Further exploration of 1-{2-[Bis-(4-fluorophenyl)methoxy]ethyl}piperazine (GBR 12909): Role of N-aromatic, N-heteroaromatic, and 3-oxygenated N-phenylpropyl substituents on affinity for the dopamine and serotonin transporter

Article abstract of DOI:10.1016/S0960-894X(03)00108-2

A series of N-aromatic, N-heteroaromatic, and oxygenated N-phenylpropyl derivatives of 1-(2-benzhydryloxyethyl)-piperazine and 1-{2-[bis-(4-fluorophenyl)methoxy]ethyl}piperazine, analogues of GBR 12909 (1a) and 12935 (1b), was synthesized and examined for their dopamine (DAT) and serotonin (SERT) transporter binding properties. One of these compounds, racemic 3-[4-(2-benzhydryloxyethyl)piperazin-1-yl]-1-(3-fluorophenyl)-propan-1-ol (33), had DAT affinity as good as, or better than, GBR 12909 and 12935, and was more selective for DAT over SERT than the GBR compounds. Both trans- (43) and cis- (47) (±)-2-(4-{2-[bis-(4-fluorophenyl)-methoxy]ethyl}piperazin-1- ylmethyl)-6-methoxy-1,2,3,4-tetrahydronaphthalen-1-ol had relatively good SERT selectivity and, as well, showed high affinity for SERT.

Full text of DOI:10.1016/S0960-894X(03)00108-2

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1385–1389
Further Exploration of 1-{2-[Bis-(4-
fluorophenyl)methoxy]ethyl}piperazine (GBR 12909): Role of
N-Aromatic, N-Heteroaromatic, and 3-Oxygenated
N-Phenylpropyl Substituents on Affinity for the Dopamine and
Serotonin Transporter
David Lewis,a,y Ying Zhang,a,{ Thomas Prisinzano,a Christina M. Dersch,b
Richard B. Rothman,b Arthur E. Jacobsona and Kenner C. Ricea,*
aLaboratory of Medicinal Chemistry, NIDDK, NIH, Bethesda, MD, 20892, USA
bPsychopharmacology Section, NIDA, Addiction Research Center, Baltimore, MD 21224, USA
Received 9 October 2002; accepted 21 November 2002
Abstract—A series of N-aromatic, N-heteroaromatic, and oxygenated N-phenylpropyl derivatives of 1-(2-benzhydryloxyethyl)-
piperazine and 1-{2-[bis-(4-fluorophenyl)methoxy]ethyl}piperazine, analogues of GBR 12909 (1a) and 12935 (1b), was synthesized
and examined for their dopamine (DAT) and serotonin (SERT) transporter binding properties. One of these compounds, racemic
3-[4-(2-benzhydryloxyethyl)piperazin-1-yl]-1-(3-fluorophenyl)-propan-1-ol (33), had DAT affinity as good as, or better than, GBR
12909 and 12935, and was more selective for DAT over SERT than the GBR compounds. Both trans- (43) and cis- (47) (ꢀ)-2-(4-{2-
[bis-(4-fluorophenyl)-methoxy]ethyl}piperazin-1-ylmethyl)-6-methoxy-1,2,3,4-tetrahydronaphthalen-1-ol had relatively good SERT
selectivity and, as well, showed high affinity for SERT.
# 2003 Elsevier Science Ltd. All rights reserved.
Dopamine reuptake inhibitors1 have been considered
for the treatment of depression and Parkinson’s disease,
and a prototypic dopamine reuptake inhibitor, GBR
12909 (Fig. 1, 1-{2-[bis-(4-fluorophenyl)methoxy]ethyl}-
4-(3-phenylpropyl)piperazine, 1b), was examined for
those purposes.2 We,3ꢁ7 and others,8ꢁ11 have suggested
that a selective high affinity dopamine reuptake inhib-
itor12 might be useful as a treatment agent for the abuse
of cocaine and other stimulants,13 and 1b is currently
being evaluated for that purpose and has satisfactorily
completed phase 1 clinical trials under the auspices of
the National Institute on Drug Abuse, NIH.11 How-
ever, although the bis(4-fluorophenyl compound 1b, has
very high affinity as a dopamine transporter protein
(DAT) ligand, it is not very selective over the serotonin
transporter (SERT).
The abuse of cocaine and other stimulants is widespread
and has unfortunate sequelae for individuals14ꢁ16 and
society.17,18 We have undertaken a program5ꢁ7,19ꢁ22 to
design and synthesize analogues of GBR 12909 (1b) and
GBR 12935 (1a) in order to study the pharmacological
effect of these ligands. In particular, we have examined
the effect of structural modifications on their affinity for
the DAT and SERT (serotonin transporter), with the
*Corresponding author. Fax: +1-301-402-0589; e-mail: kr21f@nih.gov
yCurrent address: Food and Drug Administration, 5600 Fishers Lane,
Rockville, MD 20857, USA.
{Current address: ArQule, Inc., 200 Boston Ave., Suite 4200, Med-
ford, MA 02155, USA.
Figure 1. Structure of GBR 12909 and GBR 12935 analogues.
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00108-2

1386
D. Lewis et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1385–1389
hope that new, high affinity, and more selective ligands
can be found that have potential as treatment agents
capable of reducing the incidence of drug abuse and the
concomitant HIV epidemic.
affinity compared to ketones 6–8 and desfluoro analo-
gue 9 was found to be almost as DAT-selective as 2c.
To further probe the effect of the phenyl ring in com-
pounds 2a and 2b, we sought to remove it, and in addi-
tion, attempt to limit rotational freedom we constrained
the ketone moiety by placing it in several cycloalkane
rings, 14–19. Among the cycloalkanone-containing ana-
logues, 14–19, the bisfluoro analogues 15, 17, and 19 had
higher affinity than desfluoro analogues 14, 16, and 18
respectively. Reduction of the ketones in 14–19 gave
alcohols 20–25. These modifications slightly increased
affinity for both the DAT and SERT compared to ketone
analogues 14–19. The binding data of the cycloalkane
We previously reported the synthesis and biological
activity of a series of analogues of 1a and 1b, in which
the N-phenylpropyl moiety was replaced by a hetero-
aromatic- or bicyclic aromatic-containing side chain.21
These analogues were potent and more selective DAT
ligands than 1a, as were a more recently reported series of
oxygenated analogues of 1a and 1b.22 Herein, we report
the synthesis of a novel series of 1a and 1b analogues,
which contain both of the structural modifications
reported earlier (replacement of the phenylpropyl moiety
by a heteroaromatic- or bicyclic aromatic-containing side
chain and the addition of an oxygen functionality at the
3-position of the 3-phenylpropyl moiety). A DAT selec-
tive, high affinity oxygenated compound could eventually
be formulated as an extended-action depot preparation,
enabling its examination as a second-generation ultra
long-acting agent for the treatment of cocaine abuse.23,24
One dose of the first-generation drug, the decanoate ester
of 2d, eliminated cocaine self-administration in rhesus
monkeys almost completely for nearly a month.22,25
Table 1. Binding affinities and selectivities of GBR 12909 analogues
Compda
R1
DAT Ki (nM)b
SERT Ki (nM)b
SERT/DAT
1a
1b
2a
2b
2c
2d
3c
5
6
7
8
9
H
F
H
F
H
F
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
3.7 (ꢀ0.3)
3.7 (ꢀ0.4)
208 (ꢀ4)
48 (ꢀ2)
623 (ꢀ13)
126 (ꢀ5)
168
34
6
1310 (ꢀ30)
220 (ꢀ6)
5
6.1 (ꢀ0.2)
2.1 (ꢀ0.1)
0.9 (ꢀ0.1)
300 (ꢀ9)
72 (ꢀ3)
700 (ꢀ50)
120 (ꢀ7)
115
55
158
11
5
6
1
104
15
47
8
12
12
13
18
15
11
13
29
15
16
19
22
9
13
5
2
45
20
221
14
3
135 (ꢀ7)
3300 (ꢀ120)
350 (ꢀ15)
1240 (ꢀ70)
110 (ꢀ12)
1940 (ꢀ110)
150 (ꢀ5)
The N-substituted analogues26ꢁ31 were prepared from
the previously described 1-(2-benzhydryloxy-ethyl)piper-
azine (3a) and 1-{2-[bis-(4-fluorophenyl)-methoxy]
ethyl}piperazine (3b) (Fig. 2). A three step sequence of a
modified Mannich reaction, followed by treatment of the
corresponding Mannich base with iodomethane, and
reaction with either 3a or 3b gave ketone analogues, 5–8,
14–19, 27–30, and 36–39 (Schemes 1 and 2).27 Reduction
of the ketone analogues using LAH in THF afforded
alcohols, 9–12, 20–25, 31–34, and 40–47.28 Alcohols 40–47
were separated into the corresponding trans (40–43) and
cis (44–47) isomers using preparative thin layer chroma-
tography.29 Treatment of compounds 44–47 with
p-toluenesulfonic acid gave alkenes 48–51, respectively.30
190 (ꢀ9)
84 (ꢀ4)
19 (ꢀ1)
10
11
12
14
15
16
17
18
19
20
21
22
23
24
25
27
28
29
30
31
32
33
34
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
10 (ꢀ1)
15 (ꢀ1)
690 (ꢀ14)
59 (ꢀ8)
7.2 (ꢀ0.2)
480 (ꢀ10)
84 (ꢀ2)
5700 (ꢀ350)
970 (ꢀ30)
3300 (ꢀ180)
1130 (ꢀ50)
4100 (ꢀ160)
530 (ꢀ12)
2200 (ꢀ10)
920 (ꢀ30)
3300 (ꢀ230)
580 (ꢀ20)
3900 (ꢀ170)
480 (ꢀ20)
2600 (ꢀ110)
480 (ꢀ20)
1360 (ꢀ40)
220 (ꢀ9)
250 (ꢀ10)
62 (ꢀ2)
270 (ꢀ9)
49 (ꢀ2)
170 (ꢀ6)
32 (ꢀ2)
210 (ꢀ11)
37 (ꢀ2)
210 (ꢀ6)
22 (ꢀ2)
290 (ꢀ9)
37 (ꢀ1)
Our initial investigations began with varying the
N-substituent phenyl ring in 2a and 2b with either a
furan (5, 6) or a thiophene (7, 8). The results from the
binding assay showed that neither bioisosteric replace-
ment resulted in a high affinity analogue or a highly
selective agent (Table 1). Based on the higher affinity of
alcohols, 2c and 2d compared to ketones, 2a and 2b, we
thought that reduction of the ketones in 5–8 might
increase affinity. Alcohols 9–12, did show an increase in
260 (ꢀ10)
89 (ꢀ4)
7.5 (ꢀ0.3)
2.5 (ꢀ0.1)
4.1 (ꢀ0.2)
3.9 (ꢀ0.2)
390 (ꢀ20)
57 (ꢀ4)
340 (ꢀ8)
50 (ꢀ2)
910 (ꢀ50)
53 (ꢀ2)
1050 (ꢀ50)
110 (ꢀ6)
2
78 (ꢀ2)
70 (ꢀ3)
0.9
0.1
7
2
2
0.2
14
23
3
0.4
36
41
14
4
22 (ꢀ2)
3.0 (ꢀ0.3)
440 (ꢀ40)
32 (ꢀ1)
59 (ꢀ2)
18 (ꢀ2)
40 (ꢀ4)
93 (ꢀ5)
18 (ꢀ1)
3.0 (ꢀ0.2)
500 (ꢀ30)
30 (ꢀ2)
35 (ꢀ2)
1.3 (ꢀ0.2)
32 (ꢀ1)
100 (ꢀ6)
7.3 (ꢀ0.2)
38 (ꢀ2)
3.0 (ꢀ0.3)
1360 (ꢀ60)
260 (ꢀ10)
530 (ꢀ30)
57 (ꢀ3)
H
F
H
F
6.4 (ꢀ0.6)
39 (ꢀ1)
14 (ꢀ1)
Figure 2. Structures of 1-(2-benzhydryloxy-ethyl)piperazine (3a), 1-{2-
[bis-(4-fluorophenyl)-methoxy]ethyl}piperazine (3b), and GBR 13069 (3c).
aPrepared and tested as dimaleate salt.
bValue determined as in ref 22.

D. Lewis et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1385–1389
1387
Scheme 1. (a) (CH3)2NH2ꢂ HCl, (CH2O)n, HCl, EtOH; (b) CH3I, EtOH; (c) 3a or 3b, NaHCO3, PhCH3; (d) LAH, THF.
analogues 14–19 and 20–25 confirmed our previous
finding that an aromatic ring in the side-chain appears to
be a requirement for high affinity at the DAT.32
again we saw an increase in affinity for both the DAT
and the SERT. In addition, the presence of a 4-bromo
group was better tolerated by the DAT than the SERT.
The introduction of a 3-fluoro group to 2c, that is, 33,
resulted in the best selectivity (SERT/DAT=221) of
any of the examined compounds. Curiously, this effect
was not observed in the bisfluoro analogue. The addi-
tion of a 3-fluoro group to the N-substituent in 2d, that
is, 34, decreased affinity at the DAT (Ki=3.9 nM vs
Ki=2.1 nM) and increased affinity at the SERT (Ki=53
nM vs Ki=120 nM).
Since the phenyl ring appeared to be important, we
decided to probe the effects of different substituents in
the phenyl ring on the affinity of 2a and 2b. We chose to
evaluate a 4-bromo substituent, that is, 27 and 28, and a
3-fluoro substituent, 29 and 30. The 4-bromo sub-
stituent was chosen based on our previous work that
showed a 4-methoxy group was tolerated by GBR12909
analogues.22 We were curious if that was the result of
the polar oxygen atom or from a favorable steric inter-
action. Similarly, the 3-fluoro substituent was chosen
because it most resembled the size of the H while having
different electronic characteristics. The new ligands, 27–
30, had similar affinity and selectivity for the DAT
compared to 2a and 2b. This seemed to indicate that
substitution in the 3- and 4-position was tolerated but
did not enhance affinity or selectivity. To further test the
role of the 3- and 4-positions of the N-substituent,
ketones 27–30 were reduced to alcohols, 31–34. Once
It occurred to us that to test whether the effect of the
ketone moiety is related to its conformational mobility; we
locked the ketone group in 2a and 2b into a ring, that is, 36
and 37. This resulted in a slight decrease in the affinity of
36 and 37 for the DAT compared to 2a and 2b respec-
tively. In addition, this modification increased affinity for
the SERT of 36 and 37 compared to 2a and 2b. Based on
our finding22 that addition of a methoxy group into the
4-position of the phenyl ring was tolerated, we added a
corresponding methoxy group in the locked conformers

1388
D. Lewis et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1385–1389
Scheme 2. (a) (CH3)2NH2ꢂ HCl, (CH2O)n, HCl, EtOH; (b) CH3I, EtOH; (c) 3a or 3b, NaHCO3, PhCH3; (d) LAH, THF; (e) PTLC (SiO2, CHCl3/
MeOH); (f) p-TsOH, benzene.
of 2a and 2b, giving compounds 38 and 39. This modi-
fication seems to be more favored by the SERT than the
DAT. We observed a 10-fold increase in affinity at the
SERT when a 6-methoxy group is added to 41, that is,
43. Curious if this might be also be paralleled in the case
of the corresponding alcohols, ketones 36–39 were
reduced to their trans- (40–43) and cis- (44–47) isomers.
Generally, the cis- isomers, 44–47, had higher DAT
affinity and selectivity than the trans- isomers, 40–43. In
this series, 45 had higher affinity for the DAT (Ki=1.3
nM) than 2d but was not as selective over the SERT
(SERT/ DAT=23). It occurred to us that if 47 were
dehydrated to form an alkene, that is 49, it might serve
as locked conformer of GBR 13069 (3c).21 It is inter-
esting that 49 has lower affinity (Ki=6.4 nM vs Ki=0.9
nM) and reduced DAT selectivity (41-fold vs 158-fold)
compared to 3c. This would seem to indicate that the
greater flexibility in 3c adds to its high affinity and
selectivity. Alkenes 50 and 51, which are locked con-
formers of compounds previously described,22 however,
had similar affinity and selectivity to the previous
described compounds indicating that this conformation
was preferred for DAT affinity and selectivity. This also
indicates that the 6-methoxy analogues might not be
binding like other GBR 12909 analogues.
Acknowledgements
The authors (LMC, NIDDK) thank the National Insti-
tute on Drug Abuse, NIH, for partial financial support
of our research program. We also thank Noel Whittaker
(NIDDK, NIH) for mass spectral data.
References and Notes
1. Hitri, A.; Hurd, Y. L.; Wyatt, R. J.; Deutsch, S. I. Clin.
Neuropharmacol. 1994, 17, 1.
2. Søgard, U.; Michalow, J.; Butler, B.; Laursen, A. L.;
Ingersen, S. H.; Skrumsager, B. K.; Rafaelsen, O. J. Int. Clin.
Psychopharmacol. 1990, 5, 237.
3. Rothman, R. B.; Mele, A.; Reid, A. A.; Akunne, H.; Greig,
N.; Thurkauf, A.; Rice, K. C.; Pert, A. FEBSLett. 1989, 257,
341.
4. Hsin, L. W.; Dersch, C. M.; Baumann, M. H.; Stafford, D.;
Glowa, J. R.; Rothman, R. B.; Jacobson, A. E.; Rice, K. C.
J. Med. Chem. 2002, 45, 1321.
5. Zhang, Y.; Joseph, D. B.; Bowen, W. D.; Flippen-Ander-
son, J. L.; Dersch, C. M.; Rothman, R. B.; Jacobson, A. E.;
Rice, K. C. J. Med. Chem. 2001, 44, 3937.
6. Zhang, Y.; Rothman, R. B.; Dersch, C. M.; de Costa, B. R.;
Jacobson, A. E.; Rice, K. C. J. Med. Chem. 2000, 43, 4840.
7. Gu, X. H.; Yu, H.; Jacobson, A. E.; Rothman, R. B.;
Dersch, C. M.; George, C.; Flippen-Anderson, J. L.; Rice,
K. C. J. Med. Chem. 2000, 43, 4868.
8. Newman, A. H.; Kulkarni, S. Med. Res. Rev. 2002, 22, 429.
9. Meltzer, P. C.; Wang, B.; Chen, Z.; Blundell, P.; Jayara-
man, M.; Gonzalez, M. D.; George, C.; Madras, B. K. J. Med.
Chem. 2001, 44, 2619.
10. Meltzer, P. C.; Blundell, P.; Yong, Y. F.; Chen, Z.;
George, C.; Gonzalez, M. D.; Madras, B. K. J. Med. Chem.
2000, 43, 2982.
Various N-substituted GBR 12909 and GBR 12935
analogues have been prepared. Despite their structural
similarity not all of these ligands appear to be binding in
an identical manner. Compound 33 has been identified
as a high affinity DAT ligand with good selectivity over
the SERT (SERT/DAT=221). This compound has a
hydroxyl function that might be esterified and could
possibly be converted to a long-acting drug. Further
pharmacological and SAR studies on this ligand and
others with high affinity and selectivity are in progress
and will be published in due course.
11. Preti, A. Curr. Opin. Investig. Drugs 2000, 1, 241.
12. van der Zee, P.; Koger, H. S.; Goojtes, J.; Hespe, W. Eur.
J. Med. Chem. 1980, 15, 363.

D. Lewis et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1385–1389
1389
13. Baumann, M. H.; Phillips, J. M.; Ayestas, M. A.; Ali,
S. F.; Rice, K. C.; Rothman, R. B. Ann. N.Y. Acad. Sci. 2002,
965, 92.
14. Yui, K.; Goto, K.; Ikemoto, S. K. N.; Yoshino, T.; Ishi-
guro, T. Drug Alcohol Depend. 2001, 64, 133.
15. Cross, J. C.; Johnson, B. D.; Davis, W. R.; Liberty, H. J.
Drug Alcohol Depend. 2001, 64, 191.
16. Parrott, A. C.; Sisk, E.; Turner, J. J. D. Drug Alcohol
Depend. 2000, 60, 105.
17. Surratt, H. L. Drug Alcohol Depend. 2000, 58, 267.
18. Compton, W. M. Drug Alcohol Depend. 2000, 58, 215.
19. Rothman, R. B.; Lewis, B.; Dersch, C.; Xu, H.; Radesca,
L.; de Costa, B. R.; Rice, K. C.; Kilburn, R. B.; Akunne,
H. C.; Pert, A. Synapse 1993, 14, 34.
20. Matecka, D.; Rothman, R. B.; Radesca, L.; de Costa,
B. R.; Dersch, C. M.; Partilla, J. S.; Pert, A.; Glowa, J. R.;
Wojnicki, F. H. E.; Rice, K. C. J. Med. Chem. 1996, 39, 4704.
21. Matecka, D.; Lewis, D.; Rothman, R. B.; Dersch, C. M.;
Wojnicki, F. H. E.; Glowa, J. R.; DeVries, A. C.; Pert, A.;
Rice, K. C. J. Med. Chem. 1997, 40, 705.
19.6 mmol), 3a (3.8 g, 12.8 mmol), and K2CO3 (3.3 g, 31.1
mmol) in DMF (50 mL) was heated to reflux for 1 h, cooled to
room temperature, and poured into H2O (200 mL). The aqu-
eous layer was extracted with Et2O (3ꢃ100 mL). The com-
bined Et2O portion was washed with brine, dried (Na2SO4),
and concentrated to an oil. The crude ketone were converted
to the bis-maleate salt and recrystallized from methanol to
afford 7 dimaleate, mp 166–168 ꢄC.
28. General procedure: A 1 M solution of LAH in THF (5
mL, 5 mmol) was added to a solution of 29 (1.3 g, 2.6 mmol)
in dry THF (30 mL). The reaction was stirred at room tem-
perature for 15 min and quenched the sequential addition of
1.3 mL of H2O, 1.3 mL of 15% NaOH, and 1.3 mL of H2O.
The quenched reaction mixture was filtered through a pad of
Celite, the cake washed with three portions of THF, and the
organic fractions combined and concentrated at reduced pres-
sure. The crude alcohol (racemic) was converted to the bis-
maleate salt and recrystallized from methanol to afford 33ꢂ
dimaleate, mp 175–176 ꢄC.
29. General procedure: The racemic alcohols were separated
using preparative thin layer silica gel chromatography. A
Uniplate Silica gel GF 2000 (or 1500) mM plate, with a chro-
matography system of CHCl3/MeOH, 30:1 to was used sepa-
rate the cis- from the trans-isomers. The separated isomers
were each converted to their bis-maleate salts. Recrystalliza-
tion from methanol gave the pure analogues.
30. General procedure: A mixture of 45 (0.4 g, 0.8 mmol), p-
TsOH (0.3 g, 1.8 mmol) in toluene (30 mL) was heated at
reflux for 12 h utilizing a Dean–Stark trap to remove gener-
ated water. The reaction mixture was cooled to room tem-
perature, washed with H2O and brine, dried (Na2SO4), and
concentrated to an oil. The crude olefin was converted to the
bis-maleate salt and recrystallized from methanol to afford
49ꢂ dimaleate, mp 202–204 ꢄC.
22. Lewis, D. B.; Matecka, D.; Zhang, Y.; Hsin, L. W.;
Dersch, C. M.; Stafford, D.; Glowa, J. R.; Rothman, R. B.;
Rice, K. C. J. Med. Chem. 1999, 42, 5029.
23. Glowa, J. R.; Wojnicki, F. H. E.; Matecka, D.; Rice, K. C.;
Rothman, R. B. Exp. Clin. Psychopharmacol. 1995, 3, 219.
24. Glowa, J. R.; Wojnicki, F. H. E.; Matecka, D.; Rice, K. C.;
Rothman, R. B. Exp. Clin. Psychopharmacol. 1995, 3, 232.
25. Glowa, J. R.; Fantegrossi, W. E.; Lewis, D. B.; Matecka,
D.; Rice, K. C.; Rothman, R. B. J. Med. Chem. 1996, 39,
4689.
26. Satisfactory 1H NMR and mass spectral data were
obtained for all final products. Elemental analyses were with-
inꢀ0.4% for C, H, and N.
27. General procedure: A solution of 4b (30.2 g, 239.6 mmol)
dimethylamine hydrochloride (58.7 g, 712.1 mmol), 95% para-
formaldehyde (29.2 g, 971.5 mmol) and concentrated hydro-
chloric acid (3 mL) in ethanol (200 mL) was heated at reflux
overnight. Upon cooling to room temperature, the product
crystallized from solution as the hydrochloride salt, and was
collected by filtration. The salt was converted to the free base
by treatment with aqueous ammonia and extraction with
CHCl3 to give 33.5 g of the corresponding Mannich base. A
mixture of the Mannich base (33.5 g, 182.7 mmol) and iodo-
methane (18.0 mL, 289.2 mmol) in ethanol (500 mL) was stir-
red at room temperature overnight, and the quaternary
methiodide mannich base precipitated from solution. The
precipitated product was collected by filtration and dried and
used directly in the next step without further purification. A
mixture of the quaternary methiodide Mannich base (6.4 g,
31. Uncorrected melting points of dimaleate salts: 5: 164–
167 ꢄC; 6: 164–166 ꢄC; 9: 161–162 ꢄC; 10: 162–164 ꢄC; 11: 166–
167 ꢄC; 12: 174–176 ꢄC; 14: 155–160 ꢄC (dec.); 15: 162–165 ꢄC
(dec.); 16: 164–165 ꢄC; 17: 152–154 ꢄC; 18: 165–167 ꢄC; 19:
150–152 ꢄC; 20: 185–187 ꢄC; 21: 186–188 ꢄC; 22: 174–176 ꢄC;
23: 178–179 ꢄC; 24: 169–171 ꢄC; 25: 162–164 ꢄC; 27: 177–
179 ꢄC; 28: 176–178 ꢄC; 29: 164–167 ꢄC; 30: 159–161 ꢄC; 31:
167–168 ꢄC; 32: 172–173 ꢄC; 34: 171–173 ꢄC; 36: 160–162 ꢄC;
37: 162–163 ꢄC; 38: 154–157 ꢄC; 39: 166–167 ꢄC; 40: 186–
189 ꢄC; 41: 183–184 ꢄC; 42: 181–182 ꢄC; 43: 176–177 ꢄC; 44:
180–182 ꢄC; 45: 176–177 ꢄC; 46: 181–182 ꢄC; 47: 178–179 ꢄC;
48: 197–198 ꢄC; 50: 199–201 ꢄC; 51: 179–180 ꢄC.
32. Hsin, L. W.; Prisinzano, T.; Wilkerson, C. R.; Dersch, C.
M.; Horel, R.; Jacobson, A. E.; Rothman, R. B.; Rice, K. C.
Bioorg. Med. Chem. Lett. 2003, 13, 553.