Synthesis, evaluation, and comparative molecular field analysis of 1- phenyl-3-amino-1,2,3,4-tetrahydronaphthalenes as ligands for histamine H1 receptors

Article abstract of DOI:10.1021/jm980428x

A series of 1-phenyl-3-amino-1,2,3,4-tetrahydronaphthalenes (1-phenyl- 3-aminotetralins, PATs) previously was found to modulate tyrosine hydroxylase activity and dopamine synthesis in rodent forebrain through interaction with a binding site labeled by [³H]-(-)-(1R,3S)-trans-H₂-PAT. Recently, we have discovered that PATs also bind with high affinity to the [³H]mepyramine- labeled H₁ receptor in rat and guinea pig brain. Here, we report the synthesis and biological evaluation of additional PAT analogues in order to identify differences in binding at these two sites. Further molecular modifications involve the pendant phenyl ring as well as quaternary amine compounds. Comparison of about 38 PAT analogues, 10 structurally diverse H₁ ligands, and several other CNS-active compounds revealed no significant differences in affinity at [³H]-(-)-trans-H₂-PAT sites versus [³H]mepyramine-labeled H₁ receptors. These results, together with previous autoradiographic brain receptor-mapping studies that indicate similar distribution of [³H]-(-)-trans-H₂-PAT sites and [³H]mepyramine-labeled H₁ receptors, suggest that both radioligands label the same histamine H₁ receptors in rodent brain. We also report a revision of our previous comparative molecular field analysis (CoMFA) study of the PAT ligands that yields a highly predictive model for 66 compounds with a cross-validated R² (q²) value of 0.67. This model will be useful for the prediction of high- affinity ligands at radiolabeled H₁ receptors in mammalian brain.

Full text of DOI:10.1021/jm980428x

J . Med. Chem. 1999, 42, 3041-3054
3041
Syn th esis, Eva lu a tion , a n d Com p a r a tive Molecu la r F ield An a lysis of
1-P h en yl-3-a m in o-1,2,3,4-tetr a h yd r on a p h th a len es a s Liga n d s for Hista m in e H1
Recep tor s
Ehren C. Bucholtz, Randal L. Brown, Alexander Tropsha, Raymond G. Booth, and Steven D. Wyrick*
Division of Medicinal Chemistry and Natural Products, School of Pharmacy, CB# 7360, University of North Carolina,
Chapel Hill, North Carolina 27599-7360
Received J uly 24, 1998
A series of 1-phenyl-3-amino-1,2,3,4-tetrahydronaphthalenes (1-phenyl-3-aminotetralins, PATs)
previously was found to modulate tyrosine hydroxylase activity and dopamine synthesis in
rodent forebrain through interaction with a binding site labeled by [3H]-(-)-(1R,3S)-trans-H2-
PAT. Recently, we have discovered that PATs also bind with high affinity to the [3H]-
mepyramine-labeled H1 receptor in rat and guinea pig brain. Here, we report the synthesis
and biological evaluation of additional PAT analogues in order to identify differences in binding
at these two sites. Further molecular modifications involve the pendant phenyl ring as well as
quaternary amine compounds. Comparison of about 38 PAT analogues, 10 structurally diverse
H1 ligands, and several other CNS-active compounds revealed no significant differences in
affinity at [3H]-(-)-trans-H2-PAT sites versus [3H]mepyramine-labeled H1 receptors. These
results, together with previous autoradiographic brain receptor-mapping studies that indicate
similar distribution of [3H]-(-)-trans-H2-PAT sites and [3H]mepyramine-labeled H1 receptors,
suggest that both radioligands label the same histamine H1 receptors in rodent brain. We also
report a revision of our previous comparative molecular field analysis (CoMFA) study of the
PAT ligands that yields a highly predictive model for 66 compounds with a cross-validated R2
(q2) value of 0.67. This model will be useful for the prediction of high-affinity ligands at
radiolabeled H1 receptors in mammalian brain.
Ch a r t 1. trans-H2-PAT (left) and trans-Cl,OH-PAT
(right)
In tr od u ction
The phenylaminotetralins (PATs) (()-trans-1-phenyl-
3-(dimethylamino)-1,2,3,4-tetrahydronaphthalene (trans-
H2-PAT, 1; Chart 1) and (()-trans-6-chloro-7-hydroxy-
1-phenyl-3-(dimethylamino)-1,2,3,4-tetrahydronaphtha-
lene (Cl, OH-PAT, 2; Chart 1) stimulate tyrosine hy-
droxylase activity and dopamine (DA) synthesis in
rodent brain in vitro.1 It is hypothesized that this effect
is mediated through interaction with a CNS binding site
that is labeled by [3H]-(()-trans-Cl,OH-PAT.2 This site
was designated originally as the PAT-σ3 receptor,
because although it had low affinity for σ1/σ2 ligands, it
incorporated a σ-like pharmacophore, it had a σ-like
rodent brain distribution, and its functional effects on
dopamine synthesis were blocked by the σ antagonist
BMY-14802. While (()-Cl,OH-PAT-induced stimulation
of tyrosine hydroxylase is accompanied by nonspecific
effects on dopamine release, (()-trans-H2-PAT effects
at 0.1-10 µM, are confined to stimulation of dopamine
synthesis. Consequently, the enantiomers of (()-trans-
H2-PAT were resolved, and stimulation of tyrosine
hydroxylase and DA synthesis activity at 0.1 µM was
found to reside in the (-)-trans-(1R,3S) enantiomer.2
This enantiomer also exhibited 40-fold greater affinity
for the PAT-σ3 receptor when compared to the (+)-
(1S,3R) enantiomer. Moreover, preliminary studies
indicated that at 0.1 µM, (()-cis-H2-PAT did not stimu-
late tyrosine hydroxylase and had about 5-fold less
affinity for [3H]-(()-trans-Cl,OH-PAT sites.2 These data
suggested a dependence between chirality, receptor
interaction, and functional effects. Consequently, the
stereoselective radioligand, [3H]-(-)-(1R,3S)-trans-H2-
PAT, was synthesized and used to label PAT-σ3 sites in
all subsequent radioreceptor studies.3
Recently, radioreceptor assays indicated that (-)-
trans-H2-PAT also has high affinity (K0.5 ) 1.6 nM) for
histamine H1 receptors and that the pharmacological
profile of [3H]-(-)-trans-H2-PAT sites is similar to that
of H1 receptors.4-6 We discovered, for example, that the
distribution and rank order of ligand binding at [3H]-
(-)-trans-H2-PAT sites is more comparable but not
identical to [3H]mepyramine-labeled H1 sites versus
[3H]ditolylguanidine (DTG)-labeled σ1/σ2 sites in guinea
pig brain.7 These observations, taken together with
recent molecular cloning evidence8,9 suggesting that σ
sites are not closely related to other known mammalian
neurotransmitter receptors, led us to hypothesize that
[3H]-(-)-trans-H2-PAT sites may actually represent
brain histamine H1-type receptors.
(-)-trans-H2-PAT incorporates the pharmacophore
present in the four prototypical structural classes of
histamine H1 antagonists: the ethylenediamines, the
* To whom correspondence should be addressed.
10.1021/jm980428x CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/16/1999

3042 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Bucholtz et al.
Ta ble 1. Binding Affinity of PAT Analogues Used in CoMFA/q2-GRS Studies of the [3H]-(-)-trans-H2-PAT Binding Site
K0.5 (nM)
c
compd
configa
R1
R2
R3
R4
R5
R6
R7
PATb
H1
1a
1b
1c
1d
2
3
4
5
6
(-)-trans
(+)-trans
(-)-cis
H
H
H
H
H
H
H
H
H
H
H
N(CH3)2
N(CH3)2
N(CH3)2
N(CH3)2
N(CH3)2
N(CH3)3
NH(CH3)
N(C2H5)2
NCH3(C3H5)
N(C3H5)2
H
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
C2H4
C2H4
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
o-Cl
o-CH3
p-Cl
p-CH3
p-F
H
H
H
H
H
H
H
H
H
Cl
H
H
H
H
H
H
H
H
H
H
H
H
OH
H
H
H
H
H
H
H
H
0.33 ( 0.09d
13.6 ( 2.1d
2.9 ( 0.4d
70.4 ( 8.1d
0.63 ( 0.05
35 ( 2.5
1.07 ( 0.24d
19.1( 1.6d
2.45 ( 0.13d
111 ( 3.5d
1.47 ( 0.20
120 ( 15
64 ( 10
13.3 ( 1.37
ND
ND
>2000
ND
217 ( 49
ND
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
(+)-cis
(()-trans
(()-trans
(()-trans
(()-trans
(()-trans
(()-trans
(()-cis
112 (_31
5.7 ( 0.4
3.4 ( 0.3f
10.2 ( 1.7f
1200 ( 100
940 ( 53
7
8
9
N(CH3)2
N(CH3)2
(()-trans
(()-cis
H
10
11
12
13
14
15
16
17
18
20
21
22
23
24
25
28
29
30
31
32
33
34
35
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
N(CH3)2
N(CH3)2
N(CH3)2
N(CH3)2
N(CH3)2
N(CH3)2
NCH3((CH2)2C6H5)
NCH3((CH2)3C6H5)
NCH3((CH2)4C6H5)
N(CH3)2
N(CH3)2
N(CH3)2
N(CH3)2
N(CH3)2
N(CH3)3
N(CH3)2
NH2
109 ( 13
(()-trans
(()-trans
(()-cis
H
H
20.9 ( 2.4e
60 ( 5.8
OH
OH
H
H
H
H
H
H
H
H
H
H
H
Cl
OH
Cl
H
H
H
OH
OH
H
H
H
H
H
H
H
H
H
H
H
OH
OH
OH
H
H
H
69 ( 4.2
11.8 ( 2.4
40 ( 5.0
4.4 ( 1.1
ND
9.6 ( 1.7
(()-trans
(()-cis
18.3 ( 1.9
1.90 ( 0.50
140 ( 24e
333 ( 33e
212 ( 39e
53 ( 11.2
10.2 ( 1.7
9.0 ( 2.8
(()-trans
(()-trans
(()-trans
(()-trans
(()-trans
(()-trans
(()-trans
(()-trans
(()-cis
ND
ND
42.7 ( 4.7
12.7 ( 0.6
4.4 ( 1.6
5.0 ( 0.4
2.3 ( 0.1
107 ( 4.1
ND
ND
ND
ND
ND
2.47 ( 0.23
1.53 ( 0.10
117 ( 2.8
0.54 ( 0.1f
>5000f
(()-cis
(()-cis
(()-trans
(()-trans
(()-trans
(()-trans
(()-trans
(()-trans
NH(CH3)
NH(C3H5)
NH(CH2)3C6H5
NH(CH2)4C6H5
NH2
8.1 ( 0.8f
45 ( 11f
H
H
H
H
2500 ( 400e
1500 ( 200e
>5000f
ND
ND
ND
OH
H
OH
H
NH2
1270 ( 92
a
Configuration (config) cis or trans denotes the relationship of substituents at positions 1 and 3. b Versus [3H]-(-)-trans-H2-PAT. c Versus
[3H]mepyramine. Reference 12. e Reference 10. Reference 2. ND, not determined.
d
f
aminoalkyl ethers, the semirigid 1,1-diaryl-3-(alkylami-
no)propenes, and the 1,1-diaryl-(3-alkylamino)pro-
panes.10 Functionally, H1 antagonists have been char-
acterized by their ability to block histamine-induced
contraction of guinea pig ileum. Structure-activity
relationships (SAR) for H1 antagonists have been re-
viewed elsewhere.10-14
cis- and trans-benzocycloheptyl-PATs (10 and 11, re-
spectively) which expand the tetralin ring also have
moderately reduced affinity compared to (()-trans-H2-
PAT, perhaps due to a less drastic alteration of the
3-amino position as compared to the 2-amino analogues
(8 and 9) of trans-H2-PAT. Aromatic substitutions, such
as 6-chloro-7-hydroxy (2), retain high binding affinity.
However, catechol analogues (12 and 13) were found to
exhibit reduced affinity at this site compared to 2. The
trans- and cis-1-methyl-1-phenyl-3-(dimethylamino)tet-
ralins (14 and 15) also showed reduced affinity com-
pared to (()-trans-H2-PAT by approximately 1 order of
magnitude.
Molecular modeling studies were performed on these
PATs as well as various non-PAT ligands with a wide
range of affinities for the [3H]-(-)-trans-H2-PAT binding
site.16 A pharmacophore mapping program (distance
comparison, DISCO)17 was utilized to identify structural
features that are common to ligands that bind to the
PAT binding site and to develop a ligand binding model.
The resulting alignment was utilized in a comparative
molecular field analysis (CoMFA) study to correlate the
SARs for the [3H]-(-)-trans-H2-PAT binding site have
revealed little steric tolerance for amine substituents;
a dimethyl substitution is optimal for high affinity.
N,N,N-Trimethyl (quaternary, 3), N-methyl (4), N,N-
diethyl (5), N-allyl-N-methyl (6), and N,N-diallyl (7)
PATs all have reduced affinity at the site compared to
(()-trans-H2-PAT. The effect of altering the position of
the amino group was investigated by designing PAT
analogues with the amine in position 2 of the tetrahy-
dronaphthalene and by making ring-expanded benzo-
cycloheptyl-PAT analogues.15 The cis- and trans-1-
phenyl-2-aminotetralins (8 and 9; Table 1) exhibit an
approximately 1000-fold decrease in affinity, indicating
a proper spatial orientation of the 3-amine relative to
the aromatic substituents is important for binding. The

PATs as Ligands for Histamine H1 Receptors
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3043
Ta ble 2. Binding Affinity of Non-PATs Used in CoMFA/
q2-GRS Studies of the [3H]-(-)-trans-H2-PAT Binding Site
steric and electrostatic fields with affinity for the PAT
site. A cross-validated R2 (q2) of 0.61 was obtained
indicating a model with a high degree of internal
predictability. Visualization of the volumes that con-
tribute to the affinities of the active versus less active
compounds indicated that an auxiliary binding pocket
may exist that can accommodate large phenylalkyl
nitrogen substituents. To test this model, we synthe-
sized trans-H2-PATs that included 2-phenylethyl (16),
3-phenylpropyl (17), and 4-phenylbutyl (18) nitrogen
substituents. In competition binding assays, not only
did these PAT analogues have markedly reduced affinity
for the PAT binding site (K0.5 ) 150, 300, and 280 nM,
respectively), but they also exhibited increased affinity
for dopamine and σ receptors.15 This erroneous predic-
tion of high affinity for phenylalkyl-substituted PATs
was most likely due to the fact that these new ligands
extended beyond the probed receptor space. These data
suggest that further refinement of the reported binding
model is required.
K0.5 (nM)
b
compd
PATa
H1
19, 2-(dimethylamino)tetralin
26, 3-aminofluoranthene
165 ( 24
ca. 4800
75 ( 4.7
ca. 4000
27, 3-(dimethylamino)fluoranthene ca. 4800
ca. 4000
36, diphenhydramine
37, (+)-butaclamol
38, (R)-chlorpheniramine
39, (S)-chlorpheniramine
40, triprolidine
41, atropine
42, (-)-NANM
43, doxepin
44, cis-flupenthixol
45, trans-flupenthixol
46, mazindol
4.1 ( 0.41
6.1 ( 1.7
405 ( 76
61.6 ( 6.4
1.16 ( 0.37
0.46 ( 0.05
580 ( 52
ca. 2000
147 ( 13
20.9 ( 2.9
0.31 ( 0.03
0.12 ( 0.03
1236 ( 177
>5000
0.06 ( 0.01
0.16 ( 0.04
0.86 ( 0.17c ND
5.73 ( 0.85c ND
ca. 600
ND
47, SCH-23390
48, methysergide
49, SKF-38393
101 ( 14
ND
ND
ND
ca. 3000
ca. 2500
50, amytriptyline
51, desipramine
52, 7-OH-DPAT
53, haloperidol
0.50 ( 0.03
45.4 ( 1.8
ca. 1500
215 ( 20
ca. 2000
ND
ND
ND
ND
Since reporting our previous binding model, we also
have learned more about how stereochemistry affects
binding at this site. (-)-trans-H2-PAT (1a ; Table 1) and
(-)-cis-H2-PAT (1c) have higher affinity, as compared
to their corresponding (+) isomers (1b and 1d , respec-
tively). The rank order of binding of H2-PAT isomers is
(1R,3S) (1a ) > (1S,3S) (1c) > (1S,3R) (1b) > (1R,3R)
(1d ), suggesting that chirality (S configuration) at the
C3 amine position (shared by both (-)-cis- and (-)-trans-
H2-PAT) is an important structural determinate for high
54, loperamine
ND
55, mianserin
56, rimcazole
57, spiperone
58, ketanserin
59, promethazine
60, mepyramine
61, GBR-12909
62, nortryptyline
63, (() BMY-14802
0.30 ( 0.07
>5000
ND
ND
ND
ND
ND
0.53 ( 0.12
ND
272 ( 48
1.79 ( 0.05
0.24 ( 0.06
0.58 ( 0.28
9.33 ( 1.66
15.1 ( 5.6
180 ( 14
ND
548 ( 187
a
Versus [3H]-(-)-trans-H2-PAT. b Versus [3H]mepyramine. c Ref-
18
affinity of PAT-type molecules at PAT binding sites.
erence 1.
We have undertaken further SAR studies here to
identify differences between the classical [3H]mepyr-
amine-labeled H1 sites and [3H]-(-)-trans-H2-PAT-
labeled sites to test our hypothesis that [3H]-(-)-trans-
H2-PAT may represent histamine H1-type receptors.
Using (()-trans-H2-PAT as a probe, we have compared
the SAR of the classical [3H]mepyramine-labeled H1
sites and [3H]-(-)-trans-H2-PAT-labeled sites. Structural
modifications to (()-trans-H2-PAT (analogues 19-24;
Tables 1 and 2) address steric and electrostatic require-
ments of the pendant phenyl ring. The 2-aminotetralin
19 (Table 2, Chart 2) tests if the pendent phenyl is a
requirement for binding. The substituted pendent phen-
yl ring compounds 20-24 (Table 1) address the effects
on binding of steric or electrostatic variation on the
pendent phenyl ring and provide insight as to determine
how nonrigid compounds with aromatic substituents
bind to the two sites. The cis-trimethylammonium
quaternary analogue 2 (Table 1) of 1 tests further if
quaternization of the amine reduces affinity for these
two sites. The 3-aminofluoranthenes 26 and 27 (Table
2, Chart 2) investigate the importance of the coplanarity
of the diaryl ring system and the relative position of the
amine group. We also have completed competition
binding assays to assess binding of 6,7-substituted
compounds (2, 12, and 13; Table 1) as well as numerous
other PAT ligands to the [3H]mepyramine-labeled H1
sites which previously have been analyzed at the [3H]-
(-)-trans-H2-PAT-labeled sites.2,15,18
lier2,15,18 and non-PAT compounds (36-63, Chart 2) to
develop a revised and extended CoMFA model for ligand
binding at the [3H]-(-)-trans-H2-PAT-labeled sites. The
new model is characterized by a high q2 value of 0.68
which slightly exceeds the q2 value (0.61) obtained for
a previous CoMFA model.16 This new CoMFA model for
ligand binding to [3H]-(-)-trans-H2-PAT-labeled sites of
PAT ligands provides new insights into the specificity
of PAT binding and facilitates the design of novel PAT
ligands. These SAR studies will help to increase our
understanding of brain histamine H1-type receptors and
to investigate the role histamine plays as a brain
neurotransmitter.
Ch em ica l Syn th esis
Synthesis of 2-aminotetralin 19 (Scheme 1) was
initiated by reduction of â-tetralone (64) which yielded
the 2-tetralol (65). This was converted to the tosylate
and subsequently treated with sodium azide to afford
the 2-azido-1,2,3,4-tetrahydronaphthalene (66) by the
procedure of Laus.19 The azide was reduced to the
primary amine 67 and converted to the 2-(dimethyl-
amino)-1,2,3,4-tetrahydronaphthalene (19) by the method
of Eschweiler.20 The synthetic path used for synthesis
of ortho- and para-substituted PATs was identical to
that previously reported for Cl,OH-PAT and H2-PAT.2,4
Com p u ta tion a l Meth od s
The synthesis and biological evaluation of these nine
new PATs (19-27 ) reported in this paper have been
used together with 57 (Tables 1 and 2) other PAT
analogues (1-18, 28-35) synthesized and tested ear-
SYBYL molecular modeling software21 was used for
structure generation and CoMFA.22 For this work, we
have selected 66 chemically diverse ligands that bind
to the histamine H1 receptor (Tables 2 and 3) whose

3044 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Ch a r t 2. Non-PATs Used in Molecular Modeling Studies
Bucholtz et al.

PATs as Ligands for Histamine H1 Receptors
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3045
Ta ble 3. q2 and SDEP (numbers in parentheses) Values Obtained after Performing CoMFA/q2-GRS with Different q2 Cutoff Valuesa
no. of components
dataset
q2 cutoff
lattice points
1
2
3
4
5
6
training
training
training
training
total
total
total
total
noneb
0.1
1210
700
600
0.050 (1.339)
0.136 (1.238)
0.122 (1.248)
0.134 (1.204)
0.218 (1.226)
0.282 (1.175)
0.237 (1.211)
0.184 (1.252)
0.377 (1.095)
0.441 (1.006)
0.344 (1.090)
0.333 (1.099)
0.473 (1.014)
0.510 (0.978)
0.519 (0.969)
0.492 (0.996)
0.460 (1.031)
0.451 (1.008)
0.274 (1.159)
0.272 (1.161)
0.580 (0.913)
0.594 (0.898)
0.556 (0.939)
0.481 (1.015)
0.425 (1.076)
0.472 (0.999)
0.292 (1.170)
0.308 (1.144)
0.637 (0.856)
0.590 (0.910)
0.556 (0.946)
0.487 (1.017)
0.398 (1.113)
0.503 (0.978)
0.324 (1.144)
0.324 (1.144)
0.658 (0.838)
0.617 (0.887)
0.587 (0.921)
0.470 (1.042)
0.404 (0.120)
0.513 (0.982)
0.329 (1.153)
0.329 (1.153)
0.681 (0.816)
0.634 (0.874)
0.608 (0.905)
0.427 (1.093)
0.15
0.2
400
noneb
0.1
1716
2625
1125
500
0.2
0.3
a
b
The numbers in bold represent the q2 values for the optimal number of components and the lowest SDEP. The results of conventional
CoMFA.
Ta ble 4. Summary of CoMFA/q2-GRS Results
training set (compounds 1-47)
conventional CoMFA
CoMFA/ q2-GRS
total 66
conventional CoMFA
CoMFA/ q2-GRS
q2 threshold
minimum σ
no. of small boxes
no. of lattice points
optimal no. of components
q2
none
2.0
none
1210
3
0.460
1.031
0.337
0.941
141.535a
0.000a
0.1
2.0
7
700
none
2.0
none
1716
6
0.681
0.816
0.347
0.941
192.8b
0.000b
0.1
2.0
22
2750
6
0.634
0.874
0.388
0.927
151.848b
0.000b
6
0.513
1.113
0.372
0.928
114.096a
0.000a
SDEP
SEE
R2
F values
prob of R2 ) 0
relative contributions
steric
0.648
0.352
0.553
0.447
0.663
0.337
0.635
0.365
electrostatic
a
b
n1 ) 5, n2 ) 44. n1 ) 5, n2 ) 60.
Sch em e 1
of these compounds were constructed, and configura-
tions that best matched (-)-trans-H2-PAT were used.
The enantiomer of compound 46 used was the (R)
configuration since it best matched (-)-trans-H2-PAT.
For all racemic trans-H2-PAT analogues, (-)-trans-H2-
PAT was used as a template. For all racemic cis-H2-
PAT analogues, (-)-cis-H2-PAT was used. Thus, this
strategy is likely to be valid based on the available data
with this receptor and compounds of the structural
classes we used. Structure optimization and field fit
minimization were performed using the standard Tripos
force field with the maximum iteration cutoff of 1000
steps. SYBYL random search method was used to
search for low-energy conformers. All calculations were
performed on a Silicon Graphics Indigo2 workstation.
The default SYBYL settings were used except where
otherwise noted.
receptor affinity was measured in our laboratory. This
data set includes a variety of trans- and cis-H2-PAT
analogues (Table 1), as well as representatives of several
other structural classes (Chart 2, Table 2). The competi-
tion binding activity of the compounds is expressed as
-log(K0.5) (Table 5). In the case of racemic mixtures, the
affinity of the active enantiomer was approximated by
dividing the K0.5 by 2, making the assumption that one
enantiomer was significantly less active than the other.
For those compounds with chiral centers, racemic data
was used and we modeled them as close as possible to
(-)-trans-H2-PAT. Compounds 26, 27, and 52 were
modeled such that the 3-amino group had an (S)
configuration; Compounds 42, 47, 48, and 49 were
modeled in the configurations shown in Chart 3. Since
compounds 41, 59, and 63 are extremely flexible mol-
ecules, stereochemistry at the chiral centers matters
little in alignment with (-)-trans-H2-PAT. Enantiomers
Str u ctu r e Gen er a tion a n d Align m en t Ru les
We have analyzed the SARs for the total of 66
compounds that were initially separated into two
groups: training set (compounds 1-47) and test set
(compounds 48-63). The use of chemically diverse test
compounds (as compared to the compounds in the
training set) is generally recommended for assessing the
true predictability of any QSAR study. The compounds
in the training set consisted of a wide range of struc-
tures, including PATs and non-PAT molecules that
exhibit a wide range of affinities. We also selected
ligands that vary structurally and represent different
receptor classes to form the test set.
The conformation of the template molecule, (-)-trans-
H2-PAT, was chosen from a previous modeling study of
the PAT binding site.16 All trans-PAT analogues were

3046 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Bucholtz et al.
Ta ble 5. CoMFA Actual, Calculated, and Predicted Activities for Training and Test Set Molecules
training set (1-47) or test set (48-63)
all compounds
conventional CoMFA
q2 -GRS cutoff ) 0.1
conventional CoMFA
q2 -GRS cutoff ) 0.1
compd
actual
calcd
resid
pred
resid
calcd
resid
pred
resid
calcd
resid
calcd
resid
1a
1b
1c
1d
2
3
4
5
6
9.48
7.87
8.54
7.15
9.46
7.76
7.25
8.55
8.77
8.26
6.22
6.33
7.26
7.98
7.52
8.32
8.04
9.02
7.15
6.78
6.97
7.08
7.58
8.29
8.35
8.85
9.12
7.23
5.32
5.32
9.6
9.06
7.93
8.50
7.38
8.96
8.02
6.53
8.86
8.81
8.28
6.43
6.73
7.39
7.73
7.39
8.3
8.56
8.46
6.93
6.95
7.07
7.45
7.53
8.61
8.98
9.01
8.86
6.93
5.44
5.28
9.68
6.05
8.16
7.24
5.8
0.42
-0.06
0.04
-0.23
0.5
9.56
8.44
9.01
7.41
8.33
8.42
6.44
8.61
8.93
8.32
6.61
6.81
7.19
8.12
7.64
8.27
8.29
8.60
7.03
7.10
7.13
6.76
7.37
8.30
8.77
8.99
8.58
7.13
5.10
5.17
8.98
5.71
8.45
7.42
5.93
5.87
5.65
6.47
8.52
6.88
8.19
9.48
9.71
5.92
5.30
9.85
9.12
8.05
6.90
6.58
-0.08
-0.57
-0.47
-0.26
1.13
9.12
7.70
8.67
7.11
8.44
8.37
6.57
8.52
8.97
8.14
6.67
6.85
6.98
7.74
7.39
8.22
8.52
8.17
7.23
7.02
7.05
7.24
7.51
8.45
8.35
8.46
8.34
7.16
5.83
5.53
9.46
5.87
8.09
7.60
5.75
5.85
5.42
6.52
8.37
6.66
8.23
9.75
9.95
6.01
5.53
10.16
9.3
8.32
6.87
6.71
5.35
5.26
9.78
7.46
5.79
6.52
5.72
9.34
5.40
6.72
9.41
9.67
9.31
8.08
7.48
6.9
0.36
0.17
-0.13
0.04
1.02
-0.61
0.68
0.03
-0.20
0.12
-0.45
-0.52
0.28
0.24
0.13
0.10
-0.48
0.85
-0.08
-0.24
-0.08
-0.16
0.07
-0.16
0.00
0.39
0.78
9.19
8.09
8.79
7.46
8.53
8.52
6.22
8.45
8.97
8.10
6.59
6.86
7.35
7.99
7.32
8.28
8.33
8.27
7.11
7.27
7.22
7.16
7.37
8.45
8.79
9.13
8.58
7.1
5.66
5.30
9.13
6.04
8.24
7.49
6.05
5.90
5.45
6.57
8.53
6.72
8.16
9.47
9.26
5.87
5.66
10.19
9.04
8.10
6.93
6.55
5.29
5.14
9.93
6.91
5.84
6.31
5.51
9.12
5.44
7.05
9.25
9.28
9.25
8.04
8.05
6.78
0.29
-0.22
-0.25
-0.31
0.93
-0.76
1.03
0.10
-0.26
0.72
-0.66
0.81
-0.31
-0.04
-0.02
-0.21
-0.4
-0.13
0.25
0.13
0.02
-0.52
0.56
-0.06
-0.16
-0.06
-0.39
-0.48
0.07
-0.14
-0.12
0.05
-0.25
0.42
0.12
-0.32
-0.16
0.32
-0.20
0.16
7
8
9
-0.37
-0.53
-0.09
-0.01
0.20
0.04
-0.29
0.75
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
0.22
0.04
-0.17
-0.1
-0.37
0.05
-0.32
-0.63
-0.16
0.26
0.3
-0.12
0.04
-0.08
-0.75
0.22
0.11
0.1
-0.49
-0.25
-0.08
0.21
-0.16
-0.44
-0.28
0.54
0.13
-0.34
0.02
0.47
-0.74
0.14
0.16
-0.15
0.22
-0.15
-0.37
-0.14
0.11
-0.48
0.04
0.66
0.06
-0.36
0.03
0.03
0.14
0.11
0.45
0.21
-0.01
-0.42
-0.14
0.54
0.10
0.22
0.15
0.62
-0.41
-0.07
0.23
0.07
-0.51
-0.21
0.14
-0.57
0.29
0.05
0.15
0.27
5.3
8.38
7.35
5.9
6.12
5.3
-0.03
0.25
5.6
0.52
5.62
6.60
7.93
6.82
7.72
9.75
10.04
6.01
5.46
10.07
8.87
8.21
6.69
6.43
-0.32
-0.4
0.46
-0.35
-0.27
-0.13
-0.05
-0.51
0.03
0.21
0.01
0.00
0.37
-0.32
-0.32
0.02
6.20
8.39
6.83
7.68
9.51
9.92
5.91
5.30
10.22
9.07
8.24
7.04
7.00
5.52
5.60
9.30
7.34
5.82
6.67
5.70
9.52
5.30
6.57
8.75
9.62
9.24
8.03
7.82
6.74
0.01
0.17
-0.04
-0.24
-0.12
-0.1
-0.16
0.15
0.2
0.03
0.35
0.57
-0.55
-0.24
-0.03
-0.10
-0.23
0.06
-0.23
-0.08
0.17
0.29
0.17
0.34
-0.48
-0.12
0.03
0.15
-0.02
0.18
-0.10
-0.15
-0.66
-0.05
-0.07
-0.05
0.34
-0.05
0.19
0.14
0.42
5.54
6.14
9.38
7.16
6.32
6.13
5.66
8.45
5.85
6.05
8.25
9.38
9.13
8.42
6.68
6.65
-0.02
-0.54
-0.08
0.18
-0.5
0.34
0.04
1.07
-0.55
0.52
0.50
0.24
0.11
-0.39
1.14
0.09
5.22
6.03
9.66
6.61
5.96
5.78
5.27
7.72
4.62
7.22
8.30
8.67
8.34
9.20
6.87
5.61
0.30
-0.43
-0.36
0.73
-0.14
0.89
0.43
1.80
0.68
-0.65
0.45
0.95
0.90
-1.17
0.95
1.13
0.23
0.46
-0.63
0.43
-0.02
0.36
0.19
0.40
-0.14
-0.48
-0.50
0.34
-0.01
-0.01
-0.23
-0.04
-0.16
constructed by modification of (-)-trans-H2-PAT. All cis-
PAT analogues were constructed by modification of (-)-
cis-H2-PAT.
All non-PAT analogues were constructed de novo
using the sketch option of the building component of
SYBYL. Conformational databases were developed for
all non-PAT ligands using the Random Search subrou-
tine of SYBYL. In this case, the energy cutoff for
conformers was designated as 10 kcal/mol above that
of the lowest-energy conformer found. All compounds

PATs as Ligands for Histamine H1 Receptors
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3047
Ch a r t 3. Binding Modes of Aromatic Substituents
beyond the van der Waals envelopes of all molecules by
at least 4.0 Å. The CoMFA QSAR equations were
calculated with the PLS algorithm as implemented in
Sybyl. The optimal number of components in the final
PLS model was determined by the standard error of
prediction value obtained from the leave-one-out cross-
validation technique.
q2-GRS Rou tin e.23 A conventional CoMFA is per-
formed initially using an automatically generated region
file. The rectangular region grid encompassing aligned
molecules is then broken into 125 small boxes of equal
size, and the Cartesian coordinates of the upper-right
lower-left corners of each box are calculated. The
calculated coordinates are used to create region files
with a +1 sp3 C as a probe atom. For each of the newly
created region files, a separate CoMFA is performed
with a step size of 1 Å. The regions with a q2 greater
than a specified threshold are selected for further
analysis and combined to generate a master region file.
After the master region is generated, a final CoMFA is
performed.
were modeled as the protonated cationic amines as these
analogues exist primarily in this form at pH 7.4 and it
is assumed that the cationic amine group acts as a
hydrogen bond donor for the PAT binding site. Centroids
were defined for each phenyl ring in order to obtain
alignment of PAT and non-PAT analogues.
Str u ctu r e Align m en t. (-)-trans-H2-PAT (1a ) was
used as a template onto which the PAT analogues, as
well as 26 and 27, were superimposed using SYBYL
field fit routine. Centroids were defined for all aro-
matic rings of non-PAT compounds. The centroids and
amine nitrogen of diphenhydramine (36), butaclamol
(37), doxepin (43), mazindol (46), SCH-23390 (47),
methysergide (48), SKF-8393 (49), amitriptyline (50),
desipramine (51), haloperidol (53), mianserin (55), rim-
cazole (56), spiperone (57), ketanserin (58), prometha-
zine (59), GBR-12909 (61), and nortriptyline (62) were
superimposed with the corresponding structures of the
template. Compounds 19, (-)-NANM (42), and 7-OH-
DPAT (52) were superimposed with the template using
the protonated nitrogen, the centroid of the aromatic
portion of the tetralin ring, and the C4 carbon with the
C1 carbon of the template molecule and then field fitted
to the template. Since (R)-chlorpheniramine (38) and
(S)-chlorpheniramine (39) as well as cis- and trans-
flupenthixol (44 and 45, respectively) have affinities
similar to compounds 2 (Cl,OH-PAT) and 22 (p-Cl-PAT),
respectively, 39 and 44 were aligned such that the
substituted phenyl ring centroid superimposed with the
centroid of the tetralin aromatic ring of the template.
Compounds 38 and 45 were aligned such that the
substituted phenyl ring centroid superimposed with the
centroid of the pendant phenyl ring of the template.
Triprolidine (40) and mepyramine (60) were superim-
posed such that the substituted phenyl ring is super-
imposed with the tetralin, since they have affinity
similar to that of compound 2 as compared to 22. The
centroid of the atropine (41) phenyl ring was aligned
with the centroid of the tetralin aromatic ring, the
hydroxyl oxygen was fitted with the pendant phenyl ring
centroid, and the two nitrogens were superimposed.
Loperamide (54) was superimposed on compound 16, a
PAT with a phenylethyl substituent on the amine, which
is structurally similar to 54. BMY-14802 (63) was
aligned such that the fluorophenyl ring centroid was
superimposed on the template tetralin aromatic ring,
the hydroxyl group with the pendant phenyl ring, and
the piperizine ring with the tertiary amine.
Resu lts a n d Discu ssion
The affinities of PAT analogues for both the [3H]-
(-)-trans-H2-PAT- and [3H]mepyramine-labeled sites
are shown in Table 1. We initially discuss the results
of biological experiments followed by the analysis of
molecular modeling studies.
Deter m in a tion of th e Requ ir em en t of th e P en -
d a n t P h en yl Rin g for tr a n s-H2-P AT Bin d in g a t th e
[3H]-(-)-tr a n s-H2-P AT Bin d in g Site. The majority of
clinically useful H1 antagonists contain a 1,1-diaryl-3-
aminopropne moiety where the diaryl group is hypoth-
esized to be a requirement for high affinity at the
[3H]mepyramine-labeled H1 site.24 In a study of tri-
prolidine analogues, the replacement of one aryl group
by hydrogen reduced antihistaminic activity.10 To test
the diaryl requirement at the [3H]-(-)-trans-H2-PAT
site, compound 19 was synthesized and competition
binding for both the [3H]-(-)-trans-H2-PAT- and [3H]-
mepyramine-labeled sites was evaluated (Table 2).
Consistent with previous H1 SAR studies that show the
importance of a 1,1-diaryl moiety, compound 19 exhibits
decreased affinity (∼100-fold) for both the [3H]-(-)-
trans-H2-PAT-and [3H]mepyramine-labeled sites, as
compared to (()-trans-H2-PAT (K0.5 ) 1.4 and 1.1 nM,
respectively). This indicates that the pendant phenyl
ring is required for high-affinity binding at these two
sites.
Effects of Su bstitu en ts on th e Ar om a tic Rin gs
on [3H]-(-)-tr a n s-H2-P AT Site Bin d in g. Diaryl H1
antagonists with a single ortho-substitution have been
shown to possess decreased affinity for the [3H]me-
pyramine-labeled sites. From a study of diphenhydra-
mine analogues, it was determined that small lipophilic
groups at the ortho-position reduced the antihistamin-
ergic effects.25 o-Methyl substitution on triprolidine’s
phenyl ring decreased affinity for the [3H]mepyramine-
labeled H1 receptor by more than 1000-fold.10 This
decrease is believed to be due to a conformational
change that disallows coplanarity of the diaryl ring
system. To assess the effects of ortho-subsitution in the
rigid (()-trans-H2-PAT, we synthesized o-chloro- and
o-methyl-PATs with groups on the pendant phenyl ring.
Com p a r a tive Molecu la r F ield An a lysis. Conven-
tional CoMFA was performed with the QSAR option of
Sybyl. For each cross-validated CoMFA analysis, the
minimum σ value was set to 2.0 to expedite calculations.
The steric and electrostatic field energies were calcu-
lated using sp3 carbon probe atoms with a +1 charge.
The CoMFA grid spacing was 2.0 Å in the x, y, and z
dimensions within the defined region, which extended

3048 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Bucholtz et al.
The (()-o-chloro-trans-H2-PAT (20) has approximately
50-fold decreased affinity at both radiolabeled sites
compared to (()-trans-H2-PAT (1), whereas the (()-o-
methyl-trans-H2-PAT (21) has less than a 10-fold de-
crease in affinity (Table 1). The resultant decrease in
affinity may be due to a compact hydrophobic pocket in
the active sites that could not accommodate the ortho-
substituent. The decrease in affinity of PATS with an
o-chloro- or o-methyl substituent on the pendant phenyl
ring also is consistent with the H1-like characteristics
of the [3H]-(-)-trans-H2-PAT site. The ortho-substituted
PATs also possess decreased affinity for the [3H]me-
pyramine-labeled H1 receptor (Table 1), further indicat-
ing the similarity of these two radiolabeled sites. In
comparison to the earlier reported effects of ortho-
substitution of triprolidine, o-methyl substitution on the
PAT nucleus only results in a 10-fold decrease at both
the [3H]-(-)-trans-H2-PAT-and [3H]mepyramine-labeled
sites, indicating the relative positions of the ring
systems in this rigid compound do not change signifi-
cantly by this substituent.
substituents on a phenyl ring can bind to the [3H]-(-)-
trans-H2-PAT-labeled site in a fashion similar to the
para-substituted PATs, but they bind with 10-fold
higher affinity. However, the high binding affinity of
chlorpheniramine and triprolidine correlates well with
the observed affinity of the trans- and cis-6,7-Cl,OH-
PAT analogues (2 and 28), which also have substituents
on an aromatic ring. This correlates well with antihis-
taminergic studies of a series of diarylaminopropenes
synthesized by Warringa et al.11 They found that para-
substitutions on the aromatic ring cis with respect to
the amine increased antihistaminergic activity, while
para-substitutions on the trans aromatic ring decreased
the activity. Because of our studies with the rigid PATs,
which lock the position of the amine in relation to the
aryl ring, we propose that chlorpheniramine and similar
molecules bind to both of the [3H]-(-)-trans-H2-PAT- and
[3H]mepyramine-labeled sites with their aryl ring sub-
stituents in the same molecular space as the tetralin
aryl ring of trans-Cl,OH-PAT (2) (Chart 3).
Effects of Oth er Com p ou n d s on Bin d in g a t [3H]-
(-)-tr a n s-H2-P AT- a n d [3H]Mep yr a m in e-La beled
Sites. It is generally recognized that quaternary amino
groups on ethylenediamine and aminoalkyl ethers de-
crease activity at the H1 receptor.26 Previous results15
indicated that the racemic trans-cationic quaternary
amine 3 possesses moderate (35 nM, Table 1) but
significantly less affinity for the [3H]-(-)-trans-H2-PAT
binding site as compared to (()-trans-1. The decrease
in affinity of 3 is believed to be due to inability to donate
a hydrogen bond for high-affinity binding. Compound
3 also has diminished affinity for the [3H]mepyramine-
labeled sites (120 nM, Table 1). This suggests that the
histamine H1 receptor cannot accommodate a perma-
nently charged PAT and that hydrogen bond donation
is crucial for PAT binding at this site also. To test
further if ability for hydrogen bond donation is a
possible requirement for high-affinity binding and to
test the effects of stereochemistry on permanently
charged PATs at these two sites, we synthesized com-
pound 25, the quaternary amine analogue of (()-cis-
H2-PAT. Like trans-cationic quaternary amine 3, com-
pound 25 exhibits a decrease in affinity for both the
[3H]-(-)-trans-H2-PAT- and [3H]mepyramine-labeled sites,
further suggesting the importance of hydrogen bond
donation.
Para-substitution on an aromatic ring of H1 antago-
nists such as mepyramine (60), (+)-chlorpheniramine
(39), and triprolidine (40) provides for high affinity for
the [3H]mepyramine-labeled H1 receptor (K0.5 ) 0.58,
0.31, and 0.12 nM, respectively). However, para-
substitution on the pendent phenyl ring of (()-trans-
H2-PAT results in decreased affinity at both the [3H]-
(-)-trans-H2-PAT- and [3H]mepyramine-labeled sites
(Table 1). The p-chloro-substituted PAT (22) has ap-
proximately 6-fold less affinity for the [3H]-(-)-trans-
H2-PAT-labeled site as compared to (()-trans-1 and a
similar decrease in affinity for the [3H]mepyramine-
labeled site. The (()-p-methyl-substituted PAT (23) also
exhibits a similar decrease in affinity at both radio-
labeled sites. However, p-fluoro substitution (24) does
not exhibit a decrease in affinity for these sites. Chlorine
and methyl substituents have increased van der Waals
radii as compared to hydrogen or fluorine; therefore the
decrease in affinity of 22 and 23 may be due to increased
steric bulk of their para-substituents.
Aromatic disubstituted H1 ligands have not been
analyzed previously at the [3H]mepyramine-labeled site.
Previously synthesized trans- and cis-6,7-Cl,OH-PAT
analogues (2 and 28) bind with high affinity (K0.5 ) 0.63
and 0.5 nM, respectively, Table 1) to the [3H]-(-)-trans-
H2-PAT site, whereas the cis- and trans-6,7-diOH-PAT
For maximal H1 antagonist activity, the aromatic
rings of diphenhydramine cannot be coplanar. The
fluorene analogue of diphenhydramine, which forces
both rings to be coplanar, is 100 times less active.27 In
potent phenothiazine H1 antagonists (promethazine,
59), the tricyclic ring system is not coplanar because
the thiazine ring is in the boat conformation. To test
the effects of a coplanar ring system on binding at the
[3H]-(-)-trans-H2-PAT- and [3H]mepyramine-labeled sites,
we analyzed 3-aminofluoranthene (26) and its dimeth-
ylated analogue (27). These compounds have decreased
and identical affinity at the two sites. The decrease in
affinity can be due to coplanarity of the ring systems,
consistent with previous literature, or to the change of
the relative position of the amine as compared to (()-1.
analogues (12 and 13) have decreased affinity (K0.5
)
9.6 and, 60 nM, respectively, Table 1). We now have
analyzed competition binding studies for these com-
pounds to assess the substituent effects on the [3H]-
mepyramine-labeled site. As indicated in Table 1, these
compounds have nearly identical affinity for both radio-
labeled sites. These results suggest that a 6-OH sub-
stituent, capable of forming hydrogen bonds, is detri-
mental to binding, while a 6-Cl group that is hydropho-
bic may be beneficial for binding at either radiolabeled
site.
The novel rigid aromatic ring-substituted PAT com-
pounds have provided important information regarding
the way the [3H]-(-)-trans-H2-PAT- and [3H]mepyramine-
labeled sites accommodate aromatic ring substituents.
All the proposed analogues in this study possess
nearly equal affinity for the [3H]-(-)-trans-H2-PAT- and
[3H]mepyramine-labeled sites; therefore, we analyzed
(S)-Chlorpheniramine (39) and triprolidine (40) (K0.5
0.31 and 0.12 nM, respectively), which have para-
)

PATs as Ligands for Histamine H1 Receptors
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3049
F igu r e 1. Actual versus calculated pK0.5 for the training set
(48-63) using conventional CoMFA (cf. Table 5).
F igu r e 3. Actual versus predicted pK0.5 of test set molecules
(1-47): 9 and solid line, conventional CoMFA; 4 and dashed
line, q2-GRS; R2 (conventional CoMFA) ) 0.8943, R2 (q2-GRS)
) 0.7712.
F igu r e 2. Actual versus calculated pK0.5 for all compounds
using q2-GRS (cf. Table 5).
a number of previously synthesized PAT analogues and
non-PAT compounds at the [3H]mepyramine-labeled
sites to identify possible differences in the SAR of the
radiolabeled sites. As indicated in Tables 2 and 3, none
of the tested PAT analogues nor non-PAT ligands are
able to distinguish strongly (difference in K0.5 between
[3H]-(-)-trans-H2-PAT- and [3H]mepyramine-labeled sites
is less than 5-fold) between these two radiolabeled sites.
Also, (-)-trans-H2-PAT (1a ) and mepyramine have the
same affinity at each radiolabeled site (Table 2). This
leads us to believe that both radioligands label hista-
mine H1 receptors with the same SAR in rodent brain.
Therefore, we believe that a CoMFA model of the [3H]-
(-)-trans-H2-PAT site is an important representation
of binding to histamine H1-type receptors in mammalian
brain.
F igu r e 4. Actual versus calculated pK0.5 for all compounds
using conventional CoMFA (cf. Table 5).
F igu r e 5. Actual versus calculated pK0.5 for all compounds
using q2-GRS (cf. Table 5).
CoMF A a n d CoMF A/q2-GRS of Liga n d s In clu d ed
in th e Tr a in in g Set. The results obtained after per-
forming CoMFA/q2-GRS are summarized in Table 3. The
training set (compounds 1-47) was assessed initially
by conventional CoMFA. The q2-GRS routine23 was then
applied to optimize the standard CoMFA model. Initial
conventional CoMFA produced a q2 of 0.460 and stand-
ard error of prediction (SDEP) of 1.031 with 3 compo-
nents. (Table 3) Various q2 threshold values were tried
to remove irrelevant variables (noise). The highest q2
value (0.513, 6 components) and lowest SDEP value
(0.982) were obtained with the q2 threshold value of 0.1.
Since the q2 values associated with conventional CoMFA
and q2-GRS were similar, we used both methods for
subsequent non-cross-validated PLS runs. This yielded
a conventional R2 of 0.941 and a standard error of
estimate (SEE) of 0.337 for the conventional CoMFA
and a conventional R2 of 0.928 and a SEE of 0.372 for
q2-GRS (Table 4). Using these PLS models, the activity
of each compound was calculated and compared with
the actual value (Figures 1and 2). It was found that
overall the residuals were low. The compounds with the
worst residuals for conventional CoMFA were 29 and 4
(-0.75 and 0.72, respectively) and 2 and 4 (1.13 and
0.81, respectively) for q2-GRS. The poor prediction of 4
by both models is surprising considering that other
secondary amine PATs (30-33) were predicted with low
residuals by both models. The high residual (underpre-
diction of affinity) for 2 obtained from the q2-GRS model
is most likely due to an overwhelming amount of
compounds in the model with similar substitution
patterns that have low affinity. For example, 12, 13,
29, 34, and 47 all have an equivalent 7-hydroxyl group
but all have greater than 100-fold less affinity for the
binding site.
CoMF A/q2-GRS of Liga n d s In clu d ed in th e Test
Set. The PLS models obtained for the conventional and
q2-GRS were used to predict the activities of 16 com-
pounds included in the test set. The predicted activities
for compounds 48-63 are shown in Table 5, and the
plot of actual versus predicted activities for both predic-
tion methods are shown in Figure 3. Predictive R2 of
0.8943 for conventional CoMFA and 0.7712 for q2-GRS
were obtained for the test set. This result is somewhat

3050 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Bucholtz et al.
F igu r e 6. CoMFA steric stdev*coeff contour plot. Green regions represent a contribution level of 80%, i.e., sterically favored
areas. Yellow regions represent a contribution level of 20%, i.e., sterically disfavored areas. Compounds 2 and 47 are depicted in
white and magenta, respectively.
F igu r e 7. CoMFA electrostatic stdev*coeff contour plot. Blue regions represent a contribution level of 80%, i.e., positive charge
favored areas. Red regions represent a contribution level of 20%, i.e., positive charge disfavored areas. Compounds 2 and 43 are
depicted in white and magenta, respectively.
surprising because the q2 for the training set obtained
by q2-GRS is higher than that for conventional CoMFA
(Table 3). Thus, q2-GRS results in a higher q2 while
exhibiting decreased external predictability as evi-
denced by the test set. This could be due to the fact that
q2-GRS as a variable selection procedure may have a
tendency to overfitting as was noticed recently by Ortiz
et al.,28 which results in better internal q2 but lower
external R2 than those obtained with conventional
CoMFA

PATs as Ligands for Histamine H1 Receptors
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3051
Due to the better predictability of conventional
CoMFA model, we hypothesized that it may represent
steric and electrostatic fields important to binding at
the [3H]-trans-H2-PAT site. Since this model predicted
the test set well (compounds 48-63), we added these
compounds to the training set to obtain a total of 66
compounds. The results of conventional CoMFA and q2-
GRS with different cutoffs are shown in Table 3. Initial
conventional CoMFA (sp3 carbon with a +1 charge) gave
a q2 of 0.681 and SDEP of 0.816 at 6 components, while
q2-GRS with 0.1 as a cutoff gave a q2 of 0.635 and SDEP
of 0.874 at 6 components. Since both models resulted
in a high q2, we used these analyses to obtain non-cross-
validated CoMFA results as indicated in Table 4 (con-
ventional CoMFA, R2 ) 0.941, SEE ) 0.347, F ) 192.8;
q2-GRS, R2 ) 0.927, SEE ) 0.388, F ) 151.85). The
actual, calculated, and residual activities for both
CoMFA methods are shown in Table 5. The plot of
actual versus calculated activities is shown in Figures
4and 5.
binding at the known H1 receptor. Therefore, this model
can be used also to predict high-affinity ligands for the
[3H]mepyramine-labeled H1 binding site.
Exp er im en ta l Section
All chemicals were used as received from the manufacturers.
Melting points were determined on a Mel-temp apparatus and
are uncorrected. Proton NMR spectra were obtained in CDCl3
(unless noted otherwise) using a Bruker AC300 300-MHz
spectrometer. Elemental compositions (C, H, and N) of test
compounds were determined by MHW Laboratories (Phoenix,
AZ) and agreed with calculated values (0.4%. Gas chromato-
graphic analysis was performed using a Shimadzu GC-8A
chromatograph with 2.0-m column packed with 3% OV-17 on
chromasorb. Thin-layer chromatography was performed using
silica gel 60 coated glass plates (Fisher Scientific), and column
chromatography was performed with silica gel 60 (70-230
mesh). The starting compound for the synthesis of the 1-phen-
yl-3-amino-1,2,3,4-tetrahydronaphthalenes consisted of 1-phenyl-
2-propanol which was obtained from Aldrich Chemical Co.
Oxidation of 1-phenyl-2-propanol with pyridinium chlorochro-
mate gave phenylacetone in approximately 70% yield.
2-Hyd r oxy-1,2,3,4-tetr a h yd r on a p h th a len e (65). To a
stirred slurry of NaBH4 (2.6 g, 68.5 mmol) in 50 mL of MeOH
was added dropwise a solution of â-tetralone (64) (5.0 g, 34.2
mmol) in MeOH. The reaction was refluxed for 4 h, and
volatiles were evaporated in vacuo. H2O (100 mL) was added
and extracted with CH2Cl2, which was evaporated in vacuo.
Column chromatography (CH2Cl2) afforded 2.82 g (56%) of a
bright red oil: 1H NMR (CDCl3) δ 1.8 (m, 1H, CH), 2.1 (m,
1H, CH), 2.7 (d of d, 1H, CH), 3.0 (m, 2H, CH2), 3.1 (d of d,
1H, CH), 4.3 (m, 1H, CHOH), 7.15 (s, 4H, ArH4).
2-Azid o-1,2,3,4-tetr a h yd r on a p h th a len e (66). The 2-hy-
droxy-1,2,3,4-tetrahydronaphthalene (65) (2.82 g, 19.1 mmol)
was dissolved in dry pyridine, and p-toluenesulfonyl chloride
(7.2 g, 38.1 mmol) in pyridine was added. The solution was
kept at 4 °C for 2 weeks. The solution was poured into ice
water, filtered, and yielded a gray white precipitate which was
dried in vacuo to afford 5.9 g (>100% of product). To a solution
of 2-tosyloxy-1,2,3,4-tetrahydronaphthalene (5.9 g, 19.6 mmol)
in 10 mL of dimethylformamide was added dropwise a solution
of sodium azide (2.6 g, 40 mmol) in 10 mL of water. The
reaction was heated to 45-50 °C and stirred for 4 h. The
solution was poured into ice water and extracted with CH2Cl2.
The organic layer was evaporated in vacuo to afford 3.2 g (94%)
of a brown-gray oil.
CoMF A F ield s. The CoMFA steric and electrostatic
fields obtained from the conventional CoMFA and the
structures of sample ligands are shown in Figures 6 and
7, respectively. The field values were calculated by
multiplying the â coefficient and standard deviation of
columns in the QSAR table (stdev*coeff). The green
(sterically favorable) and yellow (sterically unfavorable)
contours shown in Figure 6 represent 80% and 20% level
contributions, respectively. The green contour region
adjacent to the amine of 2 indicates that a dimethyl
substitution pattern is optimal for binding at this
radiolabeled site. The yellow contour regions at the 2
and 4 carbons indicate that steric bulk is not well-
accommodated by the site on these areas of the PAT
nucleus. The 8-hydroxy group on 47 extends into a
yellow contour region that signifies that steric bulk is
not well-accommodated by the site. Overall, the steric
contour regions well-represent binding of molecules at
the [3H]-(-)-trans-H2-PAT binding site.
The contours generated using the electrostatic field
are displayed in Figure 7. The blue (positive charge
favorable) and red (positive charge disfavorable) con-
tours represent 80% and 20% level contributions, re-
spectively. The red contour regions indicate that positive
charge of amino groups on the C2 or C4 position of the
PAT nucleus is not well-accommodated by the site. Since
many compounds with hydroxyl substituents bind with
low affinity (due to either hydrogen bond donor or
acceptor properties) we would have expected a red
contour adjacent to the 7-hydroxy group of 2. With the
poor prediction of 2, as described for both the training
set and final models, we should now design ligands that
more effectively probe this area of the pharmacophore.
2-Am in o-1,2,3,4-tetr a h yd r on a p h th a len e (67). The previ-
ous azide (66) (3.2 g,18.5 mmol) was dissolved in 200 mL of
2-propanol and shaken on a Parr hydrogenation apparatus
over 150 mg of 5% Pd/C at 50 psi overnight. The catalyst was
filtered and the filtrate evaporated in vacuo to yield 0.9 g of a
green oil. Column chromatography with CH2Cl2:methanol (9:
1) afforded 140 mg (0.95 mmol, 5.1%) of a yellow-brown oil:
1H NMR (CDCl3) δ 1.7 (m, 1H, CH), δ 2.1 (m, 1H, CH), 2.6 (d
of d, 1 H, CH), 2.9 (m, 2H, CH2), 3.1 (d of d, 1H, CH), 3.2 (m,
1H, CHN), 7.1 (s, 4H, ArH4).
2-(N,N -Dim et h yla m in o)-1,2,3,4-t et r a h yd r on a p h t h a -
len e Hyd r och lor id e Sa lt (19). The primary amine (67) (137
mg, 0.93 mmol) was dissolved in 2 mL of 37% formaldehyde
and 3 mL of 96% formic acid and refluxed for 4 h. The volatiles
were removed in vacuo, and the salt was dissolved in water.
This was extracted with methylene chloride, and the organic
layer separated and evaporated to yield 139 mg of a brown
oil. Column chromatography with CH2Cl2:methanol (9:1) af-
forded 77 mg (50%) of a light brown oil. This was dissolved in
ethereal HCl, and the salt recrystallized from ethanol to afford
This model represents improvements on our previ-
ously published model in a number of ways: an in-
creased number of compounds was used for the CoMFA,
to 66 up from 35; a greater majority of rigid molecules
have been used to define a training set for analysis; and
a test set was used to validate our initial model of 50
compounds. As discussed above, 33 of these compounds
have been tested at the [3H]mepyramine-labeled H1
binding site. Since these compounds have nearly identi-
cal affinity at the two radiolabeled sites, and we believe
that the two sites are identical, this model represents
1
15 mg (16%) of colorless crystals: mp 211-212 °C; H NMR
(CDCl3) δ 1.7 (m, 1H, CH), δ 2.1 (m, 1H, CH), 2.4 (s, 6H,
N(CH3)2) 2.6 (d of d, 1 H, CH), 2.9 (m, 2H, CH2), 3.1 (d of d,
1H, CH), 3.2 (m, 1H, CHN), 7.1 (s, 4H, ArH4). Anal. (C12H18
NCl) C,H.
-
1-(2-Ch lor op h en yl)-4-p h en yl-1-bu ten -3-on e (68). Gen -
er a l P r oced u r e. This intermediate was prepared from 2-chlo-
robenzaldehyde and phenylacetone by the method of South-

3052 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Bucholtz et al.
wick et al.29 in 72% yield as a crude brown oil (21 g) estimated
as 60% pure by gas chromatography, and used unpurified in
next synthetic step.
1-(2-Meth ylp h en yl)-4-p h en yl-1-bu ten -3-on e (69). A so-
lution of 2-CH3-bendzaldehyde and phenylacetone was reacted
as above to afford 19 g of a viscous oil in 79% yield: 1H NMR
(CDCl3) δ 7.2-7.85 (m, 9H, ArH), 6.7 (d, 1H, styryl), 3.7 (d,
1H, styryl), 4.0 (s, 2H, CH2), 2.4 (s, 3H, CH3).
1-(4-Ch lor op h en yl)-4-p h en yl-1-b u t en -3-on e (70). This
intermediate was prepared as above in 31% yield as a white
solid from phenyl acetone and 4-chlorobenzaldehyde: mp 132-
138 °C; 1H NMR (CDCl3) δ 7.2-7.85 (m, 9H, ArH), 7.0 (d, 1H,
styryl), 6.6 (d, 1H, styryl), 4.2 (s, 2H, CH2).
1-(4-Meth ylp h en yl)-4-p h en yl-1-bu ten -3-on e (71). This
intermediate was prepared as above in 6.9% (1.7 g) yield as
white solid estimated as 90% pure and used unpurified in the
next synthetic step: 1H NMR (CDCl3) δ 7.6 (d, 1H, ArH), 7.5
(m, 2H, ArH), 7.25-7.4 (m, 5H, ArH), 7.05-7.15 (t, 2H, styryl,
ArH), 6.7 (d, 1H, styryl), 3.95 (s, 2H, CH2), 2.4 (s, 1H, PhCH3).
1-(4-F lu or op h en yl)-4-p h en yl-1-bu ten -3-on e (72). This
intermediate was prepared as above in 35% (6.8 g) yield as
yellow crystals: mp 78-79 °C; 1H NMR (CDCl3) δ 7.6 (d, 1H,
ArH), 7.5 (m, 2H, ArH), 7.25-7.4 (m, 5H, ArH), 7.05-7.15 (t,
2H, styryl, ArH), 6.7 (d, 1H, styryl), 3.95 (s, 2H, CH2).
1-(2-Ch lor op h en yl)-3-oxo-1,2,3,4-t et r a h yd r on a p h t h a -
len e (73). Gen er a l P r oced u r e. Compound 68 (21 g, impure)
was dissolved in 700 mL of xylenes and added to a mechani-
cally stirred suspension of polyphosphoric acid in 240 mL of
xylenes. The reaction was heated to reflux for 7 h and
monitored by gas chromatography. The xylenes were decanted,
and the product was evaporated in vacuo to afford a crude red
oil (18.1 g) that was immediately carried through to the next
step without purification.
as above to afford 15 g of a crude orange oil. Column
chromatography over silica gel (CH2Cl2) afforded 3.0 g (28%)
of a pale yellow solid: mp: 97-105 °C; 1H NMR (CDCl3) δ
7.3-7.4 (m, 4H, ArH), 7.2-7.3 (m, 4H, ArH), 4.1 (dd, 1H,
PhCHPh), 3.2 (m, 1H, CHOH), 2.9 (m, 2H, PhCH2),2.4 (m, 2H,
CH2).
(()-cis-1-(4-Met h ylp h en yl)-3-h yd r oxy-1,2,3,4-t et r a h y-
d r on a p h th a len e (81). Tetralone 76 (7.2 mmol) was reduced
as above to afford 1.7 g of a crude orange oil. Column
chromatography over silica gel (CH2Cl2) afforded 610 mg
(35.6%) of a yellow-orange gum: 1H NMR (CDCl3) δ 6.9-7.2
(m, 8H, ArH), 4.15 (dd, 1H, PhCHPh), 3.2 (m, 1H, CHOH),
2.9 (m, 2H, PhCH2), 2.4 (m, 2H, CH2), 2.35 (s, 3H, Ph-CH3).
(()-cis-1-(4-F lu or op h en yl)-3-h yd r oxy-1,2,3,4-t et r a h y-
d r on a p h th a len e (82). The crude tetralone 77 (6.8 g, 28
mmol) was reduced as above to afford 4.0 g of a crude orange
oil. Column chromatography over silica gel (CH2Cl2) afforded
1
2.1 g (31%) of a yellow wax: mp 83-86 °C; H NMR (CDCl3)
δ 7.15-7.2 (m, 4H, ArH), 6.9-7.1 (m, 4H, ArH), 4.1 (dd, 1H,
PhCHPh), 3.2 (m, 1H, CHOH), 2.9 (m, 2H, PhCH2), 2.4 (m,
2H, CH2).
(()-tr a n s-1-(2-Ch lor op h en yl)-3-a zid o-1,2,3,4-t et r a h y-
d r on a p h th a len e (83). Gen er a l P r oced u r e. A solution of 7.2
g (38 mmol) of p-toluenesulfonyl chloride in 69 mL of dry
pyridine was added to a solution of 4.45 g of alcohol 78 (17.2
mmol) in 69 mL of dry pyridine. The reaction was allowed to
stand for 10 days at 4 °C and was then poured into ice water
with stirring. The resulting light purple gum was filtered,
dissolved in Et2O, and extracted with 1 N HCl. The solvent
was dried (Na2SO4 ) and evaporated in vacuo to afford 5.0 g
(12.1 mmol, 70%) of the tosylate. A solution of sodium azide
(2.0 g, 30.3 mmol) in 4.4 mL of water was added to the crude
tosylate in DMF (33 mL). The reaction was stirred at 50 °C
for 4.5 h. TLC (toluene) indicated complete reaction. The
solution was poured into ice water and extracted with Et2O.
The organic layer was evaporated in vacuo to near dryness,
and the solution was used immediately in the next synthetic
step.
1-(2-Met h ylp h en yl)-3-oxo-1,2,3,4-t et r a h yd r on a p h t h a -
len e (74). Compound 69 was prepared as above to afford 20 g
of a dark amber oil. Chromatography of this oil on 400 g of
silica gel (70-230 mesh) with toluene afforded 5.7 g (32%) of
a viscous oil.
(()-tr a n s-1-(2-Met h ylp h en yl)-3-a zid o-1,2,3,4-t et r a h y-
d r on a p h th a len e (84). The alcohol 79 (5.0 g, 0.022 mol) was
converted to the azide as above to afford 4.0 g (83%) of a gum
that was used immediately in the next synthetic step.
(()-tr a n s-1-(4-Ch lor op h en yl)-3-a zid o-1,2,3,4-t et r a h y-
d r on a p h th a len e (85). The alcohol 80 (3.0 g,11.6 mmol) was
converted to the azide as above. The organic layer was
evaporated in vacuo to near dryness, and the solution was used
immediately in the next synthetic step.
(()-tr a n s-1-(4-Met h ylp h en yl)-3-a zid o-1,2,3,4-t et r a h y-
d r on a p h th a len e (86). The alcohol 81 (610 mg, 2.52 mmol)
was converted to the azide as above. The organic layer was
evaporated in vacuo to near dryness, and the solution was used
immediately in the next synthetic step without purification.
(()-tr a n s-1-(4-F lu or op h en yl)-3-a zid o-1,2,3,4-t et r a h y-
d r on a p h th a len e (87). Alcohol 82 (2.1 g, 11.6 mmol) was
converted to the azide as above and was used immediately in
the next synthetic step.
(()-tr a n s-1-(2-Ch lor op h en yl)-3-a m in o-1,2,3,4-tetr a h y-
d r on a p h th a len e (88). Gen er a l P r oced u r e. The azide 83
was dissolved in 200 mL of 2-propanol and shaken on a Parr
hydrogenation apparatus over 100 mg of 10% Pd/C at 50 psi
for 2 days. The catalyst was filtered, and the filtrate evapo-
rated in vacuo to yield 3.2 g of a brown oil. Column chroma-
tography with CH2Cl2:MeOH (90:10) afforded 700 mg (2.7
mmol, 22% from tosylate) of a light brown oil: 1H NMR
(CDCl3) δ 7.4 (dd, 1H, ArH), 7.1-7.2 (m, 5H, ArH), 6.9 (d, 1H,
ArH), 6.6 (dd, 1H, ArH), 4.7 (t, 1H, PhCHPh), 3.2 (m, 1H,
PhCH), 3.1 (dd, 1H, PhCH), 2.6 (m, 1H, CH2), 2.0 (m, 1H, CH2),
1.6 (s, 2H, NH2).
(()-tr a n s-1-(2-Meth ylp h en yl)-3-a m in o-1,2,3,4-tetr a h y-
d r on a p h th a len e (89). A solution of the azide 84 (4.0 g, 0.015
mol) in 150 mL of 2-propanol was hydrogenated as above to
afford a gum which was purified by column chromatography
(CH2Cl2: CH2Cl2-methanol (9:1)) to afford 2.3 g (66%) of a
semisolid.
1-(4-Ch lor op h en yl)-3-oxo-1,2,3,4-t et r a h yd r on a p h t h a -
len e (75). Compound 70 (10.8 g, 42 mmol) was cyclized as
above to afford a crude brown oil (15.5 g, 150%) that was
immediately carried through to the next step without purifica-
tion.
1-(4-Met h ylp h en yl)-3-oxo-1,2,3,4-t et r a h yd r on a p h t h a -
len e (76). Compound 71 (1.7 g 7.2 mmol) was ring-closed as
above to afford a yellow oil (2.0 g, 118%) that was immediately
carried through to the next step without purification.
1-(4-F lu or op h en yl)-3-oxo-1,2,3,4-t et r a h yd r on a p h t h a -
len e (77). Compound 72 (6.8 g, 28 mmol) was ring-closed as
above to afford 6.8 g (100%) of an orange oil: 1H NMR (CDCl3)
δ 7.0-7.4 (m, 9H, ArH), 4.1 (t, 1H, PhCHPh), 2.9 (m, 2H,
PhCH2), 2.3 (m, 2H, CH2).
(()-cis-1-(2-Ch lor op h en yl)-3-h yd r oxy-1,2,3,4-t et r a h y-
d r on a p h th a len e (78). Gen er a l P r oced u r e. The crude tet-
ralone 73 (70 mmol) was dissolved in MeOH, and NaBH4 (247
mmol) was added in portions. The reaction was stirred
magnetically and refluxed overnight. Cautiously, 100 mL of
H2O was added to the reaction. The volatiles were evaporated
in vacuo. The residue was dissolved in CH2Cl2, extracted with
water, dried over sodium sulfate, and evaporated in vacuo to
afford 18 g of a crude orange oil. Column chromatography over
silica gel (CH2Cl2) afforded 4.5 g (25%) of a yellow-brown
gum: 1H NMR (CDCl3) δ 7.3-7.4 (m, 4H, ArH), 7.2-7.3 (m,
4H, ArH), 4.8 (dd, 1H, PhCHPh), 3.2 (dd, H, CHOH), 2.9 (m,
1H, PhCH2), 2.9 (m, 1H, PhCH2), 2.5 (m, 1H, CH2), 2.0 (m,
1H, CH2).
(()-cis-1-(2-Met h ylp h en yl)-3-h yd r oxy-1,2,3,4-t et r a h y-
d r on a p h th a len e (79). Tetralone 74 (5.7 g (0.02 mol) was
reduced as above to afford 5.0 g (88%) of semisolid.1H NMR
(CDCl3) δ 7.0-7.3 (m, 8H, ArH), 4.4 (dd, 1H, PhCHPh), 4.3
(m, 1H, CHOH), 3.2 (m, 2H, PhCH2), 2.9 (m, 2H, CH2), 2.4 (s,
3H, CH3).
(()-cis-1-(4-Ch lor op h en yl)-3-h yd r oxy-1,2,3,4-t et r a h y-
d r on a p h th a len e (80). Tetralone (75) (42 mmol) was reduced

PATs as Ligands for Histamine H1 Receptors
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3053
(()-tr a n s-1-(4-Ch lor op h en yl)-3-a m in o-1,2,3,4-tetr a h y-
d r on a p h th a len e (90). The azide 85 was dissolved in 200 mL
of 2-propanol and hydrogenated as above to afford 1.7 g of a
crude green oil. Column chromatography with CH2Cl2:MeOH
(95:5) afforded 1.3 g (5 mmol, 75% from tosylate) of a pale
yellow oil.
2.9 (dd, 1H, PhCH), 2.6 (m, 1H, CHN), 2.4 (s, 6H, N(CH3)2),
2.3 (s, 3H, PhCH3) 2.1 (t, 1H, CHCN). Anal. (C19H24NCl)
C,H,N.
(()-tr a n s-1-(4-F lu or op h en yl)-3-(N,N-d im eth yla m in o)-
1,2,3,4-tetr a h yd r on a p h th a len e (24). The primary amine 92
(345 mg, 1.28 mmol) was dimethylated as above to afford 150
mg (39%) of a pale yellow oil which was dissolved in ethereal
HCl and the salt recrystallized from ethanol/Et2O to afford
37 mg (0.12 mmol, 9.4%) of white crystals: mp: 210-212 °C;
1H NMR (CDCl3) δ 7.1-7.25 (m, 4H, ArH), 6.9 (m, 4H, ArH),
4.4 (t, 1H, PhCHPh), 3.15 (dd, 1H, PhCH), 3.0 (dd, 1H, PhCH),
2.9 (m, 1H, CHCN), 2.8 (m,1H, CHCN), 2.45 (s, 6H, N(CH3)2),
2.1 (m, 1H, CHN). Anal. (C18H21NClF) C,H,N.
(()-cis-1-P h en yl-3-(tr im eth yla m m on iu m yl)-1,2,3,4-tet-
r a h yd r on a p h th a len e Iod id e (25). Methyl iodide (1.5 mmol)
in toluene was added to a solution of (()-cis-1-phenyl-3-amino-
1,2,3,4-tetrahydronaphthalene (75 mg, 0.299 mmol) in 20 mL
of anhydrous ether, and the reaction mixture was allowed to
stand at room temperature overnight. The methiodide salt
precipitated and was filtered and dried to afford the crude
product. Recrystallization from EtOH afforded 76 mg (65%)
of pure product as yellow crystals: mp 260-262 °C; 1H NMR
(CDCl3) δ 7.0-7.4 (m, 8H, ArH), 6.7 (d, 1H, ArH), 4.4 (dd, 1H,
PhCHPh), 4.0 (m, 1H, CHN), 3.5 (t, 2H, PhCH), 3.2 (s, 9H,
(()-tr a n s-1-(4-Meth ylp h en yl)-3-a m in o-1,2,3,4-tetr a h y-
d r on a p h th a len e (91). The azide 86 was dissolved in 100 mL
of 2-propanol and hydrogenated as above to yield 250 mg of a
green oil. Column chromatography with CH2Cl2:MeOH (90:
10) afforded 184 mg (0.78 mmol, 34% from tosylate) of a pale
green oil: 1H NMR (CDCl3) δ 7.0-7.2 (m, 5H, ArH), 6.95 (d,
1H, ArH), 6.85 (dd, 1H, ArH), 4.3 (t, 1H, PhCHPh), 4.2 (dd,
1H, PhCH), 3.2 (m, 1H, PhCH), 2.8 (m, 1H, CHN), 2.3 (s, 3H,
PhCH3), 2.1 (s, 2H, NH2), 2.1 (t, 1H, CHCN).
(()-tr a n s-1-(4-F lu or op h en yl)-3-a m in o-1,2,3,4-tetr a h y-
d r on a p h th a len e (92). The azide 87 was dissolved in 100 mL
of 2-propanol and hydrogenated as above to yield 660 mg of a
crude orange oil. Column chromatography with CH2Cl2:MeOH
(95:5) afforded 356 mg (1.48 mmol, 22% from tosylate) of a
pale yellow oil: 1H NMR (CDCl3) δ 7.05-7.2 (m, 4H, ArH),
6.85-7.0 (m, 4H, ArH), 4.4 (t, 1H, PhCHPh), 3.3 (m, 1H,
PhCH), 3.2 (dd, 1H, PhCH), 2.9 (m, 1H, CH2), 2.75 (m, 1H,
CH2), 2.1 (s, 2H, NH2).
N(CH3)3), 2.7 (dd, 1H, CH2), 2.1 (q, 1H, CH2). Anal. (C19H21
NI) C,H,N.
-
(()-tr a n s-1-(2-Ch lor op h en yl)-3-(N,N-d im eth yla m in o)-
1,2,3,4-tetr a h yd r on a p h th a len e (20). Gen er a l P r oced u r e.
The primary amine 88 (0.7 mg, 2.72 mmol) was dissolved in 7
mL of 37% formaldehyde and 10.5 mL of 96% formic acid and
refluxed for 4 h. The volatiles were removed in vacuo, and the
salt was partitioned between saturated NaHCO3 and CH2Cl2
and the organic layer separated and evaporated in vacuo to
afford 0.8 g (97%) of a brown oil which was dissolved in
ethereal HCl and the salt recrystallized from ethanol to afford
100 mg (0.31 mmol) of white crystals: mp 209-210 °C; 1H
NMR (CDCl3) δ 7.4 (dd, 1H, ArH), 7.0-7.2 (m, 5H, ArH), 6.9
(d, 1H, ArH), 6.65 (dd, 1H, ArH), 4.4 (t, 1H, PhCHPh), 3.1 (dd,
1H, PhCH), 2.9 (dd, 1H, PhCH), 2.6 (m, 1H, CHN), 2.3 (s, 6H,
N(CH3)2), 2.1 (t, 1H, CHCN). Anal. (C18H21NCl2) C,H,N.
3-(Dim eth ylam in o)flu or an th en e (27). The primary amine
3-aminofluoranthene (26) (250 mg) was dissolved in 3 mL of
98% formic acid to which was added 2 mL of 37% of aqueous
formaldehyde. The slurry was stirred at reflux 12 h. The
volatiles were removed in vacuo, and the residue was parti-
tioned between CH2Cl2 and saturated NaHCO3. The organic
phase was dried (Na2SO4) and evaporated in vacuo to afford a
bright yellow powder. Recrystallization in MeOH/CHCl3 yielded
65 mg of a yellow powder. Anal. (C18H15N) C,H,N.
Ra d ior ecep tor Bin d in g Assa ys. Ch em ica ls: [3H]Me-
pyramine (23 Ci/mmol) was obtained from DuPont-NEN Corp.
(Boston, MA). Synthesis and resolution of (()-trans-H2-PAT
were as previously reported (Wyrick et al., 1993). (-)-trans-
H2-PAT subsequently was radiolabeled as previously reported
(Wyrick et al., 1994) with tritiated methyl iodide to yield [N-3H-
methyl]-(1R,3S)-(-)-trans-1-phenyl-3-(dimethylamino)-1,2,3,4-
tetrahydronaphthalene ([3H]-(-)-trans-H2-PAT) at 85 Ci/mmol.
Other biochemical reagents were purchased or donated from
Research Biochemicals Int. (RBI, Natick, MA).
Tissu e p r ep a r a tion : Frozen guinea pig brain, minus
cerebellum, was thawed and homogenized in 10 mL/g of tissue
of cold buffer solution containing 10 mM TRIS and 0.32 M
sucrose. The homogenate then was centrifuged at 1000g for
15 min at 4 °C. The supernatant was separated and recentri-
fuged at 31000g for an additional 15 min (4 °C). The P2 pellet
was resuspended in 10 mM TRIS buffer (pH 7.4, 25 °C) at 3
mL/g of original wet weight tissue and incubated for 15 min
at 25 °C. This suspension was centrifuged at 31000g for 15
min (4 °C) with the resulting pellet resuspended with 10 mM
TRIS buffer (pH 7.4, 25 °C) at a final volume of 1.5 mL/g of
original wet weight tissue with gentle vortexing. The final
tissue concentration was ca. 2.5-5 mg of protein/mL. Tissue
was stored at -80 °C until use.
Com p etition bin d in g a ssa ys: H2-PAT isomers were in-
cubated in triplicate borosilicate glass tubes (1 h, 30 °C) with
0.1 nM [3H]-trans-H2-PAT (10 mM TRIS, pH 7.4) or 2 nM [3H]-
mepyramine (50 mM Na+K+-phosphate, pH 7.4) protein from
above preparation in a total assay volume of 1.0 mL (0.5 mL
for DTG). Nonspecific binding was determined by 10.0 µM
ketanserin for [3H]-trans-H2-PAT and 20.0 µM triprolidine for
[3H]mepyramine. After incubation, assay mixtures were fil-
tered with a cell harvester through glass fiber (GF/B) sheets.
Sheets were rinsed three times with 5-10 mL of cold buffer
and counted for tritium by liquid scintillation spectrometry
at 60% efficiency. Nonlinear regression analysis of inhibition
data was used to determine IC50 values using Prism 2.0
(GraphPad, San Francisco, CA). IC50 values were converted
(()-tr a n s-1-(2-Meth ylp h en yl)-3-(N,N-d im eth yla m in o)-
1,2,3,4-tetr a h yd r on a p h th a len e (21). The primary amine 89
(2.3 g, 98 mmol) was dimethylated as above. The semisolid
was converted to the HCl salt with ethereal HCl as a methanol
solution. The solution was stripped and the 2.5 g (100%) of
residue dissolved in 5 mL of absolute ethanol and stored at
-10 °C. Crystallization began, and a small amount of ether
was added and the product again stored at -10 °C. Removal
of the supernatant and drying afforded 1.1 g (37%) of a light
1
yellow solid: mp 220-222 °C; H NMR (CDCl3) δ 7.3 (s, 1H,
ArH), 7.2 (t, 2H, ArH), 7.1 (dd, 2H ArH), 6.85 (d, 1H, ArH),
6.6 (dd, 1H, ArH), 4.6 (t, 1H, PhCHPh), 3.1 (dd, 1H, PhCH),
2.9 (dd, 1H, PhCH), 2.6 (m, 1H, CHN), 2.45 (s, 3H, PHCH3)
2.3 (s, 6H, N(CH3)2), 2.1 (t, 1H, CHCN). Anal. (C19H24NCl)
C,H,N.
(()-tr a n s-1-(4-Ch lor op h en yl)-3-(N,N-d im eth yla m in o)-
1,2,3,4-tetr a h yd r on a p h th a len e (22). The primary amine 90
(1.3 mg) was dimethylated as above to afford 1.4 g (97%) of a
yellow-brown oil which was dissolved in ethereal HCl and the
salt recrystallized from EtOH/Et2O to afford 620 mg of
colorless crystals (1.92 mmol, 38%): mp 256-258 °C; 1H NMR
(CDCl3) δ 7.2-7.4 (m, 4H, ArH), 7.1-7.18 (m, 2H, ArH), 6.8-
6.9 (m, 2H, ArH), 4.4 (t, 1H, PhCHPh), 3.0 (dd, 1H, PhCH),
2.8 (dd, 1H, PhCH), 2.6 (m, 1H, CHN), 2.2 (s, 6H, N(CH3)2),
2.0 (t, 1H, CHCN), 1.65 (m, 1H, CHCN). Anal. (C18H21NCl2)
C,H,N.
(()-tr a n s-1-(4-Meth ylp h en yl)-3-(N,N-d im eth yla m in o)-
1,2,3,4-tetr a h yd r on a p h th a len e (23). The primary amine 91
(180 mg, 0.67 mmol) was dimethylated as above to afford 180
mg (97%) of a light brown oil which was dissolved in ethereal
HCl. The salt was recrystallized from ethanol/Et2O to afford
39 mg (0.13 mmol, 19.4%) of colorless crystals: mp 252-253
°C; 1H NMR (CDCl3) δ 7.0-7.2 (m, 5H, ArH), 6.95 (d, 1H, ArH),
6.85 (dd, 1H, ArH), 4.4 (t, 1H, PhCHPh), 3.1 (dd, 1H, PhCH),
to K0.5 based upon the Cheng-Prussof equation where K0.5
)

3054 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Bucholtz et al.
IC50/(1 + L*/KD).30 Experiments were repeated at least three
times to determine mean K0.5 ( SEM.
(15) Wyrick, S. D.; Booth, R. G.; Myers, A. M.; Owens, C. E.; Bucholtz,
E. C.; Hooper, P. C.; Kula, N. S.; Baldessarini, R. G.; Mailman,
R. B. 1-Phenyl-3-amino-1,2,3,4-tetrahydronaphthalenes and Re-
lated Derivatives as Ligands for the Neuromodulatory σ3 recep-
tor: Further Structure-Activity Relationships. J . Med. Chem.
1995, 38, 3857-3864.
(16) Myers, A. M.; Charifson, P. S.; Owens, C. E.; Kula, N. S.;
McPhail, A. T.; Baldessarini, R. J .; Booth, R. G.; Wyrick, S. D.
Conformational Analysis, Pharmacophore Identification and
Comparative Molecular Field Analysis of Ligands for the Neu-
romodulatory σ3 Receptor. J . Med. Chem. 1994, 37, 4109-4117.
(17) Martin, Y. C.; Bures, M. G.; Dahaner, E. A.; Delasser, J .; Lico,
I.; Pavlik, P. A. A Fast Approach to Pharmacophore Mapping
and its Application to Dopaminergic and Benzopdiazepine
Agonists. J . Comput.-Aid. Mol. Des. 1993, 7, 83-102.
(18) Bucholtz, E. C.; Wyrick, S. D.; Owens, C. E.; McPhail, A. T.;
Booth, R. G. 1-Phenyl-3-(dimethylamino)-1,2,3,4-tetrahydronaph-
thalene (H2-PAT): Effect of Stereochemistry on Binding and
Function at Brain Histamine Receptors. Med. Chem. Res. 1998,
8, 322-332.
Ackn owledgm en t. U.S. Public Health Service Grant
NS35216 and the Pharmacy Foundation of North Caro-
lina supported this investigation. The authors are
grateful to Professor Richard B. Mailman for his interest
in this work and helpful discussions. The authors thank
Tripos Inc. for the generous provision of Sybyl 6.22 used
for various phases of this work.
Refer en ces
(1) Booth, R. G.; Wyrick, S. D.; Baldessarini, R. J .; Kula, N. S.;
Myers, A. M.; Mailman, R. B. New σ-like Receptor Recognized
by Novel Phenylaminotetralins: Ligand Binding and Functional
Studies, Mol. Pharmacol. 1993,44, 1232-1239.
(2) Wyrick, S. D.; Booth, R. G.; Myers, A. M.; Owens, C. E.; Kula,
N. S.; Baldessarini, R. J .; McPhail, A. T.; Mailman, R. B.
Synthesis and Pharmacological Evaluation of 1-phenyl-3-amino-
1,2,3,4-tetrahydronaphthalenes as Ligands for a Novel Receptor
with Neuromodulatory Sigma-like Activity. J . Med. Chem. 1993,
36, 2542-2551.
(3) Wyrick, S. D.; Myers, A. M.; Booth, R. G.; Kula, N. S.; Baldes-
sarini, R. J . Synthesis of [N-C3H3]-trans-(1R,3S)-(-)-1-phenyl-
3-N,N-(dimethylamino)-1,2,3,4-Tetrahydronaphthalene (H2-PAT).
J . Labelled Compds. Radiopharm. 1994, 34, 131-134.
(4) Owens, C. E.; Mailman, R. B.; Wyrick, S. D.; Booth, R. G.
Putative σ3 Sites in Mammalian Brain have Histamine H1
Properties. Soc. Neurosci. Abstr. 1996, 22, 2006.
(5) Choksi, N. Y.; Wyrick, S. D.; Booth, R. G. Phenylaminotetralins
Modulate Dopamine Synthesis in Rat Nucleus Accumbens In
Vivo by a Presynaptic H1 Type Receptor. Soc. Neurosci. Abs.1997,
23, 2036.
(6) Bucholtz, E. C.; Wyrick, S. D.; Brown, R. L.; Booth, R. G. Design
of Phenylaminotetralins Selective for Proposed Histamine H1
Receptor Subtypes in Mammalian Forebrain. Soc. Neurosci.
Abstr. 1997, 23, 2036.
(7) Owens, C. E.; Myers, A. M.; Lawler, C. P.; Mailman, R. B.;
Baldessarini, R. J .; Wyrick, S. D.; Booth, R. G. Characterization
and Distribution of Brain σ3 Receptors Labeled by [3H]-(-)-trans-
H2-PAT. Soc. Neurosci. Abstr. 1994, 20, 748.
(8) Keduda, R.; Prasad, P. D.; Fei, Y.; Leibach, F. H.; Ganapathy,
V. Cloning and Functional Expression of the Human Type 1
Sigma Receptor (hSigmaR1). Biochem. Biophys. Res. Commun.
1996, 229, 553-558.
(9) Hanner, M.; Moebius, F. F.; Flandorfer, A.; Knaus, H.; Striessing,
J .; Kempner, E.; Glossmann, H. Purification, Molecular Cloning,
and Expression of the Mammalian Sigma1-Binding Site. Proc.
Natl. Acad. Sci. U.S.A. 1996, 93, 8072-8077.
(10) Saxena, A. K.; Saxena, M. Developments in Antihistamines (H1)-
. In Progress in Drug Research; J ucker, E., Ed.; Birhauser
Verlag: Basel, 1992; Vol. 39 pp 35-125.
(11) Warringa, C. G.; Rekker, R. F.; Nauta, W. Th. 1,1-Diaryl-3-
aminopropenes and Some Related Compounds II. Structure-
Activity Relationships. Eur. J . Med. Chem. 1975, 10, 349-352.
(12) Ison, R. R.; Franks, F. M.; Soh, K. S. The Binding of Conforma-
tionally Restricted Antihistamines to Histamine Receptors. J .
Pharm. Pharmacol. 1973, 25, 887-894.
(13) Miyano, S.; Abe, A.; Kase, Y.; Yuizono, T.; Tachibana, K.; Miyata,
T.; Kito, G. Synthesis and Pharmacological Evaluation of Some
Pyridylmethyl Substituted Ethylenediamines. J . Med. Chem.
1970, 13, 704-708
(14) Casy, A. F.; Ganellin, C. R.; Mercer, A. D.; Upton, C. Analogues
of Triprolidine: Structural Influences upon Antihistamine Activ-
ity. J . Pharm. Pharmacol. 1992, 44, 791-795.
(19) Laus, G.; Tourwe, D.; Van Binst, G. Benzo and indoquinolizidine
derivatives XIX. Synthesis and Pharmacological Activity of
Quinolizidine Derivatives. Analogues of Butaclamol. Heterocycles
1984, 22, 311-331.
(20) Eschweiler, W. Ersatz von an Stickstoff Gebundenun Wasser-
stoffatomen Durch die Methylgruppe mit Hilfe von Formaldhyd.
Ber. 1905, 38, 880-882.
(21) The Sybyl program (version 6.22) is available from Tripos
Associates, 1699 Hanley Rd, St. Louis, MO 63144.
(22) Cramer, R. D., III; Patterson, D. E.; Bunce, J . D. Comparative
Molecular Field Analysis (CoMFA). 1. Effect of Shape on Binding
of Steroids to Carrier Proteins. J . Am. Chem. Soc. 1988, 110,
5959-5967.
(23) Cho, S. J .; Tropsha, A. Cross-Validated R2-Guided Region
Selection for Comparative Molecular Field Analysis: A Simple
Method to Achieve Consistent Results. J . Med. Chem. 1995, 38,
1060-1066.
(24) Leurs, R.; Smit, M. J .; Timmerman, H. Histaminergic Agonists
and Antagonists: Recent Developments. In Advances in Drug
Research; Testa, B., Ed.; Academic Press: London, 1991; Vol.
20, pp 217-30.
(25) Zhang, M. Q.; Leurs, R.; Timmerman, H. Histamine H1 Receptor
Antagonists. In Burger’s Medicinal Chemistry and Drug Dis-
covery, 5th ed.; Wolff, M., Ed.; J ohn Wiley & Sons Inc.: New
York, 1997.
(26) Nauta, W. T.; Rekker, R. F. Structure-Activity Relationships
of H1-Receptor Antagonists. In Handbook of Experimental
Pharmacology; Rocha e Silva, M., Ed.; Springer-Verlag: New
York, 1978.
(27) Protiva, M. Recent Progress in the Pharmaco-Chemical Research
on Antihistamine Drugs and Psychotropic Agents- Derivatives
of Tricyclic Systems Having a Seven or Eight-Membered Ring.
Farmaco Ed. Sci. 1966, 21, 76-104.
(28) Ortiz, A. R.; Pastor, M.; Palomer, A.; Cruciani, G.; Gago, F.;
Wade, R. C. Reliability of Comparative Molecular Field Analysis
Models: Effects of Data Scaling and Variable Selection Using a
Set of Human Synovial Fluid Phospholipase A2 Inhibitors. J .
Med. Chem. 1997, 40, 1136-1148.
(29) Southwick, P. L.; Sapper, D. I. 4-Styrylthialzoles. Synthesis and
Relationships Among Ultraviolet Absorption Spectra. J . Org.
Chem. 1954, 19, 1926-1937.
(30) Cheng, Y. C.; Prusoff, W. H. Relationship between the inhibition
constant (K1) and the concentration of inhibitor which causes
50% inhibition (I50) of an enzymatic reaction. Biochem. Phar-
macol. 1973, 22, 3099-3108.
J M980428X