Synthesis and pharmacology of site specific cocaine abuse treatment agents: A new synthetic methodology for methylphenidate analogs based on the Blaise reaction

Article abstract of DOI:10.1016/S0223-5234(01)01230-2

In order to make new analogs of the dopamine (DA) uptake inhibitor methylphenidate, a synthetic methodology based on the Blaise reaction was developed. The reaction between α-bromophenylacetic acid esters, zinc and α-cyano-ω-mesylates gave stable primary enamines. After reduction of the enamines with cyanoborohydride, the amines could be cyclized to methylphenidate analogs in which the amine ring size and aromatic ring were varied. These compounds were tested for inhibitory potency against [³H]WIN 35,428 binding to the cocaine recognition site and [³H]DA uptake using rat striatal tissue. When the heterocyclic ring size was varied, the six-membered ring of methylphenidate appeared to be the optimum ring size. When the aryl ring was varied the 4-trifluoromethylphenyl analog was less potent than methylphenidate, the β-naphthyl congener was considerably more potent, whereas the α-naphthyl congener was less potent. Most of the compounds tested had ratios of uptake to binding inhibition (discrimination ratio) that were similar to cocaine and were therefore not lead compounds for the development of cocaine antagonists.

Full text of DOI:10.1016/S0223-5234(01)01230-2

Eur. J. Med. Chem. 36 (2001) 303–311
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(01)01230-2/FLA
Original article
Synthesis and pharmacology of site specific cocaine abuse treatment agents: a
new synthetic methodology for methylphenidate analogs based on the Blaise
reaction
Howard M. Deutscha*, Xiaocong Yeb, Qing Shic, Zhanzhu Liud, Margaret M. Schwerie
aSchool of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332-0400, USA
bElan Pharmaceutical, South San Francisco, CA, USA
cEli Lilly Co., Indianapolis, IN, USA
dHebei Institute of Materia Medica, Shijiazhaung, Hebei, People’s Republic of China
eMercer University School of Medicine, Macon, GA 31201-1554, USA
Received 19 November 1999; revised 11 November 2000; accepted 11 November 2000
Abstract – In order to make new analogs of the dopamine (DA) uptake inhibitor methylphenidate, a synthetic methodology based
on the Blaise reaction was developed. The reaction between a-bromophenylacetic acid esters, zinc and a-cyano-v-mesylates gave
stable primary enamines. After reduction of the enamines with cyanoborohydride, the amines could be cyclized to methylphenidate
analogs in which the amine ring size and aromatic ring were varied. These compounds were tested for inhibitory potency against
[3H]WIN 35,428 binding to the cocaine recognition site and [3H]DA uptake using rat striatal tissue. When the heterocyclic ring size
was varied, the six-membered ring of methylphenidate appeared to be the optimum ring size. When the aryl ring was varied the
4-trifluoromethylphenyl analog was less potent than methylphenidate, the b-naphthyl congener was considerably more potent,
whereas the a-naphthyl congener was less potent. Most of the compounds tested had ratios of uptake to binding inhibition
(discrimination ratio) that were similar to cocaine and were therefore not lead compounds for the development of cocaine
´
antagonists. © 2001 Editions scientifiques et me´dicales Elsevier SAS
cocaine / dopamine / drug abuse / uptake inhibitor / methylphenidate / discrimination ratio
1. Introduction
transporter [9] and has been used to study the stimu-
lant binding site [10]. Because of its widespread use in
the treatment of attention-deficit disorder, there is
extensive clinical experience with this agent. We re-
cently published the synthesis and pharmacological
studies of a large number of aromatic ring-substituted
methylphenidate analogs (1b–1u) using a method
very similar to that which was developed originally
[2].
The synthetic methodology we used to make these
methylphenidate analogs is quite limiting in that the
piperidine ring portion is constructed by the alkyla-
tion of 2-bromopyridine followed by hydrogenation.
This produces several disadvantages: (1) very power-
ful nucleophiles must be used since 2-bromopyridine
is a poor electrophile; (2) only six-membered rings can
be constructed; (3) other parts of the molecule, for
example the aromatic ring, must be able to survive the
Abuse of the stimulant drug cocaine is a major
problem in the United States [1]. One of the actions of
cocaine is to block the re-uptake of dopamine (DA)
into the presynaptic neuron, thus increasing the con-
centration of DA in the synapse. Our work focuses on
the development of treatment agents [2–5] that inter-
act with the DA transporter, where the reinforcing
effect of cocaine is thought to be mediated [6–8].
( )-threo(R,R/S,S)-Methylphenidate (1a, Ritalin®,
methyl ritalinate) was selected as a promising candi-
date for drug development based on several factors. It
binds potently and somewhat selectively to the DA
* Correspondence and reprints
E-mail address: howard.deutsch@chemistry.gatech.edu
(H.M. Deutsch).

304
H.M. Deutsch et al. / European Journal of Medicinal Chemistry 36 (2001) 303–311
strong hydrogenation conditions (Pt, HOAc, 40 psi
H2). In addition, hydrogenation always leads to a
preponderance of the undesirable erythro isomer. Be-
cause of these limitations, we decided to attempt to
develop a strategy for synthesis that would be based
on the formation of the cyclic amine by ring closure.
Very recently, a new total synthesis [11], an asymmet-
ric synthesis [12], and an enantioselective synthesis
[13] of methylphenidate and analogs have been pub-
lished. After the work described in this paper was
completed, but before it was submitted for publica-
tion [14], a stereoselective synthesis of some of the
threo-methylphenidate analogs described herein (as
well as others) was published [11]. Through the cour-
tesy of these authors, several of their compounds have
been assayed in our systems and the data included in
this publication.
and is a variation of the Reformasky reaction. We were
encouraged to attempt this reaction because of a high
yield modification, which was developed by Hammick
and Kishi [15]. The direct product of this reaction is
actually an imine which, we hypothesized, would rear-
range to the enamine (figure 2). This primary enamine
would be expected to be stabilized on hydrolysis and
have a weakly nucleophilic nitrogen because of the
electron-withdrawing ester and aryl groups. Reduction
of the enamine would then allow ring closure between
the now nucleophilic amine and the leaving group X. To
test the above approach, the appropriate starting mate-
rial was synthesized (see figure 3) in order to make
threo-( )-methylphenidate (n=3). Conversion of 4-
chloro-1-butanol to 4-cyano-1-butanol was accom-
plished in low yield with KCN in water–ethanol. This
alcohol was converted to the corresponding mesylate,
which was then allowed to react with Zn–ethyl-a-bro-
mophenylacetate using the exact conditions and work-
up procedure described in Ref. [15]. A stable compound
was isolated in 79% yield (based on the mesylate), which
was assigned the Z-enamine structure 5b. Reduction of
the enamine was readily accomplished with sodium
cyanoborohydride at pH 6 using the method of Borch et
al. [16] and cyclization occurred spontaneously during
the reaction or in work-up. Unfortunately, this gave a
2. Results and discussion
2.1. Chemistry
The retrosynthetic analysis shown in figure 1 has CꢀC
bond formation via attack of an a-aryl ester enolate on
a nitrile. This reaction, when the enolate is generated
from zinc and an a-bromoester, is the Blaise reaction
Figure 1. Retrosynthetic analysis for the synthesis of methylphenidate analogs.
Figure 2. Blaise reaction to produce an enamine.

H.M. Deutsch et al. / European Journal of Medicinal Chemistry 36 (2001) 303–311
305
R
R
i
ii
iii
CN
n
CN
n
Cl
n
HO
3a, n=2
MsO
4a, n=2
HO
2a, n=2
2b, n=3
2c, n=4
n
n
CO2CH3
CO2CH3
3b, n=3
3c, n=4
4b, n=3
4c, n=4
OMs NH2
OMs NZnBr
5a, n=2, R=H 5d, n=3, R=4-CF3
5b, n=3, R=H 5e, n=3, R=3,4-benzo
5c, n=4, R=H
iv
n=2,3
R
R
R
R
v
iv
vi
n
n
5c
n
n
CO2CH3
CO2CH3
CO2CH3
CO2CH3
CO2CH3
OMs NH2
N
H
N
H
X
NH2
NH
7a, n=2, R=H, 32%
6a, n=2(96%, 31%
threo, 69% erythro)
6b, n=3 (78%, 18%
threo,82% erythro)
6d, n=3, R=4-CF3
6e, n=3, R=3,4-benzo
6c, n=4, X=OMs, R=H, 57%
8, n=4, X=Cl, R=H
7b, n=3, R=H, 65%
7c, 42%
7d, n=3, R=4-CF3, 32%
7e, n=3, R=3,4-benzo, 54%
Figure 3. Synthesis of methylphenidate analogs. Reagents: (I) KCN, EtOH, H2O; (ii) MsCl, Et3N, CH2Cl2; (iii) ArCHBrCO2CH3,
Zn, THF; (iv) (1) NaBH3CN, MeOH, add HCl–MeOH to keep pHꢀ6, (2) NaOH, H2O, pH>10; (v) (1) 50% KOH, reflux, (2)
MeOH–HCl, (3) crystallization; (vi) K2CO3, DMF, 80°C.
The synthesis of the seven-membered ring analog of
the methylphenidate was not as straightforward. The
synthesis and reduction of the corresponding enamine
(5c) proceeded without difficulty, but spontaneous cy-
clization to the seven-membered ring was not observed.
Instead, a yellow solid, shown to be a 5:1 mixture of
erythro to threo saturated primary amine, was isolated.
Washing the solid with EtOAc afforded the pure erythro
isomer as a white solid. Unfortunately, the pure threo
mixture that was predominately the undesired erythro
isomer. In addition, catalytic reduction also gave a
mixture that consisted of mostly the erythro isomer.
Other methods of reduction, such as SmCl2 or NaBH4/
acetic acid were not successful. Despite the unfavorable
stereoselectivity, epimerization of the erythro/threo-
methylphenidate mixture and isolation of the pure threo
isomer was accomplished in the standard manner [2] and
gave
a product that was identical to ( )-threo-
methylphenidate.
Encouraged by the results for the six-membered ring
analogs, we carried out the synthesis of the five-mem-
bered ring analog of methylphenidate using the same
strategy as for the six-membered ring compound. Simi-
lar yields were obtained for each step. The enamine
intermediate (5a) was highly crystalline and single-crys-
tal X-ray analysis, as shown in figure 4, verified the
tendency to form the Z stereochemistry. Again, cycliza-
tion occurred during the reduction–isolation step and
gave a mixture in which the erythro isomer predomi-
nated. Epimerization gave the desired threo isomer.
Figure 4. X-ray structure of Z-enamine (5a).

306
H.M. Deutsch et al. / European Journal of Medicinal Chemistry 36 (2001) 303–311
isomer could not be isolated. Cyclization (DMF, K2CO3,
80°C) was first carried out using the mixture of the threo
and erythro open-chain compounds as the starting mate-
rial. This gave a mixture that appeared to contain cyclized
product, but was quite complicated by 1H-NMR analysis.
The assumed threo product was less than 10%. All
attempts to isolate a pure cyclized product (erythro and/or
threo) from the crude reaction mixture were not success-
ful. This mixture was subjected to the conditions for
epimerization (50% KOH, reflux) in the hope of getting
more of the threo isomer. However, no cyclized product
could be detected. It is possible that the seven-membered
ring analog was destroyed by the harsh basic condition
during epimerization. The cyclization was also attempted
using the pure erythro open-chain compound as the
starting material. When the hydrochloride salt was used,
the crude product was a mixture of at least two com-
pounds by 1H-NMR analysis. After conversion to a
hydrochloride salt, crystallization from EtOAc–MeOH
(2:1) was performed. This gave a white solid, which
proved not to be the desired material but rather the chloro
compound (8) as determined by 1H-NMR and MS
analysis. This was, presumably, formed by the nucle-
ophilic displacement of the mesyl group by chloride
anion. The mother liquid from the isolation of 8 was
evaporated and recrystallized from EtOAc–MeOH (3:1).
This gave a different white solid, which proved to be the
erythro seven-membered ring product (7c). It was as-
signed the erythro stereochemistry based on analogy to
the known NMR patterns for erythro methylphenidate.
When the erythro seven-membered ring product (7c) was
subjected to typical epimerization conditions, neither the
erythro nor the threo product could be detected by
1H-NMR analysis of the methyl esters. The undesired
chloro compound 8 was not formed when the free base
was used as the starting material.
observed product since the more stable isomer will most
likely be formed under these conditions.
Reduction of the enamines with cyanoborohydride
under neutral conditions proceeded rapidly and in high
yield to give mixtures of products in which the erythro
isomers were always produced in greater amount (70–
80%) as compared to the threo isomers. Thus, the predom-
inant mode of reduction is the syn addition of hydrogen.
This process must be at least primarily under kinetic
control since the threo isomer is more stable (based on
epimerization experiments). Catalytic reduction, which
would be expected to give the products of syn addition
of hydrogen, also gave predominately the erythro isomers.
Several other methods of reduction were also attempted,
which either gave high recovery of starting material
(SmCl2) or complicated mixtures of products (NaBH4/
HOAc) that could not be analyzed. In most cases the pure
threo compounds could be isolated in reasonable yield by
epimerization of the methylphenidate analogs (after ring
closure).
When enamines 5a, 5b, 5d and 5e were reduced with
NaBH3CN and then worked up under basic conditions,
ring closure to form the five- or six-membered ring
compounds occurred spontaneously. This did not happen
with the precursor to the seven-membered ring analog. In
this case, the open-chain compound was isolated and then
treated with K2CO3 and DMF at 70°C. In one case, when
the hydrochloride salt of the precursor was so treated,
both the seven-membered ring analog and the chloro
compound 8 were isolated. In this case, the slow rate of
ring closure allows for the competitive nucleophilic dis-
placement of the mesyl group by the chloride anion. When
the free base of pure erythro 6c was treated with K2CO3
and DMF at 70°C, pure erythro 7c could be isolated
easily. All attempts to epimerize either pure 7c or a
mixture of 7c and the corresponding threo isomer (never
isolated) led to mixtures of products of undetermined
structures. It appeared as though the seven-membered
ring analogs were unstable under the epimerizing condi-
tions (50% KOH, 120°C). The synthesis of the threo-b-
naphthyl analog 7e and the threo-4-trifluoromethyl
analog 7d proceeded normally. All of these compounds
(7a, 7c, 7d and 7e) are of interest chemically because they
are not available by the previous methods of synthesis.
When the aromatic component of the Blaise reaction
was varied, several new analogs were made, which could
not be made by the original synthetic method. Thus, when
methyl-1-bromo-1-(4-trifluoromethylphenyl)acetate or
methyl-1-bromo-1-(2-naphthyl)acetate was treated with
Zn and 4b, and the isolated intermediates treated as
outlined above, the corresponding threo-methylphenidate
analogs 7d or 7e were isolated.
1
Although H-NMR analysis (Noe, data not shown)
indicated that these enamines had the Z stereochemistry,
it was shown unambiguously by single-crystal X-ray
crystallography. Semi-empirical calculations (CS MO-
PAC Pro) [17] indicated that the Z isomer was more
stable, dHf= −1.6 kcal mol−1; this may account for the
3. Pharmacology
The inhibition of [3H]WIN 35,428 (WIN) binding
and [3H]DA uptake are shown in table I (compounds

H.M. Deutsch et al. / European Journal of Medicinal Chemistry 36 (2001) 303–311
307
Table I. Inhibition of [3H]WIN 35,428 binding and [3H]DA uptake for methylphenidate derivatives synthesized in this study and
related compounds; value9SEM (n).
Compound
Stereochemistry
IC50 (nM)
Hill coefficient, nH,
for WIN binding
Discrimination ratio, ratio of
uptake to binding (DR)
[3H]WIN 35,428
binding
[3H]DA uptake
7a
threo
threo
erythro
threo
threo
355945(2)
83.097.9(4)
40909310(2)
615915(2)
11.092.5(3)
160915(3)
88595.0(2)
224919(4)
90309250(2)
2370985(2)
53.098.0(2)
404926(2)
1.0790.01
0.9090.09
0.8190.08
0.9790.05
1.0190.03
1.0390.01
2.5
2.7
2.2
3.8
4.8
2.5
7b(1a)
7c
7d
7e
COC a
a (−)-Cocaine.
Table II. Inhibition of [3H]WIN 35,428 binding and [3H]DA uptake for methylphenidate derivatives synthesized by Axten et al. [11]
value9SEM (n).
Compound
Stereochemistry
IC50 (nM)
Hill coefficient, nH,
for WIN binding
Discrimination ratio, ratio
of uptake to binding (DR)
[3H]WIN 35,428
binding
[3H]DA uptake
9
threo
threo
threo
threo
threo
197916(2)
62394.5(2)
12509210(2)
447091200(2)
716956(3)
701936(2)
15909150(2)
993091400(2)
84509210(2)
980994(2)
0.8290.02
0.9590.03
0.8190.05
1.0490.18
0.9890.02
3.6
2.6
7.9
1.9
1.4
10
11
12
13

308
H.M. Deutsch et al. / European Journal of Medicinal Chemistry 36 (2001) 303–311
synthesized in this study) and table II (compounds
synthesized by Axten et al. [11]).
cyano compounds). Unfortunately, reduction of the
enamines gave mostly the erythro isomers. Fortu-
nately, in most cases, threo methylphenidate analogs
could be made using an epimerization procedure.
New analogs were produced, which could not be
made using the previous procedure. However, the new
synthetic methodology of Axten et al. [11] appears to
be a more versatile way to produce these compounds.
All modifications to the piperidine ring of
methylphenidate produced less potent analogs. The
a-naphthyl analog was less active and the b-naphthyl
analog more active than methylphenidate, in agree-
ment with the previous literature results for tropanes.
While the low DR values of these compounds suggest
that they may not be useful cocaine antagonists, the
potent b-naphthyl analog may merit further investiga-
tion as substitution therapy for cocaine. The 3-mor-
pholino analog 13, although less active, may provide
a lead for cocaine antagonism because of its increased
DR value of 7.9, marginally higher than
methylphenidate, 2.7 and cocaine, 2.5.
WIN binding. All of the compounds with either
smaller (7, five-membered) or larger (11 and 12,
seven- and eight-membered) rings were somewhat less
potent than methylphenidate (1a, six-membered) by
factors of 4, 2 and 8, respectively. In agreement with
earlier work [2], the erytho isomer 9 was 21-fold less
active than the threo isomer 11. When the six-mem-
bered ring is maintained but a polar oxygen atom
inserted (13, 3-morpholino analog), the binding po-
tency is reduced substantially (15-fold). Moving the
phenyl ring one atom further away from the nitrogen,
to produce a phenylpropylamine derivative, drasti-
cally reduces binding potency 54-fold. The only aro-
matic
ring-substituted
derivative
was
the
trifluoromethyl compound 8. Surprisingly, this com-
pound was sevenfold less active (IC50 =615 nM) than
methylphenidate (IC50 =83.0 nM) and 18- and 19-
fold
less
active
than
[2],
4-fluoro
respectively.
and
4-
Most
methylpethylphenidate
interesting were the a- and b-naphthyl derivatives 15
and 10, respectively. Thus, 15 is 12-fold less active
than MP, whereas 10 is eightfold more active than
MP. This difference between the a- and b-naphthyl
analogs is in good agreement with the results of
Davies et al. [18] for the a- and b-naphthyl derivatives
of tropanes. The differences between some of the
previously reported results of Axten et al. [11] (using
a different assay system) for the a- and b-naphthyl
derivatives 15 and 10 and the results reported here
will be addressed by these authors in a future publica-
tion [19].
5. Experimental protocols
Chemistry, general. Reagents and solvents were
mostly reagent grade and were used without further
purification. Solvents or reagents that required drying
or purification were prepared according to the proce-
dures found in Ref. [20]. Column chromatography
was carried out on Fisher Scientific Co. Silica gel
(Grade 62). Extraction solvents were dried with
MgSO4. Melting points were obtained using a Labo-
ratory Devices Mel-Temp II instrument without cor-
rections. NMR spectra were recorded on a Varian
Gemini 300 (300 MHz) NMR spectrometer. Mass
spectra were measured on a VG 70-SE, 2 sector,
forward geometry instrument. IR spectra were
recorded on a Nicolet 520 FT spectrophotometer.
Microanalytical data were obtained by Atlantic Mi-
crolabs, Atlanta, GA.
[3H]DA uptake. In all cases, the inhibition of
[3H]DA uptake closely followed WIN binding. One
way to look for a divergence is to calculate the ratio
of inhibitory potencies against uptake to binding us-
ing the respective IC50 values; the term discrimination
ratio (DR) [2] has been used for this purpose. In our
hands, compound 15 had a DR of only 1.4 whereas
the value for compound 10 was 4.8. The compound
with the highest DR reported here, 7.9, was the 3-
morpholino analog 13.
4-Hydroxyl-1-butanitrile (3a). To a stirred solution
of 13.4 mL (160 mmol) of 3-chloropropan-1-ol in 76
mL of ethanol–water (3:1) was added 12.7 g (200
mmol) of KCN. The solution was then boiled for 1.5
h, cooled, filtered and the solvents were distilled off
under reduced pressure. This gave an oily residue,
4. Conclusions
1
which was distilled to give 3.40 g (25%) of 3a. H-
The Blaise reaction, using the modification of
Hammick and Kishi [15], worked quite well giving the
Z-v-mesyl-enamines in 60–75% yield (based on the
NMR (CDCl3): l 3.79 (t, 2H, J=5.8 Hz, CH2ꢀO),
2.51 (t, 2H, J=7.1 Hz, CH2ꢀCN), 1.95–1.87 (m, 2H).
MS-CI; m/z: 86 (M+H).

H.M. Deutsch et al. / European Journal of Medicinal Chemistry 36 (2001) 303–311
309
4-Mesylate-1-butanitrile (4a). With stirring, 1.0 g
(12 mmol) of compound 3a was dissolved in 60 mL of
dry CH2Cl2 and 2.5 mL (18 mmol) of triethylamine
was then added. The solution was cooled to about
−10°C, 1.0 mL (13 mmol) of methanesulfonyl chlo-
ride was slowly added over a 10-min period and it was
then stirred for an additional 15 min. The reaction
mixture was extracted with ice-water, followed by
cold 10% of HCl, saturated NaHCO3 solution and
brine. Drying and evaporation gave 1.82 g (93%) of
NaOH solution (saturated with NaCl), extracted with
EtOAc and the organic layer washed with brine and
dried. Removal of the solvent gave 337 mg (96%) of
6a, which contained 31% of threo and 69% of erythro
isomers based on 1H-NMR analysis. This crude
product was mixed with 9 mL of 50% KOH and
heated under reflux for three days. Top oil was col-
lected and dissolved in MeOH. Concentrated HCl
was added and evaporated to dryness to give a solid.
The solid was mixed with MeOH (20 mL) and thionyl
chloride (0.5 mL) and stirred at r.t. for 24 h. The
solvent was evaporated and water added. The solu-
tion was then made basic with NaOH (15%), ex-
tracted with EtOAc, washed with water and dried.
Evaporation gave 168 mg of the threo-enriched mix-
1
4a, which showed: H-NMR (CDCl3): l 4.37 (t, 2H,
J=5.8 Hz, CH2ꢀO), 3.07 (s, 3H, CH3), 2.56 (t, 2H,
J=7.1 Hz, CH2ꢀCN), 2.18 (br), 2.09 (m, 2H); MS-
CI; m/z: 164 (M=H).
Methyl-7-mesylate-3-amino-2-phenyl-2-hexenoate
(5a). To a suspension of 0.780 g (12 mmol) of acti-
vated Zn dust in 7.5 mL of refluxing dry THF under
N2 was added four 0.05-mL portions of methyl-a-bro-
mophenylacetate. The suspension was heated under
reflux for 0.5 h until a green color appeared. Then,
0.392 g (2.40 mmol) of compound 4a in 1.5 mL of dry
THF was added in one portion, followed by 1.3 mL
(total 9.6 mmol) of methyl-a-bromophenylacetate in
1.8 mL of dry THF over a 45-min period. The mix-
ture was heated under reflux for an additional 10 min,
cooled to room temperature (r.t.), diluted with 21 mL
of THF and quenched with 3.1 mL of 50% aqueous
K2CO3. Rapid stirring for 30 min gave two layers.
The upper layer was decanted and the lower layer
washed three times with THF. The combined organic
layers were dried and the solvent removed in vacuo.
This gave an oily residue, which was purified by
column chromatography (30 g of silica gel) with an
elutant of EtOAc–hexane (1:2, v/v) yielding 0.48 g
(63%) of 5a as a pale yellow solid. When washed with
EtOAc–hexane (1:1), a pure white solid was pro-
duced, which showed: m.p. 105–106°C; 1H-NMR
(CDCl3): l 7.35–7.14 (m, 5H, Ar), 4.11 (t, 2H, J=6.1
Hz, CH2ꢀOSO2), 3.59 (s. 3H, CH3O), 2.92 (s, 3H,
CH3ꢀS), 2.16 (t, 2H, J=7.4 Hz, CH2ꢀCꢁC), 1.85–
1.80 (m, 2H).
1
ture (4:1 threo–erythro by H-NMR analysis). This
mixture was converted to an HCl salt and crystallized
from a mixture of acetone and diethyl ether to give a
white solid, which was the pure threo isomer (7a) (130
1
mg, 32%): m.p. 192–194°C; H-NMR (D2O): l threo
isomer 7.31–7.26 (m, 3H, Ar), 7.20–7.17 (m, 2H, Ar),
4.06–4.01 (m, 1H, HC2¦), 3.88 (d, 1H, J=10.6 Hz,
HC2%), 3.56 (s, 3H, CH3), 3.25–3.18 (m, 2H, HC5¦),
1.87–1.47 (m, 4H); erythro isomer (not isolated)
7.32–7.25 (m, 5H, Ar), 4.07–4.04 (m, 1H, HC2¦),
3.90 (d, 1H, J=10.5 Hz, HC2%), 3.53 (s, 3H, CH3),
3.16–3.07 (m, 2H, HC5¦), 2.27–2.23 (m, 1H), 1.98–
1.90 (m, 2H), 1.70–1.66 (m, 1H); MS-EI; m/z: 220
(M+H). Anal. Calc. for C13H18ClNO2: C, H, N, Cl.
6-Hydroxy-1-hexanitrile (3c). To a stirred solution
of 9.20 g (75 mmol) of 5-chloro-1-pentanol in 40 mL
of ethanol–water (3:1) was added 6.4 g (98 mmol) of
KCN. The solution was then heated under reflux for
5.5 h. The solvents were evaporated, ether was added
and the mixture was filtered. Evaporation of the
filtrate gave an oil which was distilled at 125–145°C
1
to give 5.8 g (63%) of 3c, which showed: H-NMR
(CDCl3): l 3.66 (t, 2H, J=6.1 Hz, CH2ꢀO), 2.38 (t,
2H, J=7.1 Hz, CH2ꢀCN), 1.82–1.51 (m, 7H).
6-Mesylate-1-hexanitrile (4c). With stirring, 1.7 g
(15 mmol) of compound 3c was dissolved in 75 mL of
dry CH2Cl2 and 3.0 mL (22 mmol) of triethylamine
was then added. This solution was cooled to about
−10°C, 1.3 mL (16 mmol) of methanesulfonyl chlo-
ride was added slowly over a 10-min period and it was
stirred for an additional 15 min. The mixture was
extracted with ice-water, followed by cold 10% of
HCl, saturated NaHCO3 solution and brine. After
drying, evaporation of the solvent gave 2.8 g (97%) of
( )-threo-Methyl-2-phenyl-2-(2-azacyclopentyl)-
acetate (7a). To a solution of 0.50 g (1.6 mmol) of 5a
in 15 mL of MeOH was added a trace of bromocresol
green, followed by 0.40 g (6.3 mmol) of NaBH3CN.
Dilute HCl–MeOH solution was added until the
color turned yellow. The solution was stirred at r.t.
for 1 h. More HCl–MeOH was added dropwise dur-
ing the reaction to maintain the yellow color. The
reaction mixture was poured into 30 mL (0.1 N) of
1
4c, which showed: H-NMR (CDCl3): l 4.24 (t, 2H,

310
H.M. Deutsch et al. / European Journal of Medicinal Chemistry 36 (2001) 303–311
J=6.2 Hz, CH2ꢀO), 3.01 (s, 3H, CH3), 2.38 (t, 2H,
J=7.1 Hz, CH2ꢀCN), 1.82–1.58 (m, 6H).
using the HCl salt of compound 6c. Crystallization of
the crude product (HCl salt) from MeOH–EtOAc
(1:2) gave pure compound 8, which showed: m.p.
Methyl-8-mesylate-3-amino-2-phenyl-2-octenoate
(5c). Using the same method for the synthesis of 5a,
4.6 g (70 mmol) of activated Zn dust, 8.8 mL total (56
mmol) of methyl-a-bromophenylacetate and 2.75 g
(14 mmol) of compound 4c gave an oily residue,
which was purified by column chromatography on
120 g of silica gel using an elutant of EtOAc–hexane
(1:1, v/v). This gave 3.5 g (74%) of 5c, which crystal-
lized on standing. The solid was washed with ether to
1
221–222°C; H-NMR (D2O): l 7.35–7.23 (m, 5H,
Ar), 3.94 (d, J=7.6 Hz, 1H, HCꢀAr), 3.85–3.81 (m,
1H), 3.59 (s, 3H, CH3), 3.50–3.45 (m, 2H), 1.64–1.34
(m, 8H). MS-CI; m/z: 284 (M+H).
threo-4-Trifluoromethylmethylphenidate (7d). Using
the same procedures described in making 5c, 1.3 g
(0.02 mmol) of activated Zn dust, 5.95 g (20.0 mmol)
of
methyl-a-bromo-4-trifluoromethlyphenylacetate
1
give pure 5c, which showed: m.p. 53–55°C; H-NMR
and 0.73 g (10 mmol) of compound 4b gave an oily
residue, which was purified by column chromatogra-
phy on silica gel using an elutant of EtOAc–hexane
(1:1) to give 0.57 g (35%) of 5d. Using the same
method for the synthesis of 7a, 0.57 g of 5d and 0.36
g of NaBH3CN gave a crude product as a mixture of
threo and erythro isomers in a ratio of about 1:3 by
1H-NMR analysis. The crude product was converted
to an HCl salt and yellow solid washed with diethyl
ether several times to afford 0.37 g (75%) of 6d as a
white solid. Using the epimerization–esterification
procedure described in the synthesis of 7a, 0.37 g of
6d gave 88 mg (24%) of 7d, which showed: m.p.
(CDCl3): l 7.34–7.25 (m, 3H, Ar), 7.14 (m, 2H, Ar),
4.12 (t, 2H, J=5.9 Hz, CH2ꢀO), 3.58 (s, 3H, CH3ꢀO),
2.97 (s, 3H, CH3ꢀS), 2.02 (t, 2H, J=7.1 Hz,
CH2ꢀCꢁ), 1.60–1.25 (m, 6H).
erythro-Methyl-8-mesylate-3-amino-2-phenyl-2-oc-
tanoate (6c). Using the same method for the synthesis
of 7a, 1.62 g (4.47 mmol) of 5c and 1.2 g (19 mmol) of
NaBH3CN gave 1.77 g of crude product as a mixture
of threo and erythro isomers in a ratio of about 1:5 by
1H-NMR analysis. The crude product was converted
to an HCl salt and yellow solid washed with EtOAc
several times to afford the pure erythro isomer 6c as a
white solid, 0.93 g (57%), which showed: m.p. 158–
1
196–197°C; H-NMR (D2O): l 7.62 (d, J=8.2 Hz,
1
160°C; H-NMR (D2O): l 7.31–7.20 (m, 5H, Ar),
2H, Ar), 7.37 (d, J=8.2 Hz, 2H, Ar), 3.96 (d, J=9.4
Hz, 1H, HC2%), 3.78–3.71 (m, 1H, HC2¦), 3.59 (s, 3H,
CH3), 3.34–3.33 (m, 1H, HC6¦), 2.95 (dt, J=3.3 Hz,
12.9 Hz, HC6¦), 1.76–1.19 (m, 6H). Anal. Calc. for
C15H19ClNO2F3·0.4H2O: C, H, N, Cl.
4.17 (t, 2H, J=6.1 Hz, CH2ꢀO), 3.92 (d, 1H, J=7.2
Hz, CHꢀAr), 3.80–3.78 (m, 1H, HCꢀN), 3.55 (s, 3H,
CH3ꢀO), 3.01 (s, 3H, CH3ꢀS), 1.63–1.32 (m, 8H).
erythro - Methyl - 2 - phenyl - 2 - (2 - azacycloheptyl)-
acetate (7c). A 240-mg (7.1 mmol) portion of 6c was
added to a suspension of 570 mg of K2CO3 in 5 mL of
DMF under a nitrogen atmosphere. The mixture was
heated to 80°C for 2 h, cooled and decanted. The
solid residue was rinsed with diethyl ether (4×20 mL)
and the combined organic layer washed with water
(5×20 mL) and dried. Evaporation of the solvent
gave 160 mg crude product as a pale yellow oil which
was converted to an HCl salt and crystallized from
EtOAc–MeOH (2:1) to give 73 mg (42%) of white
solid as the pure erythro isomer (7c) which showed:
threo - Methyl - 2 - piperidyl - 2 - [(2 - naphthyl)]acetate
(7e). Using the same procedures described in making
5c, 1.0 g (16 mmol) of activated Zn dust, 4.4 g (16
mmol) of methyl-a-bromo-(2-naphthyl)acetate and
0.58 g (3.2 mmol) of compound 4b gave an oily
residue, which was purified by column chromatogra-
phy on silica gel using an elutant of EtOAc–hexane
(1:1) to give 0.72 g (59%) of 5e. Using the same
method for the synthesis of 7a, 0.61 g of 5e, and 0.36
g of NaBH3CN gave a crude product as a mixture of
threo and erythro isomers in a ratio of about 1:3 by
1H-NMR analysis. The crude product was converted
to an HCl salt and yellow solid washed with diethyl
ether several times to afford 0.31 g (68%) of 6e as a
white solid. Using the epimerization–esterification
procedure described in the synthesis of 7a, 0.31 g of
6e gave 0.13 g (54%) of 7e, which showed: m.p.
1
m.p. 178–180°C; H-NMR (D2O): l 7.39–7.29 (m,
5H, Ar), 3.99 (d, J=8.8 Hz, 1H, HC2%), 3.92 (m, 1H,
HC2¦), 3.60 (s, 3H, CH3), 3.09–3.04 (m, 2H, HC7¦),
1.93–1.90 (m, 1H), 1.71–1.41 (m, 7H). MS-CI; m/z:
248 (M+H). Anal. Calc. for C15H22ClNO2·0.25H2O:
C, H, N, Cl.
erythro-Methyl-3-amino-8-chloro-2-phenyloctano-
ate (8). This compound was obtained as a byproduct
when the above cyclization reaction was conducted
1
219–220°C; H-NMR (D2O): l 7.84–7.76 (m, 3H,
Ar), 7.70 (m, 1H, Ar), 7.47–7.43 (m, 2H, Ar), 7.26
(dd, J=1.8, 8.4 Hz, 1H, Ar), 4.00 (d, J=9.3 Hz, 1H,

H.M. Deutsch et al. / European Journal of Medicinal Chemistry 36 (2001) 303–311
311
[7] Bergman J., Madras B.K., Johnson S.E., Spealman R.D., Ef-
fects of cocaine and related drugs in nonhuman primates. III.
Self administration by squirrel monkeys, J. Pharmacol. Exp.
Ther. 219 (1989) 150–155.
HC2%), 3.81–3.75 (m, 1H, HC2¦), 3.57 (s, 3H, CH3),
3.32–3.27 (m, 1H, HC6¦), 2.93 (dt, J=3.3 Hz, 12.9
Hz, 1H, HC6¦), 1.72–1.22 (m, 6H). Anal. Calc. for
C18H22ClNO2·0.5H2O: C, H, N, Cl.
[8] Kuhar M.J., Ritz M.C., Boja J.W., The dopamine hypothesis
of the reinforcing properties of cocaine, Trends Neurosci. 14
(1991) 299–302.
[9] Taylor D., Ho B.T., Comparison of inhibition of monoamine
uptake by cocaine, methylphenidate and amphetamine, Chem.
Path. Pharm. 21 (1978) 67–75.
Pharmacology. [3H]WIN 35,428 binding and
[3H]DA uptake assays were conducted as described
previously [3].
[10] Schweri M.M., Skolnick P., Rafferty M.F., Rice K.C.,
Janowsky A.J., [3H] threo-()-Methylphenidate binding to 3,4-di-
hydroxyphenylethylamine uptake sites in corpus striatum: cor-
relation with the stimulant properties of ritalinic acid esters, J.
Neurochem. 45 (1985) 1062.
Acknowledgements
This work was supported in part by a grant from the
National Institute of Drug Abuse, RO1 DA06305
(H.M.D. and M.M.S.). The authors wish to thank Mr
Adam Eason and Ms Monica Stafford for their assis-
tance in the biochemical assays.
[11] Axten J.M., Krim L., Kung H.F., Winkler J.D., A stereoselec-
tive synthesis of DL-threo-methylphenidate: preparation and
biological evaluation of novel analogues, J. Org. Chem. 63
(1998) 9628–9629.
[12] Thai D.L., Sapko M.T., Reiter C.T., Bierer D.E., Perel J.M.,
Asymmetric synthesis and pharmacology of methylphenidate
and its para-substituted derivatives, J. Med. Chem. 41 (1998)
591–601.
References
[13] Prashad M., Kim H.Y., Lu Y.S., Liu Y.G., Har D., Repic O.,
Blacklock T.J., Giannousis P., The first enantioselective synthe-
sis of (2R,2%R)-threo-(+)-methylphenidate hydrochloride, J.
Org. Chem. 64 (1999) 1750–1753.
[1] National Household Survey on Drug Abuse: Population Esti-
mates 1990, DHHS Publ. No. (ADM) 91-1732, Washington,
DC, 1991, pp. 29–39, 59–63, and 111–115.
[14] For a preliminary account see: H.M. Deutsch, Q. Shi, Z. Liu,
M.M. Schweri, Synthesis and pharmacology of potential co-
[2] Deutsch H.M., Shi Q., Gruszecka-Kowalik E., Schweri M.M.,
Synthesis and pharmacology of potential cocaine antagonists.
2. Structure–activity relationship studies of aromatic ring-sub-
stituted methylphenidate analogs, J. Med. Chem. 39 (1996)
1201–1209.
caine
antagonists.
New
synthetic
methology
for
methylphenidate analogs, 214th American Chemical Society
Meeting, Las Vegas, NV, August 23–27, 1997.
[15] Hammick S.M., Kishi Y., Improved procedure for the Blaise
reaction: a short, practical route to the key intermediate of the
saxitoxin synthesis, J. Org. Chem. 48 (1983) 3833–3835.
[3] Deutsch
H.M.,
Structure–activity
relationships
for
methylphenidate analogs and comparisons to cocaine and
tropanes, Med. Chem. Res. 8 (1998) 91–99.
[16] Borch R.F., Bernsterin M.D., Durst H.C., J. Am. Chem. Soc.
93 (1971) 2897.
[4] Deutsch H.M., Collard D.M., Zhang L., Burnham K., Desh-
pande A.K., Holtzman S.G., Schweri M.M., Synthesis and
pharmacology of site-specific cocaine abuse treatment agents:
2-(aminomethyl)-3-phenylbicyclo[2.2.2] and [2.2.1]-alkanes do-
pamine uptake inhibitors, J. Med. Chem. 41 (1999) 882–895.
[17] CS ChemOffice Ultra 4.5, Cambridge Software, Cambridge,
MA.
[18] Davies H.M.L., Saikali E., Huby N.J.S., Gilliatt V.J., Matasi
J.J., Sexton T., Childers S.R., Synthesis of 2b-acyl-3b-aryl-8-az-
abicyclo[3.2.1]octanes and their binding affinities at dopamine
and serotonin transport sites in rat striatum and frontal cortex,
J. Med. Chem. 37 (1994) 1262–1268.
[5] Deutsch H.M., Dunn T., Ye Z., Schweri M.M., Synthesis and
pharmacology of site-specific cocaine abuse treatment agents:
the role of the phenyl group in highly modified
methylphenidate analogs as dopamine uptake inhibitors, Med.
Chem. Res. 9 (1999) 213–222.
[19] H.F. Kung, J.D. Winkler, personal communication.
[20] Furniss B.S., Hannaford A.J., Smith P.G.W., Tatchell A.R.
(Eds.), Vogel’s Textbook of Practical Organic Chemistry, 5th
ed., Wiley, New York, 1989.
[6] Ritz M.C., Lamb R.J., Goldberg S.R., Kuhar M.J., Cocaine
receptors on dopamine transporters are related to self-adminis-
tration of cocaine, Science 237 (1987) 1219–1223.
.