This article appeared in a journal published by Elsevier.

The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Biomaterials 31 (2010) 7863e7872

Contents lists available at ScienceDirect

Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials

Anti-angiogenic activity of heparin-like polysulfonated polymeric drugs in 3D

human cell culture
Luis García-Fernández a, b, *,1, Sven Halstenberg c,1, Ronald E. Unger c, María R. Aguilar a, b,
C. James Kirkpatrick c, Julio San Román a, b
a
Department of Biomaterials, Institute of Polymer Science and Technology, Juan de la Cierva 3, 28006 Madrid, Spain
b
Networking Biomedical Research Center in Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN), Spain
c
Institute of Pathology, Universitätsmedizin, Langenbeckstrasse 1, Mainz, Germany

a r t i c l e i n f o a b s t r a c t

Article history: The activity of new anti-angiogenic polymeric drugs was tested in a 3D endothelial cell culture system
Received 9 June 2010 applied as a model of angiogenesis. The assay was performed in a highly reproducible fibrin matrix that
Accepted 4 July 2010 supported endothelial cell attachment, proliferation, migration, and formation of capillary-like struc-
Available online 31 July 2010
tures. Active growth factors (FGF and/or VEGF) were added to the medium to induce the formation of
blood vessel-like structures, and the effect of the active polymers was then tested by a semi-quantitative
Keywords:
immunostaining protocol and visualized by laser-scanning confocal microscopy.
Angiogenesis
The synthetic heparin-like macromolecules that were tested for their anti-angiogenic efficacy were
Confocal microscopy
ECM (extracellular matrix)
previously characterized in terms of their anti-proliferative activity in 2D tissue culture. Two different
Endothelial cell anti-angiogenic monomers, a methacrylic derivative of 5-amino-2-naphthalenesulfonic acid (MANSA)
Fibrinogen and 2-acrylamido-2-methylpropane sulfonic acid (AMPS), were copolymerized with a hydrophilic
Fibroblast growth factor monomer (vinyl pyrrolidone, VP) or a hydrophobic monomer (butyl acrylate, BA), giving rise to different
copolymeric systems with controlled microstructure and supramolecular organization.
Both copolymeric systems have demonstrated a composition-dependent anti-mitogenic effect in “2D
in vitro” cell culture experiments using aFGF as pro-angiogenic growth factor and BALB/c 3T3 fibroblast as
cell model, as was shown in a previous publication [1,2]. These 3D experiments provide evidence for the
strong and specific modulation of angiogenesis by these systems.
The 3D experiments constitute an improvement over 2D in vitro experiments and in vivo experiments
with angiogenic drugs and may help to reduce the number of animal experiments.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction rise to functional capillaries. The time for vascular remodeling in

Upregulated angiogenic processes play an important role in the
development of various pathological processes such as psoriasis [3],
rheumatoid arthritis [4], tumor growth [5], diabetic retinopathy [6],
and others. Uncontrolled angiogenesis and neoformation of blood
vessels contribute to the progression of these diseases, especially in
the case of solid malignant tumors. When tumor dimensions exceed
1e2 mm3 tumor cells release pro-angiogenic factors to the
surrounding tissues. The endothelial cells (ECs) forming the blood
vessels that supply the surrounding tissues then switch to an angio-
genic phenotype, proliferate, migrate, and differentiate, thus giving
* Corresponding author. Department of Biomaterials, Institute of Polymer Science
and Technology, Juan de la Cierva 3, 28006 Madrid, Spain. Tel.: þ34 91 2587407;
fax: þ34 91 5719341.
E-mail address: luisgar@ictp.csic.es (L. García-Fernández).
1
These authors contributed equally to this work.
normal tissues is 7e10 years [7], while tumor endothelium prolifer-
ates 40-fold faster [8]. Tumor capillaries are normally defective and
leaky, so that small as well as large molecules have access to the
malignant tissue. Additionally, malignant cells have easier access to
the blood stream and can thus travel to other healthy organs to
produce metastasis with little impediment. In fact, Folkman et al.
demonstrated in 1971 that tumor growth and metastases were
angiogenesis-dependent [9], and also showed that tumor malignancy
and perfusion density were very closely related. As a result, inhibition
of angiogenesis (anti-angiogenic therapy) has become a very attractive
therapeutic target to inhibit tumor growth, decrease tumor mass,
induce tumor regression, and reduce the risk of metastasis [10,11].
Angiogenesis is triggered when the balance between pro-angiogenic
and anti-angiogenic signals shifts to the former. Two of the most
important families of pro-angiogenic proteins are the vascular
endothelial growth factors (VEGFs) and the fibroblast growth factors
(FGFs), both of which induce EC proliferation and migration, initiating

0142-9612/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biomaterials.2010.07.022
 Author's personal copy
7864 L. García-Fernández et al. / Biomaterials 31 (2010) 7863e7872
neo-vessel sprouting [12]. Both growth factor families are active
through their interaction with specific transmembrane signaling
receptors with intrinsic tyrosine kinase activity (VEGFRs and FGFRs,
respectively) that are responsible for initiating the biological
response. Most of the VEGF and FGF isoforms require a family of cell
surface heparin sulfate proteoglycans (HSPG) as co-receptors for
efficient signal transduction. Therefore, inhibition of the interaction of
VEGFs or FGFs with these cell surface HSPG seems an attractive
strategy to inhibit angiogenesis.
The study of angiogenesis requires the understanding of the
different steps involved in the process (proliferation, migration, and
differentiation), and numerous in vitro assays have been developed
to study the different stages of the angiogenic cascade.
Proliferation has been studied by culturing the cells in the presence
of the required growth factors and/or the test drugs, being the prolif-
eration determined by cell counts [13] or genomic DNA quantification.
Migration is normally studied by a blind-well chemotaxis
chamber. In this assay the cells are placed on the upper layer of
a filter that permits the migration of the cells in response to a test
factor placed in the medium [14].
Differentiation is frequently studied using cocultures of ECs with
fibroblasts or other relevant cell types. The formation of capillary-
like structures consisting of differentiated ECs in these cocultures
can be evaluated using specific immunohistochemical stains [14].
These assays are normally carried out on tissue culture plastic
(TCP) and the conditions differ considerably from the physiological
situation. Therefore, researchers have long tried to create defined
3D tissue-equivalents or biomimetic environments in order to
reproduce as closely as possible the in vivo conditions under which
ECs form highly branched capillary-like tubules [15,16]. The
replacement of 2D by 3D methodologies has been limited by
substrate availability, because the 3D materials should mimic the
extracellular matrix (ECM) characteristics that promote cell
proliferation, migration, and differentiation [17]. At present,
different kinds of natural and synthetic ECM scaffolds are available
for this application: Matrigel, collagen, PureColÔ, fibrin, and others.
Matrigel is a gelatinous extract from Engelbreth-Holm-Swarm
tumor cells that contains proteins, glycoproteins and growth
factors. This matrix has been extensively used for the last 15 years
[16]. It supports cell proliferation, migration and differentiation, but
it does not have a well-defined growth factor composition, and the
response of ECs can vary greatly among different batches of
Matrigel. In fact, not all commercial preparations of Matrigel
promote tube formation in vitro [14].
PureColÔ (99.9% type-I collagen) and PuraMatrixÔ (self-
assembled synthetic peptide) have also been used for this purpose.
However, these materials have long gelation times or require
extensive handling time, making their application difficult in 3D
experiments [18].
Fibrin has also been used in the preparation of 3D supports for
angiogenesis evaluation, because it provides the most relevant
elements of the neovascularization process under simplified
conditions [19,20]. Actually, a fibrin matrix is often found around
tumor-associated blood vessels [21], as it is used by ECs as
a provisional matrix that facilitates capillary sprouting. It is a very
Fig. 1. Synthesis of MANSA.
cheap and easily available product that can be combined with
different growth factors to provide the adequate environment for
ECs to mimic the angiogenesis process. It has been demonstrated
that when cultured in these 3D matrices, ECs respond to VEGF and
basic FGF (bFGF) by forming capillary-like structures [22]. Lafleur
et al. also observed a synergistic effect of these two growth factors
when added in combination [23].
In this work we used a fibrin scaffold supplemented with VEGF,
bFGF, or a combination of both to induce ECs attachment, migra-
tion, and differentiation. The effect of new families of heparin-like
polysulfonated polymeric drugs on human umbilical vein endo-
thelial cells (HUVEC) cultured in 3D fibrin scaffolds was evaluated.
Immunostaining and laser-scanning confocal microscopy were
used to visualize and quantify tubule formation, actin orientation
and the intercellular junctions in the 3D fibrin scaffolds.
The detailed synthesis of these macromolecular bioactive
systems and their anti-mitogenic activity in 2D cell culture assays
has been reported recently [1,2]. The polymers are based on
2-acrylamido-2-methylpropane sulfonic acid (AMPS) and a meth-
acrylic derivative of 5-amino-2-naphthalenesulfonic acid (MANSA).
AMPS was chosen because poly(2-acrylamido-2-methylpropane
sulfonic acid) (PAMPS) has been reported to be one of the most
potent inhibitors of angiogenesis while presenting low toxic effects
[24]. 5-amino-2-naphthalenesulfonic acid (ANSA) has been
described as a potent anti-mitogenic molecule that possesses high
activity and rather low toxicity [13].
2. Materials and methods
2.1. Polymer drugs
The detailed synthesis of both families of polymer drugs together with the in
vitro test are described in previous publications [1,2]. ANSA was modified to include
a reactive double bond in its structure, for the purpose of obtaining a pharmaco-
logically active polymer (polymeric drug) after polymerization (Fig. 1).
MANSA and AMPS were copolymerized separately with a hydrophilic monomer,
N-vinylpirrolidone (VP), or a hydrophobic monomer, butyl acrylate (BA), giving rise
to different copolymeric systems with different microstructure and supramolecular
organization. Chemical structures of the different copolymer systems and the
synthetic route of preparation are shown in the scheme of Fig. 2.
2.2. Cell culture
HUVEC were isolated from umbilical cords and cultivated according to previ-
ously published methods [25] involving passages in gelatin-coated tissue culture
flasks and the use of medium M199 (SigmaeAldrich) supplemented with 20% fetal
bovine serum (FBS, Life Technologies), 1% penicillin/streptomycin, 1% Glutamax I
(Life Technologies), 25 mg/ml sodium heparin (SigmaeAldrich), and 25 mg/ml
endothelial cell growth supplement (Becton Dickinson).
2.3. Preparation of three-dimensional fibrin gels seeded with HUVEC
Fibrinogen from human serum (w50% protein basis, Fluka) was dissolved in
sterile PBS (Gibco) at a concentration of 6 mg/ml 450 mL of cell suspension
(250,000 cells/ml) in M199 without serum (SigmaeAldrich) were mixed with 500 mL
of fibrinogen solution. Aprotinin (SigmaeAldrich) was added to a final concentration
of 200 U/ml to prevent fibrinolysis by soluble proteases. The resulting suspension of
cells was transferred to a 48-well plate (475 mL/well) and clotting was induced by
addition of 25 mL of thrombin (20 U/ml; SigmaeAldrich). Gelation took place after
approximately 5 min at 37  C. Finally, 500 mL/well of fresh culture medium M199
was added, followed by appropriate amounts of vascular endothelial growth factor
 Author's personal copy

L. García-Fernández et al. / Biomaterials 31 (2010) 7863e7872 7865

Fig. 2. Synthesis of the different copolymeric systems.

Table 1
Growth factor and polymer concentrations implemented.

VEGF bFGF Polymer Table 2

50 ng/ml e 1 mg/ml Composition of the different copolymers used in 3D assays.
e 10 ng/ml 1 mg/ml
50 ng/ml 10 ng/ml 1 mg/ml AMPS-based copolymers MANSA-based copolymers
50 ng/ml e 0.5 mg/ml Poly(BA-co-AMPS) (23% AMPS) Poly(VP-co-MANSA) (27% MANSA)
e 10 ng/ml 0.5 mg/ml Poly(BA-co-AMPS) (50% AMPS) Poly(VP-co-MANSA) (15% MANSA)
50 ng/ml 10 ng/ml 0.5 mg/ml Poly(BA-co-AMPS) (73% AMPS) Poly(VP-co-MANSA) (10% MANSA)
50 ng/ml e e Poly(BA-co-MANSA) (20% MANSA)
e 10 ng/ml e Poly(BA-co-MANSA) (18% MANSA)
50 ng/ml 10 ng/ml e Poly(BA-co-MANSA) (11% MANSA)
 Author's personal copy

7866 L. García-Fernández et al. / Biomaterials 31 (2010) 7863e7872

Fig. 3. Inverted microscopy image of HUVEC in a fibrin gel at different time points (A: 0 day, B: 2 days, C: 5 days and D: 7 days). All gels are in the presence of exogenous VEGF and
bFGF.

(VEGF, SigmaeAldrich), basic fibroblast growth factor (bFGF, SigmaeAldrich) and
the polymer under study (see Table 1).
2.4. Visualization and quantification of capillary-like structure formation and
evaluation of cell proliferation
in each sample. Image analysis was carried out using free software, Image J 1.42q,
available from the National Institute of Health’s website (http://rsb.info.nih.gov/ij/).
The analysis was made using the plugging: analyze-measure RGB, so that we
obtained the percentage of nuclei or actin in any given image.

After seven days in culture, the cells inside the fibrin scaffolds were fixed by
immersing each matrix in a solution of 3.5% paraformaldehyde in PBS. Samples were
washed with PBS and then treated with PBS-buffered 0.1% Triton X-100 to per-
meabilize cellular membranes and with PBS-buffered 1% bovine serum albumin in
order to block unspecific antibody binding to the cells and fibrin. The samples were
then incubated at 37  C with the primary antibody: mouse anti-human platelet
endothelial cell adhesion molecule (PECAM e CD31; DAKO). Following three
washing steps with PBS for 20 min each, a second incubation was performed at 37  C
with the secondary antibody, goat anti-mouse Alexa Fluor 488, to visualize EC
differentiation and zones of intercellular adhesion in green. The nuclei were coun-
terstained in blue with 1 mg/ml Hoechst 33342 (SIGMA) and the actin cytoskeleton
was visualized in red with phalloidin-TRITC (SIGMA). Scaffold specimens were then
mounted on glass slides with Fluoromount-G mounting medium (SouthernBiotech),
and the Leica True Confocal Scanner system (Leica TCS SP2) was used to conduct
laser-scanning confocal microscopy and to record all images. For quantitative
analysis, five randomly selected fields of view with a height of 50 mm were recorded
3. Results and discussion
3.1. Polymer drugs
The chemical structures of copolymer chains corresponding to
poly(BA-co-MANSA), poly(VP-co-MANSA), poly(BA-co-AMPS) and
poly(VP-co-AMPS) are described in Fig. 2. All the copolymers were
prepared by a free radical polymerization reaction in solution. The
microstructure (i.e. the distribution of comonomeric units) is rather
different according to the reactivity of each one of the monomer
species. VP-based copolymers are highly hydrophilic and homo-
geneous whereas BA-based copolymers are amphiphilic and show
a high tendency to form micellar morphologies [2].

Fig. 4. Composite confocal microscopy images of a 50 mm section of a 3D fibrin scaffold containing HUVEC cultured in the presence of VEGF and bFGF, and without addition of anti-
angiogenic copolymers. HUVEC proliferated, migrated, and formed capillary-like structures. (a) Merged image: Nuclei in blue, actin in red and CD31 in green. (b) Red channel. (c)
Green channel.
 Author's personal copy
L. García-Fernández et al. / Biomaterials 31 (2010) 7863e7872
Fig. 5. Confocal microscopy images of tubular-like structures formed in a 3D assay without anti-angiogenic copolymers.
The activity on the inhibition of aFGF-induced mitogenesis on
fibroblast BALB/c 3T3 was also tested in previous works [1,2], and
only three families show activity in this test, and these families
were used in 3D test (Table 2).
3.2. Capillary-like structure formation
HUVEC were used in all experiments. Although EC sprouting
occurs primarily from microvasculature in vivo, macrovascular EC
can also attain an angiogenic phenotype if given the appropriate
growth conditions. In fact, macrovascular EC (HUVEC) and micro-
vascular EC (HDMEC) behaved similarly in terms of capillary
formation in fibrin gels as demonstrated by Lefleur et al. [23].
In the present study fibrin gels, in which the HUVEC were
homogeneously suspended (Fig. 1 e 0th day), were supplemented
with two of the most important growth factors involved in angio-
genesis, VEGF (50 ng/ml) and bFGF (10 ng/ml). Under these
conditions the angiogenic phenotype of the cultured ECs was
triggered, leading to cell migration, proliferation, and formation of
capillary-like structures, as demonstrated by the inverted micro-
scope images shown in Fig. 3.
Further study of these structures was carried out using laser-
scanning confocal microscopy (Fig. 4). In these images the nuclei of
the HUVEC were stained blue and were used to determine the
number of cells in the sample (Fig. 4a). The actin cytoskeleton,
which is revealed in red, was aligned circumferentially within
capillary-like structures, as can be observed in Fig. 4b. The specific
transmembrane glycoprotein CD31, also known as platelet endo-
thelial cell adhesion molecule (PECAM-1), was visualized by
immunostaining in green (Fig. 4c). This adhesion molecule is
expressed in large amounts on endothelial cells at intercellular
Fig. 6. Confocal microscopy image (red ¼ actin, blue ¼ nuclei, green ¼ PECAM-CD31) and histological sections in the red channel (actin) showing an apparently continuous lumen
inside the tubule formed by HUVEC (indicated by arrows).
7867
junctions. In our cultures it was clearly associated with the tubular
structures (vessel-like) that were observed.
As seen in Figs. 4 and 5, HUVEC assembled to form tubular-like
structures. To attain such structures the cells had to migrate and
interact in complex ways to form a linear and hollow structure. All
these structures exhibited the green color (representing PECAM-
CD31) that exemplified the complex interaction of cells to form
tubular-like structures.
Fig. 6 shows the actin cytoskeleton in individual confocal
microscopy sections of a single tube-like structure. In the cross
sections of this capillary-like structure, a continuous lumen can be
observed inside the aforesaid capillary-like structure.
3.3. Inhibition of bFGF and/or VEGF-induced cell proliferation
In previous studies we demonstrated by means of 2D cell culture
experiments [1,2] that ANSA-based copolymers (poly(VP-co-
MANSA) and poly(BA-co-MANSA)) as well as poly(BA-co-AMPS)
copolymers, inhibited fibroblast mitogenesis induced by aFGF. To
investigate the effect of the synthetic polymers on EC behaviour,
the polysulfonated heparin-like copolymers were added at
different concentrations (0.5 mg/ml and 1 mg/ml). All the assayed
copolymers were water soluble and were added to the medium
before fibrinogen clotting. This ensured homogeneous distribution
of the active compounds throughout the 3D matrices.
3.3.1. AMPS-based copolymers
Liekens et al. [24] described a modelling study that predicted
the interaction between bFGF and PAMPS. During polymerization
AMPS develops an asymmetric centre at the middle carbon of each
acrylamido unit so that a stable intramolecular hydrogen bond
 Author's personal copy

7868 L. García-Fernández et al. / Biomaterials 31 (2010) 7863e7872

Fig. 7. Confocal microscopy of HUVEC in fibrin scaffolds with the addition of different concentrations of Poly(BA-co-AMPS) (50% AMPS), (a) 0 mg/ml (b) 0.5 mg/ml (c) 1 mg/ml in
presence of bFGF and VEGF.

Fig. 8. Inhibition of cell proliferation induced by bFGF (white), VEGF (gray) and bFGF-VEGF (black). The diagrams include the statistical analysis (ANOVA) for each copolymer with
respect to bFGF-VEGF (x) at a significance level of p < 0.05, VEGF (¤) at a significance level of p < 0.05 and bFGF (z) at a significance level of p < 0.05.

could be formed between acrylamido units with equivalent abso-
lute conformation. In this situation a helical structure with sulfo-
nate groups evenly spaced around 9 to 10 Å would be formed. The
authors stated that helical PAMPS might dock into the heparin-
binding site of bFGF inhibiting both the binding with low affinity of
HSPGs and the binding with high affinity of tyrosine kinase FGFRs.
For this reason we correlated the anti-proliferative activity of the
AMPS-based copolymers with their microstructure. Actually,
the higher probability of finding at least two AMPS units in the
appropriate conformation was related to the higher biological
activity of the copolymer. The poly(BA-co-AMPS) system presented
very different reactivity ratios (rBA ¼ 3.68; rAMPS ¼ 0.30) and formed
macromolecules rich in one or the other monomer. These copoly-
mers were active and inhibited fibroblast proliferation in a dose-
dependent manner. However, the VP-AMPS system presented
reactivity ratios lower than unity for both monomers (rVP ¼ 0.12;
rAMPS ¼ 0.27; azeotropic system), indicating a tendency toward
alternating polymerization and a low probability to find AMPS-rich

Fig. 9. Supramolecular organization in MANSA-based copolymers.

 Author's personal copy

L. García-Fernández et al. / Biomaterials 31 (2010) 7863e7872 7869

Fig. 10. Laser-scanning images (scale bar ¼ 300 mm) showing the effect of polymer concentration. (a) Poly(VP-co-MANSA) (23% MANSA). (b) Poly(BA-co-MANSA) (20% MANSA). (1)
Control. (2) 0.5 mg/ml (3) 1 mg/ml.

sequences. These macromolecules did not inhibit fibroblast prolif-
eration and therefore were not studied in 3D experiments [1].
The laser-scanning confocal images in Fig. 7 shows the stained
nuclei in the 3D assay in the presence of different concentrations of
polymer.
Fig. 8 shows the following quantified decrease in cell number
per picture frame as a function of polymer concentration:
All the polymers showed a significant inhibition of cell prolif-
eration (Figs. 7 and 8) when both growth factors were added.
However, the inhibition was not significant when only one growth
factor was added, except in the case of the polymer with 73% AMPS.
The polymer with 23% AMPS did not inhibit the VEGF activity, but
significantly inhibited the activity of bFGF at high concentration
(1 mg/ml).
3.3.2. MANSA-based polymers
The anti-proliferative activity of the MANSA-based copolymers
is related to their supramolecular organization (Fig. 9). The poly
(BA-co-MANSA) system (very different reactivity ratios, and with
constituent monomers of very different hydrophilicity) formed
self-assembled micellar nanoparticles with a particular distribution
of the MANSA groups on the surface [1]. These copolymers were the
most active. Poly(VP-co-MANSA) formed water-soluble homoge-
neous copolymers whose activity depended on their MANSA
content.
These two families of copolymers inhibited EC proliferation in
a dose-dependent manner as demonstrated by a series of laser-
scanning confocal microscopy images (Fig. 10) and the corre-
sponding quantitative results (Figs. 11 and 12).

Fig. 11. Inhibition of cell proliferation induced by bFGF (white), VEGF (gray) and bFGF-VEGF (black). The diagrams include the statistical analysis (ANOVA) for each copolymer with
respect to bFGF-VEGF (x) at a significance level of p < 0.05, VEGF (¤) at a significance level of p < 0.05 and bFGF (z) at a significance level of p < 0.05.
 Author's personal copy

7870 L. García-Fernández et al. / Biomaterials 31 (2010) 7863e7872

Fig. 12. Inhibition of cell proliferation induced by bFGF (white), VEGF (gray) and bFGF-VEGF (black). The diagrams include the statistical analysis (ANOVA) for each copolymer with
respect to bFGF-VEGF (x) at a significance level of p < 0.05, VEGF (¤) at a significance level of p < 0.05 and bFGF (z) at a significance level of p < 0.05.

Fig. 13. Confocal microscopy images of HUVEC in 3D bFGF-VEGF assays with poly(BA-co-AMPS) (50% AMPS). (a) Control. (b) With 1 mg/ml of polymer. (c) Control at high
magnification. (d) High magnification of culture with 1 mg/ml of polymer.
 Author's personal copy

L. García-Fernández et al. / Biomaterials 31 (2010) 7863e7872 7871

Fig. 14. Actin cytoskeleton of 3D HUVEC networks at different concentrations of anti-angiogenic polymers visualized by confocal microscopy (space bars ¼ 300 mm). (a) Poly(VP-co-
MANSA) (23% MANSA). (b) Poly(BA-co-MANSA) (20% MANSA). (1) Without polymer. (2) 0.5 mg/ml (3) 1 mg/ml.

In proportion to the overall MANSA content, all the polymers
studied demonstrated significant inhibition of cell proliferation
when both growth factors were added to the culture medium
(Fig. 11). The copolymers with the highest concentration of
MANSA (27% and 15%) also inhibited the synergistic action of both
growth factors.
Figs. 11 and 12 confirmed our previous results and demon-
strated that poly(BA-co-MANSA) copolymers are the most active
and that the higher the amount of MANSA in the copolymer, the
higher the anti-angiogenic effect. All copolymers assayed not
only inhibited the activity of the growth factors individually, but
also were even more effective when present in the culture
medium simultaneously, amounting to a powerful, synergistic
anti-angiogenic effect.
3.4. Cell distribution in 3D assay
In the 3D in vitro experiments not only cell proliferation was
studied, but cell migration and cell association in organized capil-
lary-like structures were also evaluated.
Fig. 13a and b clearly illustrate the effect of AMPS-based
copolymers on cell proliferation and cell migration. The control cell
culture (Fig. 13a) showed that cell migration, spreading and align-
ment of the actin cytoskeleton gave rise to a highly branched and
interconnected network of capillary-like structures, with lengths
between branches exceeding 300 mm when no polymer was added
to the culture medium. Fig. 13d shows clearly that in the presence of
the active polymer drug only disordered cellular domains are
formed, but without any capillary-like structures.

Fig. 15. Detail at high magnification. Formation of tubular capillary-like structures and their inhibition in the presence of MANSA-based polymers. (a) Control without polymer. (b)
1 mg/ml of poly(VP-co-MANSA) (23% MANSA). (c) 1 mg/ml of poly(BA-co-MANSA) (20% MANSA).
 Author's personal copy
7872 L. García-Fernández et al. / Biomaterials 31 (2010) 7863e7872
Similar suppression of 3D angiogenesis occurred in the presence
of MANSA-based copolymers (Fig. 14).
Fig. 14 shows the actin distribution inside the fibrin matrices in
the absence and presence of the MANSA-based copolymers (all
with exogenous bFGF and VEGF in the culture medium). In absence
of MANSA-based copolymers (Fig. 14a1 and 14b1), the HUVEC
spread and aligned, giving rise to a highly branched network of
capillary-like structures. When the active polymer was added to the
medium, the organization of the cells was inhibited. Poly(VP-co-
MANSA) copolymers were less suppressive of angiogenesis in this
model system than Poly(BA-co-MANSA) copolymers as can be seen
when comparing the effect of 0.5 mg/ml of polymer in the medium:
With 0.5 mg/ml of Poly(VP-co-MANSA), the tubular structures
observed were more abundant and interconnected (Fig. 14a2) than
in cultures at similar concentration of Poly(BA-co-MANSA)
(Fig. 14b2).
Capillary-like structure formation hardly occurred at all when
poly(VP-co-MANSA) copolymers were added at a concentration of
1 mg/ml (Fig. 14a3) and when poly(BA-co-MANSA) copolymers
were added at 0.5 mg/ml.
Fig. 15 shows this effect at higher magnification. In the
absence of MANSA-based copolymers HUVEC formed capillary-
like structures and the transmembrane glycoprotein PECAM-1
was clearly expressed (seen in green), while in the presence of
the anti-angiogenic copolymer cells did not assemble or inter-
connect, did not express PECAM-1, but continued to exhibit
a cellular morphology consistent with general viability. Due to
the compatibility with physiological fluids and lack of toxicity of
these polyacrylic systems, they are promising polymer drugs for
the modulation and inhibition of proliferative and angiogenic
processes. A special indication still to be studied is the inhibition
or modulation of metastatic processes in malignant tumors, as
well as in situations that require a control of new vessel forma-
tion process.
4. Conclusions
This study documents the usefulness of our 3D angiogenesis
model system consisting of HUVEC in a fibrin matrix and presents
strong evidence for the very low toxicity and high anti-angiogenic
activity of the new polymer drugs based on polyacrylic derivatives
carrying sulfonic acid groups. The activity of these non-toxic
polymers systems depends not only on the chemical structure of
the polyacrylic macromolecular chains, but also on their supra-
molecular structure, which is determined by the amphiphilic
character of the macromolecular segments and their microstruc-
tural arrangement.
The modulation of angiogenic processes seems to be a result of
the direct interaction of the sulfonic acid groups with the specific
growth factors (bFGF and VEGF) in a similar way to heparin, which
modulates the effect of these growth factors in the processes of
endothelial cell migration and proliferation.
Acknowledgements
The authors would like to thank Mrs. A. Sartoris for her excellent
technical assistance. Partial financial support from MAT2007-
63355, CIBER-BBN and the EU Network of Excellence EXPERTIS-
SUES is acknowledged gratefully.
Appendix
Figures with essential color discrimination. Figs. 4e7, 9, 10,
13e15 in this article are difficult to interpret in black and white.
The full color images can be found in the online version, at doi:10.
1016/j.biomaterials.2010.07.022.
References
[1] García-Fernandez L, Aguilar MR, Fernandez MM, Lozano RM, Gimenez G, San
Roman J. Antimitogenic polymer drugs based on AMPS: monomer dis-
tributionbioactivity relationship of water-soluble macromolecules. Bio-
macromolecules 2010;11(3):626e34.
[2] García-Fernandez L, Aguilar MR, Fernandez MM, Lozano RM, Giménez G,
Valverde S, et al. Structure-morphology-bioactivity of biocompatible systems
derived from functionalized acrylic polymers, based on 5-amino-2-naph-
thalenesulfonic acid. Biomacromolecules 2010;11(7):1763e72.
[3] Arbiser JL. Angiogenesis and the skin: a primer. J Am Acad Dermatol 1996;34
(3):486e97.
[4] Paleolog EM, Fava RA. Angiogenesis in rheumatoid arthritis: implications for
future therapeutic strategies. Springer Semin Immunopathol 1998;20
(1e2):73e94.
[5] Folkman J. What is the evidence that tumors are angiogenesis dependent? J
Natl Cancer Inst 1990;82(1):4e6.
[6] Xu JH, Li RX, Zhang W, Qi F. Research progress of the mechanism on intra-
ocular neovascularization. Ophthalmologica 2008;8:2496e8.
[7] Engerman RL, Pfaffenbach D, Davis MD. Cell turnover of capillaries. Lab Invest
1967;17(6):738e43.
[8] Denekamp J, Hobson B. Endothelial-cell proliferation in experimental
tumours. Br J Cancer 1982;46(5):711e20.
[9] Folkman J. Tumor angiogenesis: therapeutic implications. N Engl J Med
1971;285(21):1182e6.
[10] Conconi MT, Nico B, Guidolin D, Baiguera S, Spinazzi R, Rebuffat P, et al.
Ghrelin inhibits FGF-2-mediated angiogenesis in vitro and in vivo. Peptides
2004;25(12):2179e85.
[11] Shimizu K, ́Oku N. Cancer anti-angiogenic
́ therapy. Biol Pharm Bulĺ 2004;27
́
(5):599e605.
[12] Carmeliet P. Mechanisms of angiogenesis and arteriogenesis. Nat Med 2000;6
(4):389e95.
́
[13] Fernández-Tornero ́
C, Lozano RM, Redondo-Horcajo M, Gómez AM, López JC,
Quesada E, et al. Leads for development of new naphthalenesulfonate deriv-
atives with enhanced antiangiogenic activity. Crystal structure of acidic
fibroblast growth factor in complex with 5-amino-2-naphthalenesulfonate. J
Biol Chem 2003;278(24):21774e81.
[14] Auerbach R, Lewis R, Shinners B, Kubai L, Akhtar N. Angiogenesis assays:
a critical overview. Clin Chem 2003;49(1):32e40.
[15] Donovan D, Brown NJ, Bishop ET, Lewis CE. Comparison of three in vitro
human ‘angiogenesis’ assays with capillaries formed in vivo. Angiogenesis
2001;4(2):113e21.
[16] Lawley TJ, Kubota Y. Induction of morphologic differentiation of endothelial
cells in culture. J Invest Dermatol 1989;93(2):59Se61S.
[17] Lutolf MP, Hubbell JA. Synthetic biomaterials as instructive extracellular
microenvironments for morphogenesis in tissue engineering. Nat Biotechnol
2005;23:47e55.
[18] Serban MA, Liu Y, Prestwich GD. Effects of extracellular matrix analogues on
primary human fibroblast behavior. Acta Biomater 2008;4(1):67e75.
[19] Montesano R, Pepper MS, Vassalli JD, Orci L. Phorbol ester induces cultured
endothelial cells to invade a fibrin matrix in the presence of fibrinolytic
inhibitors. J Cell Physiol 1987;132(3):509e16.
[20] Nehls V, Drenckhahn D. A novel, microcarrier-based in vitro assay for rapid
and reliable quantification of three-dimensional cell migration and angio-
genesis. Microvasc Res 1995;50(3):311e22.
[21] Nagy JA, Brown LF, Senger DR, Lanir N, Van De Water L, Dvorak AM, et al.
Pathogenesis of tumor stroma generation: a critical role for leaky blood
vessels and fibrin deposition. Biochim Biophys Acta 1988;948(3):305e26.
[22] Anand-Apte B, Pepper MS, Voest E, Montesano R, Olsen BR, Murphy G, et al.
Inhibition of angiogenesis by tissue inhibitors of metalloproteinase-3. Invest
Ophthalmol Vis Sci 1997;38(5):817e23.
[23] Lefleur MA, Handsley MM, Knauper V, Murphy G, Edwards DR. Endothelial
tubulogenesis within fibrin gels specifically requires the activity of membrane-
type-matrix metalloproteinases (MT-MMPs). J Cell Sci 2002;115:3427e38.
[24] Liekens S, Neyts J, Dégrève B, De Clercq E. The sulfonic acid polymers PAMPS
[poly(2-acrylamido-2-methyl-1-propanesuifonie acid)] and related analogues
are highly potent inhibitors of angiogenesis. Oncol Res 1997;9:173e81.
[25] Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human endothelial
cells derived from umbilical veins. Identification by morphologic and immu-
nologic criteria. J Clin Invest 1973;52(11):2745e56.