Comparative molecular field analysis of the binding of the stereoisomers of fenoterol and fenoterol derivatives to the β2 adrenergic receptor

Article abstract of DOI:10.1021/jm070030d

Stereoisomers of fenoterol and six fenoterol derivatives have been synthesized and their binding affinities for the β₂ adrenergic receptor (K_iβ₂-AR), the subtype selectivity relative to the β₁-AR (K_iβ₁-AR/K _iβ₂-AR) and their functional activities were determined. Of the 26 compounds synthesized in the study, submicromolar binding affinities were observed for (R,R)-fenoterol, the (R,R)-isomer of the p-methoxy, and (R,R)- and (R,S)-isomers of 1-naphthyl derivatives and all of these compounds were active at submicromolar concentrations in cardiomyocyte contractility tests. The K_iβ₁-AR/K _iβ₂-AR ratios were >40 for (R,R)-fenoterol and the (R,R)-p-methoxy and (R,S)-1-naphthyl derivatives and 14 for the (R,R)-1-napthyl derivative. The binding data was analyzed using comparative molecular field analysis (CoMFA), and the resulting model indicated that the fenoterol derivatives interacted with two separate binding sites and one steric restricted site on the pseudo-receptor and that the chirality of the second stereogenic center affected K_iβ₂ and subtype selectivity.

Full text of DOI:10.1021/jm070030d

J. Med. Chem. 2007, 50, 2903-2915
2903
Comparative Molecular Field Analysis of the Binding of the Stereoisomers of Fenoterol and
Fenoterol Derivatives to the â2 Adrenergic Receptor
Krzysztof Jozwiak,†,‡ Chakir Khalid,†,§ Mary J. Tanga,| Ilona Berzetei-Gurske,| Lucita Jimenez,| Joseph A. Kozocas,|
Anthony Woo,§ Weizhong Zhu,§ Rui-Ping Xiao,§ Darrell R. Abernethy, and Irving W. Wainer*,
Department of Chemistry, Medical UniVersity of Lublin, Lublin, Poland, Laboratory of CardioVascular Science, National Institute on Aging
Intramural Research Program, Baltimore, Maryland, SRI International, Menlo Park, California, and Laboratory of Clinical InVestigation,
National Institute on Aging Intramural Research Program, Baltimore, Maryland
ReceiVed January 9, 2007
Stereoisomers of fenoterol and six fenoterol derivatives have been synthesized and their binding affinities
for the â2 adrenergic receptor (Kiâ2-AR), the subtype selectivity relative to the â1-AR (Kiâ1-AR/Kiâ2-AR)
and their functional activities were determined. Of the 26 compounds synthesized in the study, submicromolar
binding affinities were observed for (R,R)-fenoterol, the (R,R)-isomer of the p-methoxy, and (R,R)- and
(R,S)-isomers of 1-naphthyl derivatives and all of these compounds were active at submicromolar
concentrations in cardiomyocyte contractility tests. The Kiâ1-AR/Kiâ2-AR ratios were >40 for (R,R)-fenoterol
and the (R,R)-p-methoxy and (R,S)-1-naphthyl derivatives and 14 for the (R,R)-1-napthyl derivative. The
binding data was analyzed using comparative molecular field analysis (CoMFA), and the resulting model
indicated that the fenoterol derivatives interacted with two separate binding sites and one steric restricted
site on the pseudo-receptor and that the chirality of the second stereogenic center affected Kiâ2 and subtype
selectivity.
Introduction
Fenoterol, 5-[1-hydroxy-2-[[2-(4-hydroxyphenyl)-1-methyl-
ethyl]-amino]ethyl]-1,3-benzenediol (Figure 1, compound 1) is
a â2-adrenoceptor (â2-AR) agonist,1 which exists as four
stereoisomers. The clinically used drug, rac-fenoterol, is a
racemic mixture of (R,R)-1 and (S,S)-1, and this mixture is used
for the treatment of asthma2 and may also be useful in the
treatment of congestive heart failure.3,4
Initial ligand displacement binding studies using (R,R)-1 and
(S,S)-1 were carried out using rat erythrocytes. The data
demonstrated that (R,R)-1 displaced the marker ligand [3H]-
dihydroalprenolol with a Kd of 2.88 × 10-6 M, rac-1 had a Kd
of 5.79 × 10-5 M, and (S,S)-1 had no measurable specific
binding.5 The enantioselectivity of the â2-AR binding was
recently confirmed.6 In the latter study, (R,R)-1 and (S,S)-1 were
prepared using an amylose tris-(3,5-dimethylphenylcarbamate)
chiral stationary phase, and the binding studies were conducted
using membranes from HEK-293 cells that express human â2-
AR. The calculated Kd for (R,R)-1 was 460 nM and the Kd for
(S,S)-1 was >100 000 nM.
The pharmacological effects of (R,R)-1 and (S,S)-1 were also
examined in the latter study using cardiomyocyte contractility
in freshly isolated adult rat cardiomyocytes.6 In this model,
(R,R)-1 increased the maximum contractile response from 265
( 11.6% to 306 ( 11.8% of resting cell length (P < 0.05) and
reduced EC50 from -7.0 ( 0.27 log [M] to -7.1 ( 0.20 log
[M], while (S,S)-1 had no significant effect.
* To whom correspondence should be addressed. Irving W. Wainer,
National Institute on Aging, National Institutes of Health Gerontology
Research Center, 5600 Nathan Shock Drive, Baltimore, MD 21224-6825.
Phone: (410)-558-8498.Fax: (410)-558-8409.E-mail: wainerir@grc.nia.nih.gov.
† Equal contributors.
‡ Medical University of Lublin.
Figure 1. The structures of the stereoisomers of fenoterol and the
compounds 2-7 that were synthesized in this study.
§ Laboratory of Cardiovascular Science, National Institute on Aging
Intramural Research Program.
While the binding affinities and pharmacological effects of
(R,R)-1 and (S,S)-1 have been examined, to our knowledge there
have been no similar studies of (R,S)-1 and (S,R)-1. Therefore,
| SRI International.
Laboratory of Clinical Investigation, National Institute on Aging
Intramural Research Program.
10.1021/jm070030d CCC: $37.00 © 2007 American Chemical Society
Published on Web 05/17/2007

2904 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
Jozwiak et al.
Scheme 1
Figure 2. The structures of compounds 47-51.
the objectives of this study were to develop a synthetic approach
for the preparation of (R,S)-1 and (S,R)-1 and to determine their
respective binding affinities and pharmacological effects. In
addition, the study was designed to synthesize and characterize
a series of derivatives of 1, compounds 2-7 (Figure 1), to
determine the effect of altering the 4-hydroxyphenyl moiety and
the removing of the second chiral center. The aim of the project
was the development of a pharmacophore model, which could
be used as a structural guide for the design of new compounds
with â2-AR selectivity which can be tested for use in the
treatment of congestive heart failure.
Scheme 2
Synthetic Approach. The key step in the synthesis of the
four stereoisomers of 1-6 was the coupling of the epoxide
formed from either (R)- or (S)-3′,5′-dibenzyloxyphenylbromo-
hydrin with the (R)- or (S)-enantiomer of the appropriate benzyl-
protected 2-amino-3-benzylpropane (1-5) or the (R)- or (S)-
enantiomer of N-benzyl-2-aminoheptane (6), Scheme 1. The
synthesis of (R)-7 and (S)-7 was accomplished using 2-phen-
ethylamine, Scheme 2. This approach was similar to the one
developed by Trofast et al.7 for the synthesis of the stereoisomers
of formoterol, compound 47, Figure 2. The resulting compounds
were then deprotected by hydrogenation over Pd/C and purified
as the fumarate salts.
Results
Membrane Binding Studies. Binding affinities, expressed
as Ki values, were determined using membranes obtained from
a HEK 293 cell line stably transfected with cDNA encoding
human â2-AR8 with [3H]CGP-12177 as the marker ligand. The
resulting IC50 values and Hill coefficients were calculated for
each test compound using Prism software and Ki values were
calculated using the Cheng-Prusoff transformation9
The chiral building blocks used in the syntheses were
produced using Scheme 3. The (R)- and (S)-3′,5′-dibenzylox-
yphenyl-bromohydrin enantiomers were obtained by the enan-
tiospecific reduction of 3,5-dibenzyloxy-R-bromoacetophenone
using boron-methyl sulfide complex (BH3SCH3) and either
(1R,2S)- or (1S,2R)-cis-1-amino-2-indanol. The required (R)-
and (S)-2-benzylaminopropanes were prepared by enantiose-
lective crystallization of the rac-2-benzylaminopropanes using
either (R)- or (S)-mandelic acid as the counterion.
Ki ) IC50/(1 + L/Kd)
(1)
where L is the concentration of [3H]CGP-12177 and Kd is the
binding affinity of the [3H]CGP-12177. Each test compound
was assayed three times.

Binding of the Stereoisomers of Fenoterol
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2905
Scheme 3
consistent with the established enantioselective binding prefer-
ence for â2-ARs, with the R-configuration at the stereogenic
center containing the â-OH moiety, compare refs 10-13.
When just the R,R-isomers were compared, (R,R)-5 had the
highest relative affinity of the tested compounds, although the
difference between (R,R)-5 and (R,R)-1 did not reach statistical
significance, Table 1. The only other (R,R)-stereoisomer with
submicromolar affinity was (R,R)-2, which had a significantly
lower binding affinity than (R,R)-5, p ) 0.0051, and (R,R)-1, p
) 0.0291, although the mean Kd value for (R,R)-2 is only 23%
greater than that of (R,R)-1. The minimal effect of transforming
the p-OH moiety into a methyl ether is consistent with previous
data from Schirrmacher et al.14 In the previous study, rac-1 was
converted into a [18F]-fluoroethoxy ether without significant loss
of in vitro activity, and the authors concluded that, within the
accuracy of the experimental measurements, the derivatization
did not change the binding affinity of the rac-1 to the â2-AR.
Binding affinities, expressed as Ki values, for the â1-AR were
determined using rat cortical membranes, with [3H]-CGP-12177
as the marker ligand.15 The calculated Ki for (R,R)-5 was 3.35
µM and the binding affinities for the all of the remaining test
compounds were >100 µM, Table 1. Unlike the data from the
â2-AR binding studies, there was no clear trend that could be
associated with the stereochemistry of the compounds.
The relative selectivity of the compounds for the â2-AR and
â1-AR was determined using the ratio Kiâ1/Kiâ2, Table 1. Of
particular interest were the ratios for the four compounds with
submicromolar affinity for the â2-AR, (R,R)-1, (R,R)-2, (R,R)-
5, and (R,S)-5, which were 46, 43, 14, and 46, respectively.
The results for (R,R)-1 and (R,R)-2 are consistent with the
previously reported Kiâ1/Kiâ2 ratio of 53 for the â2-AR-selective
agonist (R,R)-TA-2005, compound 49, Figure 2.
Table 1. Binding Affinities to the â2-AR of the Compounds
Synthesized in this Study Calculated as Ki ( SEM (µM), n ) 3.
Comparison of â1- and â2-Adrenergic Binding Affinity of Fenoterol
Isomers
cmpd
Kiâ1
Kiâ2
Kiâ1/Kiâ2
The observed loss of â2-AR selectivity for (R,R)-5 was
unexpected, as was the 3-fold increase in selectivity displayed
by (R,S)-5 relative to (R,R)-5. Previous studies with the
stereoisomers of 47 indicated that both the (R,R)-isomer and
the (R,S)-isomer had a high degree of selectivity for the â2-AR
relative to the â1-AR, with the selectivity of the (R,R)-isomer
greater than that of the (R,S)-isomer.7 This is the case for
compounds 1 and 2, but reversed for 5. It is also interesting to
note that (S,R)-5 had a similar selectivity (19-fold) and its
affinity for the â2-AR was only 7-fold weaker than (R,R)-5,
1.78 µM and 0.24 µM, respectively.
Cardiomyocyte Contractility. The pharmacological activities
of the compounds with submicromolar affinity for the â2-AR
and (R,S)-1 and (S,S)-1 were assessed using cardiomyocyte
contractility, as previously described.6 In these studies, the
contraction amplitude indexed by electrical pacing induced
shortening of cell length was measured in single ventricular
myocytes before and after exposure to a single dose of the test
compounds. Contractile response to the agonist was expressed
as a percentage of the basal contractility, and the specificity of
the agonist toward â2-adrenergic receptor was determined by
the inhibitory effect of ICI 118 551 (10-7 mol/L), a selective
â2-AR antagonist.
(R,R)-1
(R,S)-1
(S,R)-1
(S,S)-1
(R,R)-2
(R,S)-2
(S,R)-2
(S,S)-2
(R,R)-3
(R,S)-3
(S,R)-3
(S,S)-3
(R,R)-4
(R,S)-4
(S,R)-4
(S,S)-4
(R,R)-5
(R,S)-5
(S,R)-5
(S,S)-5
(R,R)-6
(R,S)-6
(S,R)-6
(S,S)-6
(R) -7
14.8 ( 2.5
18.9 ( 2.4
>100
0.35 ( 0.03
3.69 ( 0.25
10.3 ( 1.4
27.8 ( 6.8
0.47 ( 0.04
1.93 ( 0.14
5.27 ( 0.51
15.9 ( 2.7
2.93 ( 0.17
7.94 ( 0.39
23.1 ( 2.1
28.6 ( 0.9
1.86 ( 0.18
6.04 ( 0.43
30.8 ( 3.3
28.8 ( 1.8
0.24 ( 0.04
0.34 ( 0.02
1.78 ( 0.15
2.54 ( 0.21
9.28 ( 0.90
31.4 ( 1.7
>100
43
5
NC
NC
46
16
6
NC
9
4
3
1
9
4
NC
NC
14
46
19
NC
1
NC
NC
1
>100
21.9 ( 3.1
30.8 ( 6.5
33.4 ( 9.2
>100
24.9 ( 2.1
31.3 ( 3.5
77.5 ( 3.6
31.4 ( 1.7
17.2 ( 1.3
33.1 ( 2.8
>100
>100
3.35 ( 0.13
15.8 ( 6.3
34.7 ( 9.1
>100
10.2 ( 0.5
>100
61.3 ( 5.8
52.6 ( 1.4
42.5 ( 3.5
52.2 ( 3.0
56.42 ( 5.19
10.5 ( 1.5
20.6 ( 3.7
4
3
(S)-7
The relative binding affinities to the â2-AR for the stereoi-
somers of compounds 1-4 and 6 were R,R > R,S > S,R ≈ S,S
(Table 1). This stereoselectivity is consistent with the previously
reported potencies of the formoterol stereoisomers7 and results
from binding studies with the isoproterenol derivative PTFAM,
compound 48, Figure 2.10 With compound 5, no significant
difference was found between the Kd values of the R,R- and
R,S-isomers, thus, the order was R,R ) R,S > S,R > S,S. The
Kd value for (R)-7 was greater than that of (S)-7, which is
All of the compounds tested, except for (S,S)-1, produced a
significant contractile response, which was blocked by ICI
118 551, while (S,S)-1 had no observed pharmacological effect,
Figure 3. These results were consistent with the results from
the previous study6 and with the observations that for agonist
activity at the â2-AR an R-configuration is preferred at the
stereogenic center containing the â-OH moiety, compare refs
10-13. It is of interest to note that maximum effect was elicited
with 0.1 µM (R,R)-5 and (R,S)-5, while the other active

2906 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
Jozwiak et al.
Figure 3. The effect of compounds (R,R)-1, (R,S)-1, (S,S)-1, (R,R)-3, (R,R)-5, and (R,S)-5 on cell contraction in single ventricular myocytes
isolated from 2 to 4 month old rat hearts after electrical stimulation at 0.5 Hz at 23 °C. Cell length was monitored by an optical edge-tracking
method using a photodiode array with a 3 ms time resolution. The experiments were conducted with and without the addition of the selective â2-AR
antagonist ICI 118 551 (10-7 mol/L).
compounds required 0.5 µM concentrations. In addition, while
the equivalent activities of (R,R)-5 and (R,S)-5 were suggested
by the binding data, the observed activity of (R,S)-1 was
unexpected, as previous studies of the stereoisomers of 477 and
4810 indicated that the agonist activities of the (R,R)-isomers
were significantly greater than the activities of the corresponding
(R,S)-isomers. The full pharmacological profiles of the com-
pounds synthesized in this study are under investigation and
will be reported elsewhere.
Two views of the resulting CoMFA model are presented in
Figure 4A,B. For visual clarity, only two compounds, (R,R)-2
and (R,R)-5, were included in the figures. The chiral centers
are depicted as gray balls. The fields are color coded in the
following manner: green represents a region of favorable steric
(bulk) interactions, unfavorable steric interactions are denoted
by yellow, electrostatic interactions with a positive charge (or
H-bond donors) are blue, and electrostatic interactions with a
negative charge (or H-bond acceptors) are red.
3D QSAR Modeling. The compounds used in this study were
analyzed using comparative molecular field analysis (CoMFA).
This is a 3D QSAR technique applicable to the analysis of the
relative activities of stereoisomers and/or enantiomers at a
selected target, a task usually not approached using regular 2D
QSAR.16,17 The resulting 3D QSAR model shows good statisti-
cal correlation with experimental data, R2 ) 0.920 and F )
60.380, and good prediction power, as indicated by the cross-
validated R2 value (Q2) ) 0.847 and the standard error of
prediction (SEP) ) 0.309, Table 2.
In the first stage, the model was used to identify the regions
responsible for the discrimination between the stereoisomers
(Figure 4A). The CoMFA procedure produced several distinct
asymmetric regions located in close proximity of each chiral
center. The first chiral center (carrying the â-hydroxyl group)
is surrounded by an electropositive region (large blue region)
behind the molecule, Figure 4A. An electropositive region can
be associated with hydrogen bond formation and indicates
favorable donor properties or unfavorable acceptor properties
of the pseudoreceptor. In this case, the location of the electro-

Binding of the Stereoisomers of Fenoterol
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2907
Table 2. The pKi Predicted by the CoMFA Model
able (yellow) region behind the first chiral center. This is an
additional indication that the model demonstrates a preference
for the â-hydroxyl group in the R-configuration. The preference
for the R-configuration at this center is consistent with previous
models and experimental data, which demonstrated that the
R-configuration is favored for functional activity at â-AR
receptors (cf. refs 10-13).
The CoMFA model also demonstrated the effect of the second
chiral center. The preferred configuration can be derived from
the binding data, where for compounds 1-4 and 6 the (R,R)-
isomers had the higher affinities relative to their respective (R,S)-
isomers, while the Ki values for (R,R)-5 and (R,S)-5 were
equivalent, Table 1. Thus, in this model, the more active isomers
are those with the methyl moiety on the stereogenic center on
the aminoalkyl portion of the molecules pointing out of the plane
of the figure of the CoMFA model. This is depicted by a steric
disfavoring region (yellow) behind the second chiral center of
the molecule and indicates a preference for the R-configuration
at this site, Figure 4A.
derivative
pKi measured
pKi predicted
(R,R)-1
(R,S)-1
(S,R)-1
(S,S)-1
(R,R)-2
(R,S)-2
(S,R)-2
(S,S)-2
(R,R)-3
(R,S)-3
(S,R)-3
(S,S)-3
(R,R)-4
(R,S)-4
(S,R)-4
(S,S)-4
(R,R)-5
(R,S)-5
(S,R)-5
(S,S)-5
(R,R)-6
(R,S)-6
(S,R)-6
(S,S)-6
(R)-7
6.46
5.43
4.99
4.56
6.32
5.71
5.28
4.80
5.53
5.10
4.64
4.54
5.73
5.22
4.51
4.54
6.62
6.47
5.75
5.60
5.03
4.50
4.00
4.25
4.98
4.69
5.84
5.48
5.02
4.66
6.17
5.80
5.34
4.99
5.57
5.21
4.75
4.39
5.58
5.25
4.75
4.43
6.72
6.36
5.90
5.54
5.01
4.66
4.19
3.84
5.33
4.51
In this study, only the aminoalkyl portion of the fenoterol
molecule was altered and, therefore, the key CoMFA regions
are associated with this aspect of the molecule. In the resulting
analysis, all four interacting regions were identified in the
proximity of the aromatic moiety, Figure 4B, and all can be
used to generate hypotheses concerning the mode of binding
action of the studied derivatives.
(S)-7
In the model, the large electropositive region (blue) encom-
passing the area close to the -OH or -OCH3 substituents
represents H-bond donor properties of the pseudoreceptor to
these moieties. These interactions are responsible for the
relatively high binding affinities of the O-derivatives, 1 and 2,
relative to 3 and 4, and in the latter compound, the p-amino
substituent should be positively charged under the experimental
conditions.
A large electronegative region (red) and another electropos-
itive region (blue), both located parallel on two sides of the
aromatic system, most likely represent π-π or π-hydrogen bond
interactions between the â2-AR and electron-rich aromatic
moieties, such as the naphthyl ring. This is consistent with the
increased affinity of 1, 2, and 5 relative to the other compounds
examined in this study. The relative importance of this interac-
tion is suggested by the observation that the Ki values for (R,R)-5
and (R,S)-5 were equivalent to (R,R)-1 and (R,R)-2, Table 1.
Two steric regions are located close to the electrostatic
regions, and one favors (green) and the other disfavors (yellow)
bulkiness in the respective areas. This indicates that the binding
of the aminoalkyl portions of the molecules are also sterically
restricted.
Discussion
The binding of agonists and antagonists to the â2-AR has
been extensively studied using site-directed mutagenesis and
molecular modeling techniques.11-13,18-20 There is general
agreement that the binding of the “catechol” portion of an
agonist occurs within a binding area created by the transmem-
brane (TM) helices identified as TM3, TM5, and TM6. The
binding process is a sequential event that produces conforma-
tional changes leading to G-protein activation.19 A key aspect
in this process is the interaction of the hydroxyl moiety on the
chiral carbon of the agonist with the Asn-293 residue in TM6,
and for this interaction, an R-configuration is preferable at the
chiral carbon.11,13,20 Because the “catechol” portion of the
fenoterol molecule was not altered in this study, it is not
surprising that in the CoMFA model, an R-configuration at the
first stereogenic center is preferred in most stable complexes.
Figure 4. Two views of the resulting CoMFA model. For visual clarity,
only two compounds, (R,R)-2 and (R,R)-5, were included explicitly in
the figures. The chiral centers are depicted as gray balls. The fields
are color coded in the following manner: green and yellow represents
a region of steric (bulk) interactions (favorable and unfavorable,
respectively), electrostatic interactions with a positive charge (or H-bond
donor) are blue, and electrostatic interactions with a negative charge
(or H-bond acceptor) are red.
positive field indicates that the orientation of the â-OH moiety
behind the plane of the model (the S-configuration at the chiral
center) would hinder H bond formation with the receptor. The
electropositive region is closely associated with a steric unfavor-

2908 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
Jozwiak et al.
The majority of the binding and functional studies of â2-AR
agonists have been conducted with small N-alkyl substituents
such as methyl, isopropyl, and t-butyl, compare ref 18. However,
while these compounds are active at the â2-AR, they are not
subtype selective. This is illustrated by the Kiâ1/Kiâ2 ratios
determined for 49, 50, and 51 (Figure 2), which were 53, 1.7,
and 1.3, respectively.12 The removal of the p-methoxyphenyl
moiety not only reduced the selectivity, but also the affinities,
as the respective â2Ki values were 12 nM, 170 nM, and 6300
nM.12
The role that aminoalkyl substituents play in â2-AR selectivity
has been investigated using site-directed mutagenesis and
molecular modeling techniques.12,19,20 Using (R,R)-49 as the
model ligand, Kikkawa et al. determined that hydrogen bond
formation between the p-methoxy oxygen on 49 and the
hydroxyl group of tyrosine 308 (Y308) located in the extracel-
lular end of TM7 was the source of the â2-AR selectivity.12
Furse and Lybrand developed a de novo model of the â2-AR
and investigated molecular complexes of several ligands (agonist
and antagonist) with this subtype.19 Among the structures
investigated, (R,R)-49 has the same aminoalkyl substituent as
the compound 2. Examination of the (R,R)-49/â2-AR complex
revealed that the p-methoxy group oxygen of (R,R)-49 formed
a hydrogen bond with the hydroxy group of Y308, which
supports the model proposed by Kikkawa et al.12 The distance
between the two oxygen atoms bonded in the model was 3.22
Å. However, the methoxy moiety of the ligand was also located
in close proximity to three other polar residues, histidine 296
(H296) in TM6, tryptophan 109 (W109) in TM3, and asparagine
312 (N312) in TM7, each of which can interact with an aromatic
group on the aminoalkyl portion of (R,R)-49.
In the Furse and Lybrand model, the distance between the
oxygen atom of the ligand and the hydrogen atom of H296 was
5.88 Å, and H296 was proposed as an alternative hydrogen bond
donor for interaction with the methoxy group of (R,R)-49.
Because Y308 and H296 are found only in â2-AR, with the
corresponding residues found in the â1-AR are F359 and K347,
respectively, the interaction with H296 and Y308 has been
proposed as the source of â1/â2 selectivity.19
Because the previous studies of â1/â2 selectivity utilized
(R,R)-49, the subtype selectivity of the (R,R)-stereoisomers of
the compounds synthesized in our study were compared to the
subtype selectivity of (R,R)-49. The data from this study support
the hypothesis that hydrogen bond formation between Y308 and/
or H296 and the oxygen atom on the p-substituent of the agonist
is involved in â2-AR selectivity. The interaction is possible with
(R,R)-1 and (R,R)-2 and the Kiâ1/Kiâ2 ratios for these compounds
are 43 and 46, respectively, which are comparable to the Kiâ1/
Kiâ2 ratio of 53 determined for (R,R)-49. The Kiâ1/Kiâ2 ratios
for 3, 4, 6, and 7 were <10 and reflect the fact that they do not
have the ability to form hydrogen bonds with Y308 or H296.
The hydrogen bonding interactions were also suggested by the
CoMFA model, which identified a large electropositive region
(blue) surrounding the area close to the -OH or -OCH3
substituents, representing hydrogen-bond donor properties of
the pseudoreceptor, Figure 4.
The data from this study also suggest that an aromatic moiety
on the aminoalkyl portion of the compound contributes to Ki
and subtype selectivity, even if the aromatic moiety is unable
to form a hydrogen bond with the receptor. This is demonstrated
by the comparison of the Kiâ2 values for the (R,R)-isomers of
1-5, which were <3.00 µM with Kiâ2 value of (R,R)-6, which
was 9.00 µM, and the Kiâ1/Kiâ2 ratios, which were g9 for 1-5,
while 6 displayed no subtype selectivity, Table 1. One possible
mechanism to explain the data is π-hydrogen bond formation.
The cloud of π-electrons of aromatic rings can act as hydrogen
bond acceptors, although it has been estimated that the interac-
tion would be about half as strong as a normal hydrogen bond.21
The higher affinity and subtype selectivity for (R,R)-5 relative
to (R,R)-3 and (R,R)-4 or (R)-7 is consistent with the greater
π-electron distribution in the napthyl ring relative to the other
aromatic rings.
The CoMFA model also identified a large electronegative
region (red) and another electropositive region (blue), both
located parallel to the aromatic system, which are most likely
associated with π-π or π-hydrogen bond interactions between
the â2-AR and electron-rich aromatic moieties, such as the
naphthyl ring. Using the model developed by Furse and Lybrand,
with (R,R)-49 as the interacting ligand, Y308, H296, W109,
and N312 were identified as possible sources of π-π and/or
π-hydrogen bond interactions. In the â2-AR model, the
estimated distances between the p-methoxy moiety on (R,R)-
49 and W109 and N312 were 4.80 Å and 3.45 Å, respectively.
Because W109 and N312 are fully conserved in all â-AR
subtypes, the interactions suggested by the CoMFA model may
represent the source of the increase affinities for (R,R)-1, (R,R)-
2, and (R,R)-5, relative to the other (R,R)-isomers, but not the
observed â1/â2 selectivity.
The data from this study and the resulting CoMFA model
indicate that the binding process of the compounds in this study
with the â2-AR includes the interaction of the chiral center on
the aminoalkyl portion of the agonist with a sterically restricted
site on the receptor. The existence of a sterically restricted site
has been previously suggested from the data obtained in the
development of 3D models for agonist and antagonist complexes
with the â2-AR.18 In these studies, it was observed that for (R,R)-
49 and similar compounds, substituents larger than a methyl
group at the stereogenic center on the aminoalkyl portion would
likely produce significant steric interactions that would unfavor-
ably affect the ligand-receptor complexes.
The binding of an agonist to the â2-AR has been described
as a multistep interrelated process, in which sequential interac-
tions between the agonist and the receptor produce correspond-
ing conformational changes.22 The CoMFA model reflects the
final agonist/â2-AR complex, and to discern the effect of the
steric restricted site, it is necessary to consider the effect that
interaction with this site has on the outcome of the binding
process. If one assumes that the interaction of the “catechol”
portion of the agonist with the binding area created by TM3,
TM5, and TM6 (the first binding area), then these interactions
will fix the position of the aminoalkyl portion of the agonist
relative to the steric restricted site and perhaps even create this
site. In the CoMFA model, the steric restrictions at the site force
the methyl moiety at the chiral center of the aminoalkyl portion
to point out of the plane of the model, Figure 4A.
It is important to note that due to the free rotation about the
N-atom, the configuration at the chiral center bearing the methyl
moiety should not affect the ability of the molecule to minimize
the interaction with the steric restricted site. However, in the
minimum energy conformation, that is, with the methyl group
pointing out of the plane of the CoMFA model, the orientation
of the remaining segment of the aminoalkyl portion relative to
the second binding area would be affected by the stereochem-
istry. Indeed, R- and S-configurations would produce mirror
image relationships to the second binding area. This situation
is illustrated in Figure 5, where the catechol, first chiral center,
and the methyl moieties of (R,R)-5 and (R,S)-5 have been
overlaid upon each other.

Binding of the Stereoisomers of Fenoterol
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2909
and subtype selectivity suggest the following: (1) the Kiâ2-AR
values represent the sum total of the π-hydrogen bond and π-π
interactions between the 1-naphthyl moieties and Y308 and
H296, as well as additional non-â2-AR specific interactions with
other residues such as W109 and N312; (2) the steric volume
swept out by (R,S)-5 increases the probability of interactions
of Y308 and H296 with the π cloud of the naphthyl moiety
relative to the (R,R)-5; (3) the steric volume swept out by (R,R)-5
increases the probability of interactions with non-â2-AR specific
sites relative to (R,S)-5.
The effect of the configuration at the second chiral center
and conformational-based chiral selectivity is also illustrated
by the affinities and subtype selectivities of (R,R)-3, (R,S)-3,
and (R)-7, Table 1. The inversion of the chirality at the second
chiral carbon from R to S reduced the Kiâ2 value of the (R,S)-3
/â2-AR complex relative to the (R,R)-3/â2-AR complex by ∼3-
fold, while there was no significant difference between their
Kiâ1 values. The increased subtype selectivity observed for
(R,R)-3 relative to (R,S)-3, 9 versus 4, respectively, essentially
reflects the differences in Kiâ2 values, which could be a
reflection of increased conformational energy required to bring
the aromatic portion of the aminoalkyl chain into contact with
the electropositive and electronegative regions that comprise
the second binding area or a decrease in the probability that
this interaction would occur.
The removal of the methyl moiety on the second chiral center,
and thereby the chirality at this site, that is, (R)-7, had essentially
the same effect as inverting the chirality at this site from R to
S. The Kiâ2 values for (R)-7 was 32% higher than (R,S)-3, and
there was no difference in the â2-AR selectivity, Table 1. These
results suggest that for compound 3 the primary effect of the
R-configuration at the second chiral site was to direct the
aminoalkyl chain toward the second binding area, which
increased the probability of interacting with this site and reduced
the conformational energy required to achieve this interaction.
The primary difference between compounds 3 and 5 is the
steric areas swept out by the aromatic substituents. In the case
of 3, the phenyl ring produces a smaller, more linear area, while
with 5, the 1-naphthyl ring system produces a relatively larger
and broader area. These differences will be used to guide the
synthesis of additional derivatives, which will be used in a more
extensive study of the binding of these derivatives to molecular
models of the â2-AR and â1-AR. The results of these studies
will be reported elsewhere.
Figure 5. The superimposition of the low-energy conformations for
two stereoisomers, (R,R)-5 and (R,S)-5. The methyl moiety attached
to the second stereogenic center is colored orange, the naphtyl moieties
of (R,R)- and (R,S)-configurations are green and magenta, respectively;
the rest of the molecule is atom type color coded: oxygen, red; nitrogen,
blue; carbon, cyan; and hydrogen, white. Note: These conformations
are for comparison purpose only and were not intended to recreate the
binding conformations within the binding site of the receptor.
The studies elucidating the source of â2-AR selectivity have
primarily utilized (R,R)-49 and one previous study of the effect
of chirality on subtype selectivity reported that (R,R)-47 had a
higher â2-AR selectivity than (R,S)-47.7 Thus, the observed
equivalent affinities and functional activities of (R,R)-5 and
(R,S)-5 at the â2-AR and the 3-fold increased â2-AR selectivity
of (R,S)-5 was an unexpected result. One possible explanation
of these results is that the naphthyl moiety of (R,S)-5 does not
interact with the site defined by Y308 and H296 and is directed
toward and binds to another site on the â2-AR. This interaction
also conveys or participates in subtype selectivity as well as
increased binding affinity and agonist activity. Because the
previous models of â2-AR selectivity only employed (R,R)-
isomers, it is possible that this site has been overlooked. This
possibility is currently under investigation in our laboratory.
Another explanation of the data is suggested by the “rocking
tetrahedron” chiral recognition mechanism proposed by Sokolov
and Zefirov.23 In this approach to molecular chiral recognition,
the enantiomeric ligands are secured to a chiral selector by two
binding interactions. The interactions must be nonequivalent and
directional so that only one orientation is possible. The tethered
enantiomers still have conformational mobility, and the remain-
ing moieties on the chiral center will sweep out overlapping
but not identical steric volumes. Where and to what extent the
chiral selector interacts with these steric volumes determines
the enantioselectivity of the process. If the chirality of the chiral
selector places the interaction perpendicular to the plane of the
ligand, no enantioselectivity is observed. As a deviation from
the perpendicular increases, so does the enantioselectivity
relative to the R- or S-configuration.
Based upon the results of this study, (R,R)-2 and (R,S)-5 have
been chosen as possible candidates for the development of a
new selective â2-AR agonist. These compounds may have
increased and extended systemic exposures relative to the
commercially available rac-1 due to changes in molecular
hydrophobicity, metabolic profile, and transporter interactions.
The pharmacological and clinical properties of these compounds
are under investigation, and the results will be reported
elsewhere.
Experimental Section
With (R,R)-5 and (R,S)-5, the interactions with the first
binding area and the steric restricted site of the CoMFA model
are two nonequivalent and directional interactions that place
the remaining constituents on the second chiral center in the
same, albeit mirror image, orientation relative to the second
binding area, Figure 5. As discussed above, it is reasonable to
assume that the interactions of the 1-naphthyl moieties of 5 with
Y308 and H296 are the source of the observed â2-AR selectivity.
If we assume that the 1-naphthyl rings sweep out overlapping
but not identical steric volumes, then the observed Kiâ2 values
Membrane Binding Studies. General Procedures. Compounds
were tested up to three times each to determine their binding
affinities at the â1- and â2-adrenergic receptors. Competition curves
with standard and unknown compounds included at least six
concentrations (in triplicate). For each compound, graphs were
prepared containing individual competition curves obtained for that
test compound. IC50 values and Hill coefficients were calculated
using Prism software. Ki values were calculated using the Cheng-
Prusoff transformation.9 In each experiment, a standard compound
was simultaneously run on the 96-well plate. If the standard

2910 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
Jozwiak et al.
compound did not have an IC50 value close to the established
average for that compound, the entire experiment was discarded
and repeated again.
electrostatic fields account for 48.1% of explained variance, and
steric fields account for 51.9%.
Chemistry. All reactions were carried out using commercial
grade reagents and solvents. THF was dried by refluxing over
sodium and benzophenone. Dichloromethane was dried by refluxing
over calcium hydride. Ultraviolet spectra were recorded on a Cary
50 Concentration spectrophotometer. Optical rotations were done
on a Rudolph Research Autopol IV. NMR Spectra were recorded
on a Varian Mercury VMX 300-MHz spectrophotometer using
tetramethylsilane as the internal standard. In reporting the NMR
multiplicities, we used the following abbreviations: s, singlet; d,
doublet; t, triplet; q, quartet; p, pentet; m, multiplet; and br, broad.
Low-resolution mass spectra were obtained on a Finnigan LCQDuo
LC MS/MS atmospheric pressure chemical ionization (API) quad-
rupole ion trap MS system equipped with both electrospray (ESI)
and atmospheric pressure chemical ionization (APCI) probes.
Analytical HPLC data was obtained using a Waters 2690 Separa-
tions Module with PDA detection. Method (a): ThermoHypersil
BDS 100 × 4.6 mm C18 column, H2O/CH3CN/TFA. Method (b):
Brownlee Phenyl Spheri-5 100 × 4.6 mm, water/acetonitrile/TFA.
Method (c): Vydac 150 × 4 mm C18 column, H2O /isopropanol/
TFA. Method (d): Chiralpak AD-H 250 × 10 mm, 95/5/0.05 CH3-
CN /isopropanol/diethylamine. Merck silica gel (230-400 mesh)
was used for open column chromatography.
â1-Adrenergic Receptors. â1-Adrenergic receptor binding was
done on rat cortical membrane following a previously described
procedure.15 In brief, male Sprague-Dawley rats weighing 250-
350 g were decapitated, and their brains were quickly removed.
The cerebral cortices were dissected on ice, weighed, and promptly
transferred to a 50 mL test tube containing approximately 30 mL
of 50 mM Tris-HCl, pH 7.8 (at room temperature). The tissues
were homogenized with a polytron and centrifuged at 20 000 × g
for 12 min at 4 °C. The pellet was washed again in the same manner
and resuspended at a concentration of 20 mg (original wet wt) per
1 mL in the assay buffer (i.e., 20 mM Tris-HCl, 10 mM MgCl2, 1
mM EDTA, 0.1 mM ascorbic acid at pH 7.8). To block the â2
sites present in the cortical membrane preparation, 30 nM ICI 118-
551 was also added to the assay buffer. To wells containing 100
µL of the test drug and 100 µL of [3H]CGP-12177 (1.4 nM final
concn), 0.8 mL of tissue homogenate was added. After 2 h at 25
°C, the incubation was terminated by rapid filtration. Nonspecific
binding was determined by 10 µM propranolol.
â2-Adrenergic Receptors. HEK 293 cells stably transfected with
cDNA encoding human â2-AR (provided by Dr. Brian Kobilka,
Stanford Medical Center, Palo Alto, CA) were grown in Dulbecco’s
modified Eagle medium (DMEM) containing 10% fetal bovine
serum (FBS), 0.05% penicillin-streptomycin, and 400 µg/mL G418
as previously described.8 The cells were scraped from the 150 ×
25 mm plates and centrifuged at 500 × g for 5 min. The pellet
was homogenized in 50 mM Tris-HCl, pH 7.7, with a Polytron,
centrifuged at 27 000 × g, and resuspended in the same buffer.
The latter process was repeated, and the pellet was resuspended in
25 mM Tris-HCl containing 120 mM NaCl, 5.4 mM KCl, 1.8 mM
CaCl2, 0.8 mM MgCl2, and 5 mM glucose, pH 7.4. The binding
assays contained 0.3 nM [3H]CGP-12177 in a volume of 1.0 mL.
Nonspecific binding was determined by 1 µM propranolol.
Cardiomyocyte Contractility. These experiments were carried
out using a previously described approach.6 In brief, single
ventricular myocytes were isolated from 2 to 4 month old rat hearts
by a standard enzymatic technique. The isolated cells were
resuspended in HEPES buffer solution [20 mM, pH 7.4] containing
NaCl (137 mM), KCl (5.4 mM), MgCl2 (1.2 mM), NaH2PO4 (1.0
mM), CaCl2 (1.0 mM), and glucose (20 mM). All experiments were
performed within 8 h of cell isolation.
The cells were placed on the stage of an inverted microscope
(Zeiss model IM-35, Zeiss, Thornwood, NY), perfused with the
HEPES-buffered solution at a flow rate of 1.8 mL/min and
electrically stimulated at 0.5 Hz at 23 °C. Cell length was monitored
by an optical edge-tracking method using a photodiode array (model
1024 SAQ, Reticon, Boston, MA) with a 3 ms time resolution.
Cell contraction was measured by the percent shortening of cell
length following electrical stimulation.
Comparative Molecular Field Analysis (CoMFA). In the
present work, CoMFA was performed as implemented in Sybyl
7.2. (TRIPOS Inc., St. Louis, MO). Molecular models of all
derivatives were prepared in HyperChem v. 6.03 (HyperCube Inc.,
Gainesville, FL) using ModelBuild procedure to ensure the same
conformation of the scaffold. Then the models were extracted to
Sybyl, and the partial atomic charges (Gasteiger-Huckel type) were
calculated. The models of ligands were aligned using as a common
substructure of the two asymmetric carbon atoms in the core of
the fenoterol molecule (-C*-CH2-NH-C*-CH2-). Next, two
types of molecular fields (steric and electrostatic) were sampled
on the grid (2 Å spacing) lattice surrounding each structure.
Distance-dependent dielectric constant was used in electrostatic
calculations and energetic cutoffs of 30 kcal/mol for both the steric
and the electrostatic energies were set.
3′,5′-Dibenzyloxy-R-bromoacetophenone (46). A solution of
2.4 mL (46 mmol) of Br2 in 45 mL of CHCl3 was added dropwise
over 1 h to a chilled, stirring solution of 9.66 g (29 mmol) of 3′,5′-
dibenzyloxyacetophenone 45 in 40 mL of CHCl3. The resulting
solution was allowed to warm to rt over 1 h with good stirring,
then poured into 100 mL of cold H2O and transferred to a separatory
funnel, where the CHCl3 fraction was isolated, washed with brine
solution, dried (Na2SO4), filtered, and concentrated to 10.8 g. This
material was applied to 500 g of silica gel, eluting with CHCl3 to
1
obtain 2.65 g (22%) of 46 as a white solid. H NMR (CDCl3) δ
4.39 (s, 2H), 5.08 (s, 4H), 6.85 (t, 1H, J ) 2.1 Hz), 7.20 (d, 2H, J
) 2.4 Hz), 7.31-7.44 (m, 10H).
General Procedure for the Enantioselective Reduction of 46
to 3′,5′-Dibenzyloxyphenylbromohydrins [(R)-8, (S)-8]. Under
an argon atmosphere, ∼0.06 mL (0.316 mmol, 10 mol %) of 5.0
M boron-methyl sulfide complex (BH3SCH3)in diethyl ether was
added in one portion to a solution of 25 mg (0.16 mmol, 5 mol %)
of the appropriate cis-1-amino-2-indanol in 3 mL of dry THF. This
material under argon was added over 30 min to a solution of 1.3 g
(3.16 mmol) of 3′,5′-dibenzyloxy-R-bromoacetophenone 46 in 20
mL of dry THF, while at the same time adding in ∼0.05 mL pulses
0.45 mL of 5.0 M boron-methyl sulfide complex. The resulting
solution was allowed to stir under argon for 2 h and then quenched
with 3 mL of methanol, controlling gas evolution. Solvents were
removed in vacuo, and the resulting residue was taken up in 30
mL of CHCl3 and washed with 25 mL of 0.2 M sulfuric acid
followed by 20 mL of brine and then dried (Na2SO4), filtered, and
evaporated.
(R)-(-)-3′,5′-Dibenzyloxyphenylbromohydrin [(R)-8]. (R)-8
was prepared with (1R,2S)-(+)-cis-1-amino-2-indanol as the enan-
tioselective reduction catalyst to give 1.02 g (78%) of (R)-8 as a
1
fine white powder. H NMR (CDCl3) δ 3.44 (dd, 1H, J ) 9.0,
10.5 Hz), 3.55 (dd, 1H, J ) 3.3, 10.5 Hz), 4.79 (dd, 1H, J ) 3.3,
8.7 Hz), 4.97 (s, 4H), 6.51 (t, 1H, J ) 2.4 Hz), 6.57 (d, 2H, J )
1.8 Hz), 7.21-7.38 (m, 10H); [R]D ) -12.1° (c 1.0, MeOH).
(S)-(+)-3′,5′-Dibenzyloxyphenylbromohydrin [(S)-8]. (S)-8
was prepared with (1S,2R)-(-)-cis-1-amino-2-indanol as the enan-
tioselective reduction catalyst to give 1.07 g (82%) of (S)-8 as a
1
fine white powder. H NMR (CDCl3) δ 3.43 (dd, 1H, J ) 9.0,
10.5 Hz), 3.55 (dd, 1H, J ) 3.3, 10.5 Hz), 4.78 (dd, 1H, J ) 3.3,
8.7 Hz), 4.96 (s, 4H), 6.50 (t, 1H, J ) 2.4 Hz), 6.57 (d, 2H, J )
1.8 Hz), 7.21-7.39 (m, 10H); [R]D ) +11.8° (c 0.90, MeOH).
4-Benzyloxyphenylacetone (34). To 10.0 g (41.3 mmol) of
4-benzyloxyphenylacetic acid 31 (Aldrich) was added 20 mL of
acetic anhydride and 20 mL of pyridine, which was heated to reflux
with stirring under an argon atm for 6 h. Solvents were evaporated,
and the residue was dissolved in CHCl3 (50 mL) and washed with
The partial least-square correlation procedure applied for resultant
database extracted four statistically significant components, and the
following validation parameters were obtained for the best solu-
tion: R2 ) 0.920, F (4,21) ) 60.380, standard error of estimate )
0.223, and cross-validated (leave-one-out) R2 ) 0.847. In general,

Binding of the Stereoisomers of Fenoterol
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2911
1 N NaOH (2 × 50 mL). The organic layer was dried (MgSO4),
filtered, and evaporated to 11.8 g of an amber oil. Vacuum
distillation at 0.1 mmHg in an oil bath set to 170 °C, followed by
silica gel chromatography eluting with 8/2 CH2Cl2-hexanes, gave
2.68 g (27%). 1H NMR (CDCl3) δ 2.14 (s, 3H), 3.63 (s, 2H), 5.05
(s, 2H), 6.94 (d, 2H, J ) 8.7 Hz), 7.10 (d, 2H, J ) 8.7 Hz), 7.26-
7.47 (m, 5H).
appropriate optically active mandelic acid (1 equiv) in methanol
(c ) 0.5 M) and refluxed until the solution homogenized, then
cooled to rt. The crystals were filtered, collected, and recrystallized
twice from methanol (c ) 0.3 M) to afford the optically active
2-benzylaminopropane‚mandelic acid salt. The salts were converted
to the free amine for the purpose of collecting NMR and rotation
data by partitioning the mandelic acid salt between 10% K2CO3
and CHCl3, drying organic extracts (Na2SO4), and evaporating.
Phenylacetone (35). A solution of 20.4 g (0.15 mol) of
phenylacetic acid 32 (Aldrich), acetic anhydride (70 mL), and
pyridine (70 mL) was heated to reflux with stirring under an argon
atm for 6 h. Solvents were evaporated, the residue was dissolved
in CHCl3 (100 mL) and washed with 1 N NaOH (2 × 100 mL),
and the organic layer was dried (MgSO4), filtered, and evaporated
to give 20.4 g. Vacuum distillation at 0.1 mmHg in an oil bath set
to 160 °C, followed by silica gel chromatography eluting with 1/1
(R)-(-)-1-(4′-Benzyloxy)-2-benzylaminopropane [(R)-10]. A
sample of 2.13 g (6.42 mmol) of 1-(4-benzyloxy)-2-benzylamino-
propane 37 was reacted with 972 mg (6.42 mmol) of (R)-(-)-
mandelic acid to give 295 mg (28% based on enantiomeric
1
abundance) of the free amine after workup. H NMR (CDCl3) δ
1.12 (d, 3H, J ) 6.3 Hz), 2.58-2.78 (m, 2H), 2.82-2.91 (m, 1H),
3.75 (dd, 2H, J ) 12 Hz, J ) 30 Hz)), 5.07 (s, 2H), 6.93 (d, 2H,
J ) 8.7 Hz), 7.10 (d, 2H, J ) 8.7 Hz), 7.21-7.42 (m, 10H); MS
(APCI+) m/z (rel) 332 (100); [R]D ) -19.1° (c 1.4, MeOH).
1
hexanes/CH2Cl2, gave 5.5 g (27%). H (CDCl3) δ 2.15 (s, 3H),
3.70 (s, 2H), 7.20-7.36 (m, 5H).
1-Naphthalen-1-yl-propan-2-one (36). A solution of 37.2 g (20
mmol) of naphthoic acid 33 (Aldrich), acetic anhydride (100 mL),
and pyridine (100 mL) was heated to reflux with stirring under an
argon atm for 6 h. The solvents were evaporated, the residue was
dissolved in CHCl3 (200 mL) and washed with 1 N NaOH (2 ×
150 mL), and the organic layer was dried (MgSO4), filtered, and
evaporated to give 34.6 g. Distillation at 0.5 mmHg in an oil bath
set to 170 °C, followed silica gel chromatography eluting with 1/1
(S)-(+)-1-(4′-Benzyloxy)-2-benzylaminopropane [(S)-10]. The
washes from the separation of (R)-10 were concentrated and
partitioned between 50 mL of chloroform and 50 mL of 10% K2-
CO3 in water. The organics were washed with brine, dried (Na2-
SO4), filtered, and evaporated to give 1.70 g (5.1 mmol), which
was brought to reflux with 782 mg (5.1 mmol) of (S)-(+)-mandelic
acid, as previously described and crystallized three times to obtain
670 mg of the (S)-amine‚(S)-mandelic acid salt. This compound
was triturated in ether then partitioned between 30 mL of chloroform
and 20 mL of 10% K2CO3 in water. The organic partition was
washed with brine, dried (Na2SO4), filtered, and evaporated to give
366 mg of the free amine (33% based on enantiomeric abundance).
1H NMR (CDCl3) δ 1.10 (d, 3H, J ) 6.3 Hz), 2.58-2.78 (m, 2H),
2.82-2.91 (m, 1H), 3.76 (dd, 2H, J ) 12, 30 Hz), 5.06 (s, 2H),
6.93 (d, 2H, J ) 8.7 Hz), 7.09 (d, 2H, J ) 8.7 Hz), 7.21-7.42 (m,
10H); MS (APCI+) m/z (rel) 332 (100); [R]D ) +19.2° (c 1.5,
MeOH).
1
hexanes/CH2Cl2, gave 9.7 g (26%). H (CDCl3) δ 2.11 (s, 3H),
4.12 (s, 2H), 7.40-7.53 (m, 4H), 7.81 (d, 1H, J ) 8.4 Hz), 7.87-
7.90 (m, 2H).
General Procedure for Preparation of 2-Benzylaminopro-
panes (37-39, 42, 43). To the appropriate ketone (1 equiv) in CH2-
Cl2 (c ) 0.5 M), cooled to 0 °C, was added glacial HOAc (1 equiv),
followed by benzylamine (1 equiv) and NaBH(AcO)3 (1.4 equiv),
and the mixture was then warmed to rt and stirred under argon for
20 h. The reaction mixture was cooled (ice bath), and 10% NaOH
(5 equiv) was added dropwise and then extracted into CH2Cl2,
washed with brine, dried (Na2SO4), filtered, and evaporated.
1-(4-Benzyloxy)-2-benzylaminopropane (37). Compound 37
was prepared from 4-benzyloxy-phenylacetone 34 (2.0 g, 8.3 mmol)
to afford 2.61 g (95%) as a tan solid. 1H (CDCl3) δ 1.10 (d, 3H, J
) 6.3 Hz), 2.50-2.58 (m, 1H), 2.68-2.77 (m, 1H), 2.82-2.89
(m, 1H), 3.75 (dd, 2H, J ) 12 Hz, J ) 30 Hz), 5.05 (s, 2H), 6.90
(d, 2H, J ) 8.7 Hz), 7.04 (d, 2H, J ) 8.7 Hz), 7.17-7.42 (m,
10H); MS (APCI+) m/z (rel) 332 (100).
1-Phenyl-2-benzylaminopropane (38). Compound 38 was
prepared from phenylacetone 35 (5.5 g, 41 mmol) to afford 8.4 g
(91%) as a tan solid. 1H (CDCl3) δ 1.09 (d, 3H, J ) 6.3 Hz), 2.61-
2.81 (m, 2H), 2.92 (m, 1H), 3.80 (dd, 2H), 7.14-7.30 (m, 10H);
MS (APCI+) m/z (rel) 226 (100).
1-(1′-Naphthyl)-2-benzylaminopropane (39). Compound 39
was prepared from 1-naphthalen-1-yl-propan-2-one 36 (5.0 g, 27.1
mmol) to afford 7.0 g (94%) as a tan solid. 1H (CDCl3) δ 1.14 (d,
3H, J ) 6.0 Hz), 3.02-3.18 (m, 2H), 3.27 (m, 1H), 3.80 (dd, 2H,
J ) 13.2, 43.8 Hz), 7.13-7.23 (m, 5H), 7.31-7.48 (m, 4H), 7.73
(d, 1H, J ) 7.8 Hz), 7.83-7.86 (m, 1H), 7.96-7.99 (m, 1H); MS
(APCI+) m/z (rel) 276 (100).
(R)-(-)-1-(4′-Methoxyphenyl)-2-benzylaminopropane [(R)-
11]. A sample of 3.02 g (11.8 mmol) of 1-(4′-methoxyphenyl)-2-
benzylaminopropane 42 was reacted with 1.8 g (11.8 mmol) of
(S)-(+)-mandelic acid to give 530 mg (35% based on enantiomeric
1
abundance) of the free amine after workup. H NMR (CDCl3) δ
1.10 (d, 3H, J ) 6.3 Hz), 2.57-2.76 (m, 2H), 2.88-2.94 (m, 1H),
3.79 (s, 3H), 3.72-3.88 (m, 2H), 6.82 (d, 2H, J ) 8.7 Hz), 7.07
(d, 2H, J ) 8.4 Hz), 7.15-7.31 (m, 5H); MS (APCI+) m/z (rel)
256 (100); [R]D ) -30.4° (c 1.25, MeOH).
(S)-(+)-1-(4′-Methoxyphenyl)-2-benzylaminopropane [(S)-11].
A sample of 3.36 (13.2 mmol) of the racemate 1-(4′-methoxyphe-
nyl)-2-benzylaminopropane 42 was reacted with 2.0 g (13.2 mmol)
of (R)-(-)-mandelic acid to give 740 mg (44% based on enantio-
meric abundance) of the free amine after workup. 1H NMR, (CDCl3)
δ 1.10 (d, 3H, J ) 6.2 Hz), 2.55-2.76 (m, 2H), 2.88-2.95 (m,
1H), 3.73-3.88 (m, 2H), 3.79 (s, 3H), 6.80 (d, 2H, J ) 8.7 Hz),
7.08 (d, 2H, J ) 8.4 Hz), 7.15-7.30 (m, 5H); MS (APCI+) m/z
(rel) 256 (100); [R]D ) +30.5° (c 1.1, MeOH).
(R)-(-)-1-(4′-Nitrophenyl)-2-benzylaminopropane [(R)-12]. A
sample of 2.0 g (7.3 mmol) of 1-(4′-nitrophenyl)-2-benzylamino-
propane 43 was reacted with 1.13 g (7.3 mmol) of (S)-(+)-mandelic
acid to give 486 mg (49% based on enantiomeric abundance) of
the free amine after workup. 1H NMR (CDCl3) δ 1.60 (d, 3H, J )
6.3 Hz), 2.73-2.85 (m, 1H), 3.00-3.12 (m, 2H), 3.86 (dd, 2H, J
) 26 Hz, J ) 60 Hz), 7.23-7.40 (m, 5H), 7.30 (d, 2H, J ) 9.0
Hz), 8.14 (d, 2H, J ) 8.7 Hz); MS (APCI+) m/z (rel) 271 (100);
[R]D ) -9.3° (c 1.0, MeOH).
1-(4′-Methoxyphenyl)-2-benzylaminopropane (42). Compound
42 was prepared from 4-methoxyphenylacetone (Aldrich) 40 (2.75
1
g, 13.1 mmol) to afford 2.31 g (97%). H (CDCl3) δ 1.10 (d, 3H,
J ) 6.3 Hz), 2.56-2.75 (m, 2H), 2.90 (m, 1H), 3.79 (s, 1H), 3.79
(m, 2H, J ) 13.2 Hz), 6.82 (d, 2H, J ) 8.7 Hz), 7.07 (d, 2H, J )
8.7 Hz), 7.18-7.32 (m, 5H); MS (APCI+) m/z (rel) 256 (100).
1-(4′-Nitrophenyl)-2-benzylaminopropane (43). Compound 43
was prepared from 4-nitrophenylacetone (Aldrich) 41 (4.95 g, 28
mmol) to afford 7.32 g (98%) as an amber oil. 1H (CDCl3) δ 1.60
(d, 3H, J ) 6.3 Hz), 2.73-2.85 (m, 1H), 3.00-3.12 (m, 2H), 3.86
(dd, 2H, J ) 26 Hz, J ) 60 Hz), 7.23-7.40 (m, 5H), 7.30 (d, 2H,
J ) 9.0 Hz), 8.14 (d, 2H, J ) 8.7 Hz); MS (APCI+) m/z (rel) 271
(100).
(S)-(+)-1-(4′-Nitrophenyl)-2-benzylaminopropane [(S)-12]. A
sample of 2.0 g (7.3 mmol) of 1-(4′-nitrophenyl)-2-benzylamino-
propane 43 was reacted with 1.13 g (7.3 mmol) of (R)-(-)-mandelic
acid to give 640 mg (65% based on enantiomeric abundance) of
the free amine after workup. 1H NMR (CDCl3) δ 1.60 (d, 3H, J )
6.3 Hz), 2.73-2.85 (m, 1H), 3.00-3.12 (m, 2H), 3.86 (dd, 2H, J
) 26, 60 Hz), 7.23-7.40 (m, 5H), 7.30 (d, 2H, J ) 9.0 Hz), 8.14
General Procedure for Enantiomeric Separation of 2-Ben-
zylaminopropanes [(R)-10-14, (S)-10-14]. The appropriate ra-
cemic 2-benzylaminopropane (1 equiv) was combined with the
(d, 2H, J ) 8.7 Hz); MS (APCI+) m/z (rel) 271 (100); [R]D
+8.2° (c 1.0, MeOH).
)

2912 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
Jozwiak et al.
(R)-(-)-1-Phenyl-2-benzylaminopropane [(R)-13]. A sample
of 2.62 g (11.6 mmol) of 1-phenyl-2-benzylaminopropane 38 was
reacted with 1.77 g (11.6 mmol) of (S)-(+)-mandelic acid to give
747 mg (57% based on enantiomeric abundance) of the free amine
after workup. 1H NMR (CDCl3) δ 1.13 (d, 3H, J ) 6.0 Hz), 2.62-
2.84 (m, 2H), 2.92-2.99 (m, 1H), 3.81 (dd, 2H, J ) 13.2, 34.5
Hz) 7.14-7.29 (m, 10H); MS (APCI+) m/z (rel) 226 (100); [R]D
) -24.5° (c 1.10, MeOH).
(S)-(+)-1-Phenyl-2-benzylaminopropane [(S)-13]. A sample of
5.0 g (22.2 mmol) of racemic 1-phenyl-2-benzylaminopropane 38
was reacted with 3.4 g (22.2 mmol) of (R)-(-)-mandelic acid to
give 2.15 g (86% based on enantiomeric abundance) of the free
amine after workup. 1H NMR (CDCl3) δ 1.11 (d, 3H, J ) 6.0 Hz),
2.62-2.84 (m, 2H), 2.92-2.99 (m, 1H), 3.81 (dd, 2H, J ) 13.2,
34.5 Hz), 7.14-7.29 (m, 5H); MS (APCI+) m/z (rel) 226 (100);
[R]D ) +18.2° (c 0.85, MeOH).
appropriate free benzylamine (R)- or (S)-10-15, 28 (0.95 equiv)
in a good amount of toluene and evaporated again under high
vacuum to remove trace H2O. The resulting colorless residue was
heated to 120 °C under argon for 20 h, cooled, and checked by 1H
NMR and mass spectrometry to confirm coupling. The residue was
dissolved in EtOH (c ) 0.07 M) with heat and transferred to a
Parr flask, where it was hydrogenated at 50 psi of hydrogen over
10% (wt) Pd/C (10 mg cat/65 mg bromohydrin) for 24 h. Complete
debenzylation was confirmed by mass spectrometry. The mixture
was filtered through Celite, the filter cake was rinsed with
isopropanol, and the filtrate was concentrated. The residue was
dissolved in 1:1 isopropanol/EtOH (c ) 0.2 M) and brought to
reflux for 30 min with 0.5 equiv of fumaric acid. The reaction was
cooled and the solvent was removed. The crude material was
purified by open column chromatography or preparative chroma-
tography.
(R)-(-)-1-(1′-Naphthyl)-2-benzylaminopropane [(R)-14]. The
washes recovered from the separation of [(S)-14] were concentrated
and partitioned between 40 mL of choroform and 40 mL of 10%
K2CO3 in water. The organic partition was washed with 20 mL of
brine then dried (Na2SO4) to afford 1.16 g (4.2 mmol) of the free
amine, which was reacted with 640 mg (4.2 mmol) of (S)-(+)-
mandelic acid. A total of 588 mg (46% based on enantiomeric
Column Separation of (R,R)-1 and (S,S)-1, Procedure B. A
sample of 75 mg of fenoterol HBr (Aldrich) was dissolved in 1.5
mL of 95/5/0.05 CH3CN/isopropanol/HNEt2 and applied in 100 µL
injections to a Chiralpak AD-H 10 × 250 mm 5 µm semipreparative
column using a Waters 2690 Separations Module, with PDA set to
280 nm. The eluting solvent was 95/5/0.05 CH3CN/isopropanol/
HNEt2, 5 mL/min. Retention times for (S,S)- and (R,R)-isomers
were 4.8 and 7.8 min, respectively.
1
abundance) of the free amine was obtained. H NMR (CDCl3) δ
1.07 (d, 3H, J ) 6.0 Hz), 3.02-3.18 (m, 2H), 3.27 (m, 1H), 3.74
(dd, 2H, J ) 13.2, 30.9 Hz), 7.13-7.23 (m, 5H), 7.31-7.48 (m,
4H), 7.73 (d, 1H, J ) 7.8 Hz), 7.83-7.86 (m, 1H), 7.96-7.99 (m,
1H); MS (APCI+) m/z (rel) 276 (100); [R]D ) -5.8° (c 1.0,
MeOH).
(S)-(+)-1-(1′-Naphthyl)-2-benzylaminopropane [(S)-14]. A
sample of 2.6 g (9.4 mmol) of 1-(1′-naphthyl)-2-benzylaminopro-
pane, 39, was reacted with 1.44 g (9.4 mmol) of (R)-(-)-mandelic
acid to give 420 mg (21% based on enantiomeric abundance) of
the free amine after workup. 1H NMR (CDCl3) δ 1.07 (d, 3H, J )
6.0 Hz), 3.02-3.18 (m, 2H), 3.27 (m, 1H), 3.74 (dd, 2H, J ) 13.2,
30.9 Hz), 7.13-7.23 (m, 5H), 7.31-7.48 (m, 4H), 7.73 (d, 1H, J
) 7.8 Hz), 7.83-7.86 (m, 1H), 7.96-7.99 (m, 1H); MS (APCI+)
m/z (rel) 276 (100); [R]D ) +6.3° (c 1.0, MeOH).
(R,R)-(-)-Fenoterol [(R,R)-1]. R,R-1 was obtained according
to procedure B to give 40 mg, which was collected after evapora-
1
tion. H NMR (CD3OD) δ 1.05 (d, 3H, J ) 6.3 Hz), 2.49 (q, 1H,
J ) 6.9 Hz), 2.62-2.74 (m, 2H), 2.80-2.91 (m, 2H), 4.55 (dd,
1H, J ) 5.1, J ) 3.3 Hz), 6.16 (t, 1H, J ) 2.4 Hz), 6.27 (d, 2H,
J ) 2.1 Hz), 6.68 (d, 2H, J ) 8.4 Hz), 6.94 (d, 2H, J ) 8.4 Hz);
13C NMR (CD3CN) δ 20.3, 43.2, 55.1, 55.2, 72.4, 102.2, 105.4,
116.0, 131.3, 131.8, 147.4, 156.2, 159.0; UV (MeOH) λmax 279
nm (ꢀ 2760), 225 (12 900), 204 (32 600); MS (APCI+) m/z (rel)
304 (100, M + H); [R]D ) -29.0° (c 0.2, MeOH); HPLC (a) 0.1%
diethylamine in H2O, 0.50 mL/min, 254 nm, tR 2.90 min, 99% pure;
(d) tR 7.8 min, >99% pure.
(S,S)-(+)-Fenoterol [(S,S)-1]. S,S-1 was obtained according to
1
procedure B to give 35 mg after evaporation. H NMR (CD3OD)
(R)-(-)-2-Benzylaminoheptane [(R)-15]. A sample of 0.65 mL
(4.4 mmol) of (R)-(-)-2-aminoheptane, R-44, 0.44 mL (4.4 mmol)
of benzaldehyde, and 0.1 mL of HOAc were combined in 40 mL
of CH2Cl2 and then cooled to 0 °C. To the reaction mixture was
added 2.75 mg (13 mmol) of sodium triacetoxyborohydride in one
portion, which was stirred under argon at rt for 28 h, diluted with
30 mL of CH2Cl2, and cooled in ice bath. An aliquot of 80 mL of
5% NaOH in water was added, fractions were separated, and the
organics were dried (Na2SO4) and evaporated to 638 mg (71%) of
δ 1.05 (d, 3H, J ) 6.6 Hz), 2.49 (q, 1H, J ) 7.2 Hz), 2.62-2.76
(m, 2H), 2.80-2.94 (m, 2H), 4.55 (dd, 1H, J ) 4.8, J ) 3.3 Hz),
6.16 (t, 1H, J ) 2.1 Hz), 6.27 (d, 2H, J ) 2.4 Hz), 6.68 (d, 2H, J
) 8.4 Hz), 6.94 (d, 2H, J ) 8.4 Hz); 13C (CD3CN) δ 20.3, 43.2,
55.0, 55.2, 72.4, 102.2, 105.4, 116.0, 131.3, 131.8, 147.4, 156.2,
159.0; UV (MeOH) λmax 279 nm (ꢀ 2680), 224 (12 700), 204
(32 800); MS (APCI+) m/z (rel) 304 (100, M + H); [R]D ) +28.5°
(c 0.20, MeOH); HPLC (a) 0.1% diethylamine in H2O, 0.50 mL/
min, 254 nm, tR 2.72 min, >99% pure; (d) tR 4.8 min, >99% pure.
1
(R)-15. H NMR (CDCl3) δ 0.88 (m, 3H), 1.08 (d, 3H J ) 6.6
(R,S)-(-)-Fenoterol Fumarate [(R,S)-1]. R,S-1 was prepared
Hz), 1.20-1.39 (m, 6H), 1.41-1.67 (m, 2H), 3.62-3.77 (m, 1H),
3.75 (p, 2H, J ) 12 Hz), 7.17-7.41 (m, 5H); MS (APCI+) m/z
(rel) 206 (100); [R]D ) +6.9° (c 1.0, MeOH).
from (R)-8 and (S)-10 according to procedure A to give 168 mg
1
(64%). H NMR (CD3OD) δ 1.22 (d, 3H, J ) 6.6 Hz), 2.64 (dd,
1H, J ) 9.9 Hz, J ) 13.2 Hz), 3.01-3.51 (m, 4H), 4.79 (dd, 1H,
J ) 3.0 Hz, J ) 9.9 Hz), 6.23 (t, 1H, J ) 2.4 Hz), 6.36 (d, 2H, J
) 2.1 Hz), 6.75 (s, 1H), 6.76 (d, 2H, J ) 8.4 Hz), 7.05 (d, 2H, J
) 8.1 Hz); 13C NMR (CD3OD) δ 16.2, 39.1, 52.5, 57.4, 70.4, 103.4,
105.3, 116.7, 127.8, 131.4, 135.2, 144.6, 157.7, 160.0, 168.2; UV
(MeOH) λmax 278 nm (ꢀ 2520), 205 (27 900); MS (ESI+) m/z (rel)
304 (100, M + H); [R]D ) -7.5° (c 0.75, MeOH); HPLC (a) 70/
30/0.05 1.00 mL/min, 282 nm, tR 1.35 min, >99% pure; (b) 50/
50/0.05 1.0 mL, 0.50 mL/min, 254 nm, tR 2.72 min, >99% pure;
(d) tR 4.8 min, 1.00 mL/min, 280 nm, tR 2.10 min, 97.5% pure.
(S)-(+)-2-Benzylaminoheptane [(S)-15]. A sample of 0.15 mL
(1 mmol) of (S)-(+)-2-aminoheptane, (S)-44, 0.1 mL (1 mmol) of
benzaldehyde, and 0.1 mL of HOAc were combined in 10 mL of
CH2Cl2 and cooled to 0 °C, then 650 mg (3 mmol) of sodium
triacetoxyborohydride was added in one portion. The reaction
mixture was stirred under argon at rt for 28 h, the mixture was
diluted with 10 mL of dichloromethane and cooled in ice bath,
and 20 mL of 5% NaOH in water was added. Fractions were
separated, and the organics were dried (Na2SO4) and evaporated
1
to 154 mg (70%). H NMR (CDCl3) δ 0.88 (m, 3H), 1.08 (d, 3H
J ) 6.6 Hz), 1.19-1.37 (m, 6H), 1.41-1.67 (m, 2H), 3.62-3.77
(m, 1H), 3.75 (p, 2H, J ) 12 Hz), 7.17-7.41 (m, 5H); MS (APCI+)
m/z (rel) 206 (100); [R]D ) +7.8° (c 1.0, MeOH).
Preparation of Fenoterol Analogs, Procedure A. To form the
epoxide, the appropriate 3′,5′-dibenzyloxyphenylbromohydrin (R)-8
or (S)-8 (1 equiv) was combined with K2CO3 (1.4 equiv) in 1:1
THF/MeOH (c ) 0.3 M) and stirred for 2 h under argon at rt. The
solvent was removed, and the residue was partitioned between
toluene and H2O. The toluene fraction was isolated and dried (Na2-
SO4), filtered, and evaporated. The residue was dissolved with the
(S,R)-(+)-Fenoterol Fumarate [(S,R)-1]. S,R-1 was prepared
from (S)-8 and (R)-10 according to procedure A to give 104 mg
(39%). H NMR (CD3OD) δ 1.22 (d, 3H, J ) 6.6 Hz), 2.64 (dd,
1H, J ) 9.9 Hz, J ) 13.5 Hz), 3.47-3.04 (m, 4H), 4.80 (dd, 1H,
J ) 2.7, J ) 9.6 Hz), 6.23 (t, 1H, J ) 2.4 Hz), 6.36 (d, 2H, J )
2.1 Hz), 6.75 (s, 1H), 6.76 (d, 2H, J ) 8.4 Hz), 7.05 (d, 2H, J )
8.4 Hz); 13C NMR (CD3OD) δ 16.2, 39.1, 52.5, 57.4, 70.4, 103.4,
105.3, 116.7, 127.8, 131.4, 135.2, 144.6, 157.7, 159.9, 168.2; UV
(MeOH) λmax 278 nm (ꢀ 2640), 202 (36 600); MS (ESI+) m/z (rel)
304 (100, M + H), 413 (10); [R]D ) +6.4° (c 0.50, MeOH); HPLC
1

Binding of the Stereoisomers of Fenoterol
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2913
(a) 70/30/0.05, 1.00 mL/min, 282 nm, tR 1.35 min, 95.9% pure;
(b) 50/50/0.05, 1.0 mL/min, 280 nm, tR 2.06 min, 99% pure.
(R,R)-(-)-1-p-Methoxyphenyl-2-(â-3′,5′-dihydroxyphenyl-â-
oxy)ethylamino-propane Fumarate [(R,R)-2]. R,R-2 was prepared
from (R)-8 and (R)-11 according to procedure A to give 172 mg
(38%). 1H NMR (CD3OD) δ 1.08 (d, 3H, J ) 6.3 Hz), 3.05-2.56
(m, 5H), 4.57 (dd. 1H, J ) 8.4, 5.4 Hz), 6.16 (m, 1H), 6.26 (d, 2H,
J ) 2.7 Hz), 6.81 (d, 2H, J ) 8.7 Hz), 7.03 (d, 2H J ) 8.7 Hz);
13C NMR (CD3OD) δ 18.8, 42.3, 54.5, 55.6, 56.0, 72.6, 103.0,
105.4, 115.0, 131.1, 131.2, 131.3, 146.2, 159.8, 159.9; UV (MeOH)
25.8 Hz); 13C NMR (CD3OD) δ 15.5, 39.6, 52.5, 56.7, 70.7, 103.4,
105.3, 123.3, 131.8, 132.0, 135.2, 136.9, 144.7, 160.0, 168.1; UV
(MeOH) λmax 284 nm (ꢀ 1720), 207 (28 400); MS (ESI+) m/z (rel)
303 (100, M + H), 329 (20); [R]D ) +11.1° (c 0.50, MeOH); HPLC
(a) 80/20/0.05, 0.7 mL/min, 276 nm, tR 2.01 min, <99% pure; (b)
50/50/0.05, 1.0 mL/min, 282 nm, tR 2.50 min, 99.4% pure.
(R,S)-(-)-5-{2-[2-(4-Aminophenyl)-1-methylethylamino]-1-
hydroxyethyl}-1,3-benzenediol Fumarate [(R,S)-3]. R,S-3 was
prepared from (R)-8 and (S)-12 according to procedure A to give
1
72 mg (35%). H NMR (CD3OD) δ 1.23 (m, 3H), 2.73-3.24 (m,
λmax 277 nm (ꢀ 3590), 224 (17 700), 207 (29 500); MS (ESI+) m/z
4H), 3.51 (m, 1H), 4.80 (dd, 1H, J ) 2.7, 9.6 Hz), 6.22 (t, 1H, J
) 2.1 Hz), 6.36 (d, 2H, J ) 2.4 Hz), 6.75 (s, 2H, fumarate), 7.32
(dd, 4H, J ) 8.4, 25.2 Hz); 13C NMR (CD3OD) δ 16.1, 39.12,
5.16, 56.9, 70.4, 103.4, 105.3, 123.4, 132.0, 132.0, 135.2, 136.8,
144.6, 160., 168.10; UV (MeOH) λmax 284 nm (ꢀ 1620), 205
(27 200); MS (ESI+) m/z (rel) 303 (100, M + H), 134 (14); [R]D
) -7.5° (c 0.50, MeOH); HPLC (a) 80/20/0.05, 0.7 mL/min, 276
nm, tR 2.08 min, 95.0% pure; (b) 50/50/0.05, 1.0 mL/min, 282 nm,
tR 2.51 min, 97.4% pure.
(rel) 318 (100, M + H); [R]D ) -24.9° (c 0.8, MeOH); HPLC (a)
70/30/0.05, 1.0 mL/min, 282 nm, tR 1.54 min, 96.5% pure; (b) 50/
50/0.05, 2.0 mL/min, 276 nm, tR 1.51 min, 95.9% pure.
(S,S)-(+)-1-p-Methoxyphenyl-2-(â-3′,5′-dihydroxyphenyl-â-
oxy)ethylamino-propane Fumarate [(S,S)-2]. S,S-2 was prepared
from (S)-8 and (S)-11 according to procedure A to give 318 mg
(53%). 1H NMR (CD3OD) δ 1.15 (d, 3H, J ) 6.0 Hz), 2.58-3.22
(m, 5H), 3.77 (s, 3H), 4.68 (dd, 1H, J ) 4.8, 8.4 Hz), 6.18 (t, 1H,
J ) 2.1 Hz), 6.31 (d, 2H, J ) 2.1 Hz), 2.23 (s, 0.5 H, fumarate),
6.84 (d, 2H, J ) 8.7 Hz), 7.10 (d, 2H, J ) 9.0 Hz); 13C NMR
(CD3OD) δ 16.1, 39.9, 52.4, 54.5, 55.3, 70.4, 101.9, 104.2, 114.0,
129.2, 130.1, 144.4, 158.7, 158.9; UV (MeOH) λmax 277 nm (ꢀ
2100), 224 (1100), 205 (22 700); MS (ESI+) m/z (rel) 318 (100,
M + H); [R]D ) +28.6° (c 0.95, MeOH); HPLC (a) 70/30/0.05,
1.0 mL/min, 282 nm, tR 1.67 min, 96.0% pure; (b) 50/50/0.05, 2.0
mL/min, 276 nm, tR 1.51 min, 97.1% pure.
(R,S)-(-)-1-p-Methoxyphenyl-2-(â-3′,5′-dihydroxyphenyl-â-
oxy)ethylaminopropane Fumarate [(R,S)-2]. R,S-2 was prepared
from (R)-8 and (S)-11 according to procedure A to give 160 mg
(38%). 1H NMR (CD3OD) δ 1.20 (d, 3H, J ) 6.6 Hz), 2.62-2.71
(m, 1H), 2.98-3.20 (m, 3H), 3.30-3.42 (m, 2H), 4.73-4.81 (m,
1H), 6.21 (m, 2H), 3.35 (m, 2H), 6.71 (s, 0.5H, fumarate), 6.56-
6.89 (m, 2H), 7.11-7.19 (m, 2H); 13C NMR (CD3OD) δ 15.6, 38.5,
51.8, 54.5, 55.9, 69.7, 102.1, 104.1, 114.1, 128.5, 130.2, 136.0,
143.8, 158.8, 159.1; UV (MeOH) λmax 277 nm (ꢀ 4100), 224
(21 400), 203 (50 600); MS (ESI+) m/z (rel) 318 (100, M + H);
[R]D ) -7.2° (c 1.5, MeOH); HPLC (a) 70/30/0.05, 1.00 mL/min,
282 nm, tR 1.40 min, 99% pure; (b) 50/50/0.05, 2.0 mL/min, 276
nm, tR 1.51 min, 96.1% pure.
(S,R)-(+)-1-p-Methoxyphenyl-2-(â-3′,5′-dihydroxyphenyl-â-
oxy)ethylaminopropane Fumarate [(S,R)-2]. S,R-2 was prepared
from (S)-8 and (R)-11, according to procedure A to give 200 mg
(51%). 1H NMR (CD3OD) δ 1.12 (d, 3H, J ) 6.0 Hz), 2.58-3.13
(m, 5H), 3.77 (s, 3H), 4.62 (dd, 1H, J ) 3.6, 9.0 Hz), 6.15 (m,
1H), 6.30 (d, 2H, J ) 1.8 Hz), 6.85 (d, 2H, J ) 8.7 Hz), 7.11 (d,
2H, J ) 8.7 Hz); 13C NMR (CD3OD) δ 18.2, 41.4, 54.1, 55.7,
56.5, 64.7, 103.0, 105.3, 115.1, 130.7, 131.3, 145.9, 159.8, 160.0;
UV (MeOH) λmax 277 nm (ꢀ 3150), 224 (3310), 205 (30 600); MS
(ESI+) m/z (rel) 318 (100, M + H); [R]D ) +14.1° (c 0.95,
MeOH); HPLC (a) 70/30/0.05, 1.00 mL/min, 282 nm, tR 1.42 min,
97.7% pure; (b) 50/50/0.05, 2.0 mL/min, 276 nm, tR 1.52 min,
97.8% pure.
(S,R)-(+)-5-{2-[2-(4-Aminophenyl)-1-methylethylamino]-1-
hydroxyethyl}-1,3-benzenediol Fumarate [(S,R)-3]. S,R-3 was
prepared from (S)-8 and (R)-12 according to procedure A to give
1
93 mg (42%). H NMR (CD3OD) δ 1.23 (d, 3H, J ) 6.3 Hz),
2.70-3.78 (m, 4H), 3.42-3.62 (m, 1H), 4.80 (dd, 1H, J ) 3.0, 9.9
Hz), 6.22 (t, 1H, J ) 2.1 Hz), 6.37 (d, 2H, J ) 2.1 Hz), 6.75 (s,
2H, fumarate), 7.33 (dd, 4H, J ) 8.4, 26.7 Hz); 13C NMR (CD3-
OD) δ 16.2, 39.1, 52.6, 56.9, 70.5, 103.4, 105.3, 123.5, 132.1, 133.7,
135.2, 137.1, 144.7, 160.0, 168.1; UV (MeOH) λmax 284 nm (ꢀ
8230), 207 (100 000); MS (ESI+) m/z (rel) 303 (100, M + H),
134 (18); [R]D ) +11.4° (c 0.50, MeOH); HPLC (a) 70/30/0.05,
1.00 mL/min, 280 nm, tR 1.45 min, 99% pure; (b) 50/50/0.05, 1.0
mL/min, 282 nm, tR 2.63 min, 95.33% pure.
(R,R)-(-)-5-[1-Hydroxy-2-(1-methyl-2-phenylethylamino)et-
hyl]-1,3-benzenediol Fumarate [(R,R)-4]. R,R-4 was prepared
from (R)-8 and (R)-13 according to procedure A to give 92 mg
1
(26%). H NMR (CD3OD) δ 1.22 (m, 3H), 2.68-3.28 (m, 2H),
3.10-3.28 (m, 2H), 3.53 (br-m, 1H), 4.75-4.80 (m, 1H), 6.24 (t,
1H, J ) 2.4 Hz), 6.38 (d, 2H, J ) 2.1 Hz), 6.75 (s, 1H, fumarate),
7.22-7.33 (m, 5H); 13C NMR (CD3OD) δ 15.5, 40.3, 56.9, 70.2,
103.4, 105.3, 128.3, 129.9, 130.3, 135.2, 137.3, 144.6, 144.6, 159.9,
168.1; UV (MeOH) λmax 277 nm (ꢀ 926), 204 (18 700); MS
(APCI+) m/z 288 (100, M + H); [R]D ) -21.2° (c 0.85, MeOH);
HPLC (a) 50/50/0.05, 1.00 mL/min, 282 nm; tR 1.73 min; 99%
pure; (b) 50/50/0.05, 2.0 mL/min, 276 nm, tR 1.46 min, 97.5% pure.
(S,S)-(+)-5-[1-Hydroxy-2-(1-methyl-2-phenylethylamino)et-
hyl]-1,3-benzenediol Fumarate [(S,S)-4]. S,S-4 was prepared from
(S)-8 and (S)-13 according to procedure A to give 184 mg (51%).
1H NMR (CD3OD) δ 1.21 (m, 3H), 2.70-3.13 (m, 2H), 3.15-
3.23 (m, 2H), 3.54 (br m, 1H), 4.79-4.86 (m, 1H), 6.24 (t, 1H, J
) 2.1 Hz), 6.39 (t, 2H, J ) 2.7 Hz), 6.76 (s, 1H, fumarate), 7.22-
7.32 (m, 5H); 13C NMR (CD3OD) δ 15.5, 40.3, 56.9, 70.2, 103.4,
105.3, 128.3, 129.9, 130.3, 135.1, 137.3, 144.6, 144.6, 159.9, 168.1;
UV (MeOH) λmax 278 nm (ꢀ 1510), 207 (26 600); MS (APCI+)
m/z 288 (100, M + H); [R]D ) +19.3° (c 0.90, MeOH); HPLC (a)
50/50/0.05, 1.00 mL/min, 282 nm, tR 1.49 min; 98.4% pure; (b)
50/50/0.05, 2.0 mL/min, 276 nm, tR 1.35 min, 99% pure.
(R,R)-(-)-5-{2-[2-(4-Aminophenyl)-1-methylethylamino]-1-
hydroxyethyl}-1,3-benzenediol Fumarate [(R,R)-3]. R,R-3 was
prepared from (R)-8 and (R)-12 according to procedure A to give
1
88 mg (42%). H NMR (CD3OD) δ 1.23 (m, 3H), 2.70-3.24 (m,
(R,S)-(-)-5-[1-Hydroxy-2-(1-methyl-2-phenylethylamino)et-
hyl]-1,3-benzenediol Fumarate [(R,S)-4]. R,S-4 was prepared from
(R)-8 and (S)-13 according to procedure A to give 170 mg (45%).
1H NMR (CD3OD) δ 1.22 (m, 3H), 2.68-3.28 (m, 2H), 3.13-
3.28 (m, 2H), 3.53 (br m, 1H), 4.76-4.80 (m, 1H), 6.23 (t, 1H, J
) 2.1 Hz), 6.37 (t, 2H, J ) 3.0 Hz), 6.75 (s, 1H, fumarate), 7.24-
7.37 (m, 5H); 13C NMR (CD3OD) δ 16.3, 24.2, 39.8, 57.2, 70.5,
103.4, 105.3, 128.4, 130.0, 130.4, 135.2, 137.4, 144.6, 160.1; UV
(MeOH) λmax 278 nm (ꢀ 1110), 205 (31 000); MS (APCI+) m/z
(rel) 288 (100, M + H); [R]D ) -6.9° (c 0.85, MeOH); HPLC (a)
50/50/0.05, 1.00 mL/min, 282 nm, tR 1.53 min, 99% pure; (b) 50/
50/0.05, 2.0 mL/min, 276 nm, tR 1.46 min, 98.5% pure.
4H), 3.54 (m, 1H), 4.84 (dd, 1H, J ) 3.3, 9.6 Hz), 6.23 (t, 1H, J
) 2.4 Hz), 6.38 (d, 2H, J ) 2.1 Hz), 6.75 (s, 2H, fumarate), 7.35
(dd, 4H, J ) 8.1, 21.0 Hz); 13C (CD3OD) δ 15.5, 39.6, 52.7, 56.6,
70.3, 103.4, 105.3, 123.8, 132.0, 132.1, 135.2, 137.5, 144.7, 160.0,
168.1; UV (MeOH) λmax 284 nm (ꢀ 1520), 206 (21 700); MS
(ESI+) m/z (rel) 303 (100, M + H); [R]D ) -6.8° (c 1.0, MeOH);
HPLC (a) 80/20/0.05, 0.70 mL/min, 276 nm, tR 2.07 min, 95.5%
pure; (b) 50/50/0.05, 1.0 mL/min, 282 nm, tR 2.60, 97.16% pure.
(S,S)-(+)-5-{2-[2-(4-Aminophenyl)-1-methylethylamino]-1-hy-
droxyethyl}-1,3-benzenediol Fumarate [(S,S)-3]. S,S-3 was pre-
pared from (S)-8 and (S)-12 according to procedure A to give 56
mg (25%). 1H NMR (CD3OD) δ 1.23 (m, 3H), 2.62-3.27 (m, 4H),
3.55 (m, 1H), 4.74-4.88 (m, 1H), 6.22 (t, 1H, J ) 1.8 Hz), 6.37
(d, 2H, J ) 2.4 Hz), 6.75 (s, 2H, fumarate), 7.32 (dd, 4H, J ) 8.7,
(S,R)-(+)-5-[1-Hydroxy-2-(1-methyl-2-phenylethylamino)et-
hyl]-1,3-benzenediol Fumarate [(S,R)-4]. S,R-4 was prepared from
(S)-8 and (R)-13 according to procedure A to give 212 mg (59%).

2914 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
Jozwiak et al.
1H NMR (CD3OD) δ 1.22 (m, 3H), 2.72 (dd, 1H J ) 10.2, 13.2
Hz), 3.11 (dd, 1H, J ) 10.2, 12.6 Hz), 3.18-3.27 (m, 2H), 3.48-
3.61 (m, 1H), 4.83 (dd, 1H, J ) 3.3, 9.9 Hz), 6.22 (t, 1H, J ) 2.4
Hz), 6.36 (d, 2H, J ) 2.4 Hz), 6.75 (s, 1H, fumarate), 7.24-7.37
(m, 5H); 13C NMR (CD3OD) δ 16.3, 24.2, 39.8, 57.2, 70.5, 103.4,
105.3, 128.4, 130.0, 130.4, 135.2, 137.4, 144.6, 160.1; UV (MeOH)
λmax 278 nm (ꢀ 1680), 206 (35 500); MS (APCI+) m/z (rel) 288
(100, M + H), 270 (19, M-OH); [R]D ) +9.1° (c 1.1, MeOH);
HPLC (a) 50/50/0.05, 1.00 mL/min, 282 nm, tR 1.51 min, 99%
pure; (b) 50/50/0.05, 2.0 mL/min, 276 nm, tR 1.43 min, 99% pure.
(R,R)-(-)-5-{1-hydroxy-2-[1-methyl-2-(1-naphthyl)ethylami-
no]ethyl}-1,3-benzenediol Fumarate [(R,R)-5]. R,R-5 was pre-
pared from (R)-8 and (R)-14 according to procedure A to give 135
(dd, 1H, J ) 3.3, 9.6 Hz), 6.22 (t, 1H, J ) 2.1 Hz), 6.36 (d, 2H,
J ) 2.4 Hz), 6.75 (s, 1H, fumarate); 13C NMR (CD3OD) δ 14.3,
16.0, 23.5, 26.2, 32.6, 34.2, 52.1, 55.7, 70.2, 103.3, 105.3, 135.2,
144.7, 160.0, 168.0; UV (MeOH) λmax 278 nm (ꢀ 931), 203 nm
(20 100); MS (ESI+) m/z (rel) 268 (100, M + H); [R]D ) -8.8°
(c 1.1, MeOH); HPLC (c) 70/30/0.1, 1.0 mL/min, 276 nm, tR 2.18
min, 96.6% pure; (b) 50/50/0.05, 1.0 mL/min, 279 nm, tR 2.06 min,
98.9% pure.
(S,S)-(+)-5-[1-Hydroxy-2-(1-methylhexylamino)ethyl]-1,3-
benzenediol Fumarate [(S,S)-6]. S,S-6 was prepared from (S)-8
and (S)-15 according to procedure A to give 96 mg (43%). 1H NMR
(CD3OD) δ 0.923 (t, 3H, J ) 6.6 Hz); 1.31 (d, 3H, J ) 6.6 Hz),
1.26-1.84 (m, 8H), 2.01-3.18 (m, 2H), 3.14-3.30 (m, 1H), 4.81
(dd, 1H, J ) 3.3, 9.6 Hz), 6.23 (t, 1H, J ) 2.4 Hz), 6.39 (d, 2H,
J ) 2.1 Hz), 6.76 (s, 1H, fumarate): 13C NMR (CD3OD) δ 14.2,
16.0, 23.4, 26.3, 32.6, 34.1, 52.1, 55.8, 70.2, 103.4, 105.3, 135.2,
144.7, 159.9, 168.2; UV (MeOH) λmax 278 nm (ꢀ 1340), 203
(28 800); MS (APCI+) m/z (rel) 268 (100, M + H); [R]D ) +10.8°
(c 0.50, MeOH); HPLC (c) 70/30/0.1, 1.0 mL/min, 276 nm, tR 2.16
min, 97.0% pure; (b) 50/50/0.05, 1.0 mL/min, 279 nm, tR 2.11 min,
99% pure.
(R,S)-(-)-5-[1-Hydroxy-2-(1-methylhexylamino)ethyl]-1,3-
benzenediol Fumarate [(R,S)-6]. R,S-6 was prepared from (R)-8
and (S)-15 according to procedure A to give 83 mg (38%). 1H NMR
(CD3OD) δ 0.924 (m, 3H), 1.32 (d, 3H, J ) 6.6 Hz), 1.26-1.84
(m, 8H), 2.98-3.20 (m, 2H), 3.32-3.22 (m, 1H), 4.78 (dd, 1H, J
) 3.0, 9.9 Hz), 6.23 (t, 1H, J ) 2.1 Hz), 6.37 (d, 2H, J ) 1.8 Hz),
6.76 (s, 1H, fumarate); 13C NMR (CD3OD) δ 14.2, 16.4, 23.4, 26.2,
32.6, 33.5, 52.2, 56.0, 70.4, 103.4, 105.3, 135.2, 144.7, 160.0, 168.1;
UV (MeOH) λmax 276 nm (ꢀ 2770), 203 (35 900); MS (APCI+)
m/z (rel) 268 (100, M + H); [R]D ) -15.9° (c 0.70, MeOH); HPLC
(c) 70/30/0.1, 1.0 mL/min, 276 nm, tR 2.16 min, 97.0% pure; (b)
50/50/0.05, 1.0 mL/min, 279 nm, tR 2.07 min, 96.2% pure.
(S,R)-(+)-5-[1-Hydroxy-2-(1-methylhexylamino)ethyl]-1,3-
benzenediol Fumarate [(S,R)-6]. S,R-6 was prepared from (S)-8
and (R)-15 according to procedure A to give 81 mg (38%). 1H NMR
(CD3OD) δ 0.920 (t, 3H, J ) 6.3 Hz), 1.32 (d, 3H, J ) 6.9 Hz),
1.30-1.77 (m, 8H), 2.99-3.17 (m, 2H), 3.23-3.26 (m, 1H), 4.76
(dd, 1H, J ) 3.0, 9.6 Hz), 6.22 (t, 1H, J ) 2.4 Hz), 6.36 (d, 2H,
J ) 2.1 Hz), 6.75 (s, 1H, fumarate); 13C NMR (CD3OD) δ 14.2,
16.5, 23.5, 26.2, 32.6, 39.5, 52.2, 56.0, 70.4, 103.4, 105.3, 135.2,
144.7, 160.0, 168.0; UV (MeOH) λmax 278 nm (ꢀ 1440), 204
(29 900); MS (APCI+) m/z (rel) 268 (100, M + H); [R]D ) +12.7°
(c 1.0, MeOH); HPLC (c) 70/30/0.1, 1.0 mL/min, 276 nm, tR 2.16
min, 99% pure; (b) 50/50/0.05, 1.0 mL/min, 279 nm, tR 2.02 min,
95.7% pure.
1
mg (46%). H NMR (CD3OD) δ 1.18-1.23 (m, 3H), 3.16-3.34
(m,1H, 2H), 3.69-3.74 (m, 2H), 4.78-4.80 (m, 1H), 6.23 (t, 1H,
J ) 2.4 Hz), 6.38 (m, 2H), 7.41-7.61 (m, 4H), 7.83 (d, 1H, J )
7.5 Hz), 7.90 (d, 1H, J ) 7.8 Hz), 8.10 (m, 1H); 13C NMR (CD3-
OD) δ 16.2, 37.2, 54.5, 56.1, 70.3, 103.4, 105.3, 124.3, 126.5, 127.0,
127.7, 129.3, 130.1, 133.2, 135.2, 135.6, 144.7, 160.1, 168.2; UV
(MeOH) λmax 282 nm (ꢀ 5860), 224 (50 900), 208 (35 500); MS
(APCI+) m/z (rel) 338 (100, M + H), 169 (15, fragment); [R]D )
-20.4 (c 0.50, MeOH); HPLC (a) 60/40/0.05, 1.00 mL/min, 282
nm, tR 2.08 min, 95.7% pure; (b) 50/50/0.05, 1.5 mL/min, 282 nm,
tR 2.20 min, 99% pure.
(S,S)-(+)-5-{1-Hydroxy-2-[1-methyl-2-(1-naphthyl)ethylami-
no]ethyl}-1,3-benzenediol fumarate [(S,S)-5]. S,S-5 was prepared
from (S)-8 and (S)-14 according to procedure A to give 118 mg
1
(40%). H NMR (CD3OD) δ 1.13-1.17 (m, 3H), 3.14-3.26 (m,
1H, 2H), 3.61-3.76 (m, 2H), 4.44-4.75 (m, 1H), 6.18 (t, 1H, J )
2.4 Hz), 6.33 (m, 2H), 7.36-7.52 (m, 4H), 7.77 (dd, 1H, J ) 1.8,
7.5 Hz), 7.84 (d, 1H, J ) 8.1 Hz), 8.04 (t, 1H, J ) 8.4 Hz); 13C
NMR (CD3OD) δ 16.1, 37.1, 54.4, 56.0, 70.3, 103.3, 105.3, 124.3,
126.5, 127.0, 127.7, 129.2, 130.0, 133.2, 135.2, 135.7, 144.5, 160.0,
168.1; UV (MeOH) λmax 282 nm (ꢀ 6210), 223 (56 400), 208
(42 700); MS (APCI+) m/z (rel) 338 (100, M + H), 169 (8,
fragment); [R]D ) +20.0° (c 1.1% MeOH); HPLC (a) 60/40/0.05,
1.00 mL/min, 282 nm, tR 2.35 min, 98.9% pure; (b) 50/50/0.05,
1.5 mL/min, 282 nm, tR 2.26 min, 97.2% pure
(R,S)-(-)-5-{1-Hydroxy-2-[1-methyl-2-(1-naphthyl)ethylami-
no]ethyl}-1,3-benzenediol Fumarate [(R,S)-5]. R,S-5 was prepared
from (R)-8 and (S)-14 according to procedure A to give 114 mg
1
(39%). H NMR (CD3OD) δ 1.08-1.11 (m, 3H), 3.02-3.24 (m,
1H, 2H), 3.54-3.68 (m, 2H), 4.45-4.75 (m, 1H), 6.11 (t, 1H, J )
1.8 Hz), 6.26 (m, 2H), 6.63 (s, 2H fumarate), 7.28-7.48 (m, 4H),
7.70 (d, 1H, J ) 7.5 Hz), 7.77 (d, 1H, J ) 7.8 Hz), 7.97 (t, 1H, J
) 7.8 Hz); 13C NMR (CD3OD) δ 16.0, 37.1, 52.5, 56.0, 70.4, 103.4,
105.3, 124.4, 126.5, 127.0, 127.6, 129.2, 130.1, 133.1, 135.2, 135.5,
144.7, 159.9, 168.2; UV (MeOH) λmax 281 nm (ꢀ 12 600), 224
(61 900), 204 (47 200); MS (APCI+) m/z (rel) 338 (100, M + H),
190 (15, fragment); [R]D ) -11.3° (c 0.85, MeOH); HPLC (a)
60/40/0.05, 1.00 mL/min, 282 nm, tr 2.30 min, 98.6% pure; (b)
50/50/0.05, 1.5 mL/min, 282 nm, tR 2.36 min, 99% pure.
(R)-(-)-5-(1-Hydroxy-2-phenethylaminoethyl)-1,3-benzene-
diol Fumarate [(R)-7]. R-7 was prepared from (R)-8 and 28
1
(Aldrich) to give 37 mg (15%). H NMR (CD3OD) δ 2.94-3.23
(m, 6H), 4.73 (dd, 1H, J ) 3.3, 9.9 Hz), 6.15 (t, 1H, J ) 2.4 Hz),
6.29 (d, 2H, J ) 1.8 Hz), 7.19-7.28 (m, 5H), 6.69 (s, 1H); UV
(MeOH) λmax 278 nm (ꢀ 1360), 205 (32 600); MS (APCI+) m/z
(rel) 274 (100, M + H); [R]D ) -13.0° (c 1.0, MeOH); HPLC (a)
80/20/0.05, 1.00 mL/min, 282 nm, tR 1.47 min, 96.7% pure; (b)
50/50/0.05, 1.0 mL/min, 272 nm, tR 2.78 min, 95.1% pure.
(S,R)-(+)-5-{1-Hydroxy-2-[1-methyl-2-(1-naphthyl)ethylami-
no]ethyl}-1,3-benzenediol Fumarate [(S,R)-5]. S,R-5 was prepared
from (S)-8 and (R)-14 according to procedure A to give 123 mg
1
(42%). H NMR (CD3OD) δ 1.18-1.22 (m, 3H), 3.10-3.28 (m,
(S)-(+)-5-(1-Hydroxy-2-phenethylaminoethyl)-1,3-benzene-
diol Fumarate [(S)-7]. S-7 was prepared from (S)-8 and 28
1H, 2H), 3.69-3.78 (m, 2H), 4.45-4.75 (m, 1H), 6.23 (t, 1H, J )
2.1 Hz), 6.39 (m, 2H), 6.73 (s, 2H fumarate), 7.39-7.59 (m, 4H),
7.80 (d, 1H, J ) 7.5 Hz), 7.88 (d, 1H, J ) 7.8 Hz), 8.01 (t, 1H, J
) 9.0 Hz); 13C NMR (CD3OD) δ 16.4, 37.4, 52.5, 56.2, 70.6, 103.4,
105.3, 124.4, 126.5, 127.0, 129.3, 130.1, 133.1, 133.4, 135.6, 136.3,
144.8, 160.0, 171.4; UV (MeOH) λmax 282 nm (ꢀ 7740), 224
(70 900), 206 (55 800); MS (ESI+) m/z (rel) 338 (100, M + H);
[R]D ) +15.5 (c 1.0, MeOH); HPLC (a) 60/40/0.05, 1.0 mL/min,
282 nm, tR 1.95, 95.7% pure; (b) 50/50/0.05, 1.5 mL/min, 282 nm,
tR 2.29 min, 95.7% pure
1
(Aldrich) to give 51 mg (17%). H NMR (CD3OD) δ 2.87-3.21
(m, 6H), 4.68 (dd, 1H, J ) 3.6, 9.9 Hz), 6.10 (t, 1H, J ) 2.4 Hz),
6.24 (d, 2H, J ) 2.1 Hz), 6.63 (s, 1H), 7.12-7.21 (m, 5H); UV
(MeOH) λmax 278 nm (ꢀ 1280), 204 (33 700); MS (APCI+) m/z
(rel) 274 (100, M + H); [R]D ) +14.64° (c 1.1, MeOH); HPLC
(a) 80/20/0.05, 1.00 mL/min, 282 nm, tR 1.47 min, 98.6% pure;
(b) 50/50/0.05, 1.0 mL/min, 272 nm, tR 2.74 min, 98.8% pure.
Acknowledgment. This work was supported in part by funds
from the National Institute on Aging, Intramural Research
Program, NIA Contract N01-AG-3-1009 and the Foundation
for Polish Science (FOCUS 4/2006 programme). K.J. would
like to acknowledge Professor D. Matosiuk from the Department
(R,R)-(-)-5-[1-Hydroxy-2-(1-methylhexylamino)ethyl]-1,3-
benzenediol Fumarate [(R,R)-6]. R,R-6 was prepared from (R)-8
and (R)-15 according to procedure A to give 45 mg (29%). 1H NMR
(CD3OD) δ 0.920 (t, 3H, J ) 6.9 Hz), 1.30 (d, 3H, J ) 6.9 Hz),
1.29-1.64 (m, 8H), 3.01-3.18 (m, 2H), 3.14-3.30 (m, 1H), 4.80

Binding of the Stereoisomers of Fenoterol
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2915
and in activation of the â2-adrenergic receptor. Proc. Natl. Acad.
Sci. U.S.A. 1996, 93, 9276-9281.
of Synthesis and Chemical Technology of Pharmaceutical
Substance, Medical University of Lublin, for letting him use
the QSAR software.
(12) Kikkawa, H.; Isogaya, M.; Nagao, T.; Kurose, H. The role of the
seventh transmembrane region in high affinity binding of a â2-
selective agonist TA-2005. Mol. Pharmacol. 1998, 53, 128-134.
(13) Zuurmond, H. M.; Hessling, J.; Blu¨ml, K.; Lohse, M.; Ijzerman, A.
P. Study of interaction between agonists and Asn293 in helix VI of
human â2-adrenergic receptor. Mol. Pharmacol. 1999, 56, 909-916.
(14) Schirrmacher, E.; Schirrmacher, R.; Thews, O.; Dillenburg, W.;
Helisch, A.; Wessler, I.; Buhl, R.; Hohnemann, S.; Buchholz, H. G.;
Bartenstein, P.; Machulla, H. J.; Rosch, F. Synthesis and preliminary
evaluation of (R,R) (S,S) 5-(2-(2-[4-(2-[(18)F] fluoroethoxy)phenyl]-
1-methylethylamino)-1-hydroxyethl)-benzene-1,3-diol ([(18)F]FEFE)
for the in vivo visualization and quantification of the beta2-adrenergic
receptor status in lung. Bioorg. Med. Chem. Lett. 2003, 13, 2687-
2692.
References
(1) O’Donnell, S. R. A selective â-adrenoceptor stimulant (Th1165a)
related to orciprenaline. Eur. J. Pharmacol. 1970, 12, 35-43.
(2) Heel, R. C.; Brogden, R. C.; Speight, T. M.; Avery, G. S. Fenerotol:
A review of its pharmacological properties and therapeutic efficacy
in asthma. Drugs 1978, 15, 3-32.
(3) Xiao, R.-P.; Zhang, S.-J.; Chakir, K.; Avdonin, P.; Zhu, W.; Bond,
R. A.; Balke, C. W.; Lakatta, E. G.; Cheng, H. Enhanced Gi signaling
selectively negates â2-adrenergic receptor (AR)sBut not â1AR-
mediated positive inotropic effect in myocytes from failing rat hearts.
Circulation 2003, 108, 1633-1639.
(4) Xiao, R.-P.; Zhu, W.; Zheng, M.; Chakir, K.; Bond, R.; Lakatta, E.
G.; Cheng, H. Subtype-specific â-adrenoceptor signaling pathways
in the heart and their potential clinical implications. Trends Phar-
macol. Sci. 2004, 25, 358-365.
(5) Kaiser, G.; Wiemer, G.; Kremer, G.; Dietz, J.; Palm, D. Identification
and quantification of â-adrenoceptor sites in red blood cells from
rats. Naunyn-Schmiedeberg’s Arch. Pharmacol. 1978, 305, 41-50.
(6) Beigi, F.; Bertucci, C.; Zhu, W.; Chakier, K.; Wainer, I. W.; Xiao,
R.-P.; Abernethy, D. A. Enantioselective separation and online affinity
chromatographic characterization of R,R- and S,S-fenoterol. Chirality
2006, 18, 822-827.
(7) Trofast, J.; O¨ sterberg, K.; Ka¨llstro¨m, B.-R.; Waldek, B. Steric aspects
of agonism and antagonism at â-adrenoceptors: Synthesis of and
pharmacological experiments with the enantiomers of formoterol and
their diastereomers. Chirality 1991, 3, 443-450.
(8) Pauwels, P. J.; Van Gompel, P.; Leysen, J. E. Human â1- and â2-
adrenergic receptor binding and mediated accumulation of cAMP in
transfected Chinese hamster ovary cells. Biochem. Pharmacol. 1991,
42, 1683-1689.
(15) Beer, M.; Richardson, A.; Poat, J.; Iversen L. L.; Stahl, S. M. In
Vitro selectivity of agonists and antagonists for beta1- and beta2-
adrenoceptor subtypes in rat brain. Biochem. Pharmacol. 1988, 37,
1145-1151.
(16) Cramer, R. D., III; Patterson, D. E.; Bunch, 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.
(17) Kubinyi, H. Comparative Molecular Field Analysis (COMFA).
Encyclopedia of Computational Chemistry; John Wiley & Sons,
Ltd.: New York, 1998; http://www.wiley.com//legacy/wileychi/ecc/.
(18) Kontoyianni, M.; DeWeese, C.; Penzotti, J. E.; Lybrand, T. P. Three-
dimensional models for agonist and antagonist complexes with â2
adrenergic receptor. J. Med. Chem. 1996, 39, 4406-4420.
(19) Furse, K. E.; Lybrand, T. P. Three-dimensional models for â-adr-
energic receptor complexes with agonists and antagonists. J. Med.
Chem. 2003, 46, 4450-4462.
(20) Swaminath, G.; Xiang, Y.; Lee, T. W.; Steenhuis, J.; Parnot, C.;
Kobilka, B. K. Sequential binding of agonists to the â2 adrenoceptor.
J. Biol. Chem. 2004, 279, 686-691.
(21) Levitt, M.; Perutz, M. F. Aromatic rings act as hydrogen bond
acceptors. J. Mol. Biol. 1988, 201, 751-754.
(22) Kobilka, B. Agonist binding: A multistep process. Mol. Pharm. 2004,
65, 1060-1062.
(9) Cheng,Y. C.; Prusoff, W. H. Relationship between the inhibition
constant (Ki) and the concentration of inhibitor which causes 50
percent inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol.
1973, 22, 3099-3108.
(10) Eimerl, S.; Schramm, M.; Lok, S.; Goodman, M.; Khan, M.; Melmon,
K. The four stereoisomers of a high potency congener of isoproter-
enol: Biological activity and the relationship between the native and
the chemically inserted asymmetric carbon. Biochem. Pharmacol.
1987, 36, 3523-3527.
(23) Sokolov, V. I.; Zefirov, N. S. Enantioselectivity in two-point
binding: Rocking tetrahedron model. Dokl. Akad. Nauk SSSR 1991,
319, 1382-1383.
(11) Wieland, K.; Zuurmond, H. M.; Krasel, C.; Ijzerman, A. D. P.; Lohse,
M. J. Involvement of Asn-293 in stereospecific agonist recognition
JM070030D