Antitumor effects of curcumin and structurally β-diketone modified analogs on multidrug resistant cancer cells

Article abstract of DOI:10.1016/j.bmcl.2007.11.021

Using concepts of bioisostery a series of curcumin analogs were synthesized: the diketonic system of the compound was elaborated into enaminones, oximes, and the isoxazole heterocycle. The cell growth inhibitory and apoptosis inducing effects of the new analogs were evaluated by in vitro assays in the hepatocellular carcinoma HA22T/VGH cells, as well as in the MCF-7 breast cancer cell line and in its multidrug resistant (MDR) variant MCF-7R. Increased antitumor activity on all cell lines was found with the isoxazole analog and especially with the benzyl oxime derivative; in the HA22T/VGH cell model, the latter compound inhibited constitutive NF-κB activation.

Full text of DOI:10.1016/j.bmcl.2007.11.021

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 845–849
Antitumor effects of curcumin and structurally b-diketone
modified analogs on multidrug resistant cancer cells
Daniele Simoni,a,* Michele Rizzi,a Riccardo Rondanin,a Riccardo Baruchello,a
Paolo Marchetti,a Francesco Paolo Invidiata,b Manuela Labbozzetta,c Paola Poma,c
Valeria Carina,c Monica Notarbartolo,c Alessandra Alaimoc and Natale D’Alessandroc
a
`
Dipartimento di Scienze Farmaceutiche, Universita di Ferrara, Via Fossato di Mortara 17/19, 44100 Ferrara, Italy
b
`
Dipartimento Farmacochimico Tossicologico e Biologico, Universita di Palermo
`
Dipartimento di Scienze Farmacologiche, Universita di Palermo, Via del Vespro 129, 90127 Palermo, Italy
c
Received 18 September 2007; revised 2 November 2007; accepted 8 November 2007
Available online 13 November 2007
Abstract—Using concepts of bioisostery a series of curcumin analogs were synthesized: the diketonic system of the compound was
elaborated into enaminones, oximes, and the isoxazole heterocycle. The cell growth inhibitory and apoptosis inducing effects of the
new analogs were evaluated by in vitro assays in the hepatocellular carcinoma HA22T/VGH cells, as well as in the MCF-7 breast
cancer cell line and in its multidrug resistant (MDR) variant MCF-7R. Increased antitumor activity on all cell lines was found with
the isoxazole analog and especially with the benzyl oxime derivative; in the HA22T/VGH cell model, the latter compound inhibited
constitutive NF-jB activation.
Ó 2007 Elsevier Ltd. All rights reserved.
Drug resistance, either innate or acquired and especially
in its multiple form (multidrug resistance, MDR), re-
mains a major and difficult problem to resolve in the
therapy of many cancer types. Today, it is recognized
that it is most often a multifactorial and heterogeneous
process, due to many possible molecular alterations,
including the over-expression of multidrug efflux trans-
porters (like, e.g., P-glycoprotein [ABCB1, P-gp]) and
of factors (like BCL-2, Bcl-XL, and Inhibitor of Apopto-
sis Proteins [IAPs]), which prevent tumor cell death
from drugs and other therapeutic stimuli.1,2
Among these agents, curcumin (diferuloylmethane),
extracted from Curcuma longa L. and present in curry
spice, has a long story of use in Indian medicine for
anti-inflammatory and other therapeutic purposes; it
has exhibited definite tumor suppressive and preventive
activities in many in vitro or in vivo models (for a re-
view, see, e.g., Ref.3). The ability of curcumin to selec-
tively induce apoptosis in cancer cells contributes to
underline its high antitumor potential; moreover, the
molecule has been described to efficiently induce cell
growth inhibition and apoptosis even in different
MDR tumor models.4–7 Thus, many research groups
have considered curcumin as a very good lead com-
pound to design a large variety of analogs as potential
new antitumor drugs; nevertheless, the results of the
structural modifications have been only seldom investi-
gated within the tumor drug resistance aspect.4,8
The multiplicity of the drug resistance determinants
poses the question of the optimal strategies to face them;
reasonably, multi-targeted agents might be less likely to
encounter problems of drug resistance than single-tar-
geted ones. In this respect, there is an increasing evi-
dence that the dietary polyphenols are endowed with a
remarkable number of different antitumor mechanisms,
yet accompanied by a limited toxicity for normal cells.
Curcumin is endowed with a diketone function, which
appears to be important for its antitumor activity.9–11
For example, depending on the dose, the compound
may show either pro-oxidant or anti-oxidant effects,
which both may, at least in part, be linked to this struc-
tural moiety.12 In the lower concentration, ‘chemopre-
Keywords: Curcumin oxime derivatives; Cell growth inhibition; MDR
breast cancer cells; NF-jB inhibition.
Corresponding author. Tel.: +39 0532 455923; fax: +39 0532
455953; e-mail: smd@unife.it
*
ventive’, range, curcumin is
a documented anti-
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.11.021

846
D. Simoni et al. / Bioorg. Med. Chem. Lett. 18 (2008) 845–849
oxidant, although this activity may possibly provide
negative interactions with other chemotherapeutic
agents.13 At higher concentrations, the a, b-unsaturated
ketone, as a Michael acceptor, can form adducts with
the –SH groups and generate reactive oxygen species.14
This may lead to induction of apoptosis through differ-
ent possible mechanisms involving endoplasmic reticu-
lum stress, loss of mitochondrial membrane potential,
activation of terminal caspases or also other mitochon-
dria- and caspase-independent pathways.6,15,16 Inhibi-
tion of the signaling by the transcription factor
NF-jB, and suppression of the Bcl-2 and IAP family
proteins, is also considered to be important for the
pro-apoptotic activities of curcumin.16
in contrast to MCF-7, MCF-7R lacks estrogen receptor
alpha expression, is estrogen-insensitive and over-ex-
presses P-gp, different IAPs and COX-2.20 Curcumin
and its derivatives were tested also in the hepatocellular
carcinoma HA22T/VGH cell line, which is known to in-
nately produce remarkable amounts of drug resistance
and anti-apoptotic factors.16
Scheme 1 summarizes the procedures used to prepare
the curcumin analogs described in this paper. The
enaminone analogs 2–6 were prepared in 25–35% yield,
adapting literature procedures.21 The previously pre-
pared acetate salt of the opportune amine was sus-
pended in benzene and to the mixture were added
curcumin and acetic acid. The mixture was heated at re-
flux with water removal using a Dean–Stark apparatus.
Spectroscopic NMR data of compounds 2–6 in dimeth-
ylsulfoxide-d6 and CDCl3 showed an exchangeable 1H
peak at 11.8 together with a CH singlet at 5.6–5.7 ppm
as confirmation of an intramolecular hydrogen bond
due to enaminone tautomerism.22
Here we planned to evaluate the consequences of bio-
isosteric substitutions at the 1,3 dicarbonyl fragment
of curcumin with enamino- and oxime functions with re-
gard to its antitumor activity on some cancer cell lines.
Curcumin may exist in equilibrium between the diketo
and keto-enol forms; the keto-enol form is strongly fa-
vored by intramolecular H-bonding.17 Thus, a series of
structurally related b-enaminoketone derivatives 2–6 of
curcumin were envisaged as suitable bioisostere analogs
also stabilized by intramolecular H-bonding (Fig. 1).18
Additionally, we utilized the bioisosteric substitution
of the ketone moieties with oximes to obtain derivatives
8 and 9.
Since curcumin appears to have the possibility to react
with both the 1,3-diketone moiety and the a, b-unsatu-
rated portions, the isoxazole analog 7 was prepared
mainly to confirm the exact structure of the synthesized
enaminones. Compound 7 was prepared by adding
hydroxylamine hydrochloride to a solution of curcumin
in ethanol and heating the resulting mixture at reflux
temperature for 4 h.23 Spectral data were in agreement
with data reported in the literature.
Finally, the isoxazole derivative 7, which was used as a
precursor of the enaminone 2, was also included in the
pool of derivatives tested in our biological assays. The
compound has been previously reported as a good anti-
oxidant and cyclooxygenase (COX) inhibitory agent.19
Ring opening of the isoxazole system of compound 7 by
treatment with molybdenum hexacarbonyl (Mo(CO)6)
in acetonitrile and some drops of water afforded in
appreciable yield (75%) the b-enaminoketone 2, identi-
cal to compound obtained with previous procedure.
Similarly, having this information, we assigned the same
structure to the derivatives obtained with more complex
amino reagents.
We examined the cytotoxic and pro-apoptotic effects of
curcumin and of its new analogs in the breast carcinoma
line MCF-7 and its MDR variant MCF-7R. The latter
cell line was established treating the wild type cells with
gradually increasing concentrations of doxorubicin. The
concentration of doxorubicin which causes 50% inhibi-
tion of cell growth (IC50) in MCF-7R is 17.1 lM, which
is about 92 times higher than the original IC50. Further,
The dioxime analogs 8 and 9 were synthesized by the addi-
tion of a solution of curcumin and O-methyl or O-benzyl-
hydroxylamine hydrochloride (2 equiv)in 25% water–eth-
anol to a solution of potassium carbonate in water. The
mixture was refluxed for 25 min, cooled, and concen-
trated in vacuo to remove most of the ethanol to give a res-
idue. Chromatographic separation from unreacted
curcumin led us to obtain a mixture of analogs 8 or 9 in
40% and 32% yield, respectively.24 Structural evaluation
of isolated mixtures was complicated by a multiplicity
of isomers deriving from possible different stereochemis-
try ofboth oxime functions. A univocal ESI mass determi-
nation confirmed the expected molecular weight but, at
the same time, HPLC analysis showed nearly unseparable
three peaks, and NMR data give signal multiplicity.25
Moreover, all evidences from NMR spectra in dimethyl-
sulfoxide-d6 and CDCl3 are suggesting that, differently
from compounds 2–6, no intramolecular hydrogen bond-
ing was present in 8 and 9 structures.
H
O
O
HO
OH
OCH3
OCH3
Curcumin (1)
R
H
O
N
HO
OH
OCH3
OCH3
2-6
Figure 1. Curcumin exists in solution as an equilibrium mixture of the
symmetrical dienone and the keto-enol tautomer stabilized by intra-
molecular H-bonding. Enaminones are also tautomers stabilized by
intramolecular H-bonding.
Curcumin 1 and its derivatives 2–9 were examined for
their activity toward the MCF-7 breast cancer cell line

D. Simoni et al. / Bioorg. Med. Chem. Lett. 18 (2008) 845–849
847
R
O
HN
1) NH2-R; CH3COOH;
MeOH
2) Benzene; Δ
HO
OH
OCH3
2. R = H;
OCH3
3. R = CH2Ph
Mo(CO)6;
Acetonitrile/H2O
4. R = CH(CH3)2;
5. R = CH(CH3)Ph
6. R = CH2CH3
O
OH
N
O
NH2OH * HCl;
EtOH
HO
OH
HO
OH
OCH3
Curcumin (1)
OCH3
OCH3
OCH3
7
OR OR
N
N
NH2-O-R * HCl;
K2CO3; H2O/EtOH
HO
OH
8. R = CH2Ph
9. R = CH3
OCH3
OCH3
Scheme 1.
and its MDR variant MCF-7R. Through cell growth
inhibition assays,16 we observed that the b-enaminoke-
tone compounds bearing a N-alkyl substituent 2, 4,
and 6 had comparable or somewhat lower cytotoxic
activity than curcumin in MCF-7 and in MCF-7R
(Table 1). On the other hand, compounds 3 and 5, bear-
ing an aromatic portion at the amine substituent, were
found to be not active. Similar results were observed
also in the hepatocellular carcinoma cell line HA22T/
VGH. Differently, the results obtained with the isoxaz-
ole 7 were of more interest. The IC50 and IC70 values
showed its greater potency, which was about twice that
of curcumin both in the MCF-7 and in the MCF-7R cell
line. For the dioxime derivatives 8 and 9, the results of
Table 1 showed a difference between the methyl deriva-
tive 9 and the benzyl compound 8, being the latter con-
sistently more active, in particular the most active
among all the compounds tested in the cell lines. Appar-
ently, there seems to be an inverse correlation between
activity of enaminones versus oximes and their alkyl/
aryl substituent nature.
Evaluations of cell death induction by flow cytometry
analysis16 yielded results in good agreement with the
cytotoxicity data. In particular, Figures 2 and 3 show
that in all the cell lines the order of potency in determin-
ing cell death was: compound 8> compound 7> curcu-
min. There was also further evidence that curcumin
and compounds 7 and 8 exert in MCF-7R cells antitu-
mor effects comparable to those achieved in MCF-7
(Fig. 3). The results on PI staining of DNA were very
well confirmed by flow cytometry assessments of Annex-
in V binding to cell surface (not shown).
As said before, inhibition of NF-jB signaling may be
important for the proapoptotic activities of curcumin.
Since, in contrast to MCF-7 and MCF-7R cells,
HA22T/VGH, show constitutive activation of NF-
Table 1. Cell growth inhibitory effects of the compounds evaluated after 72 h of treatment by MTS assays
Compound
HA22T/VGH
MCF-7
MCF-7R
IC50 (lM)
IC70 (lM)
IC50 (lM)
IC70 (lM)
IC50 (lM)
IC70 (lM)
Curcumin (1)
17.4 1.2
33.0 1.4
>100
24.6 1.8
41.1 2.8
>100
29.3 1.7
49.9 4.6
>100
51.7 2.2
68.1 5.2
>100
26.2 1.6
54.7 4.6
>100
46.2 3.9
81.8 6.7
>100
2
3
4
5
6
7
8
9
26.3 2.9
>100
29.8 3.1
>100
70.3 8.0
>100
93.5 7.2
>100
>100
49.8 6.6
>100
79.5 8.4
>100
26.1 3.3
12.8 1.5
5.9 1.2
32.9 1.2
64.4 5.5
18.3 1.7
8.4 0.6
40.4 2.6
93.0 6.7
13.1 1.6
7.1 0.2
67.6 7.4
12.0 2.0
9.3 1.7
88.8 5.4
24.7 2.7
12.9 1.9
52.8 6.8
23.2 2.9
9.2 0.6
43.4 1.9
36.3
2
39
5
Data are expressed as the concentrations which inhibit 50% (IC50) or 70% (IC70) cell growth and are means SE of at least three separate
experiments. Data of curcumin (1) compared with compounds 7 and 8 have been highlighted with bold characters.

848
D. Simoni et al. / Bioorg. Med. Chem. Lett. 18 (2008) 845–849
Figure 2. Induction of cell death in HA22T/VGH cells. (A) The cells were incubated with the compounds for 48 h and thereafter cell death was
evaluated by flow cytometry analysis of cell DNA stained with propidium iodide. Data are means SE of two separate experiments; (B) A
representative analysis. The percentages of the events in the preGo-G1 position are indicated in the panels.
Figure 4. Western blot analysis of the levels of survivin, XIAP, Bcl-2
and Bcl-XL in HA22T/VGH cells treated for 24 h with the compounds.
Lane 1: control; lane 2: curcumin (25 lM); lane 3: compound 7
(25 lM); lane 4: compound 8 (20 lM). A repeat experiment gave very
Figure 3. Induction of cell death in MCF-7 and MCF-7R cells. The
similar results.
cells were incubated with the compounds for 48 h and thereafter cell
death was evaluated by flow cytometry analysis of cell DNA stained
with propidium iodide. Data are means SE of three separate
experiments.
an isoxazole or pyrazole ring increases the inhibitory ef-
fects of curcumin on COX-2.19 This enzyme, strongly in-
volved in many aspects of tumor progression and drug
jB,16,20 we examined in the latter cells the effects of cur-
cumin and of compounds 7 and 8 on the DNA-binding
activity of NF-jB (p65 subunit). At the concentration of
20 lM, only compound 8 consistently reduced (33.7%
and 64.3% of the control after 1.5 and 4 h, respectively)
NF-jB activation in HA22T/VGH cells. Further, NF-
jB is able to up-regulate different anti-apoptotic factors,
including Bcl-2, Bcl-XL and IAP family members like
survivin and XIAP.16,26 Western blot analyses (Fig. 4)
showed that after 24 h of treatment of HA22T/VGH
cells, curcumin and compounds 7 and 8 reduced the lev-
els of these factors; the decreases were particularly pro-
nounced in the case of compound 8.
resistance, was previously found to be abundantly ex-
pressed in our cell lines.16,20 In other studies, the pyra-
zole analog of curcumin inhibited with high selectivity
and potency the proliferation of bovine aortic endothe-
lial cells, thus candidating as a potential anti-angiogenic
agent,27 or strongly increased the cytotoxicity against
various tumor cell lines.28
Our data underline that curcumin, as well as the deriva-
tives studied herein, including the potent compounds 7
and 8, exert unaltered antitumor activity on breast can-
cer cells with a MDR phenotype due to different mech-
anisms together. In fact, it is noteworthy that in the
HA22T/VGH model the marked pro-apoptotic activity
of compound 8 appeared to match with its ability to in-
hibit constitutive NF-jB activation as well as the expres-
sion of anti-apoptotic factors like survivin, XIAP, Bcl-2,
and Bcl-XL. Also curcumin and compound 7 decreased,
although less strongly, the same factors, but not NF-jB,
The data of this work show that small modifications can
give access to compounds presenting antitumor activi-
ties superior to those of curcumin. Further, they suggest
that the diketone fragment of curcumin is not indispens-
able for these activities. Interestingly, others have re-
ported that replacement of the diketone fragment by

D. Simoni et al. / Bioorg. Med. Chem. Lett. 18 (2008) 845–849
849
13. Mitchell, T. M. Cancer Res. 2003, 63, 5165.
14. Atsumi, T.; Tonosaki, K.; Fujisawa, S. Arch. Oral Biol.
2006, 51, 913.
suggesting a mechanism different from the interference
with the transcription factor.
15. Pae, H. O.; Jeong, S. O.; Jeong, G. S.; Kim, K. M.; Kim,
H. S.; Kim, S. A.; Kim, Y. C.; Kang, S. D.; Kim, B. N.;
Chung, H. T. Biochem. Biophys. Res. Commun. 2007, 353,
1040.
16. Notarbartolo, M.; Poma, P.; Perri, D.; Dusonchet, L.;
Cervello, M.; D’Alessandro, N. Cancer Lett. 2005, 224, 53.
17. Litwinienko, G.; Ingold, K. U. J. Org. Chem. 2004, 69,
5888.
Overall, the structure of curcumin may represent an
important basis for the development of very effective
anticancer agents even in MDR tumors, also without
retaining the diketone moiety. Moreover, some modifi-
cations at this portion have led to obtain more informa-
tion about different biological activities on several
cancer cell lines. A better understanding of these mech-
anisms may be of interest to evaluate the importance of
novel curcumin derivatives within the tumor drug resis-
tance aspect.
18. Gilli, P.; Bertolasi, V.; Ferretti, V.; Gilli, G. J. Am. Chem.
Soc. 2000, 122, 10405.
19. Selvam, C.; Jachak, S. M.; Thilagavathi, R.; Chakraborti,
A. K. Bioorg. Med. Chem. Lett. 2005, 15, 1793.
20. Poma, P.; Notarbartolo, M.; Labbozzetta, M.; Maurici,
A.; Carina, V.; Alaimo, A.; Rizzi, M.; Simoni, D.;
D’Alessandro, N. Int. J. Mol. Med. 2007, 20, 329.
21. Baraldi, P. G.; Simoni, D.; Manfredini, S. Synthesis 1983,
902.
Acknowledgment
This paper was supported by PRIN MIUR 2005051211
2005.
22. As an example, the spectroscopic characterization of
compound 3: 1H NMR (400 MHz, CDCl3): d 3.91 and
3.93 (2s, 6H); 4.62 (d, J = 6.4 Hz, 2H); 5.60 (s, 1H); 5.80
and 5.84 (2s br, 2H); 6.63 (d, J = 16 Hz, 1H); 6.65 (d,
J = 16 Hz, 1H); 6.88–6.91 (m, 3H); 6.99 (d, J = 8 Hz, 1H);
7.06–7.12 (m, 2H); 7.21 (d, J = 16 Hz, 1H); 7.35 (m, 1H);
References and notes
1. D’Alessandro, N.; Flugy, A.; Borsellino, N.; Travali, S.
Oncol. Rep. 1997, 4, 207.
1
7.36 (m, 4H); 7.47 (d, J = 16 Hz, 1H); 11.77 (br, 1H). H
NMR (DMSO-d6): d 3.82 (s, 6H); 4.73 (d, J = 6.4 Hz, 2H);
5.71 (s, 1H); 6.72–6.80 (m, 3H); 6.97–7.01 (m, 2H); 7.08 (d,
J = 8 Hz, 1H); 7.22-7.38 (m, 9H); 9.34 and 9.48 (2s br,
2H); 11.76 (t, J = 6 Hz, 2H). 13C NMR (CD3OD): d 48.2,
55.4, 111.3, 111.4, 116.4, 117.8, 123.2, 123.5, 127.3, 128.4,
128.7, 129.0, 129.3, 129.4, 138.7, 140.5, 149.3, 165.3, 187.1.
ESI Mass [M+1]. Found: 458.5.
2. Reed, J. C. J. Clin. Oncol. 1999, 17, 2941.
3. Manson, M. M. Eur. J. Cancer 2005, 41, 1842.
4. Ramachandran, C.; You, W. Breast Cancer Res. Treat.
1999, 54, 269.
5. Chearwae, W.; Anuchapreeda, S.; Nandigama, K.;
Ambdukar, S. V.; Limtrakul, P. N. Biochem. Pharmacol.
2004, 68, 2043.
23. Simoni, D.; Invidiata, F. P.; Rondanin, R.; Grimaudo, S.;
Cannizzo, G.; Barbusca, E.; Porretto, F.; D’Alessandro,
N.; Tolomeo, M. J. Med. Chem. 1999, 42, 4961.
6. Bielak-Mijewska, A.; Piwocka, K.; Magalska, A.; Sikora,
E. Cancer Chemother. Pharmacol. 2004, 53, 179.
7. Lin, Y. G.; Kunnumakkara, A. B.; Nair, A.; Merritt, W.
M.; Han, L. Y.; Armaiz-Pena, G. N.; Kamat, A. A.;
Spannuth, W. A.; Gershenson, D. M.; Lutgendorf, S. K.;
Aggarwal, B. B.; Sood, A. K. Clin. Cancer Res. 2007, 13,
3423.
8. Wortelboer, H. M.; Usta, M.; van Zanden, J. J.; van
Bladeren, P. J.; Rietjens, I. M.; Cnubben, N. H. Biochem.
Pharmacol. 2005, 69, 1879.
9. Simon, A.; Allais, D. P.; Duroux, J. L.; Basly, J. P.;
Durand-Fontanier, S.; Delage, C. Cancer Lett. 1998, 129,
111.
10. Robinson, T. P.; Ehlers, T.; Hubbard, R. B.; Bai, X.;
Arbiser, J. L.; Goldsmith, D. J.; Bowen, J. P. Bioorg. Med.
Chem. Lett. 2003, 13, 115.
24. Buchi, G.; Vederas, J. C. J. Am. Chem. Soc. 1972, 94,
¨
9128.
25. Characterization of 8 as isolated: 1H NMR (CDCl3): d
3.80–3.87 (m, 8H); 5.12, 5.23 and 5.26 (3s, 4H); 5.69 and
5.75 (2s br, 2H); 6.54 and 6.62 (2d, J = 16 Hz each, 1H);
6.67–6.84 (m, 4H); 6.88–6.95 (m, 3H); 7.01–7.08 (m, 1H);
7.20 (s, 1H); 7.24 (m, 2H); 7.30–7.42 (m, 8H).IR: 1601,
1592, 1510, 1452, 1427, 1365, 1259, 1012, 696. ESI Mass
[M+1]. Found 579.3.
26. Sethi, G.; Ahn, K. S.; Xia, D.; Kurie, J. M.; Aggarwal, B.
B. J. Immunol. 2007, 179, 1926.
27. Shim, J. S.; Kim, D. H.; Jung, H. J.; Kim, J. H.; Lim, D.;
Lee, S. K.; Kim, K. W.; Ahn, J. W.; Yoo, J. S.; Rho, J. R.;
Shin, J.; Kwon, H. J. Bioorg. Med. Chem. 2002, 10, 2987.
28. Ishida, J.; Ohtsu, H.; Tachibana, Y.; Nakanishi, Y.;
Bastow, K. F.; Nagai, M.; Wang, H. K.; Itokawa, H.; Lee,
K. H. Bioorg. Med. Chem. 2002, 10, 3481.
11. Mosley, C. A.; Liotta, D. C.; Snyder, J. P. Adv. Exp. Med.
Biol. 2007, 595, 77.
12. Ahsan, H.; Parveen, N.; Khan, N. U.; Hadi, S. M. Chem.
Biol. Interact 1999, 121, 161.