Effect of ring size or an additional heteroatom on the potency and selectivity of bicyclic benzylamine-type inhibitors of phenylethanolamine N- methyltransferase

Article abstract of DOI:10.1021/jm9508292

In the search for potent and selective inhibitors of the enzyme phenylethanolamine N-methyltransferase (PNMT; EC 2.1.1.28), we examined the effect of ring size or an additional heteroatom in the conformationally- restricted benzylamine-type PNMT inhibitors. Based on semiempirical calculations (MNDO) and molecular modeling studies, PNMT-inhibitory activity of these compounds seemed to be dependent on (a) the torsion angle between the plane of the aromatic ring and the endo N atom lone pair (τ₂ angle), with the optimal value of τ₂ being about -75°, and (b) the amount of steric bulk about the 3-position of 1,2,3,4-tetrahydroisoquinoline (5, THIQ). 2,3,4,5-Tetrahydro-1H-2-benzazepine (6) was found to have the highest selectivity (PNMT K(i) = 3.34 μM, α₂ K(i) = 11 μM, selectivity = 3.2) as compared to other homologues of THIQ (PNMT K(i) = 9.67 μM, α₂ K(i) = 0.35 μM, selectivity = 0.036). The higher PNMT-inhibitory activity of 6 was attributed to favorable steric interactions of the puckered methylene groups in the putative bioactive conformation of 6 at the PNMT active site, whereas unfavorable interactions of these puckered methylene groups at the α₂- adrenoceptor were thought to be the cause of reduced α₂ affinity of 6. No further enhancement of the selectivity of the benzazepine ring system could be obtained via introduction of a second heteroatom (N, O, S) at the 5- position in this ring system.

Full text of DOI:10.1021/jm9508292

J . Med. Chem. 1996, 39, 3539-3546
3539
Effect of Rin g Size or a n Ad d ition a l Heter oa tom on th e P oten cy a n d Selectivity
of Bicyclic Ben zyla m in e-Typ e In h ibitor s of P h en yleth a n ola m in e
N-Meth yltr a n sfer a se1a
Gary L. Grunewald,* Vilas H. Dahanukar, Piao Ching,1b and Kevin R. Criscione
Department of Medicinal Chemistry, University of Kansas, Lawrence, Kansas 66045
Received November 9, 1995X
In the search for potent and selective inhibitors of the enzyme phenylethanolamine N-
methyltransferase (PNMT; EC 2.1.1.28), we examined the effect of ring size or an additional
heteroatom in the conformationally-restricted benzylamine-type PNMT inhibitors. Based on
semiempirical calculations (MNDO) and molecular modeling studies, PNMT-inhibitory activity
of these compounds seemed to be dependent on (a) the torsion angle between the plane of the
aromatic ring and the endo N atom lone pair (τ2 angle), with the optimal value of τ2 being
about -75°, and (b) the amount of steric bulk about the 3-position of 1,2,3,4-tetrahydroiso-
quinoline (5, THIQ). 2,3,4,5-Tetrahydro-1H-2-benzazepine (6) was found to have the highest
selectivity (PNMT Ki ) 3.34 µM, R2 Ki ) 11 µM, selectivity ) 3.2) as compared to other
homologues of THIQ (PNMT Ki ) 9.67 µM, R2 Ki ) 0.35 µM, selectivity ) 0.036). The higher
PNMT-inhibitory activity of 6 was attributed to favorable steric interactions of the puckered
methylene groups in the putative bioactive conformation of 6 at the PNMT active site, whereas
unfavorable interactions of these puckered methylene groups at the R2-adrenoceptor were
thought to be the cause of reduced R2 affinity of 6. No further enhancement of the selectivity
of the benzazepine ring system could be obtained via introduction of a second heteroatom (N,
O, S) at the 5-position in this ring system.
Epinephrine represents about 5-10% of the total
brain catecholamine content,2 and immunohistochemi-
cal techniques3,4 have unambiguously provided evidence
for the presence of epinephrine-containing neurons in
the central nervous system (CNS), although their exact
function is currently unknown. Based on the anatomi-
cal localization of these neurons in the CNS, it was
postulated that central epinergic neurons might be
involved in the regulation of R-adrenoceptors, blood
pressure, respiration, and the secretion of pituitary
hormones.4-7 Epinephrine is the final product in the
catecholamine biosynthetic pathway. Thus, its role in
the CNS can be studied without affecting the levels of
other catecholamines through the inhibition of the en-
zyme phenylethanolamine N-methyltransferase (PNMT;
EC 2.1.1.28)8 which catalyzes the N-methylation of
norepinephrine to epinephrine using S-adenosyl-L-meth-
ionine (AdoMet) as the cofactor. It has been suggested
that there is an ordered sequential binding of AdoMet
followed by norepinephine9,10 and that most inhibitors
are competitive with either norepinephrine or AdoMet.11
The suppositions regarding the role of epinephrine were
later substantiated by pharmacological studies per-
formed with PNMT inhibitors. Most notably, the ability
of PNMT inhibitors to lower blood pressure in sponta-
neously hypertensive rats was widely investigated.12
Reduction in the levels of central epinephrine by PNMT
inhibitors was initially believed to be the cause of this
antihypertensive effect. Later, it was found that most
of the widely-studied PNMT inhibitors were nonselec-
tive and had R2-adrenoceptor binding affinity,13 which
would complicate an interpretation of the effects of these
compounds on blood pressure. We have undertaken the
task of designing a selective and potent inhibitor of
PNMT so that the role of central epinephrine can be
defined unambiguously.
Benzylamines are a class of potent PNMT inhibitors.
Conformational restriction of the aminomethyl side
chain in benzylamine (1) in the form of 1,2,3,4-tetrahy-
droisoquinoline14 (5) or 2,3,4,5-tetrahydro-1H-2-benz-
azepine15 (6) results in increased inhibitory potency.
Some well-studied inhibitors of the benzylamine class
are SK&F 64139 (2a , 7,8-dichloro-1,2,3,4-tetrahydroiso-
quinoline),16 SK&F 29661 (2b, 7-(aminosulfonyl)-1,2,3,4-
tetrahydroisoquinoline),17 and LY 134046 (3, 8,9-dichloro-
2,3,4,5-tetrahydro-1H-2-benzazepine).18a Compound 3
(PNMT Ki ) 0.264 µM, R2 Ki ) 4.1 µM, selectivity )
16) is more selective than 2a (PNMT Ki ) 0.202 µM, R2
Ki ) 0.022 µM, selectivity ) 0.11).18b The 2,3-dichlo-
robenzylamine moiety in 2a and 3 confers potent
PNMT-inhibitory activity. Since phenylethylamines
also form a class of R2-adrenoceptor agonists,19 it was
suggested by Fuller that the R2-adrenoceptor affinity of
2a might result from the presence of a conformationally-
restricted phenylethylamine skeleton in 2a . Further-
more, the lack of the phenylethylamine moiety in 3
might make it a more selective inhibitor than 2a .13,20
Correlations of inhibitor potency with pKa indicate that
benzylamine-type inhibitors bind in the neutral form.21,22
In addition, studies on some conformationally-restricted
benzylamines from this laboratory have demonstrated
the dependence of PNMT-inhibitory activity on the
* Author to whom correspondence should be addressed. Phone: (913)
864-4497. Fax: (913) 864-5326.
X Abstract published in Advance ACS Abstracts, August 1, 1996.
S0022-2623(95)00829-6 CCC: $12.00 © 1996 American Chemical Society

3540 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Grunewald et al.
Sch em e 2
orientation of the amino group lone pair and the steric
bulk at the active site.23 It is possible that the confor-
mational effects resulting from the increase of hetero-
cyclic ring size from 2a to 3 change the orientation of
the amino group lone pair and the steric bulk and might
be responsible for the observed increase in selectivity.
Such conformational effects could be extended by fur-
ther changes in the ring size (4 and 7) or by the addition
of a second heteroatom (8-10) at the 5-position of the
benzazepine ring. Such changes in conformationally-
restricted benzylamines might help in the identification
of a heterocycle that lacks affinity for the R2-adreno-
ceptor but retains affinity for the active site of PNMT.
4:2:2:1) (Scheme 2).29 Instead of separating the crude
reaction mixture by chromatography,29 it was reduced
with LAH to afford a mixture of amines from which
2,3,4,5-tetrahydro-1H-2-benzazepine (6) was readily
separated by taking advantage of the pKa difference
between the aliphatic and aromatic amine (1,3,4,5-
tetrahydro-2H-1-benzazepine, 24).
Ch em istr y
A similar approach based on the Schmidt reaction
failed to provide the amide precursor to 7 from 1-ben-
zosuberone28 and 2-benzosuberone. Hence, N-methyl-
1,2,3,4,5,6-hexahydro-2-benzazocine (25) was prepared
according to the procedure described by Hauser30 and
then N-demethylated using the protocol reported by
Rice.31 Pictet-Spengler cyclocondensation32 of either
2,3-Dihydroisoindole (4) was prepared by reduction
of phthalimide with borane as reported by Dubois et al.24
The Schmidt reaction was used to prepare the seven-
membered heterocyclic amides from the corresponding
aromatic ketones. The mixture of regioisomeric amides
resulting from aryl and alkyl group migration was
readily separated by chromatography. Even though a
mixture of regioisomers was obtained and the yields
were low due to the low reactivity of these aromatic
ketones, this approach provided a convenient way of
synthesizing benzazepines containing a second hetero-
atom. The isolated amides were readily reduced with
borane to yield amines 8-10 (Scheme 1). In the case
racemic or enantiomerically pure 1-phenyl-2-aminopro-
pane33 with formaldehyde gave 26 in an average yield
of 43% (Scheme 3).
Sch em e 1
Sch em e 3
of 4-chromanone (13), the undesired regioisomer 19
hydrolyzed faster than 20 under the reaction conditions,
thus facilitating the isolation of the desired regioiso-
mer.25,26 With other aromatic ketones (11, 12, and 14),
both regioisomers formed were stable under the reaction
conditions and could be isolated.
The Schmidt reaction on 1-tetralone (11) under strong
nonaqueous acidic conditions such as sulfuric acid25 or
polyphosphoric acid27 produced the amide 16 from an
aryl ring migration as the major product. Hjelte and
Agback28 reported the tetrazole as the major product of
the Schmidt reaction on 1-tetralone in concentrated
hydrochloric acid. In our laboratory, the desired 15 was
obtained as the major product, while 23, unreacted
starting material (11), and 16 were also obtained (ratio
Bioch em istr y
All the compounds were evaluated as their hydro-
chloride salts for in vitro activity as inhibitors of PNMT.
In vitro PNMT activity was assessed by use of a
standard radiochemical assay that has been described
previously for both substrates34 and inhibitors.35 Bovine
adrenal PNMT used for the in vitro assay was purified
according to the procedure of Connett and Kirshner36
through the isoelectric precipitation step. Inhibition
constants were determined by using three different
concentrations of inhibitor, as previously described,35
with phenylethanolamine as the variable substrate. R2-
Adrenergic receptor binding assays were performed

Bicyclic Benzylamine Inhibitors of PNMT
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18 3541
Ta ble 1. In Vitro Activities of Benzylamine Analogues as
Inhibitors of PNMT and Binding of [3H]Clonidine at the
R2-Adrenoceptor
F igu r e 2. Three points used in generating root mean square
fit illustrated for the template 5 to which the other analogues
were fitted.
A, PNMT
Ki ( SEM, µM
B, R2
Ki ( SEM, µM
B/A
selectivity
compd
n
Ta ble 3. Summary of Calculations for the Conformations of
Benzylamine Analogues 4-7
1
4
5
6
7
0
1
2
3
4
179 ( 9
20 ( 1
0.11
231 ( 17
0.82 ( 0.11
0.35 ( 0.10
11 ( 1
0.0035
0.036
3.2
9.67 ( 0.37
3.34 ( 0.24
21.3 ( 0.8
compound
6
7
16 ( 1
0.73
4
5
(global)a (local)a (global)a,b (local)a,b
τ1
-1.70 -15.3 -61.2
-117 -75.1 -36.1 -75.4
5.16
21.70 16.71
0.34
19.2
29.9
-56.0
4.27
20.64
0.39
37.0
-72.2
4.61
20.36
0.37
Ta ble 2. In Vitro Activities of Benzylamine Analogues as
Inhibitors of PNMT and Binding of [3H]Clonidine at the
R2-Adrenoceptor
τ2
Dcent-lp (Å)
Hf (kcal/mol)
rms fit
4.53
3.63
17.80
0.49
4.69
20.17
0.25
a
The global minimum was obtained by using the random search
option in the Sybyl molecular modeling program, and the energies
were calculated using the Tripos molecular mechanics force field.
The energies of local minima were about 0.4 (6) and 1.4 (7) kcal/
A, PNMT
Ki ( SEM, µM
B, R2
Ki ( SEM, µM
B/A
selectivity
b
compd
X
mol higher than the corresponding global minima. The heat of
formation (Hf) was calculated using the MNDO method in the
semiempirical molecular orbital package (MOPAC, version 5.0 at
the Sybyl interface). Based on the Hf, the local minima of 7 might
actually be the global minimum conformation.
6
8
9
CH2
NH
O
3.34 ( 0.24
28.0 ( 1.9
21.2 ( 2.7
4.10 ( 0.33
11 ( 1
3.2
2.6 ( 0.1
4.1 ( 0.2
4.6 ( 0.5
0.093
0.19
1.1
10
S
formers of 1 as evidenced by its low activity in compari-
son with THIQ (5). Conformational restriction of 1 to
form a five-membered ring showed surprisingly lowered
PNMT-inhibitory activity. However, other homologues
of 4 were more potent than 1 itself.
To explain the effect of ring size on the biological
activity, conformational analysis studies were done on
ring systems using the random search option in the
Sybyl molecular modeling program. The random search
option in Sybyl enables one to identify the global
minimum energy conformation as well as several local
minima. The geometries of the supposed global mini-
mum conformations (calculated using the Tripos mo-
lecular mechanics force field) were further optimized by
using MNDO, a semiempirical quantum mechanical
method implemented in MOPAC version 5.0 at the
Sybyl interface. For 1,3-dihydroisoindole (4) and THIQ
(5), only a single conformer was located by the random
search conformational analysis. As 5 was relatively
rigid and active as compared to the other homologues,
it was used as a template in the rigid fitting procedure.
In this three-point-fitting procedure (Figure 2), the two
ends of a normal (2 Å long) passing through the centroid
of the aromatic ring and an axial lone pair on the N
atom (2.4 Å long, representing a possible H-bonding
distance) were used. Earlier studies on tricyclic con-
formationally-restricted benzylamines had indicated the
dependence of PNMT-inhibitory activity on the endo
(axial) lone pair.23 Better root mean square (rms) fits
were produced when an axial lone pair on the N atom,
instead of an equatorial lone pair, was used for the fit.
The energy difference between the two forms (axial vs
equatorial lone pair) was within 1 kcal/mol for analogues
5-7. However, use of these global minimum conforma-
tions resulted in a poor fit for 6 and 7. The results of
the modeling studies are presented in Table 3.
F igu r e 1. Conformational descriptors (τ1 and τ2) defined for
(a) benzylamine and (b) its conformationally-restricted ana-
logues with axial orientation of the N atom lone pair. The low-
energy conformation of benzylamine has the side chain
oriented perpendicular to the plane of the aromatic ring. The
figure was partly adapted from ref 39.
using cortex obtained from male Sprague-Dawley
rats.37 [3H]Clonidine was used as the radioligand to
define the specific binding, and phentolamine was used
to define the nonspecific binding.
Resu lts a n d Discu ssion
The results of biochemical evaluation are grouped in
Table 1 (effect of ring size) and Table 2 (effect of
heteroatom). All the compounds evaluated exhibited
competitive kinetics indicative of inhibition of the bind-
ing of phenylethanolamine (substrate) at the active site.
The moderately low activity of 1 as a PNMT inhibitor
may be due to a conformational effect. Conformations
of 1 have been studied using mass resolved excitation
spectroscopy, and the low-energy conformation of 1 as
defined in terms of conformational descriptors (torsion
angles τ1 and τ2) is shown in Figure 1.38,39 In this low-
energy conformation the aminomethyl side chain was
oriented perpendicular to the ring (τ1 ) 90°), while the
N-lone pair electrons were oriented away from the ring
(τ2 ) 180°). A slightly higher energy (∆E < 0.01 kcal/
mol) conformer of 1 had τ2 ) 60°. These low-energy
conformers may not necessarily be the bioactive con-
Examination of the data shown in Table 3 revealed
that the important factors determining the PNMT-
inhibitory activity were not only the rms fit but, more

3542 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Grunewald et al.
F igu r e 4. Computer graphics-generated intersection of the
volume of the puckered methylene groups in 6 and the methyl
group in enantiomers of 26 in the background of the MNDO-
optimized local minimum conformation of 6. The volume of
the puckered methylene groups was obtained by subtraction
(MVOL command) of the volume between 6 and modified 6 in
which these two methylene groups were deleted. The MNDO-
optimized conformations of 5 and PNMT active (S)-(+)-26 were
fitted using the three-point rms-fitting procedure. Intersection
of the volume occupied by the puckered methylene groups in
6 with the volume difference between 5 and (S)-(+)-26 (volume
occupied by the methyl group) is displayed as a dotted
contoured surface (upper left volume), and it represents the
area of steric bulk tolerance at the PNMT active site. Thus,
the puckered methylene groups in 6 partly occupy the same
region of space corresponding to this volume difference.
Favorable hydrophobic interaction about this compact region
of steric bulk tolerance leads to enhanced potency (as in 6) at
the PNMT active site. Conversely, the intersection between
the volume of the puckered methylene groups in 6 and the
volume difference between (R)-(-)-26 and 5 represents the
area of steric bulk intolerance at the R2-adrenoceptor and is
shown as a bold contoured surface (lower right volume). The
puckered methylene groups in 6 are projected more closely in
the region of acute steric bulk intolerance at the R2-adreno-
ceptor which results in lowered affinity.
F igu r e 3. MNDO-optimized geometries of homologues 4, 5,
6 (local), and 7 (local).
Ta ble 4. In Vitro Activities of Enantiomers of
3-Methyl-1,2,3,4-tetrahydroisoquinoline as Inhibitors of PNMT
and Binding of [3H]Clonidine to the R2-Adrenoceptor
A, PNMT
Ki ( SEM, µM
B, R2
Ki ( SEM, µM
B/A
selectivity
compd
26 (()
R-(-)
S-(+)
2.13 ( 0.11
38.3 ( 1.7
1.01 ( 0.05
0.84 ( 0.08
6.2 ( 0.3
0.52 ( 0.06
0.39
0.16
0.51
The eight-membered homologue 7 had steric bulk
comparable to 3-ethyl-THIQ in the region of limited
steric bulk tolerance. Negative steric interaction was
the likely cause of decreased PNMT-inhibitory activity
of 7 as compared to 6.
importantly, the τ2 angle and the distance from the
centroid of the aromatic ring to the N atom lone pair
(Dcent-lp). The low-energy conformer of 1 had a τ2 angle
different from that of the template 5. Thus, the low
activity of 1, as well as that of 4, may be due to a
different τ2 angle as compared with its other analogues.
This was reflected as an increased value of Dcent-lp in
4. At the active site the N atom lone pair in 4 may not
be able to directionally interact optimally at a specific
amino acid residue where the other active homologues
5-7 can interact. Homologues 5-7 have similar τ2
angles and Dcent-lp, but they differ largely in steric bulk.
While there does not appear to be a strong correlation
between the τ2 angle and PNMT activity, the optimal
τ2 angle required for binding at the PNMT active site
appears to be approximately -75° as this torsion angle
maintains a more constant value of Dcent-lp. The puta-
tive bioactive conformations of the homologues are
shown in Figure 3.
Introduction of a methyl group at the 3-position in 5
resulted in increased PNMT-inhibitory activity (26, see
Table 4) compared to 5.23,40 Increased potency of the
more active (S)-(+)-26 possibly resulted from a favorable
but directional hydrophobic interaction at the PNMT
active site about the 3-position of the THIQ nucleus. In
the local minimum conformation of homologue 6, the
puckered methylene groups partly occupied the same
steric region of space about the 3-position of THIQ at
the PNMT active site and therefore had a Ki lower than
that for 5 (Figure 4). However, this region at the PNMT
active site has a limited steric bulk tolerance, and a
further increase in steric bulk, such as in (()-3-ethyl-
THIQ, resulted in decreased activity (Ki ) 23.9 µM).40
The R2-adrenoceptor affinity for 1 and its analogues
showed a different trend. Conformational restriction of
1 as in 4 resulted in about a 25-fold increased affinity
toward the R2-adrenoceptor. Although homologues 4
and 5 have comparable affinity toward the R2-adreno-
ceptor, they have very different τ2 angles. In contrast,
5 and 6 have similar τ2 angles but differed widely in
their abilities to bind with the R2-adrenoceptor. These
results implied a nondependence of the τ2 angle on the
R2-adrenoceptor affinity and a nonrigid interaction mode
for the N atom lone pair at the R2-adrenoceptor. The
lowered affinity of (R)-(-)-26 probably resulted from
negative steric interactions of the methyl group at the
R2-adrenoceptor. Based on the active analogue ap-
proach developed by Marshall et al.,41 the volume
occupied by the methyl group in (R)-(-)-26 represented
the area of steric bulk intolerance at the R2-adrenocep-
tor. The 30-fold decreased affinity of the seven-
membered homologue 6 as compared to 5 can be
attributed to nonfavorable interaction of the puckered
extra methylene group which was projected closer to the
region of steric bulk intolerance. The acute nature of
the steric bulk intolerance was further defined by
decreased affinity of the higher homologue 7.
Lesser involvement of puckered methylene groups in
6 at the PNMT active site was reflected in a small (3-
fold) gain in potency with respect to 5. In contrast, the
observed 30-fold loss in affinity of (R)-(-)-26 with
respect to 5 at the R2-adrenoceptor was in accordance
with the proposed proximity of the puckered methylene

Bicyclic Benzylamine Inhibitors of PNMT
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18 3543
are reported in parts per million (ppm) relative to tetrameth-
ylsilane (TMS, 0.00 ppm). Carbon nuclear magnetic resonance
spectra (13C NMR) were recorded on a Varian XL-300 spec-
trometer with CDCl3 as the solvent, and the chemical shifts
are reported in ppm relative to CDCl3 (77.0 ppm). For the
hydrobromide salts of amines, NMR spectra were recorded in
deuterated dimethyl sulfoxide (DMSO-d6) and the chemical
shifts are reported relative to DMSO (2.49 ppm for 1H NMR
and 39.5 ppm for 13C NMR). Multiplicity abbreviations: s,
singlet; d, doublet; t, triplet; q, quartet; m, multiplet; bs, broad
singlet; e, exchangeable. Infrared spectra were obtained on a
Perkin Elmer 1420 infrared spectrophotometer. Flash chro-
matography46 was performed using silica gel 60 (230-400
mesh) supplied by Universal Adsorbents, Atlanta, GA. Pre-
parative centrifugal thin layer chromatography (PCTLC) was
performed on a Harrison model 7924 Chromatotron (Harrison
Research, Palo Alto, CA) using Merck silica gel 60 PF254/
CaSO4‚0.5H2O binder on 1, 2, or 4 mm thickness plates.
Analytical TLC was performed by using silica gel with a
fluorescent indicator coated on 1 × 3 in. glass plates in 0.2
mm thickness (Whatman MKGF silica gel 200 µm). Optical
rotations were recorded on a Perkin-Elmer polarimeter using
the sodium D line as the source. Bulb-to-bulb distillations were
carried out on a Kugelrohr distillation apparatus (Aldrich
Chemical Co., Milwaukee, WI), and oven temperatures were
recorded. Combustion analyses were performed on a Hewlett-
Packard model 185B CHN analyzer at the University of
Kansas by Dr. Tho Ngoc Nguyen. Amine hydrochloride salts
were prepared by adding a solution of methanolic HCl to the
methanolic solution of the amine followed by crystallization
of the resulting hydrochloride from MeOH-Et2O. S-Adenosyl-
L-methionine was obtained from Sigma Chemical Co. [methyl-
3H]-S-Adenosyl-L-methionine used in the radiochemical assay
was purchased from New England Nuclear Corp. (Boston, MA).
Bovine adrenal glands were obtained from Stinson Meat
Processing (Ottawa, KS). [3H]Clonidine used in the R2-
adrenoceptor binding assay was purchased from Amersham
Corp. (Arlington Heights, IL).
groups to the region of steric bulk intolerance at that
site. The region about the 3-position of 26 at the PNMT
active site and the R2-adrenoceptor seemed to be comple-
mentary in terms of their steric bulk tolerance (see
Figure 4).
Because the seven-membered ring showed maximum
selectivity, a further improvement in selectivity was
sought via the introduction of an additional heteroatom
into the ring. The conformations of heterocycles having
two heteroatoms would be considerably different from
those containing a single heteroatom (nitrogen). The
additional heteroatom could influence the overall con-
formation mainly through electrostatic interactions and
intramolecular H-bonding.42 A search of the Cambridge
Structural Database43 for structures similar to com-
pounds 8-10 found one compoundsthe 7-acetyl deriva-
tive of 9. The Sybyl-generated low-energy conformation
of 9 was quite similar to the X-ray structure found.44
Based strictly on qualitative considerations, it was
observed that bioisosteric replacement of the methylene
group in the benzazepine ring with electronegative and
strong H-bonding heteroatoms like N and O (analogues
8 and 9) decreased the affinity for PNMT. This effect
may be the result of an altered preferred conformation
due to the presence of an additional heteroatom or from
a different binding mode at the PNMT active site via
the interaction of the additional heteroatom with an
amino acid residue. The PNMT-inhibitory activity for
the heterocycle 10 bearing the S atom was comparable
to that of analogue 6. The better PNMT-inhibitory
activity of the sulfur-containing analogue 10 was per-
haps the consequence of the lower electronegativity (and
increased lipophilicity) of the sulfur atom or its ability
to form weaker H-bonds than nitrogen or oxygen
atoms.45 The R2-adrenoceptor affinity of the analogues
containing an additional heteroatom was higher (lower
Ki) than for 6. Again, as proposed above, the confor-
mational changes induced by the other heteroatom or
an altered binding mode might be responsible for this
outcome.
2,3,4,5-Tetr a h yd r o-1H-2-ben za zep in e (6). The Schmidt
reaction on 1-tetralone (11; 2.10 g, 14.1 mmol) was done
according to the procedure described by Hjelte and Agback.28
After the usual workup, the crude reaction mixture (2.30 g)
was dissolved in dry THF (50 mL) and cooled in an ice bath,
and LiAlH4 (0.88 g, 23 mmol) was added. The suspension was
heated to reflux under N2 for 12 h and cooled in an ice bath,
and to the suspension was cautiously added water (2.2 mL)
followed by 15% NaOH (2.2 mL) and again water (6.9 mL) to
quench the reaction. The reaction mixture was warmed to
room temperature, stirred for 1 h, and filtered through Celite.
The filter cake was washed with CH2Cl2 and water. The
filtrate was extracted with CH2Cl2 (thrice), and the combined
CH2Cl2 extracts were extracted with 2 N HCl (thrice). The
acidic aqueous extract was carefully basified to pH 6.5-7
(using a pH meter to measure the pH) and washed with CH2-
Cl2 (thrice) to remove the less basic side products, which were
isolated and identified previously.29 The more basic desired
amine was obtained by further basification of the aqueous
layer with KOH pellets followed by extraction with CH2Cl2
(thrice). The combined CH2Cl2 extracts were dried over
anhydrous K2CO3, evaporated to yield a yellow oil (1.44 g),
and distilled bulb-to-bulb (75-80 °C, 0.25 mmHg) to afford a
colorless oil (0.83 g, 41%). The hydrochloride salt was obtained
as small colorless prisms: mp 220-221 °C dec (lit.28 mp 223-
225 °C; IR (KBr, HCl salt) 2940, 2800, 1570, 1445, 1070, 1020,
Su m m a r y a n d Con clu sion
The PNMT-inhibitory activity in the conformation-
ally-restricted analogues of 1 seems to be dependent on
the τ2 angle and the amount of steric bulk. In contrast,
only steric bulk was considered the main factor govern-
ing the affinity toward the R2-adrenoceptor. Steric bulk
was tolerated only to a limited extent about the 3-posi-
tion of THIQ at the PNMT active site. However, the
addition of steric bulk in the same region of space at
the R2-adrenoceptor resulted in significant loss of affin-
ity. Introduction of an additional heteroatom in the
benzazepine ring system failed to improve selectivity.
The benzazepine ring system thus appears to be an
optimal skeleton for further structural modifications to
improve selectivity. This suggestion is consistent with
the observed 5-fold increase in selectivity of 3 vs 6.
1
760 cm-1; H NMR (CDCl3) δ 7.14-7.09 (m, 4 H, ArH), 3.92
(s, 2 H, H-1), 3.19 (t, 2 H, J ) 5.3 Hz, H-3), 2.95-2.91 (m, 2 H,
H-5), 1.84 (bs, e, 1 H, NH), 1.73-1.68 (m, 2 H, H-4); 13C NMR
(CDCl3) δ 142.6, 142.6, 128.9, 128.1, 126.8, 125.8, 54.8 (C-1),
53.4 (C-3), 35.9 (C-4), 30.7 (C-5). Anal. (C10H13N‚HCl) C, H,
N.
Exp er im en ta l Section
All reagents and solvents were reagent grade or purified
by standard methods before use. Melting points were deter-
mined in open capillaries on a Thomas Hoover melting point
apparatus calibrated with known compounds but are otherwise
uncorrected. Proton nuclear magnetic resonance spectra (1H
NMR) were recorded on a Varian XL-300 or a GE QE-300
spectrometer with CDCl3 as the solvent, and chemical shifts
Gen er a l P r oced u r e for th e Sch m id t Rea ction for th e
Syn th esis of Com p ou n d s 15, 18, 20, a n d 22.25 The ketone
was dissolved in 2 times its weight of concentrated sulfuric
acid, and sodium azide (1.3 mol relative to the ketone) was
added in portions at ice bath temperature under a positive
pressure of N2. Additional concentrated sulfuric acid was

3544 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Grunewald et al.
added to facilitate stirring. After stirring for 24 h at room
temperature, the resulting brown-red solution was poured over
ice water and cautiously basified with potassium hydroxide
pellets. The solution was extracted with ether (thrice), and
the combined ether extracts were dried over MgSO4.
1,2,3,4,5,6-Hexa h yd r o-2-ben za zocin e (7). To an ice-cold
solution of N-methyl-1,2,3,4,5,6-hexahydro-2-benzazocine30 (25;
1.20 g, 6.85 mmol) in dry toluene (10 mL) was cautiously added
phenyl chloroformate (5.62 g, 4.50 mL, 36.5 mmol), and the
solution was heated to reflux under N2 for 36 h. The yellow
reaction mixture was diluted with ice water and extracted with
CHCl3 (thrice). The combined CHCl3 extracts were washed
with 3 N NaOH (thrice) and brine (once) and evaporated to
afford a yellow oil. Hydrazine (95%, 5 mL) and hydrazine
solution in water (64%, 5 mL) were carefully added to this oil,
and the solution was heated to reflux under N2 for 6 h. After
cooling, the reaction mixture was concentrated to yield a
yellow-white solid that was basified with 40% NaOH and
extracted with CH2Cl2 (thrice). The combined CH2Cl2 extracts
were washed with brine (once), dried over anhydrous Na2SO2,
and evaporated to give a viscous yellow oil (0.97 g). Bulb-to-
bulb distillation (85-95 °C, 0.25 mmHg) gave a colorless oil
(0.31 g, 28%). Because the hydrochloride salt was difficult to
crystallize, the oxalate salt was prepared and crystallized from
MeOH-Et2O: mp 193-196 °C dec; 1H NMR (CDCl3) δ 7.22-
7.19 (m, 2 H, ArH), 7.14-7.10 (m, 2 H, ArH), 3.92 (s, 2 H,
H-1), 3.35 (s, e, 1 H, NH), 2.86-2.82 (m, 2 H, H-3), 2.70 (t, 2
H, J ) 5.0 Hz, H-6), 1.71-1.60 (m, 2 H, H-5), 1.59-1.52 (m, 2
H, H-5); 13C NMR (CDCl3) δ 141.3, 138.0, 129.5, 129.1, 127.7,
126.6, 48.8 (C-1), 45.8 (C-3), 32.0, 30.7, 28.2; EIMS m/ z 162
(M + 1, 4), 161 (M+, 17), 160 (M+ - 1, 14), 146 (14), 132 (40),
118 (31), 104 (63), 91 (39), 78 (31), 77 (29), 44 (100). Anal.
(C11H15N‚C2H2O4) C, H, N.
Gen er a l P r oced u r e for th e Red u ction of Sch m id t
Am id es for th e Syn th esis of Com p ou n d s 6 a n d 8-10. To
the ice-cold solution of amide (3 mmol) in dry THF (10 mL)
under N2 was added BH3‚THF complex (1 M solution in THF,
6 mL, 6 mmol). After stirring for 15 min at ice bath temper-
ature, the reaction mixture was heated to reflux for 12 h. The
reaction mixture was cooled in an ice bath, and excess borane
was destroyed by careful addition of MeOH. After the removal
of the solvents on a rotary evaporator, the residue was treated
with methanolic HCl (15 mL) and heated to reflux for 3 h to
destroy the amine-borane complex. The residue obtained
after the evaporation of MeOH was suspended in water (15
mL), cooled, basified with solid KOH (pH about 12), and
extracted with CH2Cl2 (thrice). The combined CH2Cl2 extracts
were washed with brine (once), dried over anhydrous K2CO3,
and evaporated to give an oil which was distilled bulb-to-bulb
to furnish the amine.
2,3,4,5-Tetr a h yd r o-5H-1,4-ben zod ia zep in e (8). 2,3-Di-
hydro-1,4-benzodiazepin-5(4H)-one (18; 0.24 g, 20%) was
prepared by the Schmidt reaction on 1,2,3,4-tetrahydroquino-
lin-4-one47 (12; 1.1 g, 7.6 mmol). The crude product was
purified by PCTLC (2 mm, silica gel; CH2Cl2-MeOH-NH4-
OH, 250:20:1, as the eluent) and then recrystallized from
EtOAc to afford small pale yellow prisms, mp 133-134 °C (lit.48
mp 158 °C). Amide 18 (0.51 g, 3.1 mmol) was reduced to afford
a colorless oil (bp 110 °C, 0.5 mmHg, 0.29 g, 63%). The
dihydrochloride salt was obtained as a yellow-green crystalline
solid: mp (2HCl) 238-240 °C dec (lit.49 mp (2HCl) 243-244
°C); IR (KBr, 2HCl) 2900-2400 (b), 1570, 1465, 1400, 1310,
(()-3-Meth yl-1,2,3,4-tetr a h yd r oisoqu in olin e (26). The
procedure described by Potapov et al.32 was slightly modified
to improve the yield. 1-Phenyl-2-aminopropane (2.10 g, 15.5
mmol) was treated with 37% formaldehyde (2.10 mL) yielding
a colorless precipitate. Concentrated HCl (12.6 mL) was added
to it, and the suspension was refluxed for 6 h. The brown
solution was concentrated on a rotary evaporator, and the
residue obtained was taken up in water (50 mL). The aqueous
solution was washed with CH2Cl2 (thrice) to remove brown-
colored impurities. The yellow aqueous extract was cooled,
basified with KOH pellets, extracted with CH2Cl2 (thrice),
dried over anhydrous Na2SO2, and evaporated to provide a red
viscous oil (1.58 g). Bulb-to-bulb distillation (75-80 °C, 0.5
mmHg) gave a colorless oil (1.04 g). Purification of this oil by
flash chromatography using CH2Cl2-MeOH-NH4OH (250:18:
1) as the eluent provided the desired compound as a colorless
oil (0.98 g, 43%). The hydrochloride salt was crystallized as a
colorless solid from MeOH-Et2O: mp 234-236 °C dec; IR
(film) 3240 (NH), 3000, 2940, 2910, 2880, 1490, 1440, 1370,
1
1170, 930, 770 cm-1; H NMR (DMSO-d6) δ 10.21-9.98 (m, 4
H, 2NH2+), 7.64 (d, 1 H, J ) 7.8 Hz, ArH), 7.57 (d, 1 H, J )
7.6 Hz, ArH), 7.48-7.43 (m, 1 H, ArH), 7.34-7.29 (m, 1 H,
ArH), 4.52 (s, 2 H, H-1), 3.59-3.53 (m, 4 H, H-3, H-4); 13C NMR
(DMSO-d6) δ 141.5, 132.4, 130.1, 126.6, 126.3, 122.1, 48.3, 45.6,
44.6. Anal. (C9H12N2‚2HCl) C, H, N.
2,3,4,5-Tetr a h yd r o-5H-1,4-ben zoxa zep in e (9). 2,3-Di-
hydro-1,4-benzoxazepin-5(4H)-one (20) was prepared by the
Schmidt reaction on chroman-4-one (13; 5.0 g, 34 mmol)
according to the above described procedure of Lockhart et
al.25,26 (2.0 g, 36% after crystallization from EtOAc-hexanes),
mp 115-117 °C (lit.26 mp 114-116 °C). Reduction of amide
20 (1.0 g, 6.1 mmol) by BH3‚THF followed by bulb-to-bulb
distillation (75 °C, 0.10 mmHg) furnished the desired amine
as a colorless oil (0.67 g, 74%). The hydrochloride was
crystallized to afford well-defined colorless needles: mp (HCl)
188-190 °C; IR (KBr, HCl salt) 2890, 2730, 1490, 1440, 1230,
1
1310, 740 cm-1; H NMR (CDCl3) δ 7.12-6.98 (m, 4 H, ArH),
4.06 and 3.99 (AB q, J AB ) 16.1 Hz, H-1), 3.00-2.94 (m, 1 H,
H-3), 2.75 (dd, 1 H, J ) 16.1, 3.9 Hz, H-4), 2.51-2.42 (m, 1 H,
H-4), 1.43 (bs, e, NH), 1.21 (d, 3 H, J ) 6.3 Hz); 13C NMR
(CDCl3) δ 135.3, 134.8, 129.0, 125.9, 125.6, 49.1, 48.5, 37.1,
22.4. Anal. (C10H13N‚HCl) C, H, N.
1
1050, 970, 760 cm-1; H NMR (CDCl3) δ 7.11-7.05 (m, 2 H,
ArH), 6.99-6.93 (m, 2 H, ArH), 3.95-3.92 (m, 2 H, H-4), 3.86
(s, 2 H, H-1), 3.11 (t, 2 H, J ) 4.3 Hz, H-3), 1.58 (s, e, 1 H,
NH); 13C NMR (CDCl3) δ 159.6, 134.7, 128.8, 127.8, 122.8,
120.5, 74.6, 52.7, 51.9. Anal. (C9H11NO‚HCl) C, H, N.
2,3,4,5-Tetr a h yd r o-5H-1,4-ben zoth ia zep in e (10). 3,4-
Dihydro-1,4-benzothiazepin-5(2H)-one (22) was prepared by
the Schmidt reaction on thiochroman-4-one (14; 5.0 g, 30
mmol). The crude product was subjected to flash chromatog-
raphy (silica gel; EtOAc-hexanes, 2:1, as the eluent) to afford
2,3-dihydro-1,5-benzothiazepin-4(5H)-one (21; 1.1 g, 41%, mp
210-212 °C (lit.50 mp 218-219 °C)) and the desired amide 22
(0.45 g, 16%, crystallization from EtOAc-hexanes gave color-
less crystals, mp 191-192 °C (lit.51 mp 190-191 °C)). Amide
22 (0.45 g, 2.5 mmol) was reduced by BH3‚THF, and the pure
amine was obtained as a colorless oil (0.35 g, 83%) after bulb-
to-bulb distillation (115 °C, 0.45 mmHg). The hydrochloride
salt was crystallized to afford well-defined colorless flakes: mp
237-239 °C dec (lit.52 237-238 °C); IR (KBr, HCl salt) 2930,
2800, 1570, 1440, 1390, 875, 750 cm-1; 1H NMR (CDCl3) δ 7.54
(d, 1 H, J ) 7.8 Hz, H-6), 7.21-7.11 (m, 3 H, ArH), 4.10 (s, 2
H, H-1), 3.36-3.33 (m, 2 H, H-3), 2.74-2.70 (m, 2 H, H-4),
1.57 (bs, e, 1 H, NH); 13C NMR (CDCl3) δ 146.5, 136.6, 132.4,
129.1, 127.5, 126.9, 55.1, 52.8, 36.3. Anal. (C9H11NS‚HCl) C,
H, N.
(R)-(-)-3-Met h yl-1,2,3,4-t et r a h yd r oisoq u in olin e ((R)-
(-)-26). Using the procedure described above for the race-
mate, (R)-(-)-26 was synthesized from (2R)-(-)-phenyl-2-
aminopropane:33 mp (HCl) 273-275 °C dec; [R]23 ) -75° (c
D
) 0.40, MeOH). Anal. (C10H13N‚HCl) C, H, N.
(S)-(+)-3-Met h yl-1,2,3,4-t et r a h yd r oisoq u in olin e ((S)-
(+)-26). Using the procedure described above for the race-
mate, (S)-(+)-26 was synthesized from (2S)-(+)-phenyl-
2-aminopropane: mp (HCl) 270-273 °C dec (lit.32 mp 276-
277.5 °C); [R]23 ) 78° (c ) 0.40, MeOH) (lit.32 [R]20 ) 76° (c
D
D
) 0.92, EtOH)). Anal. (C10H13N‚HCl) C, H, N.
Ra d ioch em ica l Assa y for P NMT Activity. The assay
used in this study has been described elsewhere.35 Briefly, a
typical assay mixture consisted of 50 µL of 0.5 M phosphate
buffer (pH 8.0), 25 µL of 10 mM unlabeled AdoMet, 5 µL of
[methyl-3H]AdoMet, containing approximately 3 × 105 dpm
(specific activity approximately 15 mCi/mmol), 25 µL of
substrate solution (phenylethanolamine), 25 µL of inhibitor
solution, 25 µL of the enzyme preparation, and sufficient water
to achieve a final volume of 250 µL. After the mixture was
incubated for 30 min at 37 °C, the reaction was quenched by
addition of 250 µL of 0.5 M borate buffer (pH 10.0), and the
mixture was extracted with 2 mL of toluene-isoamyl alcohol

Bicyclic Benzylamine Inhibitors of PNMT
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18 3545
(8) Axelrod, J . Purification and Properties of Phenylethanolamine
N-Methyltransferase. J . Biol. Chem. 1962, 237, 1657-1660.
(9) Pendleton, R. G.; Snow, I. B. The Binding Order of Substrates
to Phenylethanolamine N-Methyltransferase. Mol. Pharmacol.
1973, 9, 718-725.
(10) Lee, H.-S.; Schulz, A. R.; Fuller, R. W. Product Inhibition Studies
and the Reaction Sequence of Rabbit Adrenal Norepinephrine
N-Methyl Transferase Isozymes. Arch. Biochem. Biophys. 1978,
185, 239-250.
(11) Grunewald, G. L.; Monn, J . A.; Sall, D. J . Approaches to the
Design of Selective Inhibitors of Phenylethanolamine N-Methyl-
transferase: Development of an Active Site Model. In Epineph-
rine in the Central Nervous System; Stolk, J . M., U’Prichard, D.
C., Fuxe, K., Eds.; Oxford University Press: New York, 1988;
pp 117-140 and references cited therein.
(12) Hieble, J . P.; Roesler, J . M.; McCafferty, J . P.; Fujita, T.;
Pendleton, R. G. Contributions of CNS Effects to the Cardio-
vascular Effects of Phenylethanolamine N-Methyltransferase.
In Epinephrine in the Central Nervous System; Stolk, J . M.,
U’Prichard, D. C., Fuxe, K., Eds.; Oxford University Press: New
York, 1988; pp 322-329.
(13) Toomey, R. E.; Horng, J . S.; Hemrick-Luecke, S. K.; Fuller, R.
W. R2-Adrenoceptor Affinity of Some Inhibitors of Norepineph-
rine N-Methyltransferase. Life Sci. 1981, 29, 2467-2472.
(14) Bondinell, W. E.; Chapin, R. W.; Girard, G. R.; Kaiser, C.; Krog,
A. J .; Pavloff, A. M.; Schwartz, M. S.; Silvestri, J . S.; Vaidya, P.
D.; Lam, B. L.; Wellman, G. R.; Pendleton, R. G. Inhibitors of
Phenylethanolamine N-Methyltransferase and Epinephrine Bio-
synthesis. 1. Chloro-Substituted 1,2,3,4-Tetrahydroisoquinolines.
J . Med. Chem. 1980, 23, 506-511.
(15) Fuller, R. W.; Molloy, B. B.; Hemrick, S. K. Inhibition in vitro
of Rabbit Adrenal Norepinephrine N-Methyltransferase by
2,3,4,5-Tetrahydro-1H-2-benzazepines. Biochem. Pharmacol. 1979,
28, 528-530.
(7:3). A 1 mL portion of the organic layer was removed, trans-
ferred to a scintillation vial, and diluted with cocktail for
counting. The mode of inhibition was ascertained to be com-
petitive in all cases reported in Tables 1, 2, and 4 by inspection
of the 1/V vs 1/S plots of the data. All assays were run in du-
plicate with three inhibitor concentrations over a 5-fold range.
Ki values were determined by a hyperbolic fit of the data.
r2-Ad r en ocep tor Ra d ioliga n d Bin d in g Assa y. The ra-
dioligand receptor binding was performed according to the
method of U’Prichard et al.37 Male Sprague-Dawley rats were
decapitated, and the cortexes were dissected out and homog-
enized in 20 vol (w/v) of ice-cold 50 mM Tris/HCl buffer (pH
7.7 at 25 °C). Homogenates were centrifuged thrice for 10 min
at 50000g with resuspension of the pellet in fresh buffer
between spins. The final pellet was homogenized in 200 vol
(w/v) of ice-cold 50 mM Tris/HCl buffer (pH 7.7 at 25 °C).
Incubation tubes containing [3H]clonidine (specific activity ca.
19.2 mCi/mmol, final concentration 4.0 nM), various concen-
trations of drugs, and an aliquot of freshly resuspended tissue
(800 µL) in a final volume of 1 mL were used. Tubes were
incubated at 25 °C for 30 min, and the incubation was
terminated by rapid filtration under vacuum through GF/B
glass fiber filters. The filters were rinsed with three 5 mL
washes of ice-cold 50 mM Tris buffer (pH 7.7 at 25 °C). The
filters were counted in vials containing premixed scintillation
cocktail. Nonspecific binding was defined as the concentration
of bound ligand in the presence of 2 µM phentolamine. All
assays were run in quadruplicate with five inhibitor concen-
trations over a 16-fold range. IC50 values were determined
by a log-probit analysis of the data, and Ki values were
determined by the equation Ki ) IC50/(1 + [clonidine]/KD), as
all Hill coefficients were approximately equal to 1.
(16) Pendleton, R. G.; Kaiser, C.; Gessner, G. Studies on Adrenal
Phenylethanolamine N-Methyltransferase (PNMT) with SK&F
64139, a Selective Inhibitor. J . Pharmacol. Exp. Ther. 1976, 197,
623-632.
Molecu la r Mod elin g Stu d ies. Molecular modeling was
performed using the Sybyl software package (version 6.00,
Tripos Assoc., Inc., St. Louis, MO) on an IBM RS/6000 (models
560 and 350) workstation. The MNDO method in the semiem-
pirical molecular orbital package (MOPAC, version 5.0 at the
Sybyl interface) and molecular mechanics (Sybyl, Tripos force
field) were used in optimizing geometries. The FIT command
was used in fitting the molecules to template 5, and the
MVOLUME command was used to generate the volume
difference and intersection maps.
(17) Pendleton, R. G.; Gessner, G.; Weiner, G.; J enkins, B.; Sawyer,
J .; Bondinell, W.; Intoccia, A. Studies on SK&F 29661, an Organ
Specific Inhibitor of Phenylethanolamine N-Methyltransferase.
J . Pharmacol. Exp. Ther. 1979, 208, 24-30.
(18) (a) Fuller, R. W.; Hemrick-Luecke, S.; Toomey, R. E.; Horng, J .-
S.; Ruffolo, R. R., J r.; Molloy, B. B. Properties of 8,9-Dichloro-
2,3,4,5-tetrahydro-1H-2-benzazepine, an Inhibitor of Norepineph-
rine N-Methyltransferase. Biochem. Pharmacol. 1981, 30, 1345-
1352. (b) PNMT and R2 Ki’s for these literature compounds were
determined in our laboratory for consistent internal comparison.
(19) Ruffolo, R. R., J r.; DeMarinis, R. M.; Wise, M.; Hieble, J . P.
Structure-Activity Relationships for Alpha-2 Adrenergic Recep-
tor Agonists and Antagonists. In The Alpha-2 Adrenergic Recep-
tors; Limbird, L. E., Ed.; Humana Press: Clifton, NJ , 1988; pp
115-186.
(20) Fuller, R. W. Pharmacology of Brain Epinephrine Neurons.
Annu. Rev. Pharmacol. Toxicol. 1982, 22, 31-55.
(21) Fuller, R. W.; Molloy, B. B.; Day, W. A.; Roush, B. W.; Marsh,
M. M. Inhibition of Phenylethanolamine N-Methyltransferase
by Benzylamines. 1. Structure-Activity Relationships. J . Med.
Chem. 1973, 16, 101-106.
Ack n ow led gm en t. This research was supported by
USPHS Grant HL 34193. We would also like to thank
Mr. Christopher Gunn, supervisor of the KU Molecular
Modeling Lab, for his assistance with the Sybyl molec-
ular modeling calculations.
(22) Menon, S.; Grunewald, G. L. Determination of Ionization
Constants of Benzylamines by Potentiometry. Presented at the
25th Midwest Regional Meeting of the American Chemical
Society, Manhattan, KS, Nov. 7-9, 1990; Abstract 20.
(23) Grunewald, G. L.; Sall, D. J .; Monn, J . A. Conformational and
Steric Aspects of the Inhibition of Phenylethanolamine N-
Methyltransferase by Benzylamines. J . Med. Chem. 1988, 31,
433-444.
(24) Dubois, R. J .; Lin, C. L.; Beisler, J . A. Synthesis and Antitumor
Properties of Some Isoindolylalkylphosphonium Salts. J . Med.
Chem. 1978, 21, 303-306.
(25) Evans, D.; Lockhart, I. M. The Schmidt Reaction with Aromatic
Ketones. J . Chem. Soc. 1965, 4806-4812.
(26) Huckle, D.; Lockhart, I. M.; Wright, M. The Preparation of Some
2,3-Dihydro-1,4-benzoxazepine-5(4H)-ones and Related Com-
pounds. J . Chem. Soc. 1965, 1137-1141.
(27) Conley, R. T. Schmidt Reactions in Polyphosphoric Acid. I.
Rearrangement of Ketones. J . Org. Chem. 1958, 23, 1330-1333.
(28) Hjelte, N. S.; Agback, T. Benzocycloalkanones in the Schmidt
Reaction. Acta Chem. Scand. 1964, 18, 191-194.
Refer en ces
(1) (a) Taken in part from the Ph.D. Dissertation of V.H.D.,
University of Kansas, 1994. (b) Summer undergraduate research
participant,Department of Medicinal Chemistry, University of
Kansas, 1991.
(2) Vogt, M. The Concentration of Sympathin in Different Parts of
the Central Nervous System Under Normal Condition and After
the Administration of Drugs. J . Physiol. 1954, 123, 451-481.
(3) Ho¨kfelt, T.; Fuxe, K.; Goldstein, M.; J ohansson, O. Evidence for
Adrenaline Neurons in the Rat Brain. Acta Physiol. Scand. 1973,
89, 286-288.
(4) Ho¨kfelt, T.; Fuxe, K.; Goldstein, M.; J ohansson, O. Immunohis-
tochemical Evidence for the Existence of Adrenaline Neurons
in the Rat Brain. Brain Res. 1974, 66, 235-251.
(5) Goldstein, M.; Lew, J . Y.; Matsumoto, Y.; Ho¨kfelt, T.; Fuxe, K.
Localization and Function of PNMT in the Central Nervous
System. In Pyschopharmacology: A Generation of Progress;
Lipton, M. A., DiMascio, A., Killam, K. F., Eds.; Raven Press:
New York, 1978; pp 261-269.
(29) Grunewald, G. L.; Dahanukar, V. H. Synthesis of 3-Alkyl-8-
substituted- and 4-Hydroxy-8-substituted-2,3,4,5-tetrahydro-1H-
2-benzazepines. J . Heterocycl. Chem. 1994, 31, 1609-1617.
(30) J ones, G. C.; Hauser, C. R. Ortho Substitution Rearrangement
of 1,1-Dimethyl-2-phenylpyrrolidinium Ion by Sodium Amide.
Ring Enlargement to Form a 2-Benzazocine. J . Org. Chem. 1962,
27, 3572-3576.
(31) Rice, K. C. An Improved Procedure for the Demethylation of 6,7-
Benzomorphan, Morphine and Codeine. J . Org. Chem. 1975, 40,
1850-1851.
(6) Perry, B. D.; Stolk, J. M.; Vantini, G.; Guchhait, R. B.; U’Prichard,
D. C. Strain Differences in Rat Brain Epinephrine Synthesis:
Regulation of R2-Adrenergic Receptor Number by Epinephrine.
Science 1983, 221, 1297-1299.
(7) Stolk, J. M.; Vantini, G.; Perry, B. D.; Guchhait, R. B.; U’Prichard,
D. C. Assessment of the Functional Role of Brain Adrenergic
Neurons: Chronic Effects of Phenylethanolamine N-Methyl-
transferase Inhibitors and Alpha Adrenergic Receptor Antago-
nists on Brain Norepinephrine Metabolism. J . Pharmacol. Exp.
Ther. 1984, 230, 577-586.

3546 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 18
Grunewald et al.
(32) Potapov, V. M.; Dem’yanovich, V. M.; Soifer, V. S.; Terent’ev, A.
P. Stereochemical Investigations XXII. Rotary Dispersions of
Some 3,4-Dihydro- and 1,2,3,4-Tetrahydroisoquinolines. J . Gen.
Chem., USSR (Engl. Transl.) 1967, 37, 2550-2555.
(33) Repke, D. B.; Bates, D. K.; Ferguson, W. J . Synthesis of
Dextroamphetamine Sulfate and Methamphetamine Hydro-
chloride from D-Phenylalanine. J . Pharm. Sci. 1978, 67, 1167-
1168.
(34) Grunewald, G. L.; Grindel, J . M.; Vincek, W. C.; Borchardt, R.
T. Importance of the Aromatic Ring in Adrenergic Amines.
Nonaromatic Analogues of Phenylethanolamine as substrates
for Phenylethanolamine N-Methyltransferase Mol. Pharmacol.
1975, 11, 694-699.
(35) Grunewald, G. L.; Borchardt, R. T.; Rafferty, M. F.; Krass, P.
Conformationally Defined Adrenergic Agents. 5. Conformational
Preferences of Amphetamine Analogues for Inhibition of Phen-
ylethanolamine N-Methyltransferase. Mol. Pharmacol. 1981, 20,
377-381.
(36) Connett, R. J .; Kirshner, N. Purification and Properties of Bovine
Phenylethanolamine N-Methyltransferase. J . Biol. Chem. 1970,
245, 329-334.
(37) U’Prichard, D. C.; Greenberg, D. A.; Synder, S. H. Binding
Characteristics of Radiolabelled Agonists and Antagonists at
Central Nervous System Alpha Noradrenergic Receptors. Mol.
Pharmacol. 1977, 13, 454-473.
(38) Li, S.; Bernstein, E. R. Supersonic J et Studies of Benzylamines.
Geometry of their Minimum Energy Conformations. Tetrahedron
Lett. 1991, 32, 3945-3948.
(39) Li, S.; Bernstein, E. R.; Seeman, J . I. Stable Conformations of
Benzylamine and N, N-Dimethylbenzylamine. J . Phys. Chem.
1992, 96, 8808-8813.
Design: The Active Analogue Approach. In Computer-Assisted
Drug Design; Olson, E. C., Christoffersen, R. E., Eds.; ACS Symp.
Series 112; American Chemical Society: Washington, DC, 1979;
pp 205-226.
(42) Riddell, F. G. The Conformational Analysis of Heterocyclic
Compounds; Academic Press: London, 1980; pp 22-32.
(43) Allen, F. H.; Kennard, O. 3D Search and Research Using the
Cambridge Structural Database. Chem. Design Automation
News 1993, 8, 1, 31-37.
(44) Ishihara, Y.; Tanaka, T.; Miwatashi, S.; Fujushima, A.; Goto,
G. Regioselective Friedel-Crafts Acylation of 2,3,4,5-tetrahydro-
1H-2-benzazepine and Related Nitrogen-heterocycles. J . Chem.
Soc., Perkin Trans. I 1994, 207, 2993-2999.
(45) Stenlake, J . B. Foundations of Molecular Pharmacology Vol. 2.
The Chemical Basis of Drug Action; The Athlone Press: London,
1979; pp 48-49.
(46) Still, W. C.; Mitra, A.; Kahn, M. Rapid Chromatographic
Technique for Preparative Separations with Moderate Resolu-
tion. J . Org. Chem. 1978, 43, 2923-2924.
(47) J ohnson, W. S.; Woroch, E. L.; Buell, B. G. Cyclization Studies
in the Quinoline Series. A New Synthesis of 4-Aminoquinolines.
J . Am. Chem. Soc. 1949, 71, 1901-1905.
(48) Wu¨nsch, K.-H.; Stahnke, K. H.; Gomoll, P. 1,2,3,4-Tetrahydro-
5H-1,4-benzodiazepinon-(5) durch Schmidt-Reaktion aus 1,2,3,4-
Tetrahydrochinolinon-(4). Z. Chem. 1970, 10, 219-220.
(49) Uskokovic, M.; Iacobelli, J .; Wenner, W. 3H-1,4-Benzodiazepine-
2,5(1H,4H)-dione and Related Compounds. J . Org. Chem. 1962,
27, 3606-3608.
(50) Bose, A. K.; Hoffman, W. A.; Manhas, M. S. Studies on â-Lac-
tams. Part 46. Synthesis of Nine-Membered Heterocycles via
â-Lactams. J . Chem. Soc., Perkin Trans. I 1976, 2343-2348.
(51) J akob, F.; Schlack, P. 5-Oxo-1-thia-4-aza-cycloheptan-Synthesen.
Chem. Ber. 1963, 96, 88-92.
(52) Wu¨nsch, K.-H.; Ehlers, A.; Beyer, H. 2,3-Dihydro-1,4-benzothi-
azepine-5(4H)-one. Z. Chem. 1967, 7, 185-186.
(40) Grunewald, G. L.; Sall, D. J .; Monn, J . A. Synthesis and
Evaluation of 3-Substituted Analogues of 1,2,3,4-Tetrahydroiso-
quinoline as Inhibitors of Phenylethanolamine N-Methyltrans-
ferase. J . Med. Chem. 1988, 31, 824-830.
(41) Marshall, G. R.; Barry, C. D.; Bosshard, H. E.; Dammkoehler,
R. A.; Dunn, D. A. The Conformational Parameters in Drug
J M9508292