Synthesis, antimycobacterial activity evaluation, and QSAR studies of chalcone derivatives

Article abstract of DOI:10.1016/j.bmcl.2006.12.112

In order to develop relatively small molecules as antimycobacterial agents, twenty-five chalcones were synthesized, their activity was evaluated, and quantitative structure-activity relationship (QSAR) was developed. The synthesis was based on the Claisen

Full text of DOI:10.1016/j.bmcl.2006.12.112

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1695–1700
Synthesis, antimycobacterial activity evaluation,
and QSAR studies of chalcone derivatives
P. M. Sivakumar,a S. Prabu Seenivasan,b Vanaja Kumarb and Mukesh Doblea,*
aDepartment of Biotechnology, Indian Institute of Technology Madras, Chennai, Tamilnadu 600036, India
bTuberculosis Research centre (ICMR), Chetpet, Chennai, Tamilnadu 600031, India
Received 12 July 2006; revised 7 December 2006; accepted 22 December 2006
Available online 12 January 2007
Abstract—In order to develop relatively small molecules as antimycobacterial agents, twenty-five chalcones were synthesized, their
activity was evaluated, and quantitative structure–activity relationship (QSAR) was developed. The synthesis was based on the Cla-
isen-Schimdt scheme and the resultant compounds were tested for antitubercular activity by luciferase reporter phage (LRP) assay.
Compound C24 was found to be the most active (ꢀ99%) in this series based on the percentage reduction in Relative Light Units at
both 50 and 100 lg/ml levels, followed by compound C21. Four compounds at the 50 lg/ml and eight compounds at the 100 lg/ml
showed activity above 90% level. QSAR model was developed between activity and spatial, topological, and ADME descriptors for
the 50 lg/ml data. The statistical measures such as r, r2, q2, and F values obtained for the training set were in acceptable range
and hence this relationship was used for the test set. The predictive ability of the model is satisfactory (q2 = 0.56) and it can be used
for designing similar group of compounds.
Ó 2007 Elsevier Ltd. All rights reserved.
Tuberculosis is a major and challenging health problem
around the world and the current survey says that about
two million deaths are caused annually due to the highly
infective acid fast bacilli Mycobacterium tuberculosis.1–3
Tuberculosis is mostly asymptomatic and is aggravated
when impairment of immunity arises due to conditions
like malnutrition, diabetes, malignancy, and AIDS.
The last condition is susceptible to Mycobacterium
avium complex (MAC).4 Treatment of tuberculosis is a
complex process due to several factors which include
patient’s inability to persist with combined treatment
regimen, the spreading ability of the nontubercular
mycobacteria (NTM) like M. avium complex (MAC),
the ineffectiveness of the drugs on immunosuppressed
patients, and MDR (Multi Drug Resistance).5 Strepto-
mycin was introduced in 1944 for the treatment of tuber-
culosis which was followed by first line tuberculosis
drugs such as isoniazid, rifampicin, ethambutol, and
pyrizinamide, and later by second line of drugs. Multi-
drug regimen was found to be effective for the treatment
of tuberculosis (Isoniazid + Rifampicin + Pyrizina-
mide + Ethambutol). MDR is observed in all front line
and second line of drugs. Resistance toward isoniazid
has arisen due to mutation of katG and inhA, resistance
toward rifampicin due to mutation of rpoB-gene, resis-
tance toward pyrizinamide due to mutation of pncA
gene, etc.6 Over the past twenty years design of new
drugs to address drug resistance is in progress, with
few successes.
Use of small molecules in the current treatment regimen
encouraged us to work on chalcone derivatives.7 More-
over licochalcone, a natural product obtained from Glyc-
yrrhiza inflata, showed low MICs (between 5 and 20 mg/
L) against three of the species such as M. tuberculosis,
M. avium, and Mycobacterium bovis.8 Chalcones and
flavanoids have also been found to be effective against
M. tuberculosis H37Rv which proved that 20-hydroxy
substitution favors antimycobacterial activity.9 Chal-
cones are known to posses antibacterial activity and,
dimethylamino chalcones have been reported as iNOS
inhibitor10 (literature evidence points that NOS2 gene
expression is involved in the production of NO in
primary TB).11 This has encouraged us to include
methylthio and dimethyl amino substitutions at the R3
position. In addition, effect of methoxy substitution at
various positions in the A- and B-rings has also been
studied in this paper (see Scheme 1).
The following synthetic procedure has been adopted for
the preparation of twenty-five prop-2-en-1-one
derivatives.12
Keywords: Chalcones; Antimycobacterial activity; QSAR.
*
Corresponding author. Tel.: +91 44 2237 4107; fax: +91 44 2237
4102; e-mail: mukeshd@iitm.ac.in
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.112

1696
P. M. Sivakumar et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1695–1700
O
O
O
2M NaOH
H
+
MeOH
3 hrs
Scheme 1. Synthetic route for prop-2-en-1-one derivatives.
Aldehyde (10 mmol) and acetophenone (10 mmol) deriv-
atives were added and stirred in methanol (10 ml) at
room temperature. When 2 M NaOH was added under
rapid stirring to the above mixture, in most of the cases
an almost pale yellowish solid precipitated out. The solid
was washed with 50% alcohol, followed by water and
then dried. In majority of the cases the yield was more
than 80%. Synthesized compounds were characterized
by FT-IR, NMR, and mass spectrometry. Spectral data
for compound 17 (C17) given in reference.28
inhibition is tabulated in Table 1. A compound is con-
sidered to be an antimycobacterial agent if fifty percent
reduction in the Relative Light Units (RLU) is observed
when compared to the control using a luminometer.
Reports indicate that substituted (either with electron-
donating or electron-withdrawing groups) or unsubsti-
tuted A-ring did not affect the biological activity.14
But A-ring with hydrophobic substitution showed in-
creased anti-tubercular activity.9 In the present study,
introduction of substitutions like –NMe2, –OMe,
–SMe irrespective of their positions in the A-ring exhib-
ited higher activity when compared to the parent com-
pound C4 (without any substitution in the A-ring).
Hydrophilic substituents such as –NO2 or –OH in B-ring
Antimycobacterial activity of the synthesized com-
pounds was evaluated by luciferase reporter phage assay
method13 against M. tuberculosis H37Rv at two concen-
trations (50 and 100 lg/ml) and the observed percentage
Table 1. Structures and experimental antitubercular activity of chalcone derivatives
R1
R5
O
R2
R6
A
B
R3
R7
R4
Compound
A-ring
R3
B-ring
R6
% Reduction in RLU
ClogP
R1
R2
R4
R5
R7
50 lg/ml (data)
100 lg/ml (data)
C1
C2
NMe2
SMe
SMe
70.02
46.29
30.95
41.93
31.59
7.36
70.18
66.51
48.06
50.64
98.86
17.38
90.52
51.59
48.83
0
3.789
4.183
4.088
3.624
4.057
4.403
4.517
4.682
4.972
4.088
3.543
5.451
3.543
5.122
2.924
4.057
4.550
3.537
3.187
3.694
3.448
5.057
4.042
3.417
3.448
C3
C4
NO2
C5
C6
SMe
SMe
SMe
SMe
SMe
SMe
OMe
SMe
OH
OMe
C7
C8
OH
Cl
68.41
16.88
24.06
0
Me
Cl
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25
NO2
58.75
0
65.08
0
Cl
Br
OMe
66.07
0
72.98
21.87
63.9
SMe
OMe
SMe
OMe
OMe
OMe
OMe
43.16
29.25
0
OH
Cl
55.16
12.17
96.45
89.27
96.22
98.15
33.8
OMe
OMe
OMe
NMe2
Cl
OMe
OMe
NO2
NO2
NO2
92.19
64.61
90.06
94.30
31.55
60.24
98.90
84.16
NO2
Cl
NMe2
OMe
OMe
Cl
Me
98.52
99.04
97.94
OH
NO2
OMe

P. M. Sivakumar et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1695–1700
1697
enhanced the antimycobacterial activity. Eight com-
pounds namely C5, C7, C18, C20, C21, C23, C24, and
C25 exhibited greater than 90% inhibition activity at
100 lg/ml level. In this group of compounds except for
C7, C21, and C23, other five compounds have substitu-
tion at the R6 position. Compound C19 with 89.3%
activity with –NO2 at the R6 position just falls outside
the above group. Nine compounds show inhibition
activity between 50% and 90%, and hence can also be
considered active. In this group, except for compounds
C8, C16, and C19 other six compounds do not have
any substitution in the B-ring. Halogen substitution at
the R7 position inactivates the chalcone derivatives (C9
and C14), and introduction of an additional halogen in
the B-ring increases the hydrophobicity of the B-ring
reducing the activity further (C12, C17, and C22). More-
over, nitro and methoxy substitution in the B ring at
R7 position yields inactive compounds (C3 and C6) at
the 100 lg/ml level. In brief at 50 lg/ml level, 11 com-
pounds are active in nature, out of which four com-
pounds (C18, C20, C21, and C24) have more than 90%
activity and the rest of the 14 compounds are inactive
(less than 50% inhibition).
introduction of hydrophobic substituents in the B-ring.
Cluster 4 has compounds whose activity is in the range
of 0–10%. From these observations it is evident that
antimycobacterial activity totally depends on the substi-
tution in the B-ring and substitution in the A-ring has
little impact.
The structure of the various molecules as listed in Table 1
was drawn and the minimum energy conformation was
determined using Cerius2 softwareÒ using Universal
force field (Acceryls Inc., USA). Two hundred and forty
nine descriptors that included topological, charge, geo-
metrical, aromaticity indices, constitutive properties,
quantum mechanics, and thermodynamics were evaluat-
ed for each compound. Several literature reports give a
very detailed description of descriptors.17–19 Equations
were developed between the observed activity and
the descriptors. The set of descriptors that would give
the statistically best models (r, r2, q2 > 0.5) were selected
from the large pool using a Genetic function approach.
While r2 is an indication of the model data fit, q2 is an
indication of the predictive capability of the model.
QSAR was developed between the observed activity
(Table 2) and various molecular descriptors for the
50 lg/ml data set. Twenty-one compounds from this
data set were divided into training and test sets, the
former set consisting of 16 randomly chosen compounds
and the remaining in the latter set. Four compounds
with RLU equal to zero were not included for develop-
ing the QSAR. The model developed using the training
set was used to predict the activity of the compounds in
the test set. The multi-linear regression model with 16
compounds gave very good fit. The best model was
selected based on the r, r2, F-ratio, and q2. The predic-
tive capability of the equation is determined using
leave-one-out cross-validation method. The relation
for q2 is as shown below,
It is reported that there is a relationship between % inhi-
bition and ClogP.7,15 Percentage inhibition increased
with increase in ClogP. Lipinski’s rule states that com-
pounds with ClogP greater than 5 will not be active,
which is also observed in our studies (C12, C14, and
C22 are inactive). ClogP is an indication of the lipophil-
icity of molecules, and more lipophilic molecules can
easily enter the lipid-enriched mycobacterial cell wall.16
Cluster analysis for antimycobacterial activity at 50 lg/
ml is shown in Figure 1. The figure shows four clusters
based on their activity at 50 lg/ml. Clusters 1 and 2 con-
tain moderate and highly active compounds. Cluster 2
has compounds with activity in the range of 80–100%
reduction in RLU due to their hydrophilic substituents
at the B-ring. Fifty percent of compounds in Cluster 1
does not have any substitution in B-ring. Clusters 3
and 4 have inactive compounds, which is due to the
q2 ¼ 1 ꢁ PRESS=TOTAL
P
2
ðY predicted ꢁ Y observed
Þ
is the predictive error sum of
P
2
squares (=PRESS).
sum of squares (=TOTAL), where Ypredicted, Yobserved
ðY observed ꢁ Y mean
Þ
is the total
,
and Ymean are the predicted, observed, and mean values
of activity, respectively. The activity is defined as log(p/
100 ꢁ p), where p is the percent reduction in RLU at the
50 lg/ml.
The best QSAR is given in Table 3 with corresponding
r2, q2, F ratio, and standard error between the model
predictions and data. Figure 2 shows a comparison of
the experimental and predicted activity for the 16 com-
pounds taken in the training set at 50 lg/ml concentra-
tion level. Table 2 lists the values of the selected
descriptors and the model-predicted activity of the train-
ing set (16 compounds). Cross-correlation between these
selected descriptors is given in Table 4. A low cross-cor-
relation indicates that the least correlated descriptors are
chosen for developing the QSAR, which is generally
desired. Figure 3 shows a comparison of the experimen-
tal and predicted activity for the five compounds select-
ed in the test set with the 95% confidence regions. Here
Figure 1. Cluster analysis of compounds at 50 lg/ml concentration
level.

1698
P. M. Sivakumar et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1695–1700
Table 2. Values of the selected descriptors and the predicted activity of the training set (16 compounds)
Compound
ADME_solubility_level
CHI-V-1
Shadow-Zlength
Observed activity = log(p/100 ꢁ p)
Predicted activity
Residual
C2
C3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
2
6.93398
7.43342
5.18617
7.06827
7.07425
7.34466
5.70923
7.06827
6.74369
6.74369
6.72031
6.21465
7.17614
6.11991
5.84352
6.21465
5.24383
5.3824
ꢁ0.065
ꢁ0.349
ꢁ0.141
ꢁ0.336
0.336
ꢁ0.283
ꢁ0.536
0.292
ꢁ0.317
0.63
0.218
0.187
C4
C5
4.02785
5.3608
7.6019
ꢁ0.433
ꢁ0.019
ꢁ0.294
ꢁ0.134
0.089
C7
C8
5.20032
4.26311
6.10955
7.41913
7.31911
6.47091
7.47266
6.10791
5.10662
5.11937
5.44659
ꢁ0.692
0.154
ꢁ0.558
0.065
0
C11
C16
C18
C19
C20
C21
C22
C23
C24
C25
ꢁ0.384
1.072
0.261
ꢁ0.384
0.313
0.759
0.716
0.371
1.112
ꢁ0.068
0.167
1.954
0.252
ꢁ0.455
0.586
0.107
0.957
1.219
ꢁ0.336
0.18
1.954
ꢁ0.268
0.013
0
0.473
0.725
Table 3. QSAR of 50 lg/ml data
n
r2
r2adj
q2
F
Standard error
ꢁ1.398 + 1.609 ADME_solubility_level ꢁ 0.624 CHI-V-1 + 0.424 Shadow-Zlength 16 0.81 0.76 0.56 17.08 0.325
*
*
*
n, number of molecules in the data set; r2, squared correlation coefficient; q2, cross-validated squared correlation coefficient.
2.5
2
1.5
1
0.5
0
-1
-0.5
0
0.5
1
1.5
2
2.5
-0.5
-1
Observed activity
Figure 3. Comparison of observed and predicted activities of the
compounds in the test set (dotted lines indicate 95% confidence limits).
Figure 2. Comparison of observed and predicted activities of the
compounds in the training set.
Table 4. Matrix of correlation between the selected descriptors
ADME_solubility_level
CHI-V-1
Shadow-Zlength
Activity
ADME_solubility_level
CHI-V-1
1
ꢁ0.311042
ꢁ0.173082
0.616005
1
Shadow-Zlength
Activity
0.430808
ꢁ0.447219
1
0.326398
1
the activity was predicted using the QSAR developed
with the training set. The r2 (squared correlation coeffi-
cient between this observed and the predicted activity)
was found to be 0.70.
have been observed by us for chalcones, chalcone-like
compounds, flavones, and flavonones.20 ADME_solu-
bility_level indicates the aqueous solubility of the mole-
cule. It is reported that aqueous solubility of organic
compounds can be predicted from molecular descriptors
such as number of H-bond donor and acceptors, dipole
moment, number of rotatable bonds, and surface area.21
Aqueous solubility plays an important role in drug
absorption. Palm et al. explained that the dynamic polar
ADME_solubility_level, CHI-V-1, and shadow-Zlength
are the three descriptors that appear in the QSAR at
50 lg/ml concentration level. Contributions of ADME
properties and connectivity indices toward TB activity

P. M. Sivakumar et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1695–1700
1699
S.; Brennan, P. J. Antimicrob. Agents Chemother. 1995, 39,
2484; (c) Telenti, A.; Philipp, W. J.; Sreevatsan, S.;
Bernasconi, C.; Stockbauer, K. E.; Wieles, B.; Musser, J.
M., ; Jacobs, W. R., Jr. Nat. Med. 1997, 3, 567; (d)
Sreevatsan, S.; Pan, X.; Zhang, Y.; Kreiswirth, B. N.;
Musser, J. M. Antimicrob. Agents. Chemother. 1997, 41,
636.
molecular surface area (PSAd) is also an ideal surface
descriptor for the differentiation of drugs which in turn
considers three-dimensional shape and flexibility of the
molecules22. Tangallapally et al. observed that hydro-
philic substitutions improved the solubility of cyclic sec-
ondary
amine-substituted
phenyl
and
benzyl
nitrofuranyl amides. The improved solubility improved
the antimycobacterial activity.23
7. Ballell, L.; Field, R. A.; Duncan, K.; Young,
R. J. Antimicrob. Agents. Chemother. 2005, 49,
2153.
CHI-V-1 is the valence modified connectivity index,
which is an indication of the number of bonds in the
molecule. Importance of connectivity indices in virtual
screening of new active agents against M.avium complex
is explained by Garcia et al.24 Shadow indices are a set
of geometric descriptors that characterizes the shape of
the molecule and Shadow-Zlength is the length of the
molecule in the Z dimension. This shows positive contri-
bution in the QSAR equation, which means that
increasing the length of the molecule (also related to
hydrophobicity) increases the antimycobacterial activity
of the compound.
8. Kanokmedhakul, S.; Kanokmedhakul, K.; Phonkerd, N.;
Soytong, K.; Kongsaeree, P.; Suksamrarn, A. Planta Med.
2002, 68, 834.
9. Lin, Y. M.; Zhou, Y.; Flavin, M. T.; Zhou, L. M.; Nie,
W.; Chen, F. C. Bioorg. Med. Chem. 2002, 10, 2795.
10. Rojas, J.; Dominguez, J. N.; Charris, J. E.; Lobo, G.;
Paya, M.; Ferrandiz, M. L. Eur. J. Med. Chem. 2002, 37,
699.
11. Macmicking, J. D.; North, R. J.; Lacourse, R.; Mudgett,
J. S.; Shah, S. K.; Nathan, C. F. Proc. Natl. Acad. Sci.
U.S.A. 1997, 94, 5243.
12. Lin, J.; Rivett, D. E.; Wilshire, J. F. K. Aust. J. Chem.
1977, 30, 629.
13. Riska, P. F.; Su, Y.; Bardarov, S.; Freundlich, L.; Sarkis,
G.; Hatfull, G.; Carriere, G.; Vanaja Kumar; Chan, J.;
Jacobs, W. R., Jr. J. Clin. Microbiol. 1999, 37, 1144, This
procedure is optimized and modified by Dr. Vanaja
Kumar Tuberculosis Research Centre, Chetpet, Chennai,
India. Fifty-microliter bacterial suspension equivalent to
MacFarlands No.2 standard was added to 400 ll of G7H9
with and without the test compound. For each sample,
two drug-free controls and two drug concentrations were
prepared and this set up was incubated for 72 h at 37 °C.
After incubation 50 ll of the high titer Luciferase reporter
phage (phAE129) and 40 ll of 0.1 M CaCl2 were added to
all the vials and this setup was incubated at 37 °C for 4 h.
After incubation 100 ll of the mixture was taken from
each tube into a luminometer cuvette and equal amount of
working D-luciferin (0.3 mM in 0.05 M sodium citrate
buffer, pH 4.5) solution was added. The RLU was
measured after 10 s of integration in the Luminometer
(Monolight 2010). Duplicate readings were recorded for
each sample and the mean was calculated. The percentage
reduction in the RLU was calculated for each test sample
and compared with control. The experiment was repeated
when the mean RLU of the control was less than 1000.
14. Liu, M.; Wilairat, P.; Croft, S. L.; Tan, A. L.; Go, M. L.
Bioorg. Med. Chem. 2003, 11, 2729.
The above short-listed descriptors describe the molecu-
lar size, degree of branching, flexibility, overall shape,
and aqueous solubility and it is known that they in turn
are related to the hydrophlicity–hydrophobicity ratio of
the molecule (logp). Our studies corroborate the previ-
ous research that hydrophobicity25–27 and solubility of
the compound play an important role in the antituber-
cular ctivity.
Acknowledgment
Authors thank Dr. Moni (SAIF, IIT) for providing
spectral analysis.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.
2006.12.112.
15. Patole, J.; Shingnapurkar, D.; Padhye, S.; Ratledge, C.
Bioorg. Med. Chem. Lett. 2006, 16, 1514.
References and notes
16. Smith, I. Clin. Microbiol. Rev. 2003, 16, 463.
17. Todeschini, R.; Lasagni, M.; Marengo, E. J. Chemom.
1994, 8, 263.
18. Todeschini, R.; Consonni, V. Handbook of Molecular
Descriptors; Wiley-VCH: Weinheim, 2000.
19. Karelson, M. Molecular Descriptors in QSAR/QSPR;
Wiley Interscience: New York, 2000.
20. Sivakumar, P. M.; Geetha Babu, S. K.; Doble, M. Chem.
Pharm. Bull. 2007, 55, 44.
21. Chen, X. Q.; Cho, S. J.; Li, Y.; Venkatesh, S. J. Pharm.
Sci. 2002, 91, 1838.
22. Palm, K.; Luthman, K.; Ungell, A. L.; Strandlund, G.;
Beigi, F.; Lundahl, P.; Artursson, P. J. Med. Chem. 1998,
41, 5382.
23. Tangallapally, R. P.; Yendapally, R.; Lee, R. E.; Lenaerts,
A. J.; Lee, R. E. J. Med. Chem. 2005, 48, 8261.
24. Garcia, A. G.; Galvez, J.; Julian, J. V.; Domenech, R. G.;
Munoz, C.; Guna, R.; Borras, R. J. Antimicrob. Chemo-
ther. 2004, 53, 65.
1. Blumberg, H. M.; Michael, K.; Leonard, J.; Jasmer, R. M.
J. Am. Med. Assoc. 2005, 293, 2776.
2. Deangelis, C. D.; Flanagin, A. J. Am. Med. Assoc. 2005,
293, 2793.
3. Dye, C.; Watt, C. J.; Bleed, B. M.; Hosseini, S. M.;
Raviglione, M. C. J. Am. Med. Assoc. 2005, 293, 2767.
4. World Health Organization. Global tuberculosis. 2002
Doc. WHO/CDS/TB/2000, 275.
5. (a) Sensi, P.; Grassi, G. In Burger’s Medicinal Chemistry
and Drug Discovery; 5th ed.; John Wiley & Sons. Inc.,
1996; Vol. 2, pp 575–635; (b) Inderlied, C. B.; Kemper, C.
A.; Bermudez, L. E. M. Clin. Microb. Rev. 1993, 6, 266; (c)
Mandell, G. L.; Petri, W. A., Jr. In Goodman and Gilman’s
The Pharmacological Basis of Therapeutics; 9th ed.;
McGraw Hill, 1996, pp 1155–1174.
6. (a) Hoffner, S. E.; Kallenius, G. Eur. J. Clin. Infect. Dis
1988, 7, 188; (b) Mikusova, K.; Slayden, R. A.; Besra, G.

1700
P. M. Sivakumar et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1695–1700
25. Ragno, R.; Marshall, G. R.; Santo, R. D.; Costi, R.;
Massa, S.; Rompei, R.; Artico, M. Bioorg. Med. Chem.
2000, 8, 1423.
26. Matyk, J.; Waisser, K.; Drazkova, K.; Kunes, J.; Klimes-
ova, V.; Palat, K. J.; Kaustova, J. IL Farmaco. 2005, 60,
399.
28. selected spectral data for C17: IR (KBr) 1659 cmꢁ1
(Carbonyl), 1H NMR (400 MHz, CDCl3): d 3.92 (s,
6 H), d 7.26–7.48 (m, 3H), d 6.88 (d, 1H), d 6.96 (d, 1H)
J = 16 Hz, d 7.08 (s, 1H), d 7.14 (d, 1H). 13C NMR
(100 MHz, CDCl3): d 192.71, 151.89, 149.31, 146.98,
137.67, 136.61, 132.24, 130.23, 130.11, 127.21, 127.17,
123.98, 123.58, 111.09, 109.98, 56.01, and 55.92. ESI MS:
[M]+ 337.20.
27. Klimesova, V.; Palat, K.; Waisser, K.; Klimes, J. Int.
J. Pharm. 2000, 207, 1.