Synthesis and QSAR study of novel cytotoxic spiro[3H-indole-3, 2′(1′H)-pyrrolo[3,4-c]pyrrole]-2,3′,5′(1H,2′aH, 4′H)-triones

Article abstract of DOI:10.1016/j.ejmech.2011.10.058

1,3-Dipolar cycloaddition reaction of 1-aryl-1H-pyrrole-2,5-diones 1a-e with non-stabilized azomethine ylides, generated in situ via decarboxylative condensation of isatins 2a-c and sarcosine (3) in refluxing ethanol, afforded 4′-aryl-5′a,6′-dihydro-1′-me

Full text of DOI:10.1016/j.ejmech.2011.10.058

European Journal of Medicinal Chemistry 47 (2012) 312e322
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis and QSAR study of novel cytotoxic spiro[3H-indole-3,20(10H)-pyrrolo
[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-triones
,
b
,
c
a
Adel S. Girgis a , Jacek Stawinski
, Nasser S.M. Ismail , Hanaa Farag
*
**
a Pesticide Chemistry Department, National Research Centre, Dokki, Cairo 12622, Egypt
b Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University, S-106 91 Stockholm, Sweden
c Pharmaceutical Chemistry Department, Faculty of Pharmacy, Ain Shams University, Cairo, Egypt
a r t i c l e i n f o
a b s t r a c t
Article history:
1,3-Dipolar cycloaddition reaction of 1-aryl-1H-pyrrole-2,5-diones 1a-e with non-stabilized azomethine
ylides, generated in situ via decarboxylative condensation of isatins 2aec and sarcosine (3) in refluxing
ethanol, afforded 40-aryl-50a,60-dihydro-10-methyl-spiro[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-
2,30,50(1H,20aH,40H)-triones 4aeo in good yields. Compound 4l exhibited high anti-tumor activity against
Received 5 August 2011
Received in revised form
27 October 2011
Accepted 28 October 2011
Available online 7 November 2011
HEPG2 (liver cancer) cell line (IC50 ¼ 12.16
m
M) compared to that of Doxorubicin (IC50 ¼ 7.36
mM), and the
other synthesized compounds revealed moderate anti-tumor properties against HCT116 (colon), MCF7
(breast) and HEPG2 (liver) human tumor cell lines. 3D-Pharmacophore modeling and quantitative
structure-activity relationship (QSAR) analysis were combined to explore the structural requirements
controlling the observed anti-tumor properties. It was found that the major structural factors affecting
potency of these compounds were related to their basic skeleton.
Keywords:
1H-pyrrole-2,5-diones
Azomethine ylides
Spiro[3H-indole-3,20(10H)-pyrrolo[3,4-c]
Ó 2011 Elsevier Masson SAS. All rights reserved.
pyrrole]
Anti-tumor
QSAR
1. Introduction
triones byutilizing1,3-dipolarcycloadditionof1-aryl-1H-pyrrole-2,5-
dione derivatives with non-stabilized azomethine ylides generated in
In the last few decades increasing interest has been directed
towards construction of spiropyrrolidinyl-oxindole containing
compounds [1]. This could be attributed to the fact that many
spiropyrrolidinyl-oxindolyl analogues have been isolated from
natural sources and identified as promising bio-active agents, e.g.,
spirotryprostatine A and spirotryprostatine B that were found to be
inhibitorsofmammaliancellcycle atG2/Mphase, fromthesecondary
metabolites of Aspergillus fugimatus [2,3]. Elacomine [4] was isolated
from Eleagnus commutata, and horsfiline [5e10], from Horsfieldia
superba, a small Malaysian tree, extracts of which have found use in
indigenous medicine. Mitraphylline was isolated from Uncaria
tomentosa (cat’s claw) and identified as anti-tumor active agent
against human brain cancer cell lines, neuroblastoma SKN-BE(2), and
malignant glioma GAMG [11]. Generally, several oxindole derivatives
are well known as anti-tumor active agents due to their kinase
inhibitory properties [12e16], especially, tyrosine kinase [12e14].
In the present work, we intended to investigate synthesis of spiro
[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-
situ via decarboxylative condensation of isatins with a-amino acids
according to the previously described procedure [17], and to evaluate
theiranti-tumor properties against avarietyof human tumorcell lines
(HCT116 “colon”, MCF7 “breast” and HEPG2 “liver” cancers). One of
the driving forces for initiating thiswork wasour previousobservation
that compounds with alkaloid heterocyclic system skeletons, such as
dispiro[1H-indene-2,30-pyrrolidine-20,300-[3H]indole]-1,200(100H)-dio-
nes and dispiro[3H-indole-3,20-pyrrolidine-30,300-piperidine]-2(1H),
400-diones, revealed promising anti-tumor properties against SK-MEL-
2 (melanoma) cell line [18], and colon (HCT-116), breast (T-47D),
leukemia [HL-60 (TB), MOLT-4, RPMI-8226] and prostate (PC-3) cell
line cancers [19]. In this work, quantitative structure-activity rela-
tionship (QSAR) studies were also performed, not only for validating
the observed pharmacological properties but also, for investigating
the most important parameters controlling these properties.
2. Results and discussion
2.1. Chemistry
* Corresponding author. Tel.: þ20 2 01220447199; fax: þ20 2 33370931.
** Corresponding author. Fax: þ46 8 15 49 08.
Reaction of 1-aryl-1H-pyrrole-2,5-diones 1aee with non-
stabilized azomethine ylides, generated in situ via decarboxylative
E-mail addresses: girgisas10@yahoo.com (A.S. Girgis), js@organ.su.se
(J. Stawinski).
0223-5234/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2011.10.058

A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
313
condensation of isatins 2aec and sarcosine (3) as a representative
example of -amino acid, afforded in refluxing ethanol only one
product as indicated by TLC analysis.The structures of the isolated
compounds were established to be 40-aryl-50a,60-dihydro-10-methyl-
spiro[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)
Single crystal X-ray studies of 4a and 4m add a conclusive
support for the assigned structures. They reveal that both the
pyrrolidinyl nuclei appear as typical five-membered heterocycles
(pentagonal structure) where, each of them seems like an
opened-envelope and fused together with angles C(12)eC(10)e
C(21) ¼ 113.5(1), C(7)eC(14)eC(11) ¼ 112.4(1)ꢀ (in case of
compound 4a) and C(14)eC(17)-C(26) ¼ 111.6(2), C(13)eC(23)e
C(15) ¼ 114.3(2)ꢀ (in case of compound 4m). The methine
protons (H-20a, H-50a) were observed in cis-configuration
[C(12)eC(10)eH(10) ¼ 111.3(1), C(14)eC(10)eH(10) ¼ 109.4(1),
C(21)eC(10)eH(10) ¼ 111.6(1), C(7)eC(14)eH(14) ¼ 112.0(2),
C(10)eC(14)eH(14) ¼ 109.5(1), C(11)eC(14)eH(14) ¼ 112.4(2)ꢀ
(in case of compound 4a) and C(14)eC(17)eH(17) ¼ 113.1(2),
C(23)eC(17)eH(17) ¼ 109.4(2), C(26)eC(17)eH(17) ¼ 112.3(2),
C(13)eC(23)-H(23) ¼ 109.0(2), C(15)eC(23)-H(23) ¼ 109.0(2),
C(17)eC(23)eH(23) ¼ 114.5(2)ꢀ (in case of compound 4m)]. The
2-oxindole ring appears as a nearly planar system linked with
the fused pyrrolo[3,4-c]pyrrol residue through the spiro-carbon
connection, “C-3 (C-20)”, (Figs. 2 and 3).
a
-triones 4aeo (Scheme 1) based on spectroscopic (IR, 1H NMR, 13
C
NMR, 1H,1H-COSY, HSQC, MS) and elemental analyses data. 1H NMR
spectrum of 4a, as a representative example of the synthesized
compounds, reveals the presence of four signal sets, due to the dia-
stereotopic methylene protons H2C-60 at
d
¼ 3.36, 3.63 ppm, and two
methine protons, H-20a and H-50a, at
d
¼ 3.55, 3.75 ppm, respectively.
1H,1H-COSY spectrum of compound 4a supports these interpreta-
tions.13C NMR (on-resonance) spectrum of 4a exhibits the methylene
(H2C-60) and methine (HC-50a, HC-20a) carbons at
d
¼ 54.1, 44.6 and
52.0 ppm, respectively. A signal from the quaternary spiro-carbon, “C-
3 (C-20)”, appears at
d
¼ 72.2 ppm, and those from the carbonyl
carbons, at
d
¼ 174.0, 177.8 and 178.1 ppm 1H,13C-heteronuclear single
quantum coherence (HSQC) spectrum of compound 4a supports
these assignments (Fig. 1). The other synthesized analogues revealed
similar spectral features (c.f. experimental section). The structural
assignments for the synthesised compounds have been substantiated
by mass spectral data (EI-MS) that reveal the corresponding parent
ion peaks as either base peaks or high relative intensity signals.
Theoretical calculations using both AM1 and PM3 methods were
carried out to compare the observed geometric parameters obtained
from single crystal X-rayanalysis with the calculated ones (Table 1 of
supplementary material). Geometries of compounds 4a and 4m
H3C
O
N+
CH2
R'
O
_
H
N
R'
O
EtOH
+
H3C
OH
O
N
H
(3)
N
H
(2)
O
O
O
N
R
(1)
CH3
N
H
N
1
1'
4'
2
5'a
2'a
O
O
N
R'
R
(4)
1a, R = Ph
4a; R = Ph, R' = H
1b, R = 4-ClC6H4
1c, R = 4-FC6H4
4b; R = Ph, R' = Cl
4c; R = Ph, R' = OCH3
1d, R = 4-H3CC6H4
1e, R = 4-H3COC6H4
4d; R = 4-ClC6H4, R' = H
4e; R = 4-ClC6H4, R' = Cl
4f; R = 4-ClC6H4, R' = OCH3
4g; R = 4-FC6H4, R' = H
2a, R' = H
2b, R' = Cl
2c, R' = OCH3
4h; R = 4-FC6H4, R' = Cl
4i; R = 4-FC6H4, R' = OCH3
4j; R = 4-H3CC6H4, R' = H
4k; R = 4-H3CC6H4, R' = Cl
4l; R = 4-H3CC6H4, R' = OCH3
4m; R = 4-H3COC6H4, R' = H
4n; R = 4-H3COC6H4, R' = Cl
4o; R = 4-H3COC6H4, R' = OCH3
Scheme 1. Synthetic route of compounds 4aeo.

314
A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
Fig. 1. 1H, 13C-Heteronuclear single quantum coherence (HSQC) spectrum of 4a.
were optimized by the molecular mechanics force field (MMþ),
followed byeither semi-empirical AM1 [20] or PM3 [21,22] methods
implemented in the HyperChem 8.0 package. The structures were
fully optimized without fixing any parameters, thus bringing all
geometric variables to their equilibrium values. The energy mini-
mization protocol employed the PolakeRibiere conjugated gradient
algorithm. Convergence to a local minimum was achieved when the
energy gradient was ꢁ0.01 kcal molꢂ1. The RHF method was used in
the spin pairing for the two semi-empirical tools [23e25].
Rhodamine-B (SRB) standard method [26]. From the observed
results (Table 1, Figs. 1e3 of supplementary material), it has been
concluded that most of the synthesized compounds had moderate
anti-tumor properties against the investigated human tumor cell
lines. In contrast, compound 4l exhibited considerable anti-tumor
activity against HEPG2 (liver cancer) cell line with potency
(IC50 ¼ 12.16
of cell growth compared to control experimental) comparable to
M), that was used as a reference
mM; concentration required to produce 50% inhibition
that of Doxorubicin (IC50 ¼ 7.36
m
standard during this study. In order to understand the observed
pharmacological properties, quantitative structure-activity rela-
tionship (QSAR) study was initiated.
2.2. Anti-tumor properties
The synthesized compounds 4a-o were screened for their anti-
tumor properties against HCT116 (colon), MCF7 (breast) and
HEPG2 (liver) human tumor cell lines utilizing the Sulfo-
2.3. QSAR study
2.3.1. 3D-QSAR pharmacophore modeling
This study was performed using Discovery Studio 2.5 software
(Accelrys Inc., San Diego, CA, USA), which permits pharmacophore
generation, structural alignment, activity prediction and 3D data-
base creation [27e29]. 3D-QSAR pharmacophore protocol was used
to generate predictive pharmacophores via aligning different
conformations in which the molecules are likely to bind with the
receptor. A given hypothesis could be combined with a known
activity data to create a 3D-QSAR model that identifies overall
aspects of molecular structure governing activity. 3D-QSAR based
on pharmacophore was constructed using collections of molecules
with activities ranging over a number of orders of magnitude.
Pharmacophores explain the variability of bioactivity with respect
to the geometric localization of the chemical features present in the
molecules. The observed HYPOGEN identifies a 3D array of
a maximum of four chemical features common to the training set
4aef,hek,meo that provides relative alignment for each input
molecule, consistent with its binding mode to a proposed common
receptor site. The chemical features considered were: hydrogen
bond donor (HBD), hydrogen bond acceptor (HBA), hydrophobic
(H) and positive ionizable features (PosIon) (Fig. 4, Table 2 exhibits
Fig. 2. ORTEP projection of single crystal X-ray diffraction of 4a.

A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
315
Fig. 3. ORTEP projection of single crystal X-ray diffraction of 4m.
constraint distances and angles between features of the generated
pharmacophores, Tables 3e5 exhibit fit values and estimated
activities of the training set due to the generated 3D-pharmaco-
phore models). The conformational flexibility of the training set
was modelled by creating multiple conformers that covered
a specified energy range for each input molecule [28,30,31].
Through the pharmacophore mapping study, it was found that the
major structural factors affecting the potency of these compounds
were related to their basic skeleton. In case of HCT116 (colon
cancer) cell line pharmacophore model, the controlling features
were two hydrogen bond acceptor sites together with a positive
charged site, while for MCF7 (breast cancer) cell line, a hydrogen
bond donor and a positive ionizable. On the other hand, both
a hydrogen bond donor and an acceptor were required for the
generated pharmacophore due to HEPG2 (liver cancer) cell line. In
addition, hydrophobic features were recognized in all the gener-
ated 3D-pharmacophore models of the tested cancer types.
Comparison of 3D-pharmacophore models generated for HCT116
(colon) and MCF7 (breast) cancers revealed that common phar-
macophoric features recognized are the positive ionizable and
hydrophobics with root mean square division (RMSD) value of
0.058. However, in case of HCT116 (colon) and HEPG2 (liver)
cancers, the hydrogen bond acceptor and hydrophobics are the
common features between them (RMSD ¼ 0.334). For tumor lines
MCF7 (breast) and HEPG2 (liver), the common pharmacophoric
features were hydrogen bond donor and hydrophobics
(RMSD ¼ 0.329) (Fig. 4DeF). Mapping of the training set analogues
into the generated 3D-QSAR pharmacophore based model is shown
in Figs. 4e6 of the supplementary material.
2.3.2. QSAR modeling
Despite of the significance of pharmacophoric hypotheses for
understanding ligand molecule affinity and 3D search queries, their
predictive value as 3D-QSAR models is generally limited by steric
shielding and bioactivity-modulating auxiliary groups (electron-
donating or withdrawing functionalities) [32e34]. Thus, a classical
Table 1
Anti-tumor properties of the synthesized compounds.
Entry
Compd.
R
R0
IC50, mg/ml (mM)
Colon cancer (HCT116)
Breast cancer (MCF7)
Liver cancer (HEPG2)
1
2
3
4
5
6
7
8
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
4k
4l
4m
4n
4o
Ph
Ph
Ph
H
Cl
OCH3
H
Cl
OCH3
H
Cl
OCH3
H
Cl
OCH3
H
Cl
OCH3
e
22.14 (63.73)
12.62 (33.05)
12.86 (34.08)
14.05 (36.80)
14.76 (35.46)
19.52 (47.40)
15.48 (42.37)
29.29 (73.26)
12.38 (31.31)
16.67 (46.13)
>50.00 (>126.31)
18.10 (46.24)
20.24 (53.63)
16.90 (41.03)
16.67 (40.92)
3.73 (6.86)
12.38 (35.64)
14.76 (38.66)
16.90 (44.78)
16.43 (43.03)
16.90 (40.60)
15.95 (38.73)
12.38 (33.88)
35.95 (89.92)
17.38 (43.96)
13.57 (37.55)
19.76 (49.92)
18.33 (46.83)
18.10 (47.96)
15.71 (38.14)
17.38 (42.66)
2.97 (5.46)
11.43 (32.90)
12.62 (33.05)
16.67 (44.17)
12.38 (32.42)
12.14 (29.16)
17.86 (43.37)
31.90 (87.31)
16.43 (41.09)
10.95 (27.69)
27.86 (77.09)
25.00 (63.16)
4.76 (12.16)
27.38 (72.55)
11.90 (28.89)
20.24 (49.68)
4.00 (7.36)
4-ClC6H4
4-ClC6H4
4-ClC6H4
4-FC6H4
4-FC6H4
4-FC6H4
4-H3CC6H4
4-H3CC6H4
4-H3CC6H4
4-H3COC6H4
4-H3COC6H4
4-H3COC6H4
e
9
10
11
12
13
14
15
16
Doxorubicin

316
A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
Fig. 4. (AeC), the generated 3D-pharmacophore model due to HCT116 (colon), MCF7 (breast) and HEPG2 (liver) cancer cell lines, respectively; (DeF) comparison between the
generated pharmacophore models of the tested cancer cell lines.
QSAR analysis was employed to search for the best combination of
orthogonal pharmacophores using a fit value and other structural
descriptors (connectivity, topological, etc.) capable of explaining
bioactivity variation across a collected list of the descriptors,
allowing different pharmacophoric models competing within the
3D-QSAR framework.
A set of 13 compounds (4aef,hek,meo) was used as a training
set for a QSAR modeling. The remaining 2 compounds (4g and 4l)
were adopted as an external test subset for validating the QSAR
models. Many molecular descriptors were calculated for each
compound employing a calculate molecular properties module.
The calculated descriptors including various simple and valence
connectivity indices, electro-topological state indices, single point
quantum-mechanical descriptors (via the AM1 model) and other
molecular descriptors, were considered. Furthermore, the training
set compounds were fitted against the corresponding pharmaco-
phore hypotheses generated by the HYPOGEN automatic runs and
their fit values (produced by the best-fit command) were added as
additional molecular descriptors. Genetic function approximation
(GFA) was employed to search for the best possible QSAR regres-
sion equation capable of correlating the variations in biological
activities of the training set compounds with variations in the
generated descriptors, i.e. multiple linear regression modeling
(MLR) [30,35]. Equations (1)e(3) show our best-performing QSAR
models (Figs. 5e7 exhibit the corresponding scatter plots of
experimental versus estimated bioactivity values for the training
set compounds against HCT116, MCF7 and HEPG2 tumor cell lines,
respectively). The goodness of the model was validated by squared
correlation coefficient (R2) and residuals between the predicted
and experimental activity of the test set and training set
(Tables 6e8). R2 values for HCT116 (colon), MCF7 (breast) and
HEPG2 (liver) cancer QSAR models are 0.906, 0.856 and 0.923,
respectively.
Equation (1)
Potency (IC50) against HCT116 (colon cancer) cell line (N “number of
molecules in the training set” ¼ 13, R2“squared correlation coeffi-
cient value” ¼ 0.906)
Table 2
Constraint distances (Å) and angles (ꢀ) between features of the generated
pharmacophores.
IC50 ¼ 1320.335e160.262 [fit value] - 14.284 [LUMO-Eigen
Cancer cell Constraint distances (Å)
line
Constraint angles (ꢀ)
value-VAMP]
HCT116
(colon)
(PosIon)-(HBA-1), 3.66;
(PosIon)-(H), 4.76;
(PosIon)-(HBA-2),
(H)-(PosIon)-(HBA-1), 25.94;
(PosIon)-(HBA-2)-(H), 109.01;
(PosIon)-(HBA-1)-(HBA-2), 48.79
Equation (2)
Potency (IC50) against MCF7 (breast cancer) cell line (N ¼ 13,
4.25; (HBA-1)-(HBA-2), 4.58;
(HBA-2)-(H), 4.37
(PosIon)-(HBD), 3.58;
(PosIon)-(H-1),
6.46; (PosIon)-(H-2), 4.60;
(HBD)-(H-2),
R2 ¼ 0.856)
MCF7
(breast)
(H-2)-(HBD)-(PosIon), 49.75;
(H-2)-(PosIon)-(H-1), 36.92;
(H-2)-(HBD)-( H-2), 22.25
IC50 ¼ 1504.55 þ 37.71 [shadow-nu] - 192.14 [fit value]
3.36; (H-1)-(H-2), 7.64
(HBD)-(H-1), 3.08;
(HBD)-(HBA), 7.20;
(HBD)-(H-2), 9.19;
(HBA)-(H-2), 4.17;
(HBA)-(H-1), 8.05
Equation (3)
HEPG2
(liver)
(H-1)-(HBD)-(HBA), 94.48;
(H-2)-(HBA)-(HBD), 66.51;
(H-1)-(HBD)-(H-2), 19.59
Potency (IC50) against HEPG2 (liver cancer) cell line (N ¼ 13,
R2 ¼ 0.923)
IC50 ¼ 882.47e13.38 [pKa] - 106.68[fit value]

A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
317
Table 3
Table 5
Best fit values and estimated activities for compounds of the training set
(4aef,hek,meo) mapped with the generated 3D-pharmacophore model due to
HCT-116 (colon) cancer cell line.
Best fit values and estimated activities for compounds of the training set
(4aef,hek,meo) mapped with the generated 3D-pharmacophore model due to
HEPG2 (liver) cancer cell line.
Entry
Compd.
Estimated
activity
Observed
activity
Relative
energy
Fit
value
Entry
Compd.
Estimated
activity
Observed
activity
Relative
energy
Fit
value
1
2
3
4
5
6
7
8
4a
4b
4c
4d
4e
4f
4h
4i
4j
4k
4m
4n
4o
64.78
40.62
42.17
43.79
37.25
35.31
60.08
34.97
64.87
83.08
42.29
53.86
35.75
63.73
33.05
34.08
36.80
35.46
47.40
73.26
31.31
46.13
>126.31
53.63
41.03
40.92
0.000
1.203
0.397
4.733
3.732
19.095
16.268
7.402
0.045
0.032
8.105
6.441
5.232
7.804
8.007
7.990
7.974
8.044
8.067
7.837
8.072
7.803
7.696
7.989
7.884
8.062
1
2
3
4
5
6
7
8
4a
4b
4c
4d
4e
4f
4h
4i
4j
4k
4m
4n
4o
38.15
33.90
35.49
51.74
35.13
33.36
45.58
31.94
94.47
58.96
57.09
25.71
37.34
32.90
33.05
44.17
32.42
29.16
43.37
41.09
27.69
77.09
63.16
72.55
28.89
49.68
6.638
1.264
1.204
6.612
0.796
14.250
0.040
0.654
0.000
4.649
8.621
14.558
2.704
7.597
7.648
7.628
7.465
7.633
7.655
7.520
7.674
7.203
7.408
7.422
7.768
7.606
9
9
10
11
12
13
10
11
12
13
2.3.3. Validation of QSAR
Searching for set descriptors (D), containing D descriptors of
optimal subset (d), where d << D ones with minimum standard
deviation (S), by means of multivariable linear regression
technique:
External validation of the determined QSAR equations was
performed utilizing two of our synthesized analogues exhibiting
mild (4g) and promising (4l) anti-tumor properties. The observed
activities and those provides by QSAR study, are presented in
Table 10.
N
X
1
S ¼
resi
(1)
ðN ꢂ d ꢂ 1Þ i ¼ 1
3. Conclusion
Where; N, is the number of molecules of the training set; resi, is the
residual for molecule; i, is the difference between the experimental
property (p) and predicted property (ppred).
1,3-Dipolar cycloaddition reaction of 1-aryl-1H-pyrrole-2,5-
diones 1aee with non-stabilized azomethine ylides, generated
in situ via decarboxylative condensation of isatins 2aec and
More precisely, the Kubinyi function (FIT) [36,37] is a statistical
parameter which is closely related to the Fisher ratio (F), but
avoids the main disadvantage of the latter of being too sensitive to
changes in small d values, and poorly sensitive to changes in large
d values. The FIT (d) criterion has a low sensitivity to changes in
small d values and a substantially increasing sensitivity for large
d values. The greater the FIT value the better the linear equation
[38]. It is given by the following equation, “where R (d) is the
correlation coefficient for a model with (d) descriptors”. The
observed FIT values are 3.24, 2.95 and 3.45 corresponding to
models due to HCT116, MCF7 and HEPG2 cancer cell lines,
respectively (Table 9).
sarcosine (3) as a representative example of a-amino acid affor-
ded 40-aryl-50a,60-dihydro-10-methyl-spiro[3H-indole-3,20(10H)-
pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-triones 4aeo in good
yields (71e90%). Most of the synthesized compounds revealed
moderate anti-tumor properties against HCT116 (colon), MCF7
(breast) and HEPG2 (liver) human tumor cell lines, while
compound 4l exhibited considerable anti-tumor activity against
HEPG2 (liver cancer) cell line (IC50 ¼ 12.16
mM), comparable to
that of Doxorubicin (IC50 ¼ 7.36 M). 3D-QSAR pharmacophore
m
modeling utilizing 13 compounds (4aef,hek,meo) as a training
set afforded an HYPOGEN of four chemical features: two
hydrogen bond acceptors, one hydrophobic and one positive
ionizable, in case of HCT116 (colon) cancer cell line. For MCF7
(breast) cancer cell line two hydrophobics, one positive ionisable,
and one hydrogen bond donor features were observed and for
HEPG2 (liver) cancer cell line two hydrophobics, one hydrogen
bond donor and one hydrogen bond acceptor features, were
recognized.
RðdÞ2ðN ꢂ d ꢂ 1Þ
FIT ¼
(2)
ꢀ
ꢁꢀ
ꢁ
N þ d2 1 ꢂ R2
Table 4
Classical QSAR studies delivered equations of two descriptors
with R2 ¼ 0.906, 0.856, 0.923, in case of HCT116 (colon), MCF7
(breast) and HEPG2 (liver) cancer cell lines, respectively. The
most important descriptor in all the equations was the fit value
derived from mapping of the synthesized analogue into the
generated pharmacophore. The other controlling descriptors
were LUMO-Eigen value-VAMP (in case of HCT116 cell line),
shadow-nu (in case of MCF7 cell line) and pKa (in case of HEPG2
cell line).
External validation of the established QSAR models utilizing two
of our synthesized analogues exhibiting mild (4g) and promising
(4l) anti-tumor properties, revealed good agreement between the
experimental and the calculated data. It can be concluded that
combination of 3D-pharmacophore modeling and QSAR provides
as an effective technique for understanding the observed phar-
macological properties and thus could be adopted for developing
effective lead structures.
Best fit values and estimated activities for compounds of the training set
(4aef,hek,meo) mapped with the generated 3D-pharmacophore model due to
MCF-7 (breast) cancer cell line.
Entry
Compd.
Estimated
activity
Observed
activity
Relative
energy
Fit
value
1
2
3
4
5
6
7
8
4a
4b
4c
4d
4e
4f
4h
4i
4j
4k
4m
4n
4o
40.82
43.25
44.51
38.43
44.64
40.07
66.43
52.69
38.34
40.93
38.49
45.16
43.07
35.64
38.66
44.78
43.03
40.60
38.73
89.92
43.96
37.55
49.92
47.96
38.14
42.66
0.000
0.252
0.617
0.000
0.091
0.848
0.040
0.319
0.000
0.226
0.000
0.108
3.914
7.997
7.972
7.959
8.023
7.958
8.005
7.785
7.886
8.024
7.996
8.022
7.953
7.974
9
10
11
12
13

318
A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
Fig. 5. Predicted versus experimental IC50 of the tested compounds against HCT116 (colon) human tumor cell line.
4. Experimental protocols
O),1618,1496. 1H NMR:
d 2.00 (s, 3H, NCH3), 3.36 (d,1H, upfield H of
CH2, J ¼ 9.9 Hz), 3.55 (d, 1H, H-20a, J ¼ 8.4 Hz), 3.63 (pseudo t, 1H,
downfield H of CH2, J ¼ 8.25 Hz), 3.75 (pseudo t, 1H, H-50a,
J ¼ 7.95 Hz), 6.84e7.56 (m, 9H, arom. H), 10.57 (s, 1H, NH). 13C NMR:
Melting points were recorded on a Stuart SMP3 melting point
apparatus. IR spectra (KBr) were recorded on JASCO 6100 and
Shimadzu 8400S FT-IR spectrophotometers. 1H NMR, as well as 2D-
NMR spectra were recorded on a Varian MERCURY VX300 (1H:
300,13C: 75 MHz) spectrometer in DMSO-d6. 13C NMR spectra were
recorded on Varian MERCURY 300BB (75 MHz) and JEOL AS 500
(125 MHz) spectrometers in DMSO-d6. Mass spectra were recorded
on a Shimadzu GCMS-QP 1000 EX (EI, 70 eV) spectrometer. The
starting compounds 1aee [39e41] were prepared according to the
previously reported procedures.
d
34.3 (NCH3), 44.6 (HC-50a), 52.0 (HC-20a), 54.1 (H2C-60), 72.2 [spiro
C-3 (C-20)], 109.7, 121.7, 125.2, 125.9, 126.8, 128.4, 129.0, 129.6, 132.3,
142.7 (arom. C), 174.0, 177.8, 178.1 (C]O). MS: m/z (%) 347 (M, 100),
318 (87), 317 (23), 227 (18), 199 (62), 198 (28), 172 (76), 171 (70).
Anal. Calcd. for C20H17N3O3 (347.38): C, 69.15; H, 4.93; N, 12.10.
Found: C, 69.36; H, 5.04; N, 12.18.
4.1.2. 5-Chloro-50a,60-dihydro-10-methyl-40-phenyl-spiro[3H-
indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-trione
(4b)
4.1. Synthesis of 40-aryl-50a,60-dihydro-10-methyl-spiro[3H-indole-
3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-triones 4aeo
(general procedure)
Reaction time 5 h, colorless crystals from n-butanol, mp
262e264 ꢀC, yield (1.7 g) 89%. IR: nmax/cmꢂ1 3283 (NH), 1705 (br,
C]O), 1620, 1477. 1H NMR:
d 2.03 (s, 3H, NCH3), 3.38 (d, 1H, upfield
H of CH2, J ¼ 9.0 Hz), 3.60 (d, 1H, H-20a, J ¼ 8.1 Hz), 3.77 (pseudo t,
1H, downfield H of CH2, J ¼ 7.95 Hz), 3.90 (br s, 1H, H-50a),
6.85e7.57 (m, 8H, arom. H), 10.73 (s, 1H, NH). MS: m/z (%) 381 (M,
100), 383 [(Mþ2), 36], 352 (67), 351 (7), 261 (15), 233 (54), 232 (17),
206 (70), 205 (43). Anal. Calcd. for C20H16ClN3O3 (381.82): C, 62.91;
H, 4.22; N, 11.01. Found: C, 62.98; H, 4.31; N, 11.13.
A mixture of equimolar amounts of the appropriate 1aee
(5 mmol), the corresponding 2aec and sarcosine (3) in absolute
ethanol (25 ml) was boiled under reflux for the appropriate time.
The separated solid while refluxing was collected and crystallized
from a suitable solvent affording the corresponding 4bei,keo. In
case of 4a,j the clear reaction mixture was stored at room
temperature (20e25 ꢀC) overnight and the separated solid was
collected and crystallized from a suitable solvent.
4.1.3. 50a,60-Dihydro-5-methoxy-10-methyl-40-phenyl-spiro[3H-
indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-trione
(4c)
4.1.1. 50a,60-Dihydro-10-methyl-40-phenyl-spiro[3H-indole-
3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-trione (4a)
Reaction time 6 h, colorless crystals from n-butanol, mp
259e261 ꢀC, yield (1.3 g) 75%. IR: nmax/cmꢂ1 3166 (NH),1709 (br, C]
Reaction time 9 h, colorless crystals from n-butanol, mp
216e218 ꢀC, yield (1.5 g) 80%. IR: nmax/cmꢂ1 3159 (NH), 1708 (br, C]
O),1604,1495. 1H NMR:
d 2.01 (s, 3H, NCH3), 3.37 (d,1H, upfield H of
Fig. 6. Predicted versus experimental IC50 of the tested compounds against MCF7 (breast) human tumor cell line.

A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
319
Fig. 7. Predicted versus experimental IC50 of the tested compounds against HEPG2 (liver) human tumor cell line.
CH2, J ¼ 7.8 Hz), 3.55 (d, 1H, H-20a, J ¼ 8.4 Hz), 3.60 (s, 3H, OCH3),
3.75 (br s, 1H, downfield H of CH2), 3.85 (pseudo t, 1H, H-50a,
J ¼ 9.6 Hz), 6.75e7.56 (m, 8H, arom. H), 10.37 (s, 1H, NH). MS: m/z
(%) 377 (M, 38), 257 (21), 229 (31), 228 (21), 202 (19). Anal. Calcd.
for C21H19N3O4 (377.40): C, 66.83; H, 5.07; N, 11.13. Found: C, 66.67;
H, 5.03; N, 11.29.
1H, NH). MS: m/z (%) 415 (M, 100), 419 [(Mþ4), 11], 417 [(Mþ2), 60],
386 (76), 385 (66), 261 (26), 233 (87), 232 (66), 206 (97), 205 (60).
Anal. Calcd. for C20H15Cl2N3O3 (416.27): C, 57.71; H, 3.63; N, 10.09.
Found: C, 57.65; H, 3.53; N, 9.95.
4.1.6. 40-(4-Chlorophenyl)-50a,60-dihydro-5-methoxy-10-methyl-
spiro[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-
2,30,50(1H,20aH,40H)-trione (4f)
4.1.4. 40-(4-Chlorophenyl)-50a,60-dihydro-10-methyl-spiro[3H-
indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-trione
(4d)
Reaction time 9 h, colorless crystals from n-butanol, mp
245e247 ꢀC, yield (1.6 g) 78%. IR: nmax/cmꢂ1 3150 (NH), 1711 (br,
Reaction time 12 h, colorless crystals from n-butanol, mp
C]O), 1610, 1493. 1H NMR:
d 2.00 (s, 3H, NCH3), 3.36 (d, 1H,
272e274 ꢀC, yield (1.4 g) 73%. IR: nmax/cmꢂ1 3173 (NH), 1712 (br, C]
upfield H of CH2, J ¼ 8.4 Hz), 3.48 (d, 1H, H-20a, J ¼ 8.7 Hz), 3.53
(pseudo t, 1H, downfield H of CH2, J ¼ 9.15 Hz), 3.75 (s, 3H, OCH3),
3.85 (pseudo t, 1H, H-50a, J ¼ 9.0 Hz), 6.75e7.59 (m, 7H, arom. H),
10.37 (s, 1H, NH). MS: m/z (%) 411 (M, 71), 413 [(Mþ2), 23], 382
(29), 381 (3), 257 (25), 229 (71), 228 (16), 202 (60), 201 (26). Anal.
Calcd. for C21H18ClN3O4 (411.85): C, 61.24; H, 4.41; N, 10.20. Found:
C, 61.40; H, 4.51; N, 10.01.
O), 1617, 1481. 1H NMR:
d 1.99 (s, 3H, NCH3), 3.35 (d, 1H, upfield H of
CH2, J ¼ 9.3 Hz), 3.54 (d, 1H, H-20a, J ¼ 8.4 Hz), 3.63 (pseudo t, 1H,
downfield H of CH2, J ¼ 8.4 Hz), 3.75 (pseudo t,1H, H-50a, J ¼ 7.8 Hz),
6.83e7.61 (m, 8H, arom. H), 10.57 (s, 1H, NH). 13C NMR:
d 34.9
(NCH3), 45.2 (HC-50a), 52.5 (HC-20a), 54.6 (H2C-60), 72.7 [spiro C-3
(C-20)], 110.3, 122.2, 125.6, 126.5, 129.1, 129.7, 130.2, 131.6, 133.5,
143.3 (arom. C), 174.4, 178.3, 178.5 (C]O). MS: m/z (%) 381 (M, 89),
383 [(Mþ2), 27], 352 (81), 351 (2), 227 (23), 199 (87), 198 (24), 172
(100), 171 (92). Anal. Calcd. for C20H16ClN3O3 (381.82): C, 62.91; H,
4.22; N, 11.01. Found: C, 63.11; H, 4.38; N, 11.10.
4.1.7. 50a,60-Dihydro-40-(4-fluorophenyl)-10-methyl-spiro[3H-indole-
3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-trione (4g)
Reaction time 10 h, colorless crystals from n-butanol, mp
234e236 ꢀC, yield (1.3 g) 71%. IR: nmax/cmꢂ1 3176 (NH),1709 (br, C]
4.1.5. 5-Chloro-40-(4-chlorophenyl)-50a,60-dihydro-10-methyl-spiro
[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-
trione (4e)
O), 1615, 1510. 1H NMR:
d 1.99 (s, 3H, NCH3), 3.33 (d, 1H, upfield H of
CH2, J ¼ 8.4 Hz), 3.45 (d, 1H, H-20a, J ¼ 8.4 Hz), 3.56 (pseudo t, 1H,
downfield H of CH2, J ¼ 8.85 Hz), 3.73 (pseudo t, 1H, H-50a,
J ¼ 7.8 Hz), 6.83e7.38 (m, 8H, arom. H), 10.55 (s, 1H, NH). MS: m/z
(%) 365 (M, 100), 336 (80), 335 (13), 227 (16), 199 (63), 198 (43), 172
(81), 171 (74). Anal. Calcd. for C20H16FN3O3 (365.37): C, 65.75; H,
4.41; N, 11.50. Found: C, 65.64; H, 4.28; N, 11.43.
Reaction time 7 h, colorless crystals from n-butanol, mp
280e282 ꢀC, yield (1.8 g) 87%. IR: nmax/cmꢂ1 3157 (NH),1712 (br, C]
O),1616,1488. 1H NMR:
d 2.00 (s, 3H, NCH3), 3.35 (d,1H, upfield H of
CH2, J ¼ 7.8 Hz), 3.59 (d, 1H, H-20a, J ¼ 8.4 Hz), 3.87 (br s, 2H,
downfield H of CH2 þ H-50a), 6.83e7.64 (m, 7H, arom. H), 10.71 (s,
Table 6
Table 7
Estimated activity data of the training set against HCT116 (colon cancer) cell line and
calculated descriptors governing activity according to Equation (1).
Estimated activity data of the training set against MCF7 (breast cancer) cell line and
calculated descriptors governing activity according to Equation (2).
Entry
Compd.
Estimated
activity
Observed
activity
Residuals
LUMO
Fit
value
Entry
Compd.
Estimated
activity
Observed
activity
Residuals
Shadow
-nu
Fit
value
1
2
3
4
5
6
7
8
4a
4b
4c
4d
4e
4f
4h
4i
4j
4k
4m
4n
4o
71.48
41.94
40.95
44.87
37.84
29.58
69.44
30.48
71.38
92.18
40.79
62.41
29.48
63.73
33.05
34.08
36.80
35.46
47.40
73.26
31.31
46.13
>126.31
53.63
41.03
40.92
ꢂ7.752
ꢂ8.887
ꢂ6.872
ꢂ8.068
ꢂ2.379
17.824
3.823
ꢂ0.127
ꢂ0.332
ꢂ0.081
ꢂ0.172
ꢂ0.468
ꢂ0.150
ꢂ0.350
ꢂ0.260
ꢂ0.113
ꢂ0.363
ꢂ0.056
ꢂ0.391
ꢂ0.082
7.804
8.007
7.990
7.974
8.044
8.067
7.837
8.072
7.803
7.696
7.989
7.884
8.062
1
2
3
4
5
6
7
8
4a
4b
4c
4d
4e
4f
4h
4i
4j
4k
4m
4n
4o
38.42
36.56
45.09
34.45
51.66
35.26
84.41
49.47
42.37
44.18
47.94
40.50
40.32
35.64
38.66
44.78
43.03
40.60
38.73
89.92
43.96
37.55
49.92
47.96
38.14
42.66
ꢂ2.784
2.096
1.820
1.691
1.804
1.990
1.922
1.866
1.830
1.787
1.979
1.908
2.064
1.990
1.978
7.997
7.972
7.959
8.023
7.958
8.005
7.785
7.886
8.024
7.996
8.022
7.953
7.974
ꢂ0.308
8.577
ꢂ11.057
3.474
5.507
ꢂ5.507
ꢂ4.818
4.818
0.835
9
ꢂ25.250
33.822
12.845
ꢂ21.379
11.438
9
10
11
12
13
10
11
12
13
0.023
ꢂ2.359
2.339

320
A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
4.1.11. 5-Chloro-50a,60-dihydro-10-methyl-40-(4-methylphenyl)-
spiro[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-
2,30,50(1H,20aH,40H)-trione (4k)
Table 8
Estimated activity data of the training set against HEPG2 (liver cancer) cell line and
calculated descriptors governing activity according to Equation (3).
Entry
Compd.
Estimated
activity
Observed
activity
Residuals
PKa
Fit
value
Reaction time
266e268 ꢀC, yield (1.6 g) 81%. IR: nmax/cmꢂ1 3275 (NH), 1707 (br, C]
O), 1617, 1511. 1H NMR:
2.03 (s, 3H, NCH3), 2.36 (s, 3H, ArCH3), 3.37
8 h, colorless crystals from n-butanol, mp
1
2
3
4
5
6
7
8
4a
4b
4c
4d
4e
4f
4h
4i
4j
4k
4m
4n
4o
31.82
26.44
46.71
43.12
27.12
43.53
39.18
29.60
79.71
60.38
65.46
33.26
48.74
32.90
33.05
44.17
32.42
29.16
43.37
41.09
27.69
77.09
63.16
72.55
28.89
49.68
1.080
6.612
ꢂ2.540
ꢂ10.701
2.043
1.90
2.23
2.23
1.90
1.90
2.23
1.93
1.90
1.90
2.62
1.93
1.90
2.62
7.597
7.648
7.628
7.465
7.633
7.655
7.520
7.674
7.203
7.408
7.422
7.768
7.606
d
(d,1H, upfield H of CH2, J ¼ 9.0 Hz), 3.58 (d,1H, H-20a, J ¼ 8.4 Hz), 3.74
(pseudo t, 1H, downfield H of CH2, J ¼ 7.8 Hz), 3.87 (br s, 1H, H-50a),
6.82e7.47 (m, 7H, arom. H), 10.72 (s, 1H, NH). MS: m/z (%) 395 (M,
64), 397 [(Mþ2), 42], 366 (52), 365 (48), 261 (10), 233 (46), 232 (22),
206 (56), 205 (32). Anal. Calcd. for C21H18ClN3O3 (395.85): C, 63.72;
H, 4.58; N, 10.62. Found: C, 63.86; H, 4.61; N, 10.60.
ꢂ0.160
1.911
ꢂ1.911
ꢂ2.618
2.617
9
10
11
12
13
7.086
4.1.12. 50a,60-Dihydro-5-methoxy-10-methyl-40-(4-methylphenyl)-
spiro[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-
2,30,50(1H,20aH,40H)-trione (4l)
ꢂ4.365
0.945
Reaction time 10 h, colorless crystals from n-butanol, mp
247e249 ꢀC, yield (1.5 g) 77%. IR: nmax/cmꢂ1 3156 (NH), 1714 (br, C]
O), 1610, 1490. 1H NMR:
d 2.01 (s, 3H, NCH3), 2.36 (s, 3H, ArCH3),
4.1.8. 5-Chloro-50a,60-dihydro-40-(4-fluorophenyl)-10-methyl-spiro
[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-
trione (4h)
3.35 (d, 1H, upfield H of CH2, J ¼ 9.3 Hz), 3.51 (pseudo t, 1H, H-20a,
J ¼ 8.4 Hz), 3.61 (s, 3H, OCH3), 3.75 (br s, 2H, downfield H of
CH2 þ H-50a), 6.42e7.34 (m, 7H, arom. H), 10.39 (s, 1H, NH). MS: m/z
(%) 391 (M, 98), 362 (42), 361 (23), 257 (28), 229 (95), 228 (39), 202
(72), 201 (34). Anal. Calcd. for C22H21N3O4 (391.43): C, 67.51; H,
5.41; N, 10.74. Found: C, 67.28; H, 5.30; N, 10.67.
Reaction time 8 h, colorless crystals from n-butanol, mp
294e296 ꢀC, yield (1.8 g) 90%. IR: nmax/cmꢂ1 3114 (NH),1709 (br, C]
O),1621, 1509. 1H NMR:
d 2.03 (s, 3H, NCH3), 3.37 (d, 1H, upfield H of
CH2, J ¼ 8.4 Hz), 3.59 (d, 1H, H-20a, J ¼ 8.4 Hz), 3.75 (pseudo t, 1H,
downfield H of CH2, J ¼ 7.95 Hz), 3.89 (br s, 1H, H-50a), 6.84e7.48
(m, 7H, arom. H), 10.73 (s, 1H, NH). MS: m/z (%) 399 (M, 100), 401
[(Mþ2), 33], 370 (60), 369 (40), 261 (17), 233 (65), 232 (19), 206
(70), 205 (37). Anal. Calcd. for C20H15ClFN3O3 (399.81): C, 60.08; H,
3.78; N, 10.51. Found: C, 59.99; H, 3.72; N, 10.38.
4.1.13. 50a,60-Dihydro-40-(4-methoxyphenyl)-10-methyl-spiro[3H-
indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-
trione (4m)
Reaction time 8 h, colorless crystals from n-butanol, mp
276e278 ꢀC, yield (1.4 g) 74%. IR: nmax/cmꢂ1 3317 (NH), 1710 (br, C]
O), 1608, 1511. 1H NMR:
d 1.99 (s, 3H, NCH3), 3.34 (d, 1H, upfield H of
4.1.9. 50a,60-Dihydro-40-(4-fluorophenyl)-5-methoxy-10-methyl-
spiro[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-
2,30,50(1H,20aH,40H)-trione (4i)
CH2, J ¼ 8.1 Hz), 3.52 (d, 1H, H-20a, J ¼ 8.1 Hz), 3.61 (pseudo t, 1H,
downfield H of CH2, J ¼ 8.25 Hz), 3.72 (pseudo t, 1H, H-50a,
J ¼ 8.4 Hz), 3.80 (s, 3H, OCH3), 6.83e7.25 (m, 8H, arom. H), 10.55 (s,
Reaction time 10 h, colorless crystals from n-butanol, mp
1H, NH). 13C NMR:
d
34.9 (NCH3), 45.1 (HC-50a), 52.4 (HC-20a), 54.6
244e246 ꢀC, yield (1.7 g) 86%. IR: nmax/cmꢂ1 3158 (NH), 1706 (br,
(H2C-60), 55.9 (OCH3), 72.7 [spiro C-3 (C-20)], 110.3, 114.8, 122.2,
125.4, 125.8, 126.5, 128.6, 130.1, 143.3, 159.5 (arom. C), 174.8, 178.4,
178.9 (C]O). MS: m/z (%) 377 (M, 78), 348 (50), 347 (4), 227 (9), 199
(100), 198 (9), 172 (74), 171 (61). Anal. Calcd. for C21H19N3O4
(377.40): C, 66.83; H, 5.07; N, 11.13. Found: C, 66.77; H, 4.99; N,
11.24.
C]O), 1610, 1501. 1H NMR:
d 2.02 (s, 3H, NCH3), 3.36 (d, 1H, upfield
H of CH2, J ¼ 8.7 Hz), 3.54 (d, 1H, H-20a, J ¼ 8.1 Hz), 3.60 (s, 3H,
OCH3), 3.75 (br s, 2H, downfield H of CH2þ H-50a), 6.42e7.38 (m,
7H, arom. H), 10.40 (s, 1H, NH). MS: m/z (%) 395 (M, 65), 366 (29),
365 (8), 257 (16), 229 (52), 228 (10), 202 (47), 201 (18). Anal. Calcd.
for C21H18FN3O4 (395.39): C, 63.79; H, 4.59; N, 10.63. Found: C,
63.60; H, 4.51; N, 10.67.
4.1.14. 5-Chloro-50a,60-dihydro-40-(4-methoxyphenyl)-10-methyl-
spiro[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-
2,30,50(1H,20aH,40H)-trione (4n)
4.1.10. 50a,60-Dihydro-10-methyl-40-(4-methylphenyl)-spiro[3H-
indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-2,30,50(1H,20aH,40H)-trione (4j)
Reaction time 12 h, colorless crystals from n-butanol, mp
207e209 ꢀC, yield (1.3 g) 72%. IR: nmax/cmꢂ1 3174 (NH), 1712 (br, C]
Reaction time 8 h, colorless crystals from n-butanol, mp
283e285 ꢀC, yield (1.8 g) 88%. IR: nmax/cmꢂ1 3300 (NH), 1708 (br,
C]O), 1614, 1512. 1H NMR:
d 2.02 (s, 3H, NCH3), 3.36 (d, 1H, upfield
O),1614,1514. 1H NMR:
d 1.99 (s, 3H, NCH3), 2.34 (s, 3H, ArCH3), 3.35
H of CH2, J ¼ 9.0 Hz), 3.57 (d, 1H, H-20a, J ¼ 8.1 Hz), 3.59 (pseudo t,
1H, downfield H of CH2, J ¼ 9.0 Hz), 3.73 (pseudo t, 1H, H-50a,
J ¼ 8.1 Hz), 3.81 (s, 3H, OCH3), 6.83e7.33 (m, 7H, arom. H), 10.72 (s,
(d, 1H, upfield H of CH2, J ¼ 9.0 Hz), 3.53 (d, 1H, H-20a, J ¼ 8.1 Hz),
3.62 (pseudo t, 1H, downfield H of CH2, J ¼ 8.25 Hz), 3.73 (pseudo t,
1H, H-50a, J ¼ 7.95 Hz), 6.83e7.34 (m, 8H, arom. H),10.55 (s,1H, NH).
MS: m/z (%) 361 (M, 100), 332 (90), 227 (37), 199 (84), 198 (45), 172
(100), 171 (74). Anal. Calcd. for C21H19N3O3 (361.40): C, 69.79; H,
5.30; N, 11.63. Found: C, 69.72; H, 5.21; N, 11.47.
1H, NH). 13C NMR:
d
35.0 (NCH3), 45.0 (HC-50a), 52.6 (HC-20a), 54.8
(H2C-60), 56.0 (OCH3), 72.7 [spiro C-3 (C-20)], 111.8, 114.9, 125.3,
126.1, 126.3, 128.0, 128.5, 130.1, 142.2, 159.6 (arom. C), 174.8, 178.0,
178.7 (C]O). MS: m/z (%) 411 (M, 93), 413 [(Mþ2), 31], 382 (45), 381
(2), 261 (10), 233 (100), 232 (7), 206 (79), 205 (43). Anal. Calcd. for
C21H18ClN3O4 (411.85): C, 61.24; H, 4.41; N, 10.20. Found: C, 61.36;
H, 4.51; N, 10.14.
Table 9
Linear QSAR models established for the training set (N ¼ 13).
4.1.15. 50a,60-Dihydro-5-methoxy-40-(4-methoxyphenyl)-10-
methyl-spiro[3H-indole-3,20(10H)-pyrrolo[3,4-c]pyrrole]-
2,30,50(1H,20aH,40H)-trione (4o)
Cell lines
Descriptors
R
S
FIT
HCT116 (colon) Fit value, LUMO-Eigen value-VAMP 0.9541 0.675 3.24
MCF7 (breast)
HEPG2 (liver)
Fit value, Shadow-nu
Fit value, PKa
0.925
0.960
0.764 2.95
0.498 3.45
Reaction time 10 h, colorless crystals from n-butanol, mp
252e254 ꢀC, yield (1.5 g) 74%. IR: nmax/cmꢂ1 3276 (NH), 1711 (br, C]

A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
321
Table 10
External validation for the established QSAR models utilizing mild (4g) and promising (4l) anti-tumor active agents.
Cell lines
Compd.
3D-QSAR pharmacophore
Experimental Predicted
Classical QSAR
Experimental
Fit value
Predicted
LUMO
Shadow-nu
PKa
activity (IC50
)
activity (IC50
)
activity (IC50
)
activity (IC50)
HCT116 (colon)
MCF7 (breast)
HEPG2 (liver)
4g
4l
4g
4l
4g
4l
42.37
46.24
33.88
46.83
87.31
12.16
46.10
50.13
37.45
41.96
90.32
10.26
8.01
7.98
8.02
7.95
7.22
7.91
42.37
46.24
33.88
46.83
87.31
12.16
40.80
44.87
33.18
48.25
86.42
13.22
ꢂ0.294
e
e
ꢂ0.353
e
e
e
e
e
e
1.842
1.889
e
e
e
1.93
1.90
e
O),1607, 1509. 1H NMR:
d
2.01 (s, 3H, NCH3), 3.35 (d,1H, upfield H of
4.3. Anti-tumor activity screening
CH2, J ¼ 8.1 Hz), 3.52 (d, 1H, H-20a, J ¼ 8.4 Hz), 3.61 (s, 3H, OCH3),
3.75 (br s, 1H, downfield H of CH2), 3.79 (br s, 1H, H-50a), 3.80 (s, 3H,
OCH3), 6.42e7.22 (m, 7H, arom. H), 10.39 (s, 1H, NH). MS: m/z (%)
407 (M, 70), 378 (23), 377 (2), 257 (11), 229 (100), 228 (22), 202
(56), 201 (20). Anal. Calcd. for C22H21N3O5 (407.43): C, 64.86; H,
5.20; N, 10.31. Found: C, 64.76; H, 5.09; N, 10.23.
Anti-tumor properties of the synthesized compounds were
screened by National Cancer Institute, Cairo University, Egypt, using
the previously reported standard procedure adopting HCT116
(colon), MCF7 (breast) and HEPG2 (liver) human tumor cell lines
[26]. Cells were seeded in 96-well microtiter plates at a concen-
tration of 5 ꢄ104e105 cell/well in a fresh medium and left for 24 h
before treatment with the tested compounds to allow attachment
of cells to the wall of the plate. The tested compounds were dis-
solved in dimethylsulfoxide (DMSO) and diluted 1000-fold in the
assay. Different concentrations of the compounds under test (0, 5,
4.2. Single crystal X-ray crystallography
Full crystallographic details, excluding structure factors have
been deposited at Cambridge Crystallographic Data Centre (CCDC)
as supplementary publication numbers CCDC 846811 & 846812
corresponding to compounds 4a and 4m, respectively. The crys-
tallographic data were collected at T ¼ 298 K on a Kappa CCD Enraf
Nonius FR 590 diffractometer using a graphite monochromator
12.5, 25, and 50 mg/ml) were added to the cell monolayer. Triplicate
wells were prepared for each individual dose. The monolayer cells
were incubated with the tested compounds for 48 h at 37 ꢀC, in
atmosphere of 5% CO2. After 48 h, the cells were fixed, washed and
stained with Sulfo-Rhodamine-B (SRB) stain. Excess stain was
washed with acetic acid. The attached stain was recovered with
TriseEDTA buffer. Cell survival and drug activity were determined
by measuring the color intensity spectrophotometrically at 564 nm
using an ELISA microplate reader (Meter tech. S 960, USA). Data are
collected as mean values for experiments that performed in three
replicates for each individual dose which measured by SRB assay.
Control experiments did not exhibit significant change compared to
the DMSO vehicle. Doxorubicin was used as a standard reference
during the present in-vitro bioactivity screening assay. The
percentage of cell survival was calculated as follows:
with Mo-Ka radiation (
l
¼ 0.71073 Å). The crystal structures were
determined by direct method using SIR92 [42] and refined by
maXus [43] (Bruker Nonius, Delft and MacScience, Japan). Non-
hydrogen atoms were refined anisotropically. Hydrogen atoms
were located geometrically and were refined isotropically.
4.2.1. Compound 4a
For X-ray crystallographic studies, compound 4a was recrystal-
lized as prismatic colorless crystals from n-butanol. Chemical
formula C20H17N3O3, Mr ¼ 347.374, monoclinic, crystallizes in space
group P21/c, Cell lengths “a
c ¼ 22.8262(8) Å”, Cell angles “
¼
9.5834(2),
b
¼
7.7923(2),
a
¼ 90.00,
b
¼ 96.1695(10),
g
¼
Surviving fraction ¼ Optical densityðO:D:Þof treated
cells=O:D: of control cells
90.00ꢀ”, V ¼ 1694.71(8) Å3, Z ¼ 4, Dc ¼ 1.362 mg/m3,
q values
(3)
2.910e27.485ꢀ, absorption coefficient
m
(Mo-Ka) ¼ 0.09 mmꢂ1
,
F(000) ¼ 728. The unique reflections measured 4045 of which 2349
The IC50 (concentration required to produce 50% inhibition of
cell growth compared to control experimental) was determined
using Graph-Pad PRISM version-5 software. Statistical calculations
for determination of the mean and standard error values were
determined by SPSS 11 software. The observed anti-tumor prop-
erties were presented in Table 1 (Figs. 1e3 of the supplementary
material).
reflections with threshold expression I > 3s(I) were used in the
structural analysis. Convergence for 235 variable parameters by
least-squares refinement on F2 with w ¼ 1=½s2ðFo2Þ þ 0:10000 Fo2ꢃ.
The final agreement factors were R ¼ 0.041 and wR ¼ 0.077 with
a goodness-of-fit of 1.019.
4.2.2. Compound 4m
For X-ray crystallographic studies, compound 4m was recrys-
tallized as prismatic colorless crystals from n-butanol. Chemical
formula C21H19N3O4, Mr ¼ 377.400, monoclinic, crystallizes in space
Acknowledgment
This work is sponsored by the Swedish International Develop-
ment Cooperation Agency (SIDA) due to the International Collab-
orative Research Grant (MENA).
group P21/c, Cell lengths “a
c ¼ 12.7045(6) Å”, Cell angles “
¼
7.5625(3),
b
¼
18.7645(6),
a
¼ 90.00,
b
¼ 94.6855(14),
g
¼
90.00ꢀ”, V ¼ 1796.83(12) Å3, Z ¼ 4, Dc ¼ 1.395 mg/m3,
q values
2.910e27.485ꢀ, absorption coefficient
F(000) ¼ 792. The unique reflections measured 4670 of which 1865
m
(Mo-Ka) ¼ 0.10 mmꢂ1
,
Appendix. Supplementary data
reflections with threshold expression I > 3s(I) were used in the
structural analysis. Convergence for 253 variable parameters by
least-squares refinement on F2 with w ¼ 1=½s2ðFo2Þ þ 0:10000 Fo2ꢃ.
The final agreement factors were R ¼ 0.040 and wR ¼ 0.072 with
a goodness-of-fit of 1.151.
Supplementary data related to this article can be found online at
doi:10.1016/j.ejmech.2011.10.058. These data include MOL files and
InChiKeys of the most important compounds described in this
article.

322
A.S. Girgis et al. / European Journal of Medicinal Chemistry 47 (2012) 312e322
(b) R. Grigg, M.F. Aly, V. Sridharan, S. Thianpatanagul, J. Chem. Soc. Chem.
References
Commun. (1984) 182e183;
(c) H. Ardil, R. Grigg, V. Sridharan, S. Suerendrakumar, S. Thianpatanagul,
S. Kanajan, J. Chem. Soc. Chem. Commun. (1986) 602e604;
(d) A.R.S. Babu, R. Raghunathan, G. Gayatri, G.N. Sastry, J. Heterocycl. Chem. 43
(2006) 1467e1472.
[1] (a) N.V. Lakshmi, Y. Arun, P.T. Perumal, Tetrahedron Lett. 52 (2011) 3437e3442;
(b) R. Prasanna, S. Purushothaman, M. Suresh, R. Raghunathan, Tetrahedron
Lett. 52 (2011) 792e797;
(c) M.A. Ali, R. Ismail, T.S. Choon, Y.K. Yoon, A.C. Wei, S. Pandian, R.S. Kumar,
H. Osman, E. Manogaran, Bioorg. Med. Chem. Lett. 20 (2010) 7064e7066;
(d) R. Rajesh, R. Raghunathan, Tetrahedron Lett. 51 (2010) 5845e5848;
(e) R. Prasanna, S. Purushothaman, R. Raghunathan, Tetrahedron Lett. 51 (2010)
4538e4542;
[18] A.S. Girgis, Eur. J. Med. Chem. 44 (2009) 91e100.
[19] A.S. Girgis, Eur. J. Med. Chem. 44 (2009) 1257e1264.
[20] M.J.S. Dewar, E.G. Zoebisch, E.F. Healy, J.J.P. Stewart, J. Am. Chem. Soc. 107
(1985) 3902e3909.
[21] J.J.P. Stewart, J. Comput. Chem. 10 (1989) 209e220.
[22] J.J.P. Stewart, J. Comput. Chem. 10 (1989) 221e264.
[23] P. Machado, P.T. Campos, G.R. Lima, F.A. Rosa, A.F. C. Flores, H.G. Bonacorso,
N. Zanatta, M.A.P. Martins, J. Mol. Struct. 917 (2009) 176e182.
[24] C.P. Frizzo, M.R.B. Marzari, D.N. Moreira, P.T. Campos, N. Zanatta,
H.G. Bonacorso, M.A.P. Martins, J. Mol. Struct. 981 (2010) 71e79.
[25] (a) A.S. Girgis, N.S.M. Ismail, H. Farag, Eur. J. Med. Chem. 46 (2011)
2397e2407;
(b) A.S. Girgis, H. Farag, N.S.M. Ismail, R.F. George, Eur. J. Med. Chem. 46 (2011)
4964e4969.
[26] (a) P. Skehan, R. Storeng, D. Scudiero, A. Monks, J. McMahon, D. Vistica,
J.T. Warren, H. Bokesch, S. Kenney, M.R. Boyd, J. Natl. Cancer Inst. 82 (1990)
1107e1112;
(f) K. Karthikeyan, P.M. Sivakumar, M. Doble, P.T. Perumal, Eur. J. Med. Chem. 45
(2010) 3446e3452;
(g) A. Hazra, P. Paira, K.B. Sahu, S. Naskar, P. Saha, R. Paira, S. Mondal, A. Maity,
P. Luger, M. Weber, N.B. Mondal, S. Banerjee, Tetrahedron Lett. 51 (2010)
1585e1588;
(h) N.V. Lakshmi, P. Thirumurugan, P.T. Perumal, Tetrahedron Lett. 51 (2010)
1064e1068;
(i) R.S. Kumar, S.M. Rajesh, S. Perumal, D. Banerjee, P. Yogeeswari, D. Sriram, Eur.
J. Med. Chem. 45 (2010) 411e422;
(j) S. Kathiravan, R. Raghunathan, Tetrahedron Lett. 50 (2009) 6116e6120;
(k) R.R. Kumar, S. Perumal, P. Senthilkumar, P. Yogeeswari, D. Sriram, Eur. J. Med.
Chem. 44 (2009) 3821e3829;
(l) M. Ghandi, A. Yari, S.J.T. Rezaei, A. Taheri, Tetrahedron Lett. 50 (2009)
4724e4726;
(m) G. Periyasami, R. Raghunathan, G. Surendiran, N. Mathivanan, Eur. J. Med.
Chem. 44 (2009) 959e966;
(n) R. Murugan, S. Anbazhagan, S.S. Narayanan, Eur. J. Med. Chem. 44 (2009)
3272e3279;
(b) A.S. Girgis, H.M. Hosni, F.F. Barsoum, Bioorg. Med. Chem. 14 (2006)
4466e4476;
(c) A.R. Katritzky, A.S. Girgis, S. Slavov, S.R. Tala, I. Stoyanova-Slavova, Eur. J.
Med. Chem. 45 (2010) 5183e5199.
[27] D. Barnum, J. Greene, A. Smellie, P. Sprague, J. Chem. Inf. Comput. Sci. 36
(1996) 563e571.
[28] A. Smellie, S.L. Teig, P. Towbin, J. Comput. Chem. 16 (1995) 171e187.
[29] Y. Kurogi, O.F. Güner, Curr. Med. Chem. 8 (2001) 1035e1055.
[30] M.O. Taha, A.M. Qandil, D.D. Zaki, M.A. Al Damen, Eur. J. Med. Chem. 40 (2005)
701e727.
(o) A.R.S. Babu, R. Raghunathan, Tetrahedron Lett. 49 (2008) 4618e4620;
(p) P. Shanmugam, B. Viswambharan, K. Selvakumar, S. Madhavan, Tetrahedron
Lett. 49 (2008) 2611e2615.
[2] C.B. Cui, H. Kakeya, H. Osada, Tetrahedron 52 (1996) 12651e12666.
[3] C.B. Cui, H. Kakeya, H. Osada, J. Antibiot. 49 (1996) 832e835.
[4] M.N.G. James, G.J.B. Williams, Can. J. Chem. 50 (1972) 2407e2412.
[5] A. Jossang, P. Jossang, H.A. Hadi, T. Sévenet, B. Bodo, J. Org. Chem. 56 (1991)
6527e6530.
[6] S. Ghosal, P.K. Banerjee, Indian J. Chem. 9 (1971) 289e293.
[7] K. Jones, J. Wilkinson, J. Chem. Soc. Chem. Commun. (1992) 1767e1769.
[8] S.I. Bascop, J. Sapi, J.Y. Laronze, J. Lévy, Heterocycles 38 (1994) 725e732.
[9] C. Pellegrini, C. Strässler, M. Weber, H.J. Borschberg, Tetrahedron: Asymmetry
5 (1994) 1979e1992.
[31] M.O. Taha, A.G. Al-Bakri, W.A. Zalloum, Bioorg. Med. Chem. Lett. 16 (2006)
5902e5906.
[32] M.O. Taha, Y. Bustanji, M.A.S. Al-Ghussein, M. Mohammad, H. Zalloum, I.M. Al-
Masri, N. Atallah, J. Med. Chem. 51 (2008) 2062e2077.
[33] M.O. Taha, Y. Bustanji, A.G. Al-Bakri, A. Yousef, W.A. Zalloum, I.M. Al-Masri,
N. Atallah, J. Mol. Graph. Model. 25 (2007) 870e884.
[34] R. Abu Khalaf, G. Abu Sheikha, Y. Bustanji, M.O. Taha, Eur. J. Med. Chem. 45
(2010) 1598e1617.
[35] I.M. Al-Masri, M.K. Mohammad, M.O. Taha, Chem. Med. Chem. 3 (2008)
[10] G. Palmisano, R. Annunziata, G. Papeo, M. Sisti, Tetrahedron: Asymmetry 7
(1996) 1e4.
1763e1779.
[36] H. Kubinyi, Quant. Struct.-Act. Relat. 13 (1994) 393e401.
[37] H. Kubinyi, Quant. Struct.-Act. Relat. 13 (1994) 285e294.
[38] P.R. Duchowicz, M.G. Vitale, E.A. Castro, J.C. Autino, G.P. Romanelli,
D.O. Bennardi, Eur. J. Med. Chem. 43 (2008) 1593e1602.
[39] M.P. Cava, A.A. Deana, K. Muth, M.J. Mitchell, Org. Synth. 41 (1961) 93e95.
[40] W.R. Roderick, J. Amer. Chem. Soc. 79 (1957) 1710e1712.
[41] Q. Ahmed, T. Wagner-Jauregg, E. Pretsch, J. Seibl, Helv. Chim. Acta 56 (1973)
1646e1655.
[42] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M.C. Burla, G. Polidori,
M. Camalli, J. Appl. Cryst. 27 (1994) 435e436.
[43] S. Mackay, C.J. Gilmore, C. Edwards, N. Stewart, K. Shankland, maXus
Computer Program for the Solution and Refinement of Crystal Structures.
[11] E. Garcia Prado, M.D. Garcia Gimenez, R. De la Puerta Vázquez, J.L. Espartero
Sánchez, M.T. Sáenz Rodriguez, Phytomedicine 14 (2007) 280e284.
[12] R.D. Connell, Expert Opin. Ther. Pat. 13 (2003) 738e750.
[13] M.A. Jianguo, L.I. Shaolan, K. Reed, P. Guo, J.M. Gallo, J. Pharmacol. Exp. Ther.
305 (2003) 833e839.
[14] P. Marzola, A. Degrassi, L. Calderan, P. Farace, C. Crescimanno, E. Nicolato,
A. Giusti, E. Pesenti, A. Terron, A. Sbarbati, T. Abrams, L. Murray, F. Osculati,
Clin. Cancer Res. 15 (2004) 739e750.
[15] M.E. Lane, B. Yu, A. Rice, K.E. Lipson, C. Liang, L. Sun, C. Tang, G. McMahon,
R.G. Pestell, S. Wadler, Cancer Res. 15 (2001) 6170e6177.
[16] A.H. Abadi, S.M. Abou-Seri, D.E. Abdel-Rahman, C. Klein, O. Lozach, L. Meijer,
Eur. J. Med. Chem. 41 (2006) 296e305.
[17] (a) R. Grigg, S. Thianpatanagul, J. Chem. Soc. Chem. Commun. (1984)
180e181;
Bruker Nonius, The Netherlands, MacScience, Japan
Glasgow, 1999.
& The University of