Synthesis and rational design of anti-inflammatory compounds: N-phenyl-cyclohexenyl sulfonamide derivatives

Article abstract of DOI:10.1002/poc.1575

The synthesis of (±)-ethyl 6-[N-(2-chloro-4-fluorophenyl)sulfamoyl] cyclohex-1-ene-1-carboxylate (5n) has been reproduced from a method previously described and served as the background for the preparation of a nitro derivative, potentially useful as an a

Full text of DOI:10.1002/poc.1575

Research Article
Received: 16 October 2008,
Revised: 6 March 2009,
Accepted: 4 May 2009,
Published online in Wiley InterScience: 29 June 2009
(www.interscience.wiley.com) DOI 10.1002/poc.1575
Synthesis and rational design of
anti-inflammatory compounds:
N-phenyl-cyclohexenyl sulfonamide
derivatives
Gustavo R. Lloret , Alvaro Cunha Neto , Roberto Rittner ,
a
a
a
´
a
b
c
*
Michelle Bitencourt , Matheus P. Freitas and Nilton S. Aquino
The synthesis of (W)-ethyl 6-[N-(2-chloro-4-fluorophenyl)sulfamoyl]cyclohex-1-ene-1-carboxylate (5n) has been repro-
duced from a method previously described and served as the background for the preparation of a nitro derivative,
potentially useful as an anti-inflammatory agent. Furthermore, a structure-based QSAR analysis of a series of
N-arylsulfamoyl congeners derived a highly predictive model for the activities of novel small-molecule inhibitors
of NO and cytokine production, whose preparation may be successfully achieved according to a similar procedure as
above. Copyright ß 2009 John Wiley & Sons, Ltd.
Keywords: anti-inflammatory compounds; MIA-QSAR; N-arylsulfamoyl derivatives
INTRODUCTION
EXPERIMENTAL
Synthesis and NMR
Sepsis is a systemic inflammatory response syndrome induced by
bacterial infection, and is assumed to be one of the main causes
of death in intensive care units, with mortality rates between
30 and 70%.[1–3] It is caused by bacterial components which
induce host immune cells to activate and stimulate the
production of various kinds of mediators, such as nitric oxide
(NO) and cytokines.[4–7] Thus, regulation of proinflammatory
mediator production would be a critical process to combat
various inflammatory diseases.
Ethyl (6R)-6-[N-(2-chloro-4-fluorophenyl)sulfamoyl]cyclohex-1-ene-
1-carboxylate (TAK-242, Fig. 1) has been recently implemented as
a small molecule cytokine production inhibitor.[8] It is derived
from a series of alkyl 6-(N-substituted sulfamoyl)cyclohex-1-ene-
1-carboxylates previously reported;[9] a structure-based analysis
revealed that a chlorine atom at 2-position in the phenyl ring
causes a positive effect on bioactivity, as does a fluorine atom at
the 4-position. This suggests that a different electronegative
substituent, such as the nitro group, should lead to similar
biological activity. In addition, the congeneric series offers the
possibility of building a ligand-based QSAR model, which may be
useful to derive novel compounds with high anti-inflammatory
potency.
The synthesis of ethyl (ꢀ)-6-[N-(2-chloro-4-fluorophenyl)sulfamoyl]
cyclohex-1-ene-1-carboxylate (5n), the racemic form of TAK-242,
was reproduced from the procedure described elsewhere.[8]A
similar procedure was performed to obtain the 3-nitro derivative,
according to the following description (Scheme 1). H and 13C
1
NMR spectra were acquired on a Varian GEMINI 300 spectrometer,
operating at 300.07 and 75.45 MHz, respectively. Samples were
prepared as solutions of 20 mg of solute in 0.6 ml of DMSO-d6
solvent, and the spectra were recorded at 300 K.
Ethyl 2-mercaptocyclohex-1-ene-1-carboxylate (2)
Hydrogen sulfide (H2S) gas was bubbled through a solution of
ethyl 2-oxocyclohexanecarboxylate (1) (20 g, 117 mmol) in
*
Correspondence to: Professor, M. P. Freitas, Departamento de Qu´ımica, Uni-
versidade Federal de Lavras—UFLA, Caixa Postal 3037, 37200-000 Lavras, MG,
Brazil.
E-mail: matheus@ufla.br
´
G. R. Lloret, A. C. Neto, R. Rittner, M. Bitencourt
˜
Chemistry Institute, State University of Campinas, 13084-971 Campinas, Sao
Paulo, Brazil
a
b
c
Among the structure-based QSAR methods, MIA-QSAR (multi-
variate image analysis applied to quantitative structure–activity
relationships) has provided highly predictive models, comparable
to 2D classical and sophisticated 3D methodologies.[10–16] Thus,
the goal of this work was to synthesize a novel derivative of the
highly potent TAK-242, and also to develop a QSAR model based
on the scaffold of this compound, in order to propose new
congeners with significantly higher activity.
M. P. Freitas
Departamento de Qu´ımica, Universidade Federal de Lavras, 37200-000
Lavras, Minas Gerais, Brazil
N. S. Aquino
Faculty of Medical Sciences, State University of Campinas, 13081-970
˜
Campinas, Sao Paulo, Brazil
J. Phys. Org. Chem. 2009, 22 1188–1192
Copyright ß 2009 John Wiley & Sons, Ltd.

SYNTHESIS AND MIA-QSAR OF ARYL-SULFAMOYL COMPOUNDS
Ethyl 6-[N-(3-nitrophenyl)sulfamoyl]cyclohex-1-ene-1-
carboxylate (5)
Compound 3 (2.0 g, 8.3 mmol) obtained above was dissolved in
thionyl chloride (5.4 ml) and heated under reflux for 14 h, and
then the reaction mixture was evaporated under reduced
pressure to dryness. The residue was subjected three times to
a procedure involving an addition of hexane (7.7 ml) followed by
evaporation under reduced pressure to dryness to yield ethyl
2-(chlorosulfonyl)cyclohex-1-ene-1-carboxylate (1.2 g; 4.7 mmol).
This residue (1.2 g) was combined with ethyl acetate (4 ml), and
the resultant mixture was added to a solution of 3-nitro-aniline
(1.0 g, 6.73 mmol), Et3N (1 ml, 6.73 mmol), and ethyl acetate
(10 ml) and then stirred for 18 h. The reaction mixture was diluted
with ethyl acetate (9.5 ml), and the entire mixture was then
worked up [water (40 ml), brine (20 ml ꢂ 3)]. The residue was
crystallized from diisopropyl ether (1.5 ml), and the precipitate
was washed with diisopropyl ether (2.0 ml) and ethyl acetate
Figure 1. Structure of TAK-242
ethanol (300 ml) at ꢁ50 8C for 2 h, and then hydrogen chloride
(HCl) gas was bubbled into the reaction mixture at ꢁ20 8C for 2 h,
followed by the bubbling of H2S gas at ꢁ20 8C for 2 h. After
having been allowed to stand for 14 h, the reaction mixture was
poured into ice water (200 ml) and worked up [hexane (600 ml);
water (5 ꢂ 160 ml)]. The residue was purified by distillation (bp
136–139 8C/15–16 mmHg (bp 134–135 8C/15–16 mmHg) to give
2 (14.2 g, 65%) as a colorless oil. 1H NMR (CDCl3), d (ppm): 1.30 (3H,
t, J ¼ 7.1 Hz), 1.55–1.70 (4H, m), 2.17–2.30 (4H, m), 4.20 (2H, q,
J ¼ 7.1 Hz), 12.3 (1H, s). 13C NMR (CDCl3), d (ppm): 14.3 (CH3), 21.9
(C-6), 22.4 (C-4 and C-5), 29.1 (C-3), 60.1 (CH2-O), 97.8 (C-1), 172.0
1
(1.6 ml) to yield 5 (120 mg, 4.1%) as colorless crystals: H NMR
(CD2Cl2), d (ppm): 1.28 (3H, t, J ¼ 7.1 Hz), 1.62–1.85 (2H, m),
1.89–2.40 (4H, m), 4.20 (2H, q, J ¼ 7.1 Hz), 4.32 (1H, d, J ¼ 4.2 Hz),
5.34 (1H, s), 7.38 (1H, t, J ¼ 3.8 Hz), 7.68 (1H, ddd, J ¼ 1.0, 2.1,
—
(C-2), 172.8 (C O).
—
8.2 Hz), 8.0 (1H, ddd, J ¼ 1.0, 2.1, 8.2 Hz), 8.18 (1H, t, J ¼ 2.0 Hz). 13
C
NMR (CDCl3), d (ppm): 13.6 (CH3), 20.9 (C-12), 24.7 (C-11), 27.2
(C-10), 59.3 (C-7), 64.4 (CH2), 112.8 (C-2), 116.4 (C-4), 124.8 (C-6),
129.2 (C-5), 129.6 (C-8), 130.6 (C-9), 140.6 (C-1), 147.3 (C-3), 169.4
2-(Ethoxycarbonyl)cyclohex-1-ene-1-sulfonic acid (3)
A solution of 2 (7.0 g, 37.6 mmol) in AcOH (12.5 ml) was added
dropwise to a stirred solution of sodium borate tetrahydrate
(18.6 g, 121 mmol) in AcOH (100 ml) for 2 h at 50–55 8C. The
resulting mixture was stirred for 3 h at 50–55 8C and for 9 h at
80–85 8C, and then cooled and evaporated under reduced
pressure to dryness. Acetonitrile (165 ml) was added to the
residue, and the whole residue was stirred for 1 h followed by
filtration to remove the insoluble substances. After the filtrate
was evaporated under reduced pressure to dryness, the residue
was dissolved in MeCN (125 ml) and then stirred for 2 h. The
resulting precipitate was filtered off, and the filtrate was
evaporated under reduced pressure to dryness. Diisopropyl
ether (330 ml) was added to the residue to obtain 3 (5.6 g, 63%) as
a white powder: 1H NMR (DMSO-d6), d (ppm): 1.17 (3H, t,
J ¼ 7.1 Hz), 1.50 (4H, br), 2.08–2.09 (2H, m), 2.22–2.24 (2H, m), 3.99
(2H, q, J ¼ 7.1 Hz). 13C NMR (CDCl3), d (ppm): 13.9 (CH3), 21.2 (C-4),
21.7 (C-5), 24.3 (C-3), 27.2 (C-6), 59.7 (CH2-O), 128.2 (C-1), 139.9
—
(C( O)).
—
MIA-QSAR modeling
MIA descriptors are binaries, i.e., pixels of 2D images; in QSAR,
such images are chemical structures drawn using an appropriate
drawing program, where the substituents attached to
a
scaffold correspond to the explained variance in the data. In
this work, the 25 anti-inflammatory compounds (Table 1), a series
of N-arylsulfamoyl derivatives obtained from the literature,[9]
were drawn using the ChemSketch program.[17] Subsequently,
the chemical structures were transformed into bitmaps and saved
in a 370 ꢂ 300 pixels workspace. Since the dataset used is a
congeneric series, chemical structures contain a common
substructure, which was superimposed for the entire series. This
was achieved by taking a pixel in common among the whole
series of compounds and fitting it in a given coordinate of the
defined workspace. This 2D alignment is rapid and requires only
—
(C-2), 170.5 (C O).
—
Scheme 1. Synthesis of ethyl 6-[N-(3-nitrophenyl)sulfamoyl]-cyclohex-1-ene-1-carboxylate (5). Reagents: (a) H2S/HCl, EtOH; (b) NaBO3.4H2O, acetic acid;
(c) SOCl2; (d) 3-nitroaniline, Et3N, AcOEt
J. Phys. Org. Chem. 2009, 22 1188–1192
Copyright ß 2009 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

G. R. LLORET ET AL.
Table 1. Compounds used in the MIA-QSAR modeling, and the respective experimental, calibrated, and predicted biological
activities (pic50)
Compound
R
Exp.
Cal.
LOO CV
L-20%-O CV
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
5l
5m
5n
5o
5p
5q
5r
2-F, 4-Cl
H
2-F
3-F
4-F
2-Cl
3-Cl
4-Cl
2,3-F2
2,6-F2
2,4-F2
2,4,5-F2
2,3-Cl2
2-Cl, 4-F
2-Cl, 4-Me
2-Cl, 4-CN
2-Et
2-CO2Me
—
—
2-F, 4-Cl
H
4-F
6.796
6.585
7.125
6.824
6.959
7.921
7.180
6.398
6.854
6.796
7.796
7.523
7.699
8.495
7.387
5.796
6.886
5.959
5.000
5.000
5.770
5.086
5.854
5.387
6.638
6.885
6.674
7.152
6.757
7.410
7.540
6.848
6.432
7.241
6.819
7.958
7.669
7.291
8.262
7.479
5.752
6.868
5.851
4.834
5.105
5.624
5.490
5.649
5.433
6.681
6.935
6.729
7.142
6.647
7.621
7.267
6.440
6.658
7.395
7.116
7.930
7.515
6.974
7.828
7.323
6.809
6.881
6.375
5.210
4.677
5.818
6.017
5.619
5.266
6.986
6.898
6.891
7.195
6.316
7.466
7.036
6.194
6.962
7.381
7.142
7.948
7.373
7.078
7.885
6.939
6.956
6.717
6.384
5.398
4.725
5.931
5.864
5.631
5.387
6.973
5s
5t
6a
6b
6e
7e
8
—
—
simple manual precision. The 25 2D images were appropriately
read and converted into double arrays by using Matlab.[18] These
25 samples were then grouped to give a 25 ꢂ 370 ꢂ 300
three-dimensional array, and then unfolded to give an X-matrix
of 25 ꢂ 111 000 dimension. The MIA-QSAR model was built by
calibrating the X-matrix with the corresponding dependent
variables (inhibitory activities on NO production—pIC50, IC50 in
mol ꢃ Lꢁ1) by using the PLS regression method. Model validation
was achieved through cross-validation (LOO and L-20%-O), and
the predictive ability was statistically evaluated through the root
www.interscience.wiley.com/journal/poc
Copyright ß 2009 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 1188–1192

SYNTHESIS AND MIA-QSAR OF ARYL-SULFAMOYL COMPOUNDS
mean square errors of calibration (RMSEC) and cross-validation
(RMSECV), as well as by the squared correlation coefficients of the
regression line of experimental versus fitted (r2) or predicted (q2)
activity values.
RESULTS AND DISCUSSION
It has been found that a chlorine atom bonded to the 2-position
in the phenyl ring of the titled aryl-sulfamoyl compounds plays a
key role in enhancing the biological activities of the congeners in
Table 1; a fluorine in the 4-position activates the compound and
makes ethyl 6-[N-(2-chloro-4-fluorophenyl)sulfamoyl]cyclohex-
1-ene-1-carboxylate the most active compound of Table 1 (the
R enantiomer, TAK-242, has been separated and shown to display
a higher potency). In addition to the available data for the
anti-inflammatory activity of TAK-242, preliminary bioassays with
this drug have shown a decrease in the Jun NH2-terminal kinase
(JNK) phosphorylation in the liver and adipose and muscular
tissues, indicating that TAK-242 reduces the anti-inflammatory
sub-clinic process in obese animals. Treatment for 14 days
decreased significantly the phosphorylation in the liver (65%,
p < 0.05), skeletal muscle (67%, p < 0.05) and adipose tissue (35%,
p < 0.05). Further biological experiments with the 3-nitro
derivative and others are currently in progress.
The synthetic route applied to obtain TAK-242 was reproduced,
and then used to derive a model example, the 3-nitro derivative,
in order to show that the synthesis may be expanded to other
analogous systems. 3-Nitro-aniline was chosen as the starting
reagent for the synthesis because the nitro substituent is
electronegative and reasonably voluminous, such as the chlorine
and fluorine of TAK-242, and is supposed to affect the bioactivity
similarly. The synthetic sequence of Scheme 1 may be used to
carry out syntheses of novel compounds, which are based on the
basic structure of Table 1. Such new entities may be appropriately
rationalized by combining the substructures, in this case the
substituents, of a congeneric series of aryl-sulfamoyl com-
pounds—those of Table 1; their activities can be computationally
predicted by using a suitable structure-based QSAR method.
MIA-QSAR has been applied elsewhere to give highly predictive
models,[10–16] and then it has been used here to model the
activities of the compounds of Table 1. Moreover, some
prototypes were proposed and their activities predicted by
using the PLS regression parameters.
Figure 2. Plot of number of latent variables versus RMSEC/RMSECV
and substituents which have the greatest influence on the
biological activities. The novel compounds proposed have di- and
tri-substituted phenyl rings (substituents ¼ F and Cl, at the 2, 4,
and/or 5 ring-position), as shown in Fig. 4. These respective
positions for the substituents were calibrated, and thus the
predictions should provide reliable (within RMSECV) calculated
activities. The predicted pIC50 values for the prototype
compounds A, B, and C of Fig. 4 were 6.717, 6.847, and 7.099,
which are within the range of moderately and highly active
compounds of the dataset. This suggests that the proposed new
compounds are potentially useful as starting systems for further
structure modification.
Overall, the MIA-QSAR model built was found to be highly
predictive for the series of aryl-sulfamoyl compounds presented.
It may be applied to the prediction of novel congeners, whose
synthesis is expected to proceed successfully by using a similar
procedure as described in the literature and tested here for a
model system (the 3-nitrophenyl derivative). Furthermore,
the computational approach demonstrated its potential to
drive synthesis and, in combination with experimental pro-
cedures, is capable of diminishing time and costs during drug
development.
First, the optimum number of latent variables (linear
combinations of the original variables—the pixels) to be used
in the model was achieved by analyzing the variance in RMSEC
and RMSECV when changing the number of PLS components
(Fig. 2); the minimum error was found for five latent variables. The
calibration step for the entire dataset of Table 1 gave the fitted
pIC50 values of Table 1, corresponding to a correlation with
the experimental data of r2 0.945 (RMSEC ¼ 0.218) (Fig. 3). The
calibration model was validated through leave-one-out and
leave-20%-out cross-validations; in this latter validation pro-
cedure, samples were randomly selected into five segments. The
model was found to be satisfactorily predictive, with q2
of
LOO
0.757 (RMSECV ¼ 0.462) and q2L-20%-O of 0.694 (RMSECV ¼ 0.517),
according to the predicted values of Table 1, as illustrated in
Fig. 3.
Given the prediction ability of the QSAR model, the PLS
regression parameters were used to estimate the bioactivities of
novel structures, which are variations of the common scaffold
Figure 3. Correlation of experimental versus calibrated/predicted pIC50
J. Phys. Org. Chem. 2009, 22 1188–1192
Copyright ß 2009 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

G. R. LLORET ET AL.
Figure 4. Proposed compounds with inhibitory activities on NO production, as modeled through MIA-QSAR
[5] C. E. Hack, L. A. Aarden, L. G. Thijs, Adv. Immunol. 1997, 66, 101–195.
[6] L. C. Casey, R. A. Balk, R. C. Bone, Ann. Intern. Med. 1993, 119, 771–778.
[7] N. C. Riedemann, R. F. Guo, P. A. Ward, Nature Med. 2003, 9, 517–524.
[8] M. Ii, N. Matsunaga, K. Hazeki, K. Nakamura, K. Takashima, T. Seya, O.
Hazeki, T. Kitazaki, Y. Iizawa, Mol. Pharmacol. 2006, 69, 1288–1295.
[9] M. Yamada, T. Ichikawa, M. Ii, M. Sunamoto, K. Itoh, N. Tamura, T.
Kitazaki, J. Med. Chem. 2005, 48, 7457–7467.
Acknowledgements
FAPEMIG (grant: CEX 415/06) and FAPESP (grant: 2005/59649-0)
are gratefully acknowledged for financial support of this research,
and for a fellowship (A.C.N.), as is CNPq for studentships (to G.L.R.
and M.B.) and fellowships (to R.R. and M.P.F.).
[10] M. P. Freitas, S. D. Brown, J. A. Martins, J. Mol. Struct. 2005, 738,
151–154.
[11] M. P. Freitas, Org. Biomol. Chem. 2006, 4, 1154–1159.
[12] M. P. Freitas, Med. Chem. Res. 2007, 16, 461–467.
REFERENCES
[13] J. R. Pinheiro, M. Bitencourt, E. F. F. da Cunha, T. C. Ramalho, M. P.
Freitas, Bioorg. Med. Chem. 2008, 16, 1683–1690.
[1] M. S. Rangel-Frausto, D. Pittet, M. Costigan, T. Hwang, C. S. Davis, R. P.
Wenzel, JAMA 1995, 273, 117–123.
[2] G. S. Martin, D. M. Mannino, S. Eaton, M. Moss, N. Engl. J. Med. 2003,
348, 1546–1554.
[3] A. P. Wheeler, G. R. Bernard, N. Engl. J. Med. 1999, 340, 207–214.
[4] E. S. Van Amersfoort, T. J. C. Van Berkel, J. Kuiper, Clin. Microbiol. Rev.
2003, 16, 379–414.
[14] M. P. Freitas, Chemom. Intell. Lab. Sys. 2008, 91, 173–176.
[15] M. P. Freitas, Curr. Comput. Aided Drug Des. 2007, 3, 235–239.
[16] M. P. Freitas, R. Rittner QSAR Comb. Sci. 2007, 27, 582–585.
[17] ACD/ChemSketch Version 10. 02, Advanced Chemistry Develop-
ment, Inc., Toronto, Ontario, Canada, 2006.
[18] Matlab Version 7.5, MathWorks Inc., Natick, MA, 2007.
www.interscience.wiley.com/journal/poc
Copyright ß 2009 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 1188–1192