Haloperidol: Towards further understanding of the structural contributions of its pharmacophoric elements at D2-like receptors

Article abstract of DOI:10.1016/j.bmcl.2004.09.046

An attempt to understand the pharmacophore-relevant position of the alcoholic moiety in haloperidol and the contributions of other pharmacophoric elements led to the re-synthesis of its tropane analogue (compound 2). An analysis of the binding data sugges

Full text of DOI:10.1016/j.bmcl.2004.09.046

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5739–5742
Haloperidol: towards further understanding of the
structural contributions of its pharmacophoric
elements at D2-like receptors
Donald M. N. Sikazwe,a Shouming Li,a Leroy Mardenborough,a Vivian Cody,b
Brian L. Rothc and Seth Y. Ablordeppeya,*
aCollege of Pharmacy and Pharmaceutical Sciences, Florida A & M University, Tallahassee, FL 32307, USA
bHauptman–Woodward Medical Research Institute Inc., Buffalo, NY 14203, USA
cDepartments of Biochemistry, Psychiatry and Neurosciences and NIMH Psychoactive Drug Screening Program,
Case Western Reserve University Medical School, Cleveland, OH 44106, USA
Received 16 August 2004; revised 13 September 2004; accepted 17 September 2004
Available online 6 October 2004
Abstract—An attempt to understand the pharmacophore-relevant position of the alcoholic moiety in haloperidol and the contribu-
tions of other pharmacophoric elements led to the re-synthesis of its tropane analogue (compound 2). An analysis of the binding
data suggests that haloperidol binds to the DA receptors with the OH group in the axial position and the OH group, while not essen-
tial for binding, enhances binding especially at the D2 receptor. It also became clear that shortening the butyrophenone chain not
only reduces binding affinity at the DA receptors but eliminates subtype selectivity.
Ó 2004 Published by Elsevier Ltd.
Haloperidol (1) has long been known to be a most effec-
tive antipsychotic and until recently a drug of choice in
treating schizophrenia.1,2 Although, so much work has
been reported on the structure–activity relationship of
haloperidol,3,4 the preferred conformation and the phar-
macological role of its tertiary alcohol group in binding
to dopamine receptor subtypes5 have not been ad-
dressed. X-ray crystallographic studies indicate haloper-
idolꢀs hydroxyl group to be axial,6,7 however, the
pharmacophore-relevant conformation at the D2-like
receptors has not been identified.
The purpose of this study was threefold: to identify the
preferred conformation and pharmacophore-relevant
position of the alcoholic group at D2-like receptors, to
evaluate the contribution of the hydroxyl group to bind-
ing at the dopamine receptors, and to explore the contri-
bution of the N to phenyl distance of the butyrophenone
for selectivity at the DA D2-family receptor subtypes.
Compound 2 was previously synthesized as a haloper-
idol analogue, which could not undergo metabolism to
form the quaternary pyridinium metabolite, BCPP+.8
O
C
X
N
F
(CH2)3
Cl
OH
O
BCPP+ or HPP+
F
F
C
(CH2)3
1
N
N
Cl
Cl
Compound 2 was re-synthesized to ascertain the posi-
tion of its OH (Scheme 1). In this regard, the carba-
mate-protected intermediate, 4, was submitted for
X-ray analysis in the laboratory of Dr. V. Cody.9 The
X-ray structure analysis showed that the OH group
was fixed in the endo position (equivalent to the axial
position in haloperidol) as shown below (Fig. 1). Thus,
we confirm that Grignard reaction of 4-chlorophenyl
magnesium bromide on the tropinone, 3 results in
stereoselective synthesis of the carbamate-protected
OH
O
C
(CH2)3
2
Keywords: Haloperidol; SAR; Pharmacophoric elements; D2-like rec-
eptors; Dopamine receptors.
*
Corresponding author. Tel.: +1 850 599 3834; fax: +1 850 599
3934; e-mail: seth.ablordeppey@famu.edu
0960-894X/$ - see front matter Ó 2004 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2004.09.046

5740
D. M. N. Sikazwe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5739–5742
OH
O
OH
i
ii
HN
N
N
EtO2C
EtO2C
4
5
3
Cl
OH
Cl
iii
O
F
C
(CH2)3
2
N
Cl
Scheme 1. Reagents and conditions: (i) 4-chlorophenyl magnesium bromide, Et2O (anhydrous), N2, reflux, 12h, 42%; (ii) KOH, ethylene glycol,
150°C, 73%; (iii) 4-chloro-4-fluorobutyrophenone, KI, K2CO3, DME (anhydrous), N2, reflux, 12h, 23%.
preferred position of the OH would be provided by
the synthesis and evaluation of the exo-OH of 2 but
we were hampered by our inability to obtain the exo-
OH analogue using the usual synthetic routes.
We have previously reported the synthesis and binding
affinity of compound 6, a deoxygenated analogue of
haloperidol, to dopamine receptors (D2, Ki = 16nM;
D4, Ki = 26nM).10 The binding data suggest that deox-
Figure 1. X-ray crystallographic rendering of compound 4.
ygenation results in about 16-fold loss of affinity for the
endo-alcohol. Subsequent decarbamylation and alkyla-
D2 receptor but a less pronounced 2-fold loss of affinity
tion of the resulting amine, 5 yielded compound 2 albeit
for the D4 receptor.
in low yield. Attempts to change the Grignard reaction
conditions to obtain the exo-alcohol were unsuccessful.
We have previously shown that compound 2 has a
O
Cl
higher affinity at the D2 receptor than haloperidol. That
being the case, one can conclude that the ethylene bridge
in the tropane moiety of 2, at a minimum, does not dis-
tract from binding to the D2 receptors. If the OH in an
axial position is the preferred conformation, then com-
pound 2 with the OH in the endo position should, at a
minimum, be equipotent to haloperidol in binding to
DA receptors. On the other hand, if 2 is significantly less
potent in binding to the DA receptors then, the exo-
alcohol must be more potent and hence the OH group
must be preferred in the equatorial position in haloper-
idol. The binding data (Table 1) indicates that 2 is about
3-fold more potent than haloperidol at the D2 receptor
subtype and hence suggests that haloperidol binds to the
D2 receptor with the OH group in the axial conforma-
tion. It was hoped that additional confirmation of the
F
C (CH2)3
N
6
To further evaluate the contribution of the OH group to
binding at the dopamine receptors, it was necessary to
obtain compound 9, the deoxygenated analogue of 2.
In this regard, intermediate 5 was subjected to dehydra-
tion conditions to produce 7, which was subsequently
alkylated to obtain 8. Intermediate 7 was also hydrogen-
ated and then alkylated with the intention of obtaining
compound 9 however, subsequent hydrogenation con-
sistently resulted in dechlorination (compound 10) as
well (Scheme 2).
Table 1. Physicochemical data and binding properties of synthetic compounds at dopamine receptors
Recrystallization solvent % Yield Mp (°C) Empiricala formula D2 Ki (nM)b D3 Ki (nM)b
D4 Ki (nM)b Ki D2/Ki D4
1c
—
—
23
67
34
11
33
54
44
59
—
1.1 0.7
0.31 0.04
16.3 5.1
5.5 3.0
0.81 0.4
46.0 5.3
222.3 45
51.4 9.0
3.2 1.4
12.7 7.2
12.1 3.3
25.9 3.0
0.1
2b MeOH/Et2O
6d MeOH/Et2O
237–238 C23H26NO2Cl2FÆ0.5H2O
209–210 C21H24NOCl2F
216–217 C23H24NOCl2FÆ0.1H2O
183–184 C23H27NOClFÆ0.4H2O
128–129 C23H26NO2FÆ0.25H2O
142–143 C20H24N2OCl3FÆ0.2H2O 253.5 38.9
257–258 C21H22NO2Cl2FÆ0.2H2O 1231 144.8 >10,000
110–111 C20H22NOCl2FÆ0.25H2O 1050 209.3 172.4 33
0.03
0.6
8
MeOH/Et2O
189.8 31.6
124.4 16.5
2.3 0.28
345.3 48.7
176.8 36.0
19.1 2.3
0.6
0.7
10 MeOH/Et2O
12 None
0.12
13 MeOH/Et2O
14 MeOH/Et2O
15 MeOH/Et2O
403.9 66.0
17.5 1.98
789 363
1015 178.9
14.5
1.6
1.03
a All compounds reported are HCl salts and each passed CHN analysis within 0.4% of the theoretical values.
b Radioligand binding studies were conducted by NIMH DPSP and reported here as mean Ki SEM of four determinations. For compounds 1, 2,
and 6, data previously reported.
c Data previously reported in Ref. 8.
d Data previously reported in Ref. 10. Dopamine receptor binding assays were performed using cloned receptor preparations as previously
described.12

D. M. N. Sikazwe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5739–5742
5741
OH
i
iii
HN
HN
N
Ar
Cl
X
Cl
Cl
5
7
ii, iii
8
O
9, X = Cl
10, X = H
Ar =
F
C (CH2)3
N
Ar
Scheme 2. Reagents and conditions: (i) AcOH (17.4M), HCl (12M), reflux, 2.5h, 100%; (ii) 10% Pd/C, H2 (65psi), MeOH, 1.5h, rt, 100%; (iii) 4-
chloro-4-fluorobutyrophenone, KI, K2CO3, DME (anhydrous), N2, reflux, 12h, yields (8 = 34%, 10 = 11%).
Hence, we synthesized compound 12 using the same ap-
proach as indicated above for 2 (Scheme 3), to provide
the baseline for comparison of the deoxygenated ana-
logues, 8 and 10. The physicochemical characterization
and binding data of 8, 10, and 12 are recorded in Table 1.
methylene moiety confers
selectivity.11
a
more significant D4
N
N
Cl
N
N
H
L745,870
D2Ki = 960 nM
D4Ki = 0.43 nM
The binding data show that deoxygenation of 12 (D2,
Ki = 2.3nM; D4, Ki = 19.1nM) as expected resulted in
significant loss of activity (10, D2, Ki = 124.4nM; D4,
Ki = 176.8nM). The significant contribution of the alco-
holic function in these compounds suggests that there
may be an interaction (e.g., hydrogen bonding) between
the OH group and the dopamine receptor sites. To fur-
ther explore this possibility, the C–OH group was re-
O
F
C (CH2)3
N
N
Cl
13
D2Ki = 254 nM
D4Ki = 17 nM
placed by
a nitrogen atom. Compound 13 was
synthesized by simply alkylating the commercially avail-
able 4-(4-chlorophenyl)piperazine with the appropriate
alkyl halide as previously indicated. Evaluation of the
binding affinity of 13 indicates a substantial decrease
in affinity (D2, Ki = 254nM; D4, Ki = 17nM) at the
D2 receptor but not at the D4 receptor. The resulting
decrease confirms the lack of the N-atomꢀs capacity to
participate in a similar interaction at the D2 receptor
subtype. Comparison between 13 and the Merck com-
pound L745,870 shows that the shorter 7-azaindole
Thus, it was of interest to shorten the butyrophenone
chain in compound 2 to explore the possibility that D4
selectivity might be conferred on the tropane analogues
as well (Scheme 4). Compounds 14 and 15 were synthe-
sized to accomplish this goal according to Scheme 4
below.
Binding affinity at the dopamine receptors11,12 (Table 1)
shows that shortening of the butyrophenone chain
length not only resulted in decrease in affinity for both
OH
OH
OH
i
ii
HN
Ar
N
HN
H
12
H
5
11
Cl
O
Ar =
F
C (CH2)3
Scheme 3. Reagents and conditions: (i) 10% Pd/C, H2 (65psi), MeOH, 2h, rt, 88%; (ii) 4-chloro-4-fluorobutyrophenone, KI, K2CO3, DME
(anhydrous), N2, reflux, 18h, 33%.
OH
OH
i
HN
Ar1
Ar2
N
N
O
C
Cl
ii
5
14, Ar =
F
1
CH2
Cl
Cl
14
OH
15, Ar =
F
CH2
2
15
Scheme 4. Reagents and conditions: (i) 2-chloro-4-fluoroacetophenone, KI, K2CO3, DME (anhydrous), N2, reflux, 18h, 44%; (ii) 4-fluoro-
benzylbromide, KI, K2CO3, DME (anhydrous), N2, reflux, 18h, 59%.

5742
D. M. N. Sikazwe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5739–5742
2. Baldessarini, R. J. In Goodman and Gilmanꢀs Pharmaco-
logical Basis of Therapeutics, 7th ed.; Gilman, G. A.,
Goodman, L. S., Rall, T. W., Murad, F., Eds.; Macmillan
Publishing Company: New York, 1985; pp 385–445.
3. Sherman, M. A.; Bolger, M. B. J. Biol. Chem. 1988, 263,
4064–4074.
D2 and D4 receptors but also did not result in D4 selec-
tivity. Thus, the physicochemical characteristics of the
azaindole methylene moiety rather than the phenyl to
N distance may be a contributing factor in the affinity
and selectivity of L745,870 for the D4 receptor.
4. Sherman, M. A.; Linthicum, D. S.; Bolger, M. B. Mol.
Pharmacol. 1986, 29, 589–598.
5. Seeman, P.; Van Tol, H. H. Trends Pharmacol. Sci. 1994,
7, 264–270.
6. Reed, L. L.; Schaefer, J. P. Acta Crystallogr. 1973, B29,
1886–1890.
7. Rusig, I.; Leger, J. M.; Laguerre, M.; Saxena, A. K.;
Carpy, A. J. Chem. Crystallogr. 1995, 25(8), 443–
451.
In summary, our data suggest that the pharmacophore-
relevant conformation of the alcoholic function in
haloperidol is axial, and the OH group while not essen-
tial for binding at the D2 receptor enhances binding
affinity in a significant manner. With the limited data
available, one is led to suggest that the azaindole meth-
ylene moiety appears to confer D4 binding selectivity to
an appropriate phenyl amine moiety such as the phenyl-
piperazine moiety in L,745,870.
8. Sikazwe, D. M. N.; Lyles-Eggleston, M.; Li, S.; Ablor-
deppey, S. Y. Bioorg. Med. Chem. Lett. 2003, 13, 3779–
3782.
9. Unpublished communication.
Acknowledgements
10. Lyles-Eggleston, M.; Altundas, R.; Xia, J.; Sikazwe, D. M.
N.; Fan, P.; Yang, Q.; Li, S.; Zhang, W.; Zhu, X.;
Schmidt, A. W.; Vanase-Frawley, M.; Shrikande, A.;
Villalobos, A.; Borne, R. F.; Ablordeppey, S. Y. J. Med.
Chem. 2004, 47, 497–508.
11. Kulagowski, J. J.; Broughton, H. B.; Curtis, N. R.;
Mawer, I. M.; Ridgill, M. P.; Baker, R.; Emms, F.;
Freedman, S. B.; Marwood, R.; Patel, S.; Patel, S.;
Ragan, C. I.; Leeson, P. D. J. Med. Chem. 1996, 39, 1941–
1942.
We gratefully acknowledge the financial support of the
National Institute of General Medical Studies (NIGMS)
for MBRS grant # GM 08111, the NCRR for RCMI
grant number G12 RR 03020, Title III and the NIMH
PDSP.
References and notes
12. Shapiro, D. A.; Renock, S.; Arrington, E.; Chiodo, L. A.;
Liu, L. X.; Sibley, D. R.; Roth, B. L.; Mailman, R.
Neuropsychopharmacology 2003, 28, 1400–1411.
1. Settle, E. C.; Ayd, F. J. J. Clin. Psychiatry 1983, 44(12),
440–446.