Computer-aided design, synthesis and biological assay of p-methylsulfonamido phenylethylamine analogues

Article abstract of DOI:10.1016/S0960-894X(00)00412-1

Class III antiarrhythmic agents selectively delay the effective refractory period (ERP) and increase the transmembrance action potential duration (APD). Based on our previous studies, a set of 17 methylsulfonamido phenylethylamine analogues were investigated by 3D-QSAR techniques of CoMFA and CoMSIA. The 3D-QSAR models proved a good predictive ability, and could describe the steric, electrostatic and hydrophobic requirements for recognition forces of the receptor site. According to the clues provided by this 3D-QSAR analysis, we designed and synthesized a series of new analogues of methanesulfonamido phenylethylamine (VI(a-i)). Pharmacological assay indicated that the effective concentrations of delaying the functional refractory period (FRP) 10ms of these new compounds have a good correlation with the 3D-QSAR predicted values. It is remarkable that the maximal percent change of delaying FRP in μM of compound VI(c) is much higher than that of dofetilide. The results showed that the 3D-QSAR models are reliable. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0960-894X(00)00412-1

Bioorganic & Medicinal Chemistry Letters 10 (2000) 2153±2157
Computer-Aided Design, Synthesis and Biological Assay of
p-Methylsulfonamido Phenylethylamine Analogues
Hong Liu,a,b Min Ji,b Hualiang Jiang,a,* Ligang Liu,b
Weiyi Hua,b,* Kaixian Chena and Ruyun Jia
aShanghai Institute of Materia Medica, Chinese Academy of Sciences, 294 Taiyuan Road, Shanghai 200031, China
bDepartment of Medicinal Chemistry, China Pharmaceutical University, Nanjing 210009, China
Received 14 March 2000; accepted 11 May 2000
AbstractÐClass III antiarrhythmic agents selectively delay the e€ective refractory period (ERP) and increase the transmembrance
action potential duration (APD). Based on our previous studies, a set of 17 methylsulfonamido phenylethylamine analogues were
investigated by 3D-QSAR techniques of CoMFA and CoMSIA. The 3D-QSAR models proved a good predictive ability, and could
describe the steric, electrostatic and hydrophobic requirements for recognition forces of the receptor site. According to the clues
provided by this 3D-QSAR analysis, we designed and synthesized a series of new analogues of methanesulfonamido phenylethyl-
amine (VIa i). Pharmacological assay indicated that the e€ective concentrations of delaying the functional refractory period (FRP)
10 ms of these new compounds have a good correlation with the 3D-QSAR predicted values. It is remarkable that the maximal
percent change of delaying FRP in mM of compound VIc is much higher than that of dofetilide. The results showed that the 3D-QSAR
models are reliable. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
publication,6,7 we reported 17 methylsulfonamido
phenylethylamine analogues designed and synthesized
according to the pharmacophore of class III antiar-
rhythmic agents and the structural feature of dofetilide,
which is currently used in clinical tests. The biological
activities of these compounds for increasing the func-
tional refractory period (FRP) were also determined
in isolated animal atrium (Table 1). In order to ®nd
more potent analogues of dofetilide, 3D-QSAR analyses
have been performed on these 17 compounds by use of
Comparative Molecular Field Analysis (CoMFA)8 and
Comparative Molecular Similarity Index Analysis
(CoMSIA)9 methods. The 3D-QSAR models proved to
have a good predictive ability, and could describe the
steric, electrostatic and hydrophobic requirements for
recognition forces of the receptor site. According to the
clues provided by the 3D-QSAR analyses, we designed
and synthesized a series of new analogues of methane-
sulfonamido phenylethylamine (VIa i). Pharmacological
assay indicated that the e€ective concentrations of
delaying FRP 10 ms of these new compounds have a
good correlation with the 3D-QSAR predicted values. It
is remarkable that the maximal percent change of
delaying FRP in mM of compound VIc is much higher
than that of dofetilide. In this paper, we report the results
of the 3D-QSAR of known compounds,6,7 and design,
synthesis and pharmacological assay of new compounds.
The high incidence of sudden cardiac death (SCD) is of
continuing medical concern. SCD results from electrical
instability of the heart muscle leading to a loss of reg-
ular cardiac rhythm, and it is accepted that ventricular
arrhythmia such as sustained tachycardia (VT) and
®brillation (VF) plays the major role in these deaths.1,2
The Vanghan Williams classi®cation of antiarrhythmic
drugs recognizes four distinct categories. Class I is the
sodium channel blockade, its mechanism being of inter-
ference with the fast sodium channels in cell membranes.
Class II is b-blocking agents. Class IV is calcium channel
blocking agents. Class III agents act by delaying repo-
larization of cardiac myocytes, lengthening action
potential duration (APD) of the cell and a concomitantly
increasing ERP.3 The outward delayed recti®er potas-
sium current, IK, that contributes to repolarization,
consists of two kinetically distinct and identi®able cur-
rents, a rapidly activating, IKr, and a slowly activating
component, IKs. Selective blockade of either IKr or IKs
would lead to a prolongation of the refractory period
and be, by de®nition, a class III e€ect.4,5 In our previous
*Corresponding authors.
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00412-1

2154
H. Liu et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2153±2157
Table 1. Chemical structures of methanesulfonamido phenylethylamine compounds, their activities and CoMFA prediction
Compound
R1
R2
R3
Ca
EAb
PAc
Residued
1
2
3
4
5
6
7
8
CH3
-CH2C6H5
-COCH3
H
-CH2C6H4Cl-p
-CH3
-CH3
-CH3
-CH3
-CH3
-CH3
-CH2C6H5
-CH2C6H5
-CH2C6H5
-CH2C6H5
-CH2C6H5
-CH2C6H5
-CH2CH2OC6H4NHSO2CH3-p
-CH2CH2OC6H4NHSO2CH3-p
-CH2CH2OC6H4NHSO2CH3-p
-CH2CH2OC6H4NHSO2CH3-p
-CH2CH2OC6H4NHSO2CH3-p
-CH2C6H5
H
H
H
H
H
0.012
0.016
0.021
0.023
0.037
0.372
0.953
2.239
1.170
4.789
0.890
0.141
0.270
0.620
1.050
0.790
0.380
7.92
7.80
7.67
7.63
7.44
6.34
6.02
5.65
5.93
5.32
6.15
6.85
6.57
6.21
5.98
6.10
6.41
7.87
7.89
7.65
7.60
7.47
6.31
6.16
5.76
5.79
5.33
6.16
6.88
6.41
6.12
6.10
6.08
6.42
0.05
0.09
0.02
0.03
0.03
0.03
0.14
0.11
0.14
0.01
0.01
0.03
0.16
0.09
0.12
0.02
0.01
H
-CH2C6H4CH3-p
-CH2C6H4Cl-p
-CH2C6H4OCH3-p
-C6H11
-CH2C6H3OCH2O-3,4
-CH2C6H4OCH3-o
-CH2C6H5
-CH2C6H4CH3-p
-CH2C6H4Cl-p
-SO2CH3
-H
-SO2CH3
-H
-H
-H
-H
-SO2CH3
-H
9
10
11
12
13
14
15
16
17
-CH2C6H4OCH3-p
-CH2C6H3OCH2O-3,4
-SO2CH3
-SO2CH3
aC=the e€ective concentration of delaying FRP 10 ms (mM).
bEA= log C.
cPA=predictive activity.
dResidue=EA PA.
CoMFA and CoMSIA
useful starting point for statistical comparisons of ¯ex-
ible structures within the CoMFA and CoMSIA models.
In this study, we took dofetilide as the reference mol-
ecule due to its sca€old involved in the 17 compounds
and its high activity. We selected all the low-energy
conformations of this sca€old as template structure and
the pairs between the molecules for alignment during
CoMFA and CoMSIA analyses, and took the con-
formations corresponding to the maximum cross-vali-
dation R2 as the bioactive conformation to subject the
®nal CoMFA and CoMSIA analyses. The alignment of
the bioactive conformations for these 17 compounds is
shown in Figure 1.
Molecular 3D structure building. 3D-QSAR analyses
were performed using CoMFA and CoMSIA for the 17
compounds listed in Table 1. The 3D structures of the
entire set of these compounds were built using the
SKETCH option in Sybyl 6.510 and fully geometry
optimized using the standard Tripos force ®eld, with a
0.001 kcal/mol energy gradient convergence criterion
and a distance-dependent dielectric constant. The search
routine of Sybyl was then employed for the systematic
conformational search to ®nd out the local energy-
minimum conformations.
Alignment rule. While it is recognized that the global
energy-minimum conformation may not necessarily be
adopted in the drug±receptor complex, the use of a rea-
sonably low energy conformation in the alignment is a
CoMFA analysis. The CoMFA analysis was carried out
using the standard options of Sybyl. The analysis results
are presented in Tables 1 and 2 and Figures 1 and 2.
The optimized CoMFA of 17 compounds gave a good
Figure 1. (a) Superposition of 17 methanesulfonamido phenylethylamine compounds (cf. Table 1) for 3D-QSAR studies; (b) experimental activities
versus predictive values for the CoMFA model.

H. Liu et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2153±2157
Table 2. Summary of 3D-QSAR analysis results obtained using CoMFA and CoMSIA
2155
Contributions
Electrostatic
Method
CoMFA
PLS
R2
Optimal component
Steric
Hydrophobic
Cross-validation
Conventional
Cross-validation
Conventional
0.695
0.984
0.640
0.975
5
5
4
4
Ð
0.551
Ð
Ð
0.449
Ð
Ð
Ð
Ð
0.264
CoMSIA
0.271
0.465
Figure 2. (a) Contour maps from the ®nal CoMFA analysis with 2 A grid spacing in combination with dofetilide. Steric std*coe€ contour map.
Green contours (>80% contribution) refer to sterically favored regions; yellow contours (>20% contribution) indicate disfavored areas. Electro-
static std*coe€ contour map. Blue contours (>80% contribution) refer to regions where negatively charged substituents are disfavored; red contours
(>20% contribution) indicate regions where negatively charged substituents are favoured; (b) contour maps from the ®nal CoMSIA analysis with
2 A grid spacing in combination with dofetilide. Hydrophobic std*coe€ contour map. White contours (>80% contribution) refer to regions where
hydrophilic substituents are favored; yellow contours (>20% contribution) indicate regions where hydrophobic substituents are favored.
cross-validated value of 0.695 with an optimized com-
ponent of 5, the conventional correlation coecient is
R2=0.984, F=132.515, and the estimated standard
error is 0.130. These predictive results have been
obtained, indicating a good statistical correlation, and
we also ®nd that the resultant CoMFA model had a
fair predictive ability (Fig. 1b).
CoMSIA analysis. To estimate the hydrophobic con-
tributions to these antiarrhythmic agents, hydrophobic
similarity index ®elds were constructed by use of CoM-
SIA, which cannot be completely treated by Lennard±
Jones and Coulombic ®elds encoded in CoMFA. The
CoMSIA analysis was performed employing the stan-
dard options of Sybyl.11,12 The results of CoMSIA are
shown in Table 2 and Figure 2b. CoMSIA ®elds can
also be represented by color contour plots. In general,
yellow polyhedra indicate that hydrophobic substituents
are `good' for increasing the potency, while hydrophilic
substituents are bene®cial to the activity at the regions
of white contours.
The QSAR produced by CoMFA, with its hundreds of
thousands of terms, is usually represented as a 3D
`coecient contour' map. The CoMFA steric and elec-
trostatic ®elds for the analysis are presented as contour
plots in Figure 2(a). To aid in visualization, the potent
biological activity is displayed in the map. In general,
the color polyhedra in the map surrounded all lattice
points where the QSAR strongly associated changes in
the compounds' ®eld values with changes in biological
potency. Green polyhedra surrounded regions where
more bulk is `good' for increasing potency, while yellow
polyhedra surrounded regions where less bulk is `good'.
Red and blue contours show regions of desirable nega-
tive and positive electrostatic interactions, respectively.
A large region of green contour near the phenyl moiety
of phenoxyethane suggests that more potent analogues
may be obtained by introducing bulky substituents to
this region of dofetilide. Two red polyhedra near the
sulfonyl of methanesulfonamido phenoxyethylamine
and the O atom of the phenoxyethyl moiety indicate
that electron rich groups are bene®cial to the activity.
Two blue contours near the methanesulfonyl group of
methanesulfonamido phenylethylamine suggest that
positive charged substituents are favorable to increase
the activity.
Scheme 1. Synthesis of compounds VIa Reagents and conditions: (a)
i
K2CO3, CH3CN, re¯ux, 81.5%; (b) i=HCHO, HCOOH, re¯ux, 87.2%;
ii=R1X, CH3CN, re¯ux, 61%; iii=Ac2O, CH3CN, rt, 91%; (c) Fe, HCl,
EtOH, re¯ux, 75%; (d) CH3SO2Cl, Et3N, CH2Cl2, 0 ꢀC, 92.3%.

2156
H. Liu et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2153±2157
Table 3. The structures and biological activities of the designed new compounds
Compound
R1
R2
X
na
MaxÁFRPb
MaxÁHRc
MaxÁFCd
Dofetilide
VIa
VIb
VIc
VId
VIe
VIf
VIg
VIh
-CH3
-CH3
-CH2C6H5
-COCH3
-CH2CHˆCH2
-CH2CHˆCH2
-COC6H5
-CH2C6H5
-COCH3
-H
-H
-H
-H
-NHSO2CH3-4
5
3
3
3
2
2
2
2
2
2
24.02
20.2
Ne
58.36
9.77
Ne
5.33
Ne
Ne
30.03
28.51
31.8
46.17
15.18
8.33
17.05
20.25
25
37.87
16.67
16.67
33.33
25
Ne
Ne
-X
-X
-X
-X
-X
-X
-X
-X
-X
-H
-SO2CH3
-H
-SO2CH3
-SO2CH3
-SO2CH3
20
17.86
20.51
20.69
VIi
-CH3
12.02
an=the sample number.
bMaxÁFRP=the maximal percent change of delaying FRP in mM (%).
cMaxÁHR=the maximal percent change of heart rate in mM (%).
dMaxÁFC=the maximal percent change of force of constriction in mM (%).
eN=No e€ect in mM.
New analogues design, synthesis and bioassay. Accord-
ing to CoMFA and CoMSIA analyses (Fig. 2), we can
see that adding bulky and hydrophobic groups to the
phenyl moiety of phenoxyethane of dofetilide may
increase the bioactivity. Following this clue, we designed
nine new analogues (compounds VIa i, Table 3). Scheme 1
depicts the synthetic sequence of these compounds.
Compound I was substituted by using p-nitrophenethyl-
amine hydrobromide (II), giving the key intermediates
III. N-Methylation was performed with satisfactory
yields in formate with formaldehyde, giving compound
IVa. The analogues were prepared as described in step b.
N-Alkylation of compound III with benzyl chloride, p-
chlorobenzyl chloride, allyl bromide, benzoyl chloride
and acetic anhydride was conducted by re¯uxing with
Conclusion
In summary, 3D-QSAR analyses have been performed
on 17 methylsulfonamido phenylethylamine analogues6,7
by use of CoMFA and CoMSIA methods. The 3D-
QSAR models proved
a good predictive ability.
According to the clues provided by 3D-QSAR analyses,
we designed and synthesized a series of new analogues
of methanesulfonamido phenylethylamine (VIa i). Phar-
macological assay indicated that the e€ective concentra-
tions of delaying FRP 10 ms of these new compounds
have a good correlation to the 3D-QSAR predicted
data. It is remarkable that the maximal percent change
of delaying FRP in mM of compound VIc is much
higher than that of dofetilide, but it reduced cardiac
muscle constriction, which may be due to blocking of
complex K+/Ca2+ channel. The results showed that
experimental data correlates well with predicted activity,
indicating that the 3D-QSAR models are reliable.
K2CO3 in acetone and a€orded compounds IVb i
,
respectively, which were then reduced with Fe in hydro-
chloric acid. Finally, the analogues of general structure
VI were furnished by methylsulfonation with triethyl-
amine. Compounds of this class are thermally unstable
and should be handled in an ice bath.
Acknowledgements
With completion of the synthesis, the in vitro K+
channel inhibitory activity of these nine target com-
pounds as compared with dofetilide was measured.
According to the method of Wettwer et al.,13 the e€ec-
tive concentrations of nine methanesulfonamido phenyl-
ethylamine analogues upon increasing FRP in isolated
animal atrium were evaluated utilizing a pair-electric
stimulus technique. The results are summarized in Table 3.
It showed that compound VIa prolonged FRP, reduced
heart rate and strengthened cardiac muscle constriction.
The level of prolonging FRP is equivalent to that of
dofetilide in mM, MaxÁFRP=24.02%. It is remarkable
that the maximal percent change of delaying FRP in
mM of compound VIc is 58%, which is higher than that
of dofetilide (Table 3), but it reduced cardiac muscle
constriction, which may be due to blocking of complex
K+/Ca2+ channel. Further investigations are currently
in progress.
The authors gratefully acknowledge Dr. Xuntian Jiang,
Dr. Dongxiang Liu, Dr. Xiaojian Tan and Dr. Feng
Cheng for enthusiastic help.
References and Notes
1. Selnick, H. G.; Liverton, N. J.; Baldwin, J. J.; Butcher,
J. W.; Claremon, D. A.; Elliott, J. M.; Freidinger, R. M. J.
Med. Chem. 1997, 40, 3865.
2. Carmeliet, E.; Mubagwa, K. Prog. Biophys. Mol. Biol.
1998, 70, 1.
3. Rasmusson, R. L.; Morales, M. J.; Wang, S. Circ. Res.
1998, 82, 739.
4. Moses, C. Pharmacotherapy 1997, 20, 84.
5. Morgan, T. K.; Sullivan, M. E. Prog. Med. Chem. 1996, 29,
65.

H. Liu et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2153±2157
2157
6. Liu, H.; Ji, M.; Liu, L. G.; Jin, S.; Hua, W. Y. Journal of
China Pharmaceutical University 1999, 30, 7.
7. Ji, M.; Liu, H.; Jin, S.; Yan, B.; Liu, L. G.; Zhu, D. Y.
Journal of China Pharmaceutical University 1998, 29, 161.
8. Cramer, D. R., III; Paterson, D. E.; Bunce, J. D. J. Am.
Chem. Soc. 1988, 110, 5959.
10. SYBYL. Computer Program, Version 6.5: Tripos Associates,
St. Louis, MO, 1998.
11. Kiyama, R.; Tamura, Y.; Watanabe, F. J. Med. Chem.
1999, 42, 1723.
12. Matter, H.; Schwab, W.; Barbier, D. J. Med. Chem. 1999,
42, 1908.
9. Klebe, G.; Abraham, U.; Mietzner, T. J. Med. Chem. 1994,
37, 4130.
13. Wettwer, E.; Scholtysik, G.; Schaad, A.; Himmel, H.;
Ravens, U. J. Cardiovasc. Pharmacol. 1991, 17, 480.