Synthesis, in vitro and in vivo preliminary evaluation of anti-angiogenic properties of some pyrroloazaflavones

Article abstract of DOI:10.1016/j.bmc.2010.11.010

This work investigated the in vitro and in vivo anti-angiogenic activity of some pyrroloazaflavones, exactly 2-phenyl-1H-pyrrolo[2,3-h]quinolin-4(7H)ones, with vinblastine as reference compound. Growth inhibitory activity, migration, and capillary-like structures formation were determined in human umbilical vein endothelial cell cultures, and Matrigel plug assay was carried out to evaluate in vivo effects on angiogenesis. Collectively, our results indicate that some pyrroloazaflavone derivatives, at non-cytotoxic concentrations and like vinblastine are able: (i) to exert in vitro anti-angiogenic activity and (ii) to counteract in vitro and in vivo the pro-angiogenic effects of fibroblast growth factor-2 (FGF-2).

Full text of DOI:10.1016/j.bmc.2010.11.010

Bioorganic & Medicinal Chemistry 19 (2011) 448–457
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis, in vitro and in vivo preliminary evaluation of anti-angiogenic
properties of some pyrroloazaflavones
Maria Grazia Ferlin a, , Maria Teresa Conconi a, Luca Urbani a, Barbara Oselladore b, Diego Guidolin b,
⇑
Rosa Di Liddo a, Pier Paolo Parnigotto a
a Department of Pharmaceutical Sciences, University of Padova, via F. Marzolo, 5, 35131 Padova, Italy
b Department of Human Anatomy and Physiology, University of Padova, Anatomy Section, via A. Gabelli, 65, 35131 Padova, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
This work investigated the in vitro and in vivo anti-angiogenic activity of some pyrroloazaflavones,
exactly 2-phenyl-1H-pyrrolo[2,3-h]quinolin-4(7H)ones, with vinblastine as reference compound. Growth
inhibitory activity, migration, and capillary-like structures formation were determined in human umbil-
ical vein endothelial cell cultures, and Matrigel plug assay was carried out to evaluate in vivo effects on
angiogenesis. Collectively, our results indicate that some pyrroloazaflavone derivatives, at non-cytotoxic
concentrations and like vinblastine are able: (i) to exert in vitro anti-angiogenic activity and (ii) to coun-
teract in vitro and in vivo the pro-angiogenic effects of fibroblast growth factor-2 (FGF-2).
Ó 2010 Elsevier Ltd. All rights reserved.
Received 21 April 2010
Revised 27 October 2010
Accepted 4 November 2010
Available online 10 November 2010
Keywords:
Pyrroloazaflavones
Phenyl-pyrroloquinolinones
Angiogenesis
Anti-vascular agents
1. Introduction
Microtubule-disrupting agents such as the natural compounds
taxol, colchicines, vinca alkaloids, and flavonoids, induce extensive
In tumor pathogenesis, angiogenesis is crucial and it sustains
malignant cells with nutrients and oxygen. Tumor cells secrete var-
ious growth factors which trigger endothelium cells to form new
capillaries. Preventing the expansion of new blood vessel network
results in reduced tumor size and metastases. Since angiogenesis is
essential for tumor development and tumor vasculature is consid-
ered an optimal target for anti-cancer strategies, the research
aimed at discovering anti-angiogenic agents as a treatment of
malignancy or as an adjunct to standard chemotherapeutic regi-
ments, is very active. The targeting of tumor vasculature can be
achieved by inhibition of new vessel formation,1 through anti-
angiogenic agents, or by damaging existing blood vessel, through
vascular disrupting agents (VDAs).2
necrosis in tumors as a result of vascular collapse. Their main mech-
anism of action is believed to be impairment of tubulin assembly
properties, leading to disruption of the microtubule network of pro-
liferating endothelial cells (ECs). They have been reported to affect
certain elements of angiogenic process, such as endothelial cell
migration and in vitro capillary-like tube formation.3 Nevertheless,
these effects are elicited at dosages close to the maximum tolerated
dose (MTD). This limitation has been overcome by the development
of second-generation tubulin depolymerizing agents, like DMXAA,
CA4P, ZD6126, AVE8062A, and Oxi4503 (Fig. 1), which selectively
disrupt the tumor vasculature at doses well below their MTD.4
Very recently, some progress in understanding the molecular and
signaling mechanisms associated with endothelial activity of micro-
tubule depolymerizing/destabilizing agents have been made.3,5
Taking the natural flavone pharmacophore as a model, we
have recently developed two series of pyrroloazaflavones (Fig. 2),
namely 2-phenyl-1H-pyrrolo[2,3-h]quinolin-4(7H)ones and 7-phe-
nyl-3H-pyrrolo[3,2-f]quinolin-9(6H)-ones (2-PPyQs and 7-PPyQs,
respectively), as anti-cancer agents with an anti-microtubule
mechanism,6–8 showing that the different geometry can account
for the dissimilar effectiveness in antiproliferative activity on solid
tumor cell lines: 7-PPyQs,8 especially 3-alkyl-substituted ones, are
much more cytotoxic with nanomolar IC50 than 2-PPyQs6, among
which only some compounds were found to have low micromolar
cytotoxicity.
In the field of anti-cancer drugs, several clinically useful drugs
have been developed from natural source, by structural modifica-
tion of natural compounds or by the synthesis of new compounds
designed following a natural compound as model.
Abbreviations: ECs, endothelial cells; HUVECs, human umbilical vein ECs; MTD,
maximum tolerated dose; SARs, structure–activity relationships; VTAs, vascular
targeting agents; VDAs, vascular disrupting agents; 2-PPyQs, 2-phenyl-7H-pyrrolo-
[2,3-h]quinolinones; 7-PPyQs, 7-phenyl-3H-pyrrolo[3,2-f]quinolinones.
⇑
Corresponding author. Address: Department of Pharmaceutical Sciences, Fac-
ulty of Pharmacy, University of Padova, via Marzolo 5, 35131 Padova, Italy. Tel.: +39
0498271603; fax: +39 0498275366.
E-mail address: mariagrazia.ferlin@unipd.it (M.G. Ferlin).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.11.010

M. G. Ferlin et al. / Bioorg. Med. Chem. 19 (2011) 448–457
449
O
O
H3CO
NHAc
H3C
H3CO
OH
CH3
H3CO
O
OPO3H2
DMXAA
ZD6126
H3CO
H3CO
H3CO
H3CO
O
OCH3
OCH3
N
H
OH
NH2.HCl
OPO3Na2
OCH3
OCH3
CA4P
AVE8062
H3CO
H3CO
OPO3Na2
OCH3
OPO3Na2
OCH3
Oxi4503
Figure 1. Structures of some anti-angiogenic agents mentioned in the text.
O
O
O
NH
1
flavone
azaflavone
2-phenylquinolin-4(1H)-one
2-phenyl-4 H-chromen-4-one
1
O
O
4
9
NH
2
7
NH
NH
NH
1
2-pyrroloazaflavones
7-pyrroloazaflavones
2-phenyl-1H -pyrrolo[2,3-h]quinolin-
-4(7H)-one
7-phenyl-3 H-pyrrolo[3,2-f ]quinolin-
-9(6H)-one
Figure 2. Structures of pyrroloazaflavones and their origin from the flavone pharmacofore.
However, the potential effect of pyrroloazaflavones on angio-
genesis had not been studied. The aim of the current research
was to evaluate the in vitro and in vivo effects on angiogenesis
induced by some previously described 2-aryl-PyQs6 (20–24)
belonging to the less cytotoxic series. Furthermore, some com-
pounds with the same angular geometry 8, 12, 16, and 19 were
also prepared in order to have available a small library of differ-
ently substituted compounds (Table 1). Accordingly, the present
study addressed growth inhibitory activity, migration and capil-
lary-like tube formation in vitro on human umbilical vein ECs
(HUVECs) and the in vivo effects, evaluated by means of Matrigel
plug assay.
Vinblastine, known for its anti-angiogenic properties, was taken
as reference compound.3
2. Materials and methods
Melting points were determined on a Gallenkamp MFB 595
010M/B capillary melting point apparatus, and are uncorrected.
Infrared spectra were recorded on a Perkin–Elmer 1760 FTIR spec-

450
M. G. Ferlin et al. / Bioorg. Med. Chem. 19 (2011) 448–457
2.2. 2-Methyl-4-nitroindole (410
)
Table 1
Structure of the tested 2-aryl-pyrroloquinolinone derivatives 8,
12, 16, and 19–246
Without further purification, the imidate ester 3 was used in the
reaction: in a 100 mL flask, 3.2 g (14.4 mmol) imidate ester in
12.0 mL anhydrous DMSO were treated with 3.0 mL diethyl oxalate
(21.7 mmol, d = 1.076) in 8 mL anhydrous DMF and 1.58 g
(18.8 mmol) potassium ethoxide at 0 °C. After 72 h at refluxing,
the cooled red reaction mixture was poured onto ice/water and
the nitro-indole separated as a red amorphous solid, which was
collected and dried under vacuum at 30 °C. Subsequently it was
purified by flash chromatography (ethyl acetate/n-hexane 1:1).
Yield 25%; mp 192–94 °C; rf 0.63 (ethyl acetate/n-hexane 1:1);
1H NMR (DMSO-d6) d 2.48 (s, 3H, CH3), 6.79 (d, 1H, J3,7 = 0.76, 3-
H), 7.18 (dd, 1H, J6,7 = 8.01 Hz, J6,5 = 8.01 Hz, 6-H), 7.74 (dd, 1H,
J7,6 = 8.01 Hz, J7,5 = 0.76 Hz, 7-H), 7.98 (dd, 1H, J5,6 = 8.01 Hz,
J5,7 = 0.76 Hz, 5-H), 11.98 (br s, 1H, NH).
O
1
R2
R1
N
H
N
R3
Compd R1
R2
R3
8
12
16
19
2-Phenyl
2-Phenyl
2-Phenyl
2-Phenyl
2-Phenyl
Ethyl
7-Methylcyclopropyl
7-Ethylpropanoate
CH3
20a
21a
22a
23a
24a
30-OMe-2-phenyl
30-NO2-2-phenyl
2-Thiophen
2.3. General procedures for the synthesis of 1-substituted 4-
nitroindole derivatives 5, 9, and 13
2-Pyrrolo
a
In a round-bottomed flask, to 55.5 mmol of NaH (60% dispersion
in mineral oil)
Ref. 6.
a
solution of 4-nitro-indole derivative, 26
(18.5 mmol) in 20–30 mL dry DMF was dropped and after 30 min
at room temperature, a solution of halo-derivative in dry DMF
was added and the reaction mixture was left stirred until the end
(TLC). Then, the reaction mixture was workup in a standard way.
The raw products were purified by flash chromatography.
trometer using potassium bromide pressed disks; all values are ex-
pressed in cmꢀ1. UV–vis spectra were recorded on a Perkin–Elmer
Lambda UV/Vis spectrometer. 1H NMR spectra were recorded on
Varian Gemini (200 MHz) and Bruker (300 MHz) spectrometers,
using the indicated solvents; chemical shifts are reported in d
(ppm) downfield from tetramethylsilane as internal reference.
Coupling constants are given in Hertz. In the case of multiplets,
chemical shift was measured starting from the approximate center.
Integrals were satisfactorily in line with those expected on the ba-
sis of compound structure. Elemental analyzes were performed in
the Microanalytical Laboratory, Department of Pharmaceutical Sci-
ences, University of Padova, using a Perkin–Elmer elemental ana-
lyzer model 240B; results fell in the range of calculated values
0.4%. The analytical data are presented in detail for each final
compound in the Supplementary data. Mass spectra were obtained
with a Mat 112 Varian Mat Bremen (70Ev) mass spectrometer and
Applied Biosystems Mariner System 5220 LC/Ms (nozzle potential
250.00). Column flash chromatography was performed on Merck
silica gels (250–400 mesh ASTM); chemical reactions were moni-
tored by analytical thin-layer chromatography (TLC) using Merck
silica gel 60 F-254 glass plates with a 9:1 dichloromethane/meth-
anol mixture as eluant, unless otherwise specified.
2.3.1. 1-Ethyl-4-nitro-1H-indole (5)
Gummy brown solid; yield 78%; rf 0.62 (dichloromethane/eth-
ylacetate 8:2); 1H NMR (DMSO-d6) d 1.38 (t, 3H, J = 7.25 Hz, CH3),
4.33 (q, 2H, J = 7.25 Hz, CH2), 7.05 (dd, 1H, J3,2 = 3.25 Hz,
J3,7 = 0.76 Hz, 3-H), 7.35 (dd, 1H, J6,5 = 8.01 Hz, J6,7 = 8.01 Hz, 6-H),
7.81 (d, 1H, J2,3 = 3.25 Hz, 2-H), 8.05 (t, 2H, overlapping 5-H and
7-H); HRMS (ESI) calculated for C10H11N2O2 [M+H]+ m/z 191.078,
found 191.081.
2.3.2. 1-Cyclopropymethyl-4-nitro-1H-indole (9)
Orange liquid; yield 95%; rf 0.67 (toluene/n-hexane/ethyl ace-
tate 1:1:0.3); 1H NMR (DMSO-d6) d 0.42 (m, 2H, CH2), 0.52 (m,
2H, CH2), 1.26 (m, 1H, CH), 4.18 (d, 2H, J = 7.05, CH2), 7.05 (dd,
1H, J3,2 = 3.05 Hz, J3,7 = 0.76 Hz, 3-H), 7.34 (dd, 1H, J6,5 = 8.01 Hz,
J6,7 = 8.01 Hz, 6-H), 7.86 (d, 1H, J2,3 = 3.05 Hz), 8.26 (d, 1H,
J7,6 = 7.25 Hz), 8.01 (d, 1H, J5,6 = 8.01 Hz, 5-H); HRMS (ESI) calcu-
lated for C12H13N2O2 [M+H]+ m/z 217.093, found 217.1007.
Solutions were concentrated in a rotary evaporator under re-
duced pressure. Starting materials were purchased from Aldrich
Chimica, Acros, and solvents from Carlo Erba, Fluka, and Lab-Scan.
DMSO was obtained anhydrous by distillation under vacuum and
stored on molecular sieves.
2.3.3. 3-(4-Nitro-indol-1-yl)-propionic acid ethyl ester (13)
Brown solid; yield 77%; mp 63–65 °C; rf 0.66 (benzene/n-hex-
ane/ethyl acetate 1:1:1); 1H NMR (DMSO-d6) d 1.08 (t, 3H,
J = 7.05 Hz, CH3), 2.88 (t, 2H, J = 6.67 Hz, CH2), 3.99 (q, 2H,
J = 7.05 Hz, CH2), 4.56 (t, 2H, J = 6.67 Hz, CH2), 7.03 (dd, 1H,
J3,2 = 3.05 Hz, J3,7 = 0.76 Hz, 3-H), 7.35 (dd, 1H, J6,5 = 8.01 Hz,
J6,7 = 8.01 Hz, 6-H), 7.78 (d, 1H, J2,3 = 3.24, 2-H), 8.08 (ddd, 2H,
J5,7 = 2.48, J7,5 = 2.48 Hz, J5,6 = 8.01 Hz, J7,6 = 8.01 Hz, J7,3 = 0.76 Hz,
7-H and 5-H); HRMS (ESI) calculated for C13H15N2O4 [M+H]+ m/z
263.0987, found 263.1150.
2.1. N-(2-Methyl-3-nitro-phenyl)-acetimidic acid ethyl ester
(310
)
In a tow-necked flask equipped with a Claisen condenser were
placed commercial 2-methyl-3-nitroaniline (6.00 g, 39.4 mmol),
p-toluenesulfonic acid (0.032 g, 0.2 mmol) and anhydrous tri-
ethyl-ortoformate (10 mL, d = 0.885, 54.7 mmol); the mixture
was heated at 120 °C until all the ethanol had distilled off. Then,
following vacuum distillation at 150 °C and 0.6 mmHg, the imidate
ester (6.97 g) was collected. Yellow oily liquid; yield 92%; rf 0.80
(ethyl acetate/n-hexane 1:1); 1H NMR (DMSO-d6) d 1.29 (t, 3H,
J = 7.05 Hz, CH3), 1.77 (s, 3H, CH3), 2.15 (s, 3H, CH3), 4.23 (q, 2H,
J = 7.05 Hz, CH2), 7.04 (dd, 1H, J6,5 = 8.01 Hz, J6,4 = 0.96 Hz, 6-H),
7.34 (dd, 1H, J5,6 = 8.01 Hz, J5,4 = 8.01 Hz, 5-H), 7.55 (dd, 1H,
J4,5 = 8.01 Hz, J4,6 = 0.96, 4-H).
2.4. General procedures for the synthesis of 4-amino-indole
derivatives 6, 10, 14, and 17
A solution of a nitro-indole 5, 9, and 13 (5.87 mmol) in 400 mL
ethanol was dropped into a suspension of 10% Pd/C (125 mg) satu-
rated with H2 in ethanol 200 mL. The mixture was stirred at room
temperature with hydrogen at atmospheric pressure for 3 h,
according to TLC analysis (ethyl acetate/n-hexane 1:3). After filter-
ing off the catalyst, the filtrate was evaporated yielding the corre-
sponding aminoindoles.

M. G. Ferlin et al. / Bioorg. Med. Chem. 19 (2011) 448–457
451
2.4.1. 1-Ethyl-1H-indol-4-ylamine (6)
(m, 2H, CH2), 0.49 (m, 2H, CH2), 1.24 (m, t, 4H, overlapping,
Dense brown liquid; yield 91%; rf 0.42 (dichloromethane/ethyl-
acetate 8:2); 1H NMR (DMSO-d6) d 1.30 (t, 3H, J = 7.25 Hz, CH3),
4.07 (q, 2H, J = 7.25 Hz, CH2), 5.33 (br s, 2H, NH2), 6.16 (dd, 1H,
J7,6 = 7.43 Hz, J7,3 = 0.76 Hz, 7-H), 6.44 (dd, 1H, J3,2 = 3.24 Hz,
J3.7 = 0.76 Hz, 3-H), 6.64 (d, 1H, J5,6 = 8.01 Hz, 5-H), 6.81 (dd, 1H,
J6,7 = 7.62 Hz, J6,5 = 8.01 Hz, 6-H), 7.10 (d, 1H, J2,3 = 3.24 Hz, 2-H);
HRMS (ESI) calculated for C10H13N2 [M+H]+ m/z 161.103, found
161.1015.
J = 7.05 Hz, CH and CH3), 3.99 (d, 2H, J = 6.86 Hz, N–CH2), 4.16 (q,
2H, J = 7.05 Hz, CH2), 4.97 (s, 1H, @CH–), 5.93 (d, 1H, J7,6
=
7.72 Hz, 7-H), 6.54 (dd, 1H, J3,2 = 3.14 Hz, J3,7 = 0.76 Hz, 3-H), 6.75
(dd, 1H, J6,7 = 7.72 Hz, J6,5 = 8.20 Hz, 6-H), 7.13 (d, 1H, J5,6
8.20 Hz, 5-H), 7.33 (m, 5H, overlapping), 7.45 (d, 1H, J2,3
=
=
3.14 Hz, 2-H), 10.47 (br s, 1H, NH); HRMS (ESI) calculated for
C
23H24N2O2 [M+H]+ m/z 361.187, found 361.1928.
2.5.3. 3-[1-(2-Ethoxycarbonyl-ethyl)-1H-indole-4-ylamino]-3-
phenyl-acrylic acid ethyl ester (15)
2.4.2. 1-Cyclopropylmethyl-1H-indol-4-ylamine (10)
Dense brown liquid; yield 93%; rf 0.35 (toluene/n-hexane/ethyl
acetate 1:1:0.3); 1H NMR (DMSO-d6) d 0.33 (m, 2H, CH2), 0.47 (m,
2H, CH2), 1.19 (m, 1H, CH), 3.90 (d, 2H, CH2, J = 6.86 Hz), 5.15 (br s,
2H, NH2), 6.15 (dd, 1H, J7,6 = 7.44 Hz, J7,3 = 0.76 Hz, 7-H), 6.49 (dd,
1H, J3,2 = 3.14 Hz, J3,7 = 0.76 Hz, 3-H), 6.65 (d, 1H, J5,6 = 8.20 Hz, 5-
H), 6.80 (dd, 1H, J6,7 = 7.44 Hz, J6,5 = 8.20 Hz, 6-H), 7.14 (d, 1H,
J2,3 = 3.14 Hz, 2-H); HRMS (ESI) calculated for C12H15N2 [M+H]+
m/z 187.119, found 187.1239.
Yield was not estimated. Dense brown liquid; rf 0.51 (ethyl ace-
tate/methanol 7:3); 1H NMR (DMSO-d6) d 1.11 (t, 3H, J = 7.05 Hz,
CH3), 1.16 (t, 3H, J = 7.05 Hz, CH3), 2.80 (t, 2H, J = 6.67 Hz, CH2),
4.01 (q, 2H, J = 7.05 Hz, CH2), 4.13 (q, 2H, J = 7.05 Hz, CH2), 4.38
(t, 2H, J = 6.67 Hz, CH2), 4.98 (s, 1H, NH), 5.94 (d, 1H,
J7,6 = 7.44 Hz, 7-H), 6.54 (d, 1H, J3,2 = 3.24 Hz, 3-H), 6.75 (dd, 1H,
J6,7 = 7.82 Hz, J6,5 = 8.20 Hz, 6-H), 7.10 (d, 1H, J5.6 = 8.20 Hz, 5-H),
7.33 (m, 6H, overlapping), 10.45 (br s, 1H, NH); HRMS (ESI) calcu-
lated for C24H27N2O4 [M+H]+ m/z 407.1926, found 407.1997.
2.4.3. 3-(4-Amino-indol-1-yl)-propionic acid ethyl ester (14)
Dense brown liquid; yield 98%; rf 0.25 (ethyl acetate/methanol
7:3); 1H NMR (DMSO-d6) d 2.16 (s, 6H, –(CH3)2), 2.55 (t, 2H,
J = 6.86 Hz, CH2), 4.11 (t, 2H, J = 6.86 Hz, CH2), 5.15 (br s, 2H,
NH2), 6.14 (dd, 1H, J5,6 = 7.44 Hz, J5,7 = 0.57 Hz, 5-H), 6.47 (dd, 1H,
J3,2 = 3.24 Hz, J3,7 = 0.76, 3-H), 6.60 (d, 1H, J7,6 = 8.20 Hz, 7-H), 6.80
(dd, 1H, J6,5 = 7.63 Hz, J6,7 = 8.01 Hz, 6-H), 7.09 (d, 1H,
J2,3 = 3.24 Hz, 2H); HRMS (ESI) calculated for C13H17N2O2 [M+H]+
m/z 233.125, found 233.1245.
2.5.4. 3-(2-Methyl-1H-indole-4-ylamino)-3-phenyl-acrylic acid
ethyl ester (18)
Yield was not estimated. Brown dense liquid; rf 0.76 (ethyl ace-
tate/n-hexane 1:1); 1H NMR (DMSO-d6) d 1.24 (t, 3H, J = 7.05 Hz,
CH3), 2.39 (s, 3H, CH3), 4.03 (q, 2H, J = 7.25 Hz, CH2), 4.92 (s, 1H,
–CH), 5.97 (d, 1H, J5,6 = 7.78 Hz, 5-H), 6.22 (s, 1H, 3-H), 6.60 (dd,
1H, J6,7 = 7.78 Hz, J6,5 = 7.78 Hz, 6-H), 6.89 (d, 1H, J7,6 = 7.78 Hz, 7-
H), 7.33 (m, 5H, ar), 10.41 (br s, NH), 11.43 (br s, NH); HRMS
(ESI) calculated for
321.1573.
C
20H21N2O2 [M+H]+ m/z 321.156, found
2.4.4. 2-Methyl-4-amino-1H-indole (17)
Dense orange liquid; yield 98%; rf 0.57 (ethyl acetate/n-hexane
8:2); 1H NMR (DMSO-d6) d 2.30 (s, 3H, CH3), 4.94 (br s, 2H, NH2),
6.08 (dd, 1H, J5,6 = 7.44, J5,7 = 0.95 Hz, 5-H), 6.13 (d, 1H,
J3,7 = 0.95 Hz, 3-H), 6.30 (dd, 1H, J7,6 = 8.01 Hz, J7,5 = 0.76 Hz, 7-H),
6.65 (dd, 1H, J6,5 = 8.01 Hz, J6,7 = 8.40 Hz, 6-H), 10.57 (br s, 1H,
NH); HRMS (ESI) calculated for C9H11N2 [M+H]+ m/z 147.088, found
147.0918.
2.6. General procedures for the synthesis of 2-phenyl-
pyrroloquinolinones 8, 12, 16, and 19
In a two-necked round-bottomed flask, 50 mL of diphenyl ether
were heated to boiling temperature; 1–2 mmol of acrylates 7, 11,
15, and 18 were then added portion-wise and the mixture was re-
fluxed for 30 min. After cooling to 60 °C, the separated precipitate
was collected by filtration and washed many times with diethyl
ether. In all cases, the collected products were purified by flash
chromatography (ethyl acetate/methanol 9:1). Yields were esti-
mated from corresponding amino compound intermediates.
2.5. General procedures for the synthesis of ethyl acrylate
derivatives 7, 11, 15, and 18
In a round-bottomed flask, 3–4 mmol of 3-substituted-amino-
indole 6, 10, 14, and 17 in 10–20 mL absolute ethanol were
condensed with an equimolar quantity of commercial ethyl ben-
zoylcetate and with 1 mL glacial acetic acid and 100 mg drierite.
The mixture was refluxed for about 24 h (TLC), and after removing
the drierite; the resulting solution was evaporated to dryness and
the residue used in the following reaction without any purification.
A sample was purified for analysis by flash chromatography.
2.6.1. 7-Ethyl-2-phenyl-1H,7H-pyrrolo[2,3-h]quinolin-4-one (8)
Yield 37%; rf 0.51 (ethyl acetate/methanol 9:1), mp 250–261 °C;
IR (KBr) 1628 cmꢀ1 (CO); IR (KBr) 1628 cmꢀ1; 1H NMR (DMSO-d6) d
1.34 (t, 3H, J = 7.25 Hz, CH3), 4.22 (q, 2H, J = 7.25 Hz, CH2), 6.21 (d,
1H, J3,1 = 1.14 Hz, 3-H), 7.07 (d, 1H, J5,6 = 8.96 Hz, 5-H), 7.29 (m, 2H,
6-H and 9-H overlappping), 7.42 (d, 1H, J8,9 = 3.24, 8-H), 7.51 (m,
5H, aryl), 12.03 (s, 1H, NH quinolinic); 13C NMR (DMSO-d6) d
16.07 (CH3), 42.30 (CH2), 100.72 (C-9a), 107.77 (C-9), 17.77 (C-
4a), 113.33 (C-6), 116.90 (C-60, C-20), 117.34 (C-8), 121.15 (C-30,
C-50), 128.54 (C-40), 129.68 (C-9b e C-10), 129.82 (C-6a), 129.96
(C-5), 134.83 (C-2), 138.08 (C-4), 139.67 (C-8), 157.12 (C-30, C-50),
165.23 (C-40); HRMS (ESI) calculated for C19H16N2O [M+H]+ m/z
289.1335, found 289.1334.
2.5.1. 3-(1-Ethyl-1H-indole-4-ylamino)-3-phenyl-acrylic acid
ethyl ester (7)
Yield was not estimated. Dark red vitreous solid; rf 0.96 (dichlo-
romethane/ethyl acetate 8:2); 1H NMR (DMSO-d6) d 1.25 (t, 3H,
J = 7.05 Hz, CH3), 1.36 (t, 3H, J = 7.25 Hz, CH3), 3.17 (q, 2H,
J = 7.25 Hz, CH2), 3.99 (d, 2H, J = 7.05 Hz, CH2), 4.97 (s, 1H, –C@CH–),
5.93 (d, 1H, J5,6 = 7.63 Hz, 5-H), 6.53 (dd, 1H, J3,2 = 3.05 Hz, J3,7
=
0.76, 3-H), 6.75 (dd, 1H, J6,5 = 7.78 Hz, J6,7 = 8.20 Hz, 6-H), 7.40 (d,
1H, J2,3 = 3.05 Hz, 2-H), 7.52 (m, 6H, ar); HRMS (ESI) calculated for
C
2.6.2. 7-Cyclopropylmethyl-2-phenyl-1H,7H-pyrrolo[2,3-
h]quinolin-4-one (12)
21H23N2O2 [M+H]+ m/z 335.171, found 335.1712.
Yield 35%; rf 0.50 (ethyl acetate/methanol 9:1); mp 260–262 °C;
IR (KBr) 1627 cmꢀ1 (CO); 1H NMR (DMSO-d6) d 0.41 (m, 2H, CH2),
0.52 (m, 2H, CH2), 1.26 (m, 1H, CH), 4.14 (d, 2H, CH2), 6.23 (s,
1H, 4-H), 7.38 (d, 1H, J9,8 = 3.24 Hz, 9-H), 7.51 (d, 1H,
J8,9 = 3.24 Hz, 8-H), 7.54 (d, 1H, J6,5 = 8.86 Hz, 6-H), 7.60 (m, 3H,
aryl), 7.78 (m, 2H, aryl), 7.87 (d, 2H, J5,6 = 8.96 Hz, 5-H), 11.52 (br
2.5.2. 3-(1-Cyclopropylmethyl-1H-indole-4-ylamino)-3-phenyl-
acrylic acid ethyl ester (11)
Yield was not estimated. Yellow oily liquid; rf 0.78 (toluene/
n-hexane/ethyl acetate 1:1:0.3); 1H NMR (DMSO-d6)
d 0.36

452
M. G. Ferlin et al. / Bioorg. Med. Chem. 19 (2011) 448–457
s, NH quinolinic); 13C NMR (CDCl3): d 4.01 2(CH2), 11.35 (–CH–),
51.03 (CH2), 99.21, 107.93, 109.04, 116.36, 118.48, 119.11,
126.97, 129.06, 130.08, 134.85, 135.11, 137.48, 148.53, 178.81;
HRMS (ESI) calculated for C21H18N2O [M+H]+ m/z 315.1453, found
315.1492.
mL, Sigma) for 4 h. Formazane precipitates were dissolved in 2-pro-
panol acid (0.04 M HCl in 2-propanol; Sigma) and optical density
was measured at 570 nm using a Microplate auto reader EL 13
(BIO-TEK instruments Inc. Winooski, Vermont, USA). Results were
expressed as percent change from control non-treated cultures as
well as IC50, that is, the concentrations of the agents (lM), which in-
2.6.3. 3-(4-Oxo-2-phenyl-1,4-dihydro-pyrrolo[2,3-h]quinolin-7-
propionic acid ethyl ester (16)
duced a 50% reduction of viability in comparison with non-treated
cultures. Both IC50 values and non-cytotoxic concentrations were
obtained from dose–response curves. The linearity of absorbance
of formazan over a range of 5 ꢁ 103 to 50 ꢁ 104 cells was estab-
lished by determining the linear coefficient (0.9858).
Yield 30%; rf 0.50 (ethyl acetate/methanol 9:1); mp 260–262 °C;
IR (KBr) 1627 cmꢀ1 (CO); 1H NMR (DMSO-d6) d 0.41 (m, 2H, CH2),
0.52 (m, 2H, CH2), 1.26 (m, 1H, CH), 4.14 (d, 2H, CH2), 6.23 (s,
1H, 4-H), 7.38 (d, 1H, J9,8 = 3.24 Hz, 9-H), 7.51 (d, 1H,
J8,9 = 3.24 Hz, 8-H), 7.54 (d, 1H, J6,5 = 8.86 Hz, 6-H), 7.60 (m, 3H,
aryl), 7.78 (m, 2H, aryl), 7.87 (d, 2H, J5,6 = 8.96 Hz, 5-H), 11.52 (br
s, NH quinolinic); 13C NMR (CDCl3): d 4.01 2(CH2), 11.35 (–CH–),
51.03 (CH2), 99.21, 107.93, 109.04, 116.36, 118.48, 119.11,
126.97, 129.06, 130.08, 134.85, 135.11, 137.48, 148.53, 178.81;
HRMS (ESI) calculated for C21H18N2O [M+H]+ m/z 315.1453, found
315.1492.
3.3. Pro-apoptotic assay
HUVECs (2 ꢁ 104 cell/cm2) were seeded onto fibronectin coated
8-well chamber slides (BD) and incubated overnight at 37 °C. Then,
cells were washed and treated for 24 h with IC50 concentration of
the tested compounds dissolved in basal medium. Cultures were
fixed in 3% paraformaldehyde for 1 h at room temperature and
apoptosis was detected by TUNEL assay, using the Roche In situ
Cell Death detection kit. Briefly, cells were permeabilized with
0.1% Triton X-100 (Sigma) in 0.1% sodium citrate solution for
2 min on ice, washed twice with PBS and then incubated with a
mixture of terminal deoxynucleotidyl transferase and fluores-
cein-labeled nucleotides for 1 h at 37 °C. The total number of nuclei
were visualized by fluoresce microscopy after incubation of cul-
tures with DAPI for 10 min at room temperature. The apoptotic
rate (percent of TUNEL-positive cells) was estimated by counting
10 fields for each sample and three wells were employed for each
experimental condition. Results were expressed as percent change
from control cultures grown with basal medium. In the control
non-treated cultures, negligible apoptotic rate was seen.
2.6.4. 8-Methyl-2-phenyl-1H,7H-pyrrolo[2,3-h]quinolin-4-one
(19)
Yield 30%; rf 0.55 (ethyl acetate/methanol 9:1); mp 233–235 °C;
IR (KBr) 1628 cmꢀ1 (CO); 1H NMR (DMSO-d6) d 2.43 (s, 3H, CH3),
6.20 (s, 1H, 3-H), 7.05 (s, 1H, 9-H), 7.27 (d, 1H, J6,5 = 8.58 Hz, 6-
H), 7.57 (m, 3H, aryl), 7.76 (m, 3H, aryl), 11.38 (br s, 1H, NH pyrro-
lic), 11.46 (br s, 1H, NH, quinolinic); 13C NMR (CDCl3) d 13.54,
73.54, 99.93, 108.67, 111.54, 117.59, 118.26, 129.09, 130.05,
131.49, 136.04, 136.31, 136.79, 139.46, 152.36, 180.37, 205.57;
HRMS (ESI) [M+H]+ calculated for C18H14N2O m/z 275.1140, found
275.1179.
Elemental analysis data for new compounds are given in Sup-
plementary data.
3.4. Migration assay
3. Biology
Cell migration was evaluated using a modified Boyden chamber.
HUVECs (1.5 ꢁ 104 cell/cm2) were seeded on the upper side of
3.1. Cell cultures
5.0 lm pore Transwell insert (Corning Inc.) in basal medium sup-
plemented with 1% FCS with or without not cytotoxic concentration
of the tested compounds. Inserts were placed in a 24-well plate
containing medium supplemented with 1% FCS and growth factors
(lower chamber). After 4 h. cultures were fixed in 10% formalde-
hyde (Sigma) and the upper membrane of the insert was swabbed
to removed non-migrated cells. The membrane was cut from the in-
sert and mounted with DAPI. HUVECs migration was quantified by
counting the number of nuclei in the lower side of the membrane in
five random fields per insert (100ꢁ magnification) by fluoresce
microscopy. Experiments were repeated five times. Results were
expressed as percent change from control not treated cultures.
Primary cultures of human umbilical vein ECs (HUVECs) were
obtained by enzymatic digestion of umbilical vein endothelial
layer with 0.1% collagenase IV solution (Sigma Aldrich) and the
cells were seeded on Petri dishes previously coated with fibronec-
tin (1
MV2 (basal medium, Promocell) supplemented with 5% FCS, ascor-
bic acid (1 g/mL), hFGF-2 (10 ng/mL), hEGF (5 ng/mL), hydrocorti-
sone (0.2
g/mL), R3-IGF-1 (20 ng/mL), VEGF (0.5 ng/mL)
lg/mL) and cultured with Endothelial Cell Growth Medium
l
l
(endothelial MV2 medium kit) (Promocell), and 1% antibiotic solu-
tion containing streptomycin sulfate (10 ng/mL), amphotericin-B
(250 ng/mL), and penicillin (100 U/mL) (Sigma). Cultures were
incubated at 37 °C in a humidified atmosphere. HUVECs were used
from 2nd to 4th passage and harvested at 80% confluence.
3.5. Morphogenesis analysis
3.2. Growth inhibitory activity assay
Morphogenesis analysis was carried out seeding HUVECs on
Matrigel (BD Biosciences). Matrigel was thawed on ice overnight,
The 3-(4.5-dimethylthiazol-2-yl)-2.5-dimethyltetrazolium bro-
mide (MTT)-tetrazolium dye assay was used to evaluate the growth
inhibitory activity of the compounds at various concentrations. HU-
VECs (2 ꢁ 104 cell/cm2) were seeded into 96-well plates. After a
24 h incubation period, media were replaced with ones containing
spread evenly over each well (50 ll) of a 24-well plate, and allowed
to polymerized for 30 min at 37 °C. HUVECs (2.5 ꢁ 104 cells/cm2)
were seeded on Matrigel and cultured in basal medium supple-
mented with 1% FCS and with or without not cytotoxic concentra-
tions of the tested compounds, as above detailed, FGF-2 (50 ng/
mL), and FGF-2 plus compounds. After 18 h of incubation at 37 °C,
cultures were fixed in 2% glutaraldehyde in cacodylate buffer, pH
7.2 and then photographed (five fields for each well: the four quad-
rants and the center) at a magnification of 50ꢁ. Phase contrast
images were recorded using a digital camera and save as TIFF files.
Image analysis was carried out using the ImageJ image analysis soft-
ware and the following dimensional parameters (percent area cov-
or not various concentrations (from 10 to 10ꢀ5
lM) of the tested
compounds. Alternatively, 50 ng/mL FGF-2 (Sigma) alone or with
the compounds at the highest concentration not inducing decrease
in cell viability (10ꢀ5 M vinblastine; 10ꢀ4 M 2-PPyQ 20; 10ꢀ3
l l lM
30-NO2- 22; 10ꢀ2 lM 30-MeO- 21, 7-c-Propylmet- 12, 7-EtPropano-
ate- 16, and 7-Et- 8 2-PPyQs) were simultaneously added to the
media. After 20 h incubation, cells were treated with MTT (0.5 mg/

M. G. Ferlin et al. / Bioorg. Med. Chem. 19 (2011) 448–457
453
ered by HUVECs and total length of HUVECs network per field), and
5. Results
topologicalparameters (numberof meshes and branchingpointsper
fields) were estimated. Values were expressed as percent change
from control cultures grown with basal medium or FGF-2.
5.1. Chemistry
The new 2-PPyQs, substituted at 7N-pyrrole (8, 12, and 16) and
at position 8 (19), were synthesized according to the routes shown
in Scheme 1: in section A, 4-nitroindoles 2 and 4 were obtained by
the Sand and Bergman procedure.9,10 Commercial 3-nitro-2-
methyl-aniline was transformed into imidate esters 1 (85%) and
3 (92%) by treatment with triethyl orthoformate or orthoacetate,
respectively, in the presence of p-toluensulfonic acid. The imidates
furnished cyclization products 24 (64%) and 4 (25%) when reacted
with potassium tert-butoxide and diethyloxalate in anhydrous
DMF and DMSO. It is interesting to note that in the case of formim-
idate 1, the only one cyclized 4-nitro-indole 2 was isolated from
the reaction mixture, whereas in the case of methyl imidate 3, a
mixture was obtained of 4 and the 3-diketoethylester derivative
4a, which was isolated and characterized (Supplementary data).
This finding confirms the reaction mechanism proposed by Sand
and Bergman.9,10
3.6. In vivo Matrigel plug assay
Three-month-old C57/BL6 mice were anesthetized with isoflu-
rane (2–3%) via nose cone and 100% oxygen used as the carrier gas;
the back was shaved and disinfected with ethyl alcohol. A total of
0.5 mL of Matrigel, mixed with 12 U.I. heparin (Vister) with or with-
out 200 ng/plug FGF-2 and 0.01 lM tested compounds, was injected
subcutaneously on dorsal area. After injection, the Matrigel poly-
merized forming a plug. After 7 days, the animals were sacrificed
and the plugs were carefully removed and steeped in 300 ll/plug
Brij-35 0.1% solution in PBS at 4 °C overnight. Hemoglobin concen-
tration was analyzed using Drabkin’s Reagent kit (Sigma). The optical
density wasread at540 nm using a Microplateautoreader EL 13. The
results were expressed as mg/mL hemoglobin.
In section B, nitroindole 2 was alkylated to 1N-alkyl derivatives
5, 9, and 13 by appropriate halogenated compounds and NaH in
DMF or, in the case of 13, toluene, in order to prevent hydrolysis
of propanoate ethyl ester to the corresponding propanoic acid
derivative. Next, catalytic reductions of5, 9, 13, and 4 were all per-
4. Statistical analysis
All results were expressed as mean SD of three separate
experiments. Their statistical comparison was performed by anal-
ysis of variance followed by Student’s t-test.
A] Synthetic route to 4-nitroindoles 26 and 4
NO2
NO2
NO2
COOC2H5
O
H
CH3
C
C
b
H
a
N
C
N
H
N
CH
OC2H5
26
OC2H5
NO2
1
CH3
NH2
O
O
NO2
NO2
NO2
NO2
COOC2H5
H
C
C2H5
CH3
c
C
C
CH3
N
b
O
CH3
CH3
CH3
N
H
N
N
H
C
OC2H5
OC2H5
4
4a
3
B] Synthesis of substituted 2-phenyl-pyrroloquinolinones 8, 12, 16 and 19
O
C2H5
H
C
NO2
NH2
C
NO2
C
O
a
c
b
NH
R3
R3
R3
N
N
N
R2
R2
R2
R3
N
26
4, 5, 9, 13,
6, 10, 14, 17
R2
7, 11, 15, 18
Compounds
d
R2
R3
2
H
H
H
O
5-8
ethyl
9-12
13-16
4, 17-19
cyclopropylmethyl
ethyl propanoate
H
H
R2
N
H
N
H
methyl
R3
8, 12, 16, 19
Scheme 1. (A) Synthetic route to 4-nitroindoles 26 and 4. Reagents and conditions: (a) triethyl ortho-formate, p-toluensulfonic acid monohydride, 120 °C; (b) potassium tert-
butoxide, diethyl oxalate, anhydrous DMF/DMSO mixture, 50 °C; (c) triethyl ortho-acetate, p-toluensulfonic acid monohydride, 120 °C. (B) Synthesis of substituted 2-phenyl-
pyrroloquinolinones 8, 12, 16, and 19. Reagents and conditions: (a) bromoethane or (bromomethyl)cyclopropane, NaH/DMF or ethyl acrylate NaH/toluene, reflux; (b) H2 atm,
Pd/C 10%, ethanol, 40 °C; (c) ethyl benzoylacetate, absolute ethanol, acetic acid, drierite, reflux; (d) diphenyl ether, 250 °C.

454
M. G. Ferlin et al. / Bioorg. Med. Chem. 19 (2011) 448–457
formed with H2 and Pd/C 10%, leading to almost pure 4-amino-in-
doles 6, 10, 14, and 17 in high yields. The two final steps, conden-
sation with ethyl benzoylacetate and thermal intramolecular
cyclization, were performed following previously published proto-
cols,6–8 furnishing first enamine derivatives 7, 11, 15, and 18 and
then 2-PPyQs 8, 12, 16, and 19.
5.2. Biology
5.2.1. Growth inhibitory effects of 2-PPyQs on endothelial cells
Inhibitory effects on cell growth were determined by MTT assay
(Table 2). Results showed that derivatives, bearing a phenyl group
at the 2-position, were more cytotoxic than those with a thiophene
and pyrrolo ring 23 and 24, respectively. Compound 20 had an IC50
value of 4.3 ꢁ 10ꢀ2
lM similar to that determined for vinblastine,
4.9 ꢁ 10ꢀ2
lM (Table 2). The presence of substituents at the phenyl
group, compounds 21 and 22, decreased viability (IC50 5.2 and
Table 2
In vitro inhibitory effects on cell growth of HUVECs by compounds
0.45 lM, respectively), and, consequently, the corresponding IC50
8, 12, 16, 19–24
values were from 10- to 100-fold lower than that calculated for
compound 20. Derivative 19, bearing the methyl group at the 8-po-
sition, was almost ineffective (IC50 >10 lM). In view of the inhibi-
tory effect on HUVEC cell growth, the more active compounds, 8,
a
Compd
IC50
(lM)
7-Et-2-PPyQ 8
8.1 1.1
4.2 0.7
7-MeCypr-2-PPyQ 12
7-EtPr-2-PPyQ 16
8-Me-2-PPyQ 19
2-PPyQ 20
30-MeO-2-PPyQ 21
30-NO2-2-PPyQ 22
2-Th-2-PPyQ 23
2-Py-2-PPyQ 24
Vinblastine
5.7 0.8
>10b
12, 16, 20–22, were selected for further studies.
4.3 ꢁ 10ꢀ2 1.1 ꢁ 10ꢀ2
5.8 0.9
0.45 0.19
>10b
5.2.2. Effects on in vitro angiogenesis
It was evaluated whether the 2-PPyQ derivatives 8, 12, 16, 19–
24 could counteract the angiogenic effects of FGF-2, which repre-
sents one of the most important growth factors involved in the
angiogenic process. The increase in cell proliferation induced by
FGF-2 was completely reversed by all assayed compounds at
non-cytotoxic concentrations (see growth inhibition assay). Com-
pounds 8 and 20 appeared more, while compounds 12, 16, and
22 less effective than vinblastine (Fig. 3).
>10b
4.9 ꢁ 10ꢀ2 0.9 ꢁ 10ꢀ2
a
Values are means SD of at least three independent
experiments.
IC50 not determined, because <50% inhibition was observed
at highest tested concentration (10 M). Higher concentrations
b
l
were not used, to avoid precipitation of compounds in culture
medium.
22
*°
21
20
*°
*°
16
*°
12
°
8
*°
vinblastine
*°
FGF-2
*
-60
-40
-20
0
20
40
% cha nge from control
Figure 3. Effects of compounds 8, 12, 16, and 20–22 on HUVEC proliferation incubated for 24 h with 50 ng/mL FGF-2. Data are means SD of three separate experiments.
*
p <0.05 versus control; °p <0.05 versus FGF-2, Student’s t-test.
30
*
25
*
*
*
20
15
*
*
10
5
*
0
vinblastine
8
12
16
20
21
22
Figure 4. Pro-apoptotic effects of compounds 8, 12, 16, and 20–22 on HUVECs. Cells were treated for 24 h with compounds at IC50 concentrations. After the incubation period,
cultures were fixed and apoptosis was detected by TUNEL assay. Results, expressed as percent change from untreated control cultures (taken as 0), are means SD of at least
*
three independent experiments. p <0.05 versus control cultures, Student’s t-test.

M. G. Ferlin et al. / Bioorg. Med. Chem. 19 (2011) 448–457
455
5.2.3. Pro-apoptotic effects of 2-PPyQs
icant inhibitory effects on HUVEC migration at the highest concen-
tration not inducing decrease in cell growth (see growth inhibition
assay) compared with that determined in control cultures (Fig. 5).
Nevertheless, only compound 8 was as effective as vinblastine, the
other derivatives appeared to be less effective.
To determine the mode by which these 2-PPyQs decreased cell
growth, their effect on apoptosis was assessed by TUNEL assay. The
cytotoxic effects of 2-PPyQ 8, 12, 16, and 20–22, such as vinblas-
tine, seemed to be related to their pro-apoptotic activity. All com-
pounds induced increases in the apoptotic rate when compared
with untreated cultures (Fig. 4).
5.2.5. Effects of 2-PPyQs on capillary-like formation
The final event of angiogenesis is the organization of ECs in a
three-dimensional network of capillary-like structures. In vitro HU-
VECs seeded on a thick layer of Matrigel, rapidly align and within a
few hours form a highly branched network of capillaries, which is
complete in 24 h. Therefore, we examined whether 2-PPyQs at
non-cytotoxic concentrations (see growth inhibition assay) were
able to affect the formation of these structures. Morphogenesis on
Matrigel was inhibited by vinblastine, which significantly de-
creased both dimensional (percent area covered by HUVECs and to-
tal length of network per field, A and B in Fig. 6, respectively) and
topological parameters (number of meshes and branching points
per field, C and D in Fig. 6, respectively) of the capillary-like
network. Compounds 8, 12, 16, and 20 were more effective than
vinblastine, leading to a drastic reduction in the number of capil-
lary-like structures, while 21 and 22 only affected topological
parameters.
5.2.4. Effects on endothelial cell migration
Cell migration represents an important step in the angiogenic
process. All tested compounds 8, 12, 16, and 20–22 induced signif-
22
*
21
*
20
*
16
*
12
*
8
*
To further investigate whether 2-PPyQs exhibit anti-angiogenic
activity, their in vitro effects on capillary-like formation after treat-
ment with the pro-angiogenic FGF-2, were also evaluated. The FGF-
2-induced increases in both dimensional and topological parame-
ters were completely abolished by vinblastine, and by compounds
16 and 22 (A, B and C, D in Fig. 7, respectively). The compound 20
affected all parameters except for the numbers of branching points
per field (D in Fig. 7). The compound 21 had no effect, whereas the
compounds 8 and 12 decreased only the number of branching
points.
vinblastine
*
-80
-60
-40
-20
0
20
% change from control
Figure 5. HUVEC migration through a 5.0 lm pore after 4 h from seeding. Cells
were treated with compounds 8, 12, 16, and 20–22 at highest concentration not
inducing decrease in cell viability. Results, expressed as percent change from
untreated control cultures (taken as 0), are means SD of at least three independent
experiments. p <0.05 versus control cultures, Student’s t-test.
A
B
22
21
22
21
20
20
*
*
16
16
*
*
12
12
*
*
8
8
*
*
vinblastine
vinblastine
*
*
-100
-80
-60
-40
-20
0
20
-100
-80
-60
-40
-20
0
20
% change from control
% change from control
D
C
22
21
20
22
*
21
*
20
16
*
*
16
12
*
*
12
8
*
*
*
8
*
vinblastine
vinblastine
*
*
-100
-80
-60
-40
-20
0
20
-100
-80
-60
-40
-20
0
20
% change from control
% change from control
Figure 6. Quantitative analysis of effects of compounds 8, 12, 16, and 20–22 on dimensional (A: percent area covered by HUVECs; B: total length per field) and topological
parameters (C: numbers of meshes per field; D: numbers of branching points per field) of HUVEC meshwork. Cultures were treated with compounds at highest concentration
*
not inducing decrease in cell viability. Bars are means SD of three separate experiments. p <0.05 versus control cultures, Student’s t-test.

456
M. G. Ferlin et al. / Bioorg. Med. Chem. 19 (2011) 448–457
A
B
22
21
22
*
*
21
20
20
16
12
*
*
16
*
*
12
8
8
vinblastine
vinblastine
*
*
-80
-60
-40
-20
0
20
-80
-60
-40
-20
0
20
% change from cultures treated with FGF-2
% change from cultures treated with FGF-2
C
D
22
22
*
*
21
21
20
20
*
16
16
*
*
12
8
12
8
*
*
vinblastine
vinblastine
*
*
-80
-70
-60
-50
-40
-30
-20
-10
0
10
20
-80
-60
-40
-20
0
20
% change from cultures treated with FGF-2
% change from cultures treated with FGF-2
Figure 7. Quantitative analysis of effects of compounds 8, 12, 16, and 20–22 on dimensional (A: percent area covered by HUVECs; B: total length per field) and topological
parameters (C: numbers of meshes per field; D: numbers of branching points per field) of HUVEC meshwork. Cells were treated with 50 ng/mL FGF-2 with or without 2-PPyQs
at highest concentration not inducing decrease in cell viability. Data are expressed as percent change from control value and bars are means SD of four separate
*
experiments; p <0.05 from control or FGF-2-treated cultures, Student’s t-test.
5.2.6. Effects on in vivo angiogenesis
mentary data). According with the in vitro results, compounds 12,
16, 20, and 22 showed also in vivo a strong anti-angiogenic activity
(Fig. 8). Indeed, like vinblastine, a marked reduction of FGF-2-in-
duced angiogenic response was observed in the plugs containing
The effect of 2-PPyQs on angiogenesis was evaluated using the
Matrigel Plug assay. Subcutaneous injection of Matrigel plugs con-
taining 50 ng/mL FGF-2 induced a strong angiogenic response in
4 days, with a 2–4-fold increase in the hemoglobin content com-
pared with plugs containing Matrigel alone (see images in Supple-
0.01 lM tested compounds.
6. Discussion
Herein we describe studies undertaken to evaluate the effects of
some antimicrotubule pyrroloazaflavones on both in vitro and
in vivo angiogenesis. The in vitro assays were carried out on HU-
VECs, considering the growth inhibitory activity, the pro-apoptotic
effects, the cell migration inhibition, and the effects on morpho-
genesis. The in vivo effect was evaluated by means of Matrigel plug
assay in C57/BL6 mice. Vinblastine was taken as reference, a well
know clinically used chemotherapeutic belonging to microtu-
bule-depolymerising vinca alkaloid family. It was described as vas-
cular disrupting agent and showed to exhibit anti-angiogenic
activity at very low concentration with suppression of EC migra-
tion and capillary-like structures formation.11 Our results showed
that pyrroloazaflavones exert, albeit to different extents, anti-
angiogenic activity affecting the various step of the process that
leads to the formation of new blood vessels.
2.0
1.6
1.2
*
*
*
0.8
*
*
*
0.4
0.0
Angiogenesis is characterized by EC activation leading to EC
proliferation as well as EC migration and capillary-like tube
formation.
Growth inhibition assay revealed that 2-PPyQs 8, 12, 16, 20–
22 had inhibitory effects on both basal and FGF-2-induced cell
proliferation. The ability of these compounds to inhibit EC growth
Figure 8. The effect of 12, 16, 22, and 20 treatment on the hemoglobin content in
Matrigel plugs. 200 ng/plug FGF-2 and 0.01 lM tested compounds were injected
subcutaneously on dorsal area of C57/BL6 mice. After 7 days, hemoglobin concen-
tration in the plugs was determined. Data are expressed as mg/mL hemoglobin and
*
bars are means SD of three animals; p <0.05 from plugs containing only FGF-2,
Student’s t-test.

M. G. Ferlin et al. / Bioorg. Med. Chem. 19 (2011) 448–457
457
in the presence of the growth factor, which is reported to have a
role in the development of various tumors, prompted us to inves-
tigate the mode by which they induced reduction of the cell
number.
already shown for taxanes and cholchicine derivatives as well as
for others microtubule/tubulin agents.15–20
7. Conclusions
Our data indicate that 2-PPyQs have pro-apoptotic activity
which is likely to contribute to their anti-angiogenic effects. Sev-
eral anti-angiogenic agents such as angiostatin and endostatin
have been reported to inhibit angiogenesis in part by inducing
endothelial cell apoptosis.12 Moreover, these results match evi-
dence that almost microtubule-affecting agents promote apoptosis
in tumor cell lines.4,13 Nevertheless, the positive effect of pyrrolo-
azaflavones on apoptosis requires more investigations to asses the
mechanism of their cytotoxicity.
The 2-PPyQ derivatives, as vinblastine, have shown to inhibit EC
migration, proving an anti-angiogenic activity and at not cytotoxic
concentrations. This was further supported by finding that 2-PPyQs
prevented capillary-like structure formation in vitro. Addition of 2-
PPyQs, at the time of HUVEC seeding on Matrigel, blocked the abil-
ity of ECs to align, preventing the formation of tubular structures.
Compounds 16 and 22, due to their remarkable effects on dimen-
sional and topological parameters, greater than those of vinblas-
tine ones, turn out very attractive for further investigations. The
inhibitory effect of certain compounds on morphogenesis and cell
migration was probably due to their interaction with microtubules,
which are involved not only in the formation of the mitotic spindle
during cell division, but also play an important role during cell
movements.14 Interestingly, these effects were observed with here
2-PPyQs using in vitro experimental conditions that did not affect
cell proliferation. Previously, immunofluorescence assays evi-
denced that the here 20, 21 and the 7-methyl–PPyQ were able to
destabilize microtubules in Ovcar-3 and MCF-7 cell lines, causing
loss of the normal microtubule network likewise to that induced
by Vincristine: the microtubule network appeared to be ‘rarefied’
and fragmented.6
Collectively, our results clearly indicate that the tested pyrrolo-
azaflavone derivatives bearing the phenyl group at the 2-position,
at non-cytotoxic concentrations and like vinblastine, are able (i) to
exert in vitro anti-angiogenic activity and (ii) to counteract in vitro
and in vivo the pro-angiogenic effects of FGF-2, a well-known cyto-
kine involved in tumor growth. In view of these preliminary re-
sults, we are planning further in vivo and in vitro experiments to
evaluate whether the here described pyrroloazaflavones could
effectively inhibit tumor formation and deeply identify their spe-
cific mechanism and molecular targets.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.11.010.
References and notes
1. Donate, F. Drugs Future 2005, 30, 695.
2. Siemann, D. W.; Bibby, M. C.; Dark, G. G.; Dicker, A. P.; Esken, F. A. L. M.;
Horsman, M. R.; Maermè, D.; LoRusso, P. M. Clin. Cancer Res. 2005, 11, 416.
3. Lippert, J. W. Bioorg. Med. Chem. 2007, 15, 605.
4. Mabeta, P.; Pepper, M. S. Angiogenesis 2009, 12, 81.
5. Kanthou, C.; Tozer, G. M. Int. J. Exp. Pathol. 2009, 90, 284.
6. Ferlin, M. G.; Chiarelotto, G.; Gasparotto, V.; Barzon, L.; Palu’, G.; Castagliuolo, I.
J. Med. Chem. 2005, 48, 3417.
7. Gasparotto, V.; Castagliuolo, I.; Chiarelotto, G.; Pezzi, V.; Montanaro, G.; Brun,
P.; Palu, G.; Ferlin, M. G. J. Med. Chem. 2006, 49, 1910.
8. Gasparotto, V.; Castagliuolo, I.; Ferlin, M. G. J. Med. Chem. 2007, 50, 5509.
9. Bergman, J.; Sand, P. Org. Synth. 1987, 65, 146.
10. Bergman, J.; Sand, P.; Tilstam, U. Tetrahedron Lett. 1983, 24, 3665.
11. Vacca, A.; Lurlalo, M.; Ribatti, D.; Minischetti, M.; Nico, B.; Ria, R.; Pellegrino, A.;
Dammacco, F. Blood 1999, 94, 4143.
12. Folkman J. Cancer Biol. 2003, 13, 159.
13. Ribatti, D.; Marimpietri, D.; Pastorino, F.; Brignole, C.; Nico, B.; Vacca, A.;
Ponzoni, M. Ann. N.Y. Acad. Sci. 2004, 1028, 133.
14. Schwartz, E. L. Clin. Cancer Res. 2009, 15, 2594.
15. Belotti, D.; Vergani, V.; Drudis, T.; Borsotti, P.; Pitelli, R.; Viale, G.; Giavazzi, R.;
Taraboletti, G. Clin. Cancer Res. 1996, 2, 1843.
Our data suggest that the impairment of cytoskeleton fibers
could also act together with other mechanism, such as an inhibi-
tory effect on pro-angiogenic factors produced by both endothelial
and tumor cells. Indeed, 2-PPyQs inhibited the stimulatory effects
induced by FGF-2 on both in vitro cell proliferation and morpho-
genesis, and in vivo in Matrigel plugs.
16. Kanthou, C.; Tozer, G. M. Blood 2002, 99, 2060.
17. Kim, T. J.; Ravoori, M.; Landen, C. N.; Kamat, A. A.; Han, L. Y.; Lu, C.; Lin, Y. G.;
Merrit, W. M.; Jennings, N.; Spannuth, W. A.; Langley, R.; Gershenson, D. M.;
Coleman, R. L.; Kundra, V.; Sood, A. K. Cancer Res. 2007, 67, 9337.
18. Klauber, N.; Parangi, S.; Flynn, E.; Hamel, E.; D’Amato, R. J. Cancer Res. 1997, 57,
81.
19. Malcontenti-Wilson, C.; Chan, L.; Nikfarian, M.; Muralidharan, V.; Christophy,
C. J. Gastroenterol. Hepatol. 2008, 23, c96.
20. Micheletti, G.; Poli, M.; Borsotti, P.; Martinelli, M.; Imberti, B.; Taraboletti, G.;
Giavazzi, R. Cancer Res. 2003, 63, 1534.
Therefore, overall these preliminary studies suggest that 2-
PPyQs, although with different profiles, possess anti-angiogenic
properties in vitro a well as in vivo. Since these properties were
noted under experimental conditions that did not affect EC prolif-
eration, 2-PPyQs might conceivably affect tumor-induced angio-
genesis in vivo at local concentration lower than those necessary
to cause a cytotoxic effect on tumor cells. Such performance was