (v) NaNH2, C6H6, reflux, 24 h.
H3C
H3C
O
N
N
OR
O
OR
O
a) iii
b) iv
a) i
2
b) ii
11 R=H
R=CH3
9 R=TBDMS
10 R=H
1
Scheme 2. Reagents and conditions: (i) PhCH2Li, ether, 3 h; (ii) t-Bu4NF, THF; (iii) (a) Swern oxidation, (b) NaClO2, i-BuOH,
H2O; (iv) CH2N2.
References
constants comparable to those of cocaine (a five
line 1:2:2:2:1 multiplet at 2.85 ppm with JH2a-H3a
H3a-H4a=6 Hz, JH3a-H4b=12 Hz).7,9
=
1. Hall, W. C.; Talbert, R. L.; Ereshefsky, L. Pharmacother-
apy 1990, 10, 46.
J
2. (a) Goeders, N. E.; Smith, J. E. Science 1983, 221, 773;
(b) Kubar, M. J.; Zargin, M. A. J. Neuochem. 1975, 31,
251; (c) Kitty, J. E.; Lorang, D.; Amara, S. G. Science
1991, 254, 578.
3. (a) Landry, D. W.; Zhao, K.; Yang, G.; Glickman, M.;
Georgiadis, T. M. Science 1993, 259, 1899; (b) Yang, G.;
Chun, J.; Arakawa-Uramoto, H.; Wang, X.; Gawinow-
icz, M. A.; Zhao, K.; Landry, D. W. J. Am. Chem. Soc.
1996, 118, 5881.
Nitrile 2 is readily transformed to C-3 alkyl analogs of
cocaine, and nitrile 8 provides the corresponding
analogs of allococaine. As an illustration, we trans-
formed 2 into the desired transposed-carbonyl analog
of cocaine 1. Thus, nitrile 2 was treated with an ethereal
solution of benzyl lithium (prepared from tribenzyltin
chloride with methyllithium)8 to yield ketone 9 in 73%
yield (Scheme 2). Acidic deprotection of the primary
hydroxyl group of 9, followed by two-step oxidation
(Swern and NaClO2) gave the acid 11 and methylation
with CH2N2 provided analog 1 (50% yield for two steps
from 10).
4. Sakurai, M.; Wirsching, P.; Janda, K. D. Tetrahedron
Lett. 1996, 37, 5479.
5. (a) Chang, A.; Burges, J. P.; Carroll, F. I. J. Med. Chem.
1997, 40, 1247; (b) Lieske, S. F.; Yang, B.; Eldefrawi, M.
E.; MacKerell, A. D.; Wright, J. J. Med. Chem. 1998, 41,
864; (c) Xu, L.; Kelkar, S. V.; Lomenzo, S. A.; Izen-
wasser, S.; Katz, J. L.; Kline, R. H.; Trudell, M. L. J.
Med. Chem. 1997, 40, 858; (d) Kozikowski, A. P.; Saiah,
M. K.; Johnson, K. M.; Bergmann, J. S. J. Med. Chem.
1995, 38, 3086.
In conclusion, the versatile C-3 tropane nitriles 2 and 8
provide the first practical routes to C-3 alkyl analogs of
cocaine and allococaine, respectively. Binding studies of
analog
1 with our current anti-cocaine catalytic
antibodies3 and cocaine aptamers10 are in progress.
6. Clarke, R. L.; Daum, S. J.; Gambino, A. J. J. Med.
Chem. 1973, 16, 1260.
7. Carroll, F. I.; Coleman, M. L.; Lewin, A. H. J. Org.
Chem. 1982, 47, 13.
Acknowledgements
8. Sisido, K.; Takeda, Y.; Kinagawa, Z. J. Am. Chem. Soc.
1960, 83, 538.
9. Compound 8: mp: 73°C; H NMR (CDCl3, 300 MHz) l:
We thank Dr. Milan N. Stojanovic for helpful com-
ments. Financial support was provided by the Counter-
drug Technology Center of the Office of National Drug
Control Policy.
1
3.95 (dd, 1H, J=2, 12 Hz), 3.70 (dd, 1H, J=3, 12 Hz),
3.35 (s, broad, 1H), 3.25 (s, broad, 1H), 2.90 (d, 1H,

S. X. Deng et al. / Tetrahedron Letters 42 (2001) 6259–6261
6261
J=10 Hz, H at C-3), 2.25 (m, 1H), 2.20 (s, 3H), 2.20–
2.00 (m, 5H), 1.90 (m, 2H), 0.90 (s, 9H), 0.00 (s, 6H);
MS for C16H30N2OSi cald. 294, found 295 (M+1).
Compound 2: mp: 77°C; 1H NMR (CDCl3, 300 MHz)
l: 4.05 (dd, 1H, J=5, 12 Hz), 3.85 (dd, 1H, J=7, 12
Hz), 3.30 (broad, 1H), 3.10 (broad, 1H), 2.85 (dt, 1H,
J=6, 12 Hz, H at C-3), 2.10 (s, 3H), 2.0–1.5 (m, 7H),
0.85 (s, 9H), 0.00 (s, 6H); MS for C16H30N2OSi cald.
294, found 295 (M+1). Compound 1: 1H NMR
(CDCl3, 300 MHz) l: 7.25 (m, 5H), 3.86 (d, 1H, J=17
Hz), 3.68 (s, 3H), 3.65 (d, 1H, J=17 Hz), 3.25 (s,
broad, 1H), 3.10 (s, broad, 1H), 2.68 (dt, 1H, J=6, 12
Hz, H at C-3), 2.57 (m, 1H), 2.20 (s, 3H), 1.95 (m,
4H), 1.30 (m, 3H).
10. Stojanovic, M. N.; Prada, P. D.; Landry, D. W. J. Am.
Chem. Soc. 2000, 122, 11547.
.