Integrin-targeted nano-sized polymeric systems for paclitaxel conjugation: a comparative study

Article abstract of DOI:10.1080/1061186X.2017.1358727

The generation of rationally designed polymer therapeutics via the conjugation of low molecular weight anti-cancer drugs to water-soluble polymeric nanocarriers aims to improve the therapeutic index. Here, we focus on applying polymer therapeutics to targ

Full text of DOI:10.1080/1061186X.2017.1358727

Journal of Drug Targeting
ISSN: 1061-186X (Print) 1029-2330 (Online) Journal homepage: http://www.tandfonline.com/loi/idrt20
Integrin-Targeted Nano-Sized Polymeric Systems
for Paclitaxel Conjugation: A Comparative Study
Anat Eldar-Boock, Rachel Blau, Claudia Ryppa, Hemda Baabur-Cohen, Ariel
Many, María J. Vicent, Felix Kratz, Joaquin Sanchis & Ronit Satchi-Fainaro
To cite this article: Anat Eldar-Boock, Rachel Blau, Claudia Ryppa, Hemda Baabur-Cohen,
Ariel Many, María J. Vicent, Felix Kratz, Joaquin Sanchis & Ronit Satchi-Fainaro (2017): Integrin-
Targeted Nano-Sized Polymeric Systems for Paclitaxel Conjugation: A Comparative Study, Journal
of Drug Targeting, DOI: 10.1080/1061186X.2017.1358727
To link to this article: http://dx.doi.org/10.1080/1061186X.2017.1358727
Accepted author version posted online: 24
Jul 2017.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=idrt20
Download by: [Tufts University]
Date: 25 July 2017, At: 01:07

Integrin-Targeted Nano-Sized Polymeric Systems for Paclitaxel Conjugation: A
Comparative Study
Anat Eldar-Boock1§, Rachel Blau1§, Claudia Ryppa2, Hemda Baabur-Cohen1, Ariel Many3,
María J. Vicent4, Felix Kratz2, Joaquin Sanchis4*‡, Ronit Satchi-Fainaro1
1Department of Physiology and Pharmacology, Sackler School of Medicine, Tel Aviv
University, Tel Aviv 69978, Israel.
2Tumor Biology Center, Breisacher Strasse 117, 79106 Freiburg, Germany.
3Lis Maternity Hospital, Sourasky Medical Center, Tel Aviv, Israel.
4Centro de Investigación Príncipe Felipe, Polymer Therapeutics Lab., Av. Eduardo Primo
Yúfera 3, E-46012 Valencia, Spain.
§ Equal contribution
*Corresponding Author:
Joaquin Sanchis, Ph.D., 4Centro de Investigación Príncipe Felipe, Polymer Therapeutics Lab.
Av. Eduardo Primo Yúfera 3, E-46012 Valencia, Spain.
Tel: +61 3 99039262; Fax: +61 3 99039581;
E-mail: Joaquin.Sanchis@monash.edu
‡
Current address: Monash Institute of Pharmaceutical Sciences, Monash University, 399
Royal Parade, Parkville, VIC 3052, Australia.
Running Title: Integrin-targeted paclitaxel
Key Words: Angiogenesis, polymer therapeutics, HPMA-copolymer, PEG, PGA, paclitaxel,
RGD.

ABSTRACT
The generation of rationally-designed polymer therapeutics via the conjugation of low
molecular weight anti-cancer drugs to water-soluble polymeric nanocarriers aims to improve
the therapeutic index. Here, we focus on applying polymer therapeutics to target two cell
compartments simultaneously - tumor cells and angiogenic endothelial cells. Comparing
different polymeric backbones carrying the same therapeutic agent and targeting moiety may
shed light on any correlation between the choice of polymer and the anti-cancer activity of the
conjugate. Here, we compared three paclitaxel (PTX) bound conjugates with poly-L-glutamic
acid (PGA, 4.9 mol%), 2-hydroxypropylmethacrylamide (HPMA, 1.2 mol%) copolymer, or
polyethyleneglycol (PEG, 1:1 conjugate). PGA and HPMA copolymer are multivalent
polymers that allow the conjugation of multiple compounds within the same polymer
backbone, while PEG is a commercially-available Food and Drug Administration (FDA)-
approved polymer. We further conjugated PGA-PTX and PEG-PTX with the integrin αvβ3-
targeting moiety RGD (5.5 mol% and 1:1 conjugate respectively). We based our selection on
the overexpression of integrin αvβ3 on angiogenic endothelial cells and several types of cancer
cells. Our findings suggest that polymer structure has significant effect on the conjugate's
activity on different tumor compartments. A multivalent PGA-PTX-E-[c(RGDfK)2] conjugate
displayed a stronger inhibitory effect on the endothelial compartment, with a 50% inhibition
of cell migration in HUVEC cells, while a PTX-PEG-E-[c(RGDfK)2] conjugate possessed
enhanced anti-cancer activity in MDA-MB-231 tumor cells (IC50 = 20 nM vs IC50 300 nM for
the PGA conjugate).

INTRODUCTION
The conjugation of chemotherapeutic drugs to polymeric carriers aims to improve
drug pharmacokinetic and pharmacodynamic profiles, increase efficacy, reduce toxicity, and
enable easier administration to improve patient compliance [1]. Furthermore, these
conjugates, known as polymer therapeutics, demonstrate tumor specific accumulation via the
enhanced permeability and retention (EPR) effect [2]. However, the particular polymeric
backbone used can influence both the pharmacokinetics and pharmacodynamics of the
conjugated drug. Polymer characteristics, such as average molecular weight (Mw), molecular
weight distribution (dispersity, Đ = Mw/Mn), architecture, charge, and hydrophilicity can
affect drug solubility, biodistribution, excretion, and interaction with the immune system [3-
5]. Among the different polymers in clinical use are 2-N-hydroxypropylmethacrilamide
(HPMA) [6, 7], poly-L-glutamic acid (PGA) [8], and polyethyleneglycol (PEG) [9].
PGA is a water soluble, non-toxic, and biodegradable polymer that can be synthesized
by ring-opening polymerization (ROP) of the corresponding N-carboxyanhydrides (NCA)
[10]. PGA contains a γ-carboxyl group in each repeating unit of L-glutamic acid that offers
multivalent attachment points for drugs and can be degraded in the lysosome by cysteine
proteases such as cathepsin B [8]. These features make PGA an attractive drug carrier.
Reinforcement is given by the PGA-PTX formulation (Opaxio™, formerly known as
XYOTAX), which is currently in Phase II/III clinical trials [11-13]. Encouragingly, Opaxio™
has demonstrated promising results in different types of cancer alone or in combination with
radiotherapy [14-16].
PEG, while not biodegradable, is water-soluble, biocompatible, and is Food and Drug
Administration (FDA)-approved for clinical use [17]. PEG has been traditionally used in the
polymer therapeutics field in polymer-protein conjugates for its anti-immunogenic and
stabilizing properties as well as for being an excellent time-of-circulation enhancer. More

recently, PEG has also been conjugated to certain drugs in order to improve their
biocompatibility and pharmacokinetics profile. In fact, Naloxegol (NKTR-118 - also known
as Movantik®), the first polymer-drug conjugate to achieve market approval, uses PEG as a
carrier [18-20].
The third polymer selected for this study, HPMA copolymer, represents one of the
most studied platforms for polymer-drug conjugates. HPMA copolymer is water-soluble,
non-charged, and non-immunogenic, and has been developed for site-specific delivery of
anti-cancer drugs [21].
HPMA copolymer, PGA, and PEG have all been used in polymer conjugates with
chemotherapeutic drugs such as paclitaxel (PTX) [22-25]. PTX, an anti-microtubule and anti-
angiogenic agent [26-29], has proven to be an effective treatment for metastatic breast cancer
and ovarian cancer, among notable others [30, 31]. However, poor water solubility
necessitates a formulation with Cremophor EL, which causes anaphylactic and
hypersensitivity reactions [32]. Conjugation of PTX with PEG, PGA, or HPMA-copolymer
preserves the anti-cancer potential of PTX while also abrogating associated side effects [22-
25].
Another strategy employed to improve the efficiency of polymer-drug conjugates is
the addition of a targeting moiety that targets specific cell types within or proximal to tumor
sites. This gains significance given the synergistic anti-cancer outcomes observed when
targeting angiogenesis, the generation of new blood vessels from pre-existing vasculature
[33, 34], and cancer cells at the same time. Tumor angiogenic blood vessels overexpress
certain unique molecular markers, such as the integrin αvβ3 heterodimeric cell surface
receptors, which enhance tumor cell proliferation, migration, invasion, and survival. The
tripeptide sequence Arg-Gly-Asp (RGD) targets integrin αvβ3 [35], while cyclization of linear
RGD improves binding properties and stability [36, 37]. Synthetic RGD-containing

molecules are miss-recognized and captured by these integrins in their search for extracellular
fibrinogen, which is RGD-motive rich in its structure. The successful application of cyclic
RGD as a targeting peptide has seen success in breast [25, 38] and prostate [39] cancer
models with enhanced anti-angiogenic and anti-cancer effects and reduced toxicity compared
to the unconjugated (or free) chemotherapeutic drug used in each study. Furthermore, the
RGD-based molecule Cilengitide (EMD121974) acts as a potent integrin αvβ3/αvβ5 inhibitor
and was the first agent reported to display anti-angiogenic clinical potential [40, 41].
Cilengitide is currently being tested in several clinical trials as a monotherapy or in
combination with conventional chemotherapy [42-44].
In order to fully understand and compare each polymer-drug conjugate, we proposed
a two-step comparison. First, we conjugated PTX to three different polymers (PGA, HPMA
copolymer, and PEG) and compared each conjugate in terms of drug release and anti-cancer
activity. Following this, we compared PGA-PTX and PEG-PTX conjugates with the addition
of RGD moieties in the hope of specifically targeting integrin-expressing cancer cells and
endothelial cells, while sparing healthy tissues.
MATERIALS AND METHODS
Materials
PTX was purchased from Alcon Biosciences Ltd. (Mumbai, India; Petrus Chemicals and
Materials Ltd., Israel). Dulbecco’s modified Eagle’s medium (DMEM), Fetal bovine serum
(FBS), Penicillin, Streptomycin, Nystatin, L-glutamine, HEPES buffer, sodium pyruvate, and
fibronectin were from Biological Industries Ltd. (Kibbutz Beit Haemek, Israel). EGM-2
medium was purchased from Cambrex (Walkersville, MD, U.S.A). Matrigel matrix was
purchased from BD Biosciences, USA. E-[c(RGDfK)]2 and c(RADfK) were purchased from

Peptides International (Louisville, KY, USA); Trt-S-C2H4-NHCO-PEG-C3H6COO-NHS
(MW 11.7 kDa) and Trt-S-C2H4-NHCO-PEG-C3H6COOH (MW 9.7 kDa) were purchased
from Rapp Polymers (Germany); PGA (MW 17 kDa, Đ = 1.3) was purchased from
Polypeptide Therapeutic Solutions SL (PTS, Spain). HPMA copolymer-Gly-Phe-Leu-Gly-
ONp (Mw = 31.6 kDa, Đ = 1.66) incorporating 10 mol% of the methacryloyl-Gly-Phe-Leu-
Gly-p-nitrophenol ester monomer units were obtained from Polymer Laboratories (Church
Stretton, UK). Organic solvents were High-performance liquid chromatography (HPLC)
grade (Labscan Ltd., Dublin, Ireland; Roth, Karlsruhe, FRG; Merck, Darmstadt, FRG). All
other chemicals used were reagent grade and obtained from Sigma-Aldrich (Deisenhofen,
Germany) or Merck (Darmstadt, Germany) and used without further purification; buffers
were vacuum-filtered through a 0.2 m membrane (Sartorius, Germany) and thoroughly
degassed with nitrogen prior to use. Mass spectra were obtained using a Thermo Electron
LCQ Advantage with the associated MAT SS 200 data system using electron spray
ionization. UV/VIS-spectrophotometry was carried out with
spectrophotometer U-2000 from Hitachi.
a
double-beam
Synthesis of PGA-PTX conjugates
Synthesis of PGA-PTX, PGA-PTX-E-[c(RGDfK)2] and PGA-PTXc(RADfK) conjugates was
previously reported by Eldar-Boock et al. [25].
a) Synthesis of PGA-PTX (1)
Briefly, PGA-PTX was conjugated to PGA (Mw = 17.7 kDa, Đ = 1.3) by carbodiimide
coupling (DIC/HOBT). The reaction was allowed to proceed at room temperature (RT) for 24
h. The conjugate was precipitated in a CHCl3:Acetone mixture (1:4), filtered, and dried under
vacuum and after lyophilization, PGA-PTX (1) formed a white powder (70-80% yield)
(Figure 1). PTX content was determined by ultraviolet (UV) ( 227 nm and 230 nm in

MeOH) and HPLC (indirect analysis determining PTX content in reaction residues  = 227
nm, gradient 35- 80% (v/v) water:CH3CN) and obtained residue was dissolved in 1.0 M
NaHCO3 and purified by aqueous Size Exclusion Chromatography (SEC) in a Sephadex G25
column.
b) Synthesis of PGA-PTX-c(RADfK) (3) and PGA-PTX-E-[c(RGDfK)2] (4):
Briefly, PGA (Mw = 17.7 kDa, Đ = 1.3) was conjugated to PTX by carbodiimide coupling
(DIC/HOBT). The reaction was allowed to proceed at RT for 24 h. Then, without product
isolation, N-hydroxysuccinimide (SuccOH, 30 mol %) was added to the reaction mixture and
the reaction was allowed to proceed for another 24 h. The mixture was then poured into a
mixture of CHCl :Acetone (1:4). The resulting precipitate was collected and washed again
3
with acetone and MeOH and dried in vacuum to yield a white solid powder (2). The solid was
intermediately dissolved once more in anhydrous dimethylformamide (DMF) and then
c(RADfK) or E-[c(RGDfK)2] were added together with catalytic amounts of 4-
dimethylaminopyridine (DMAP). The pH was adjusted to 8 with DIEA. The reaction was
allowed to proceed at RT for 48 h and then the mixture was poured into CHCl3:Acetone. The
resulting precipitate was collected and dried in vacuum to yield a white powder that was then
dissolved in 1.0 M NaHCO3. This aqueous solution was run through an SEC (Sephadex G25)
column and lyophilization of the fractions yielded the product as a white powder (70-80%
yield). PTX content was analyzed as described above for conjugate 1. The total peptide
content in these polymeric conjugates was determined by UV ( = 254 nm and 260 nm in
MeOH) and HPLC (indirect analysis determining peptide content in reaction residues,
gradient: CH3CN:0.1% (v/v) trifluoroacetic acid (TFA) (10-90% (v/v))  = 220 nm). Peptide
content was also determined by amino acid analysis.

Synthesis of PEG-PTX conjugates:
a) Synthesis of PTX-EMCH (6)
In order to generate an acid sensitive linker, the non-commercial PTX-derivative Ketotaxol
(5) was synthesized following the protocol from Rodrigues et al. [45]. Afterwards, as
described by Moktan et al. [46], 400 mg of (5) (0.4 mmol) and 272 mg of 6-
maleimidocaproic acid hydrazidetrifluoroacetate (0.8 mmol) were dissolved in 12 mL
anhydrous ethanol and stirred for 90 min at RT. The solution was incubated overnight at
+4°C. The white precipitate was isolated by centrifugation, washed with 15 mL diethylether,
and isolated by centrifugation. Drying under high vacuum produced 283 mg of 6.
b) Synthesis of PTX-PEG-COOH conjugate (10)
Trt-S-C2H4-NHCO-PEG-C3H6COOH (154 mg, 16.2 µmol, 1 equiv., Mw = 9.7 kDa) in 400
µL TFA/ dichloromethane (DCM) 1/1, v/v (containing 2.5 % (v/v) iPr3SiH, 5 %
thioanisole/anisole, 1/1, (v/v); 5 % (v/v) H2O) was sonicated for approximately 30 seconds
(clear yellow solution turned into clear colorless solution) and stirred at RT for 20 min. The
solvent was removed in high vacuum (approximately 20 min) and petrol ether/ether (1.5-2
mL, 1/1, v/v) was next added to the residue and sonicated for ~30 seconds. The supernatant
was removed after centrifugation. A final wash with dry ether was performed to yield
conjugate 7. In order to remove the protecting group on the thiol, conjugate 7 was dissolved
in 500 µL of 50 mM sodium phosphate buffer pH 7. The pH of the remaining solution was
readjusted to 7-7.5 with 0.1 N NaOH. PTX-EMCH was immediately added (19 mg, 1 eq),
sonicated for 20 min, and stirred on ice for 4 h. The conjugate was purified by SEC
(Sephadex G25 in phosphate-buffered saline (PBS) pH=7) to yield conjugate 10.

c) Synthesis of HS-PEG- E-[c(RGDfK)2] (8) and HS-PEG-c(RADfK) (9)
Conjugates (8) and (9) were obtained following an already described protocol by Polyak et
al. [38]. Briefly, N,N-Diisopropylethylamine (DIEA) (10.2 or 24 L respectively) was added
to a solution of the Trt-S-C2H4-NHCO-PEG-C3H6COO-NHS (700 mg, 59.8 mmol, 1 equiv.,
Mw = 11.7 kDa) and, depending on the case, E-[c(RGDfK)2] (78.9 mg, 59.8 mmol, 1 equiv.)
or c(RADfK) (10 mg, 16.2 µmol, 1 equiv.) in dry DMF (15 mL). The solution was stirred at
RT for 48 h, then precipitated in dry diethyl ether (300 mL), washed with pentane (150 mL),
and filtered and dried in vacuum to obtain a white powder (694 mg, Mw = 13.0 kDa in the
case of conjugate 8 and 103 mg, 16 µmol Mw = 12.3 kDa for conjugate 9). The material was
stored at -20 °C. The protected Trt-S-PEG-E-[c(RGDfK)2] (25 mg) in 400 µL TFA/DCM 1/1,
v/v (containing 2.5 % (v/v) iPr3SiH, 5 % (v/v) thioanisole/anisole, 1/1, v/v; 5 % (v/v) H2O)
was sonicated for 30-60 seconds (clear yellow solution turned into clear colorless solution)
and then stirred at RT for 20 min. The solvent was removed in high vacuum (approximately
20 min). Petrol ether/ether (1.5-2 mL, 1/1, v/v) was added to the residue and sonicated for
approximately 30 seconds. The supernatant was removed after centrifugation and dry ether
was added to the residue. After centrifugation, the residue was dissolved in 500 µL of a 50
mM sodium phosphate buffer pH 7. The pH of the remaining solution was readjusted to 7-7.5
with 0.1 N NaOH. The amount of SH groups was determined by means of an Ellman assay (3
x 10 µL solution; = 412 nm,  = 13600 M-1 cm-1). The solution was used directly for
coupling to the PTX prodrug-EMCH (6).
d) Synthesis of PTX-PEG-E-[c(RGDfK)2] (11) and PTX-PEG-c(RADfK) (12)
PTX prodrug-EMCH (1.05 mg, 0.8 mol, 1 eq.) was added to a solution of freshly
deprotected HS-PEG-E-[c(RGDfK)2] (8) or HS-PEG-c(RADfK) (9) (10 and 9.3 mg, 1 eq.,
respectively) in 50 mM sodium phosphate buffer pH 7. The resulting solution was sonicated

and stirred on ice for 4 h. The PEG conjugate was isolated by SEC (G25 Sephadex in PBS
buffer, pH=7). The fractions were freeze dried to yield conjugate 11 and 12 respectively.
Synthesis of HPMA copolymer-GFLG-FK-PTX conjugate (18)
The conjugation of PTX with HPMA copolymer through a GFLG-FK peptidic linker was
performed as previously described [24]. The synthetic strategy comprises the attachment of
PTX to a FK-PABC linker (15) and then its conjugation to HPMA copolymer–GFLG-ONp
(Figure 3).
a) Synthesis of Boc-Phe-Lys (alloc)-OH (13)
L-Boc-Phe-ONp (208.6 mg, 0.54 mmol) was dissolved in 3 mL DMF and commercially
available L-lys(alloc)-OH (124 mg, 0.54 mmol) and Et3N (200 μL) were then added. The
reaction mixture was stirred for 12 h and was monitored by thin layer chromatography (TLC)
(AcOH:MeOH:EtOAc 0.5:10:89.5). Upon completion of the reaction, the solvent was
removed under reduced pressure and the crude product was purified by using column
chromatography on silica gel (AcOH:MeOH:EtOAc 0.5:10:89.5) to give conjugate 13 (214
mg) as a white solid (Figure 3).
b) Synthesis of Boc-Phe-Lys (alloc)-PABA (14)
Boc-Phe-Lys (alloc)-PABA 13 (208 mg, 0.435 mmol) was dissolved in dry tetrahydrofuran
(THF) and the solution was cooled to -15° C. Then, N-Methylmorpholine N-oxide (NMM)
(475 µL, 0.435 mmol) and isobutyl chloroformate (675 µL, 0.522 mmol) were added. The
reaction was stirred for 20 min and a solution of 4-aminobenzyl alcohol (80.46 mg, 0.65
mmol) in dry THF was added. The reaction mixture was stirred for 2 h and was monitored by
TLC (EtOAc). Upon completion of the reaction, the solvent was removed under reduced

pressure and the crude product was purified by using column chromatography on silica gel
(EtOAc) to give conjugate 14 (208 mg) as a yellow solid (Figure 3).
c) Synthesis of Boc-Phe-Lys (alloc)-PABC-ONp (15)
Boc-Phe-Lys (alloc)-PABA (14) (186.1 mg, 0.315 mmol) was dissolved in dry THF and the
solution was cooled to 0° C. Then N, N-Diisopropylethylamine (DIPEA) (221 µL, 1.273
mmol), PNP-chloroformate (193.1 mg, 0.526 mmol), and a catalytic amount of pyridine were
added. The reaction was stirred for 2 h and monitored by TLC (EtOAc:Hex 3:1). Upon
completion of the reaction, the solvent was removed under reduced pressure. The crude
product was diluted with EtOAc and washed with saturated NH4Cl. The organic layer was
dried over magnesium sulfate and the solvent was removed under reduced pressure. The
crude product was purified by using column chromatography on silica gel (EtOAc:Hex 3:1)
to give conjugate 15 (238.5 mg) as a white solid (Figure 3).
d) Synthesis of Boc-Phe-Lys (alloc)-PABC-PTX (16)
Boc-Phe-Lys (alloc)-PABC-ONp 15 (212 mg, 0.28 mmol) was dissolved in dry DCM. Then
PTX (123.5 mg, 0.33 mmol) and DMAP (17.65 mg, 0.33 mmol) were added. The reaction
mixture was stirred for 8 h at RT and monitored by TLC (EtOAc). Upon completion of the
reaction, the solvent was removed under reduced pressure and the crude product was purified
by using column chromatography on silica gel (EtOAc) to give conjugate 16 (389.4 mg) as a
white solid (Figure 3).

e) Synthesis of HPMA copolymer-GFLG-Phe-Lys(alloc)-PABC-PTX conjugate (17)
Boc-Phe-Lys (alloc)-PABC-PTX (Conjugate 16) (12 mg, 7.57 μmol) was dissolved in 0.5
mL TFA and stirred for 2 min at 0°C. The excess of acid was removed under reduced
pressure and the crude amine salt was dissolved in 0.5 mL DMF. HPMA copolymer-GFLG-
ONp (26.3 mg, ONp = 8.32 μmol, 31.6 kDa, Đ =1.66) was added followed by the addition of
Et3N (3 μL). The reaction mixture was stirred for 12 h and the solvent was removed under
reduced pressure. Free FK-PTX and ONp were removed by fast protein liquid
chromatography (FPLC) using an XK26/70 column with Sephadex LH20 column (MeOH, 1
mL/1 min) to give the alloc protected HPMA copolymer-GFLG-FK-PTX (17) as a white
solid (20 mg) (Figure 3).
f) Synthesis of HPMA copolymer –GFLG-FK-PTX (18)
Alloc protected HPMA copolymer –GFLG-FK(alloc)-PTX (30 mg, alloc = max. 9.9 μmol)
was dissolved in DMF (1 mL). Then, acetic acid (2.71 μL, 47.4 μmol), Bu3SnH (30.6 μL, 113
μmol), and a catalytic amount of Pd (PPh3)4 were added. The reaction mixture was stirred for
2 h at RT and was concentrated under reduced pressure, followed by addition of 10 mL of
acetone. The precipitate was filtered out and washed with acetone several times. The crude
product was purified by FPLC using XK26/70 column with Sephadex LH20 (MeOH, 1 mL/1
min) to give HPMA copolymer-GFLG-FK-PTX (20 mg) as a white solid (Figure 3). Content
of PTX in the conjugate was determined by HPLC analysis according to calibration curve of
free PTX-FK.
Determination of mean hydrodynamic diameter of the conjugates
The mean hydrodynamic diameter of the nanocarriers was evaluated by dynamic light
scattering (DLS) using a ZetaSizer Nano ZS instrument with an integrated 4mW, He-Ne laser

(λ=633 nm; Malvern Instruments Ltd., Malvern, Worcestershire, UK). Samples were
prepared by dissolving 1 mg of polymer conjugate in 1 ml of 155 mM phosphate buffer,
pH=7.4. All measurements were performed at 25°C using polystyrol/polystyrene (10×4×45)
mm cell.
Drug release profile and polymer degradation kinetics
a) Degradation in the presence of cathepsin B
Degradation of PTX catalyzed by cathepsin B was studied for PGA-PTX and HPMA
copolymer-GFLG-FK-PTX. In the case of PGA-PTX, a previously reported protocol was
followed [25]. Briefly, cathepsin (5 U) was added to a solution of conjugate or free drug (3
mg/mL) in a fresh prepared buffer made of 20 mM sodium acetate, 2 mM EDTA and 5 mM
DTT, pH 5.5) and incubated at 37°C. Aliquots (100 μL) were taken at different times up to 72
h, immediately frozen in liquid nitrogen, and stored in darkness. The amount of released
compound was assayed by HPLC (analysis after extraction procedure by Poros50 resin drug
content analysis (λ=227 nm)) using doxycycline as internal standard. The release of PTX
from HPMA-PTX-FK conjugate was monitored by reversed phase (RP) HPLC using
Phenomenex Jupiter 5 µm 250 X 4.60 mm C-18 300A column. Samples (50 µL) were
collected every 24 h, until a plateau was observed (up to 72 h). For PTX extraction, sodium
carbonate buffer solution (0.2 M, pH=9.6) was added to each sample, followed by ethyl
acetate. Samples were vigorously vortexed and centrifuged and the organic layer carefully
removed and evaporated. The residue was dissolved in MeCN and analyzed by HPLC. For
non-enzymatic hydrolytic cleavage assessment, polymers were incubated in buffer alone (pH
5.5) in the absence of cathepsin B. Additionally, an LC-MS analysis of the released
compounds was carried out to determine the major metabolites released.

b) Degradation in acidic pH
PEG-PTX or PTX-PEG-RGD conjugate were dissolved at a concentration of 5 mg/mL in
PBS at pH 5.5 or pH 7.4. Samples were incubated at 37°C and aliquots (50 µL) were taken at
several time-points up to 72 h. Samples were stored at -20°C in darkness until analysis. The
amount of released drug was assayed by HPLC against a calibration curve of free
PTX. Samples (50 µL) were injected to HPLC equipped with C18 column (Jupiter, 300A,
250x4.6 mm, 5 micron).
Cell culture
The MDA-MB-231 human mammary adenocarcinoma cell line and 4T1 murine breast cancer
cell lines were purchased from the American Type Culture Collection (ATCC). MDA-MB-
231 cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with
10% (v/v) fetal bovine serum (FBS), 100 U/mL Penicillin, 100 µg/mL Streptomycin, 12.5
U/mL Nystatin, and 2 mM L-glutamine. 4T1 cells were grown in RPMI-1640 supplemented
with 10% (v/v) FBS, 100 U/mL Penicillin, 100 µg/mL Streptomycin, 12.5 U/mL Nystatin, 2
mM L-glutamine, 1% (w/v) HEPES 1M, 22.5% (w/v) Glucose and 1 mM Sodium Pyruvate.
Human umbilical vein endothelial cells (HUVEC) were isolated in our laboratory and grown
in EGM-2 medium (Cambrex). Cells were grown at 37°C; 5% CO2.
Isolation of endothelial cells from human umbilical cord veins (HUVEC)
Umbilical cords were collected at Lis Maternity Hospital, Sourasky Medical Center, Tel
Aviv, Israel. The protocol was approved by the Institutional Review Board (IRB), as
previously described [47]. Umbilical cords were washed with 50 mL PBS containing 100
U/mL Penicillin, 100 µg/mL Streptomycin ,12.5 U/mL Nystatin. A blunt end needle was
inserted into one edge and both edges were tightened using a clip. Following 15 min

incubation with 0.25 mg/mL collagenase II (Sigma) at 37˚C, HUVEC were washed into tubes
using 50 mL PBSx1. Cells were centrifuged for 10 min at 1100 rpm, resuspended in EGM-2,
and cultured on fibronectin-coated tissue culture plates. Cells were grown according to the
manufacturer’s protocol in EGM-2 MV medium (Cambrex). All Cells were grown at 37°C;
5% CO2.
Cell viability assay
HUVEC were plated onto 24-well plate (1.5 x 104 cells/well) in growth factor-reduced media
(EBM-2, Cambrex, USA) supplemented with 5% (v/v) FBS. Following 24 h of incubation
(37°C; 5% CO2), medium was replaced with EGM-2 (Cambrex, USA). 4T1 cells (3,000
cells/well), were plated onto 24-well culture plates in RPMI supplemented with 5% (v/v)
FBS and incubated for 24 h (370C; 5% CO2). The medium was then replaced with RPMI
1640 supplemented with 10% (v/v) FBS. MDA-MB-231 cells were plated onto 96 well plate
(5 x 103 cells/well) in DMEM supplemented with 5% (v/v) FBS and incubated for 24 h
(37°C; 5% CO2). Following 24 h of incubation, medium was replaced with full medium.
Cells were exposed to PTX and PTX bounded conjugates at serial dilutions, at equivalent
dose of the free PTX for up to 72 h. Following incubation, HUVEC and 4T1 cells were
counted using a Coulter Counter (Beckman Coulter®). MDA-MB-231 cell viability was
measured using MTT assay.
αvβ3 integrin expression
As was previously described [25], cells were harvested with 2.5 mM
ethylenediaminetetraacetic acid (EDTA), re-suspended in serum-free medium and incubated
for 30 min. Cells were then re-suspended in PBS (containing Mg2+ and Ca2+) with an
MAB1976-anti-V3 integrin antibody (Chemicon) and incubated for 1 h at RT. Control

samples were antibody-free. Cells were then washed and incubated with FITC-donkey anti-
mouse IgG antibody (Jackson) for 30 min at RT in the dark. Cells (1x105) were collected by
fluorescence-activated cell sorter (FACS) and analyzed using Cyplogic software (6Cytek
DxP 6-Color Upgrade, Facscan™).
Capillary-like tube formation assay
The surface of 24-well plates was coated with Matrigel matrix (50 µL/well) (BD Biosciences,
USA) on ice and was then allowed to polymerize at 37°C for 30 min. HUVEC (3 x 104 cells)
were challenged with free PTX (20 nM), PTX produg, PTX-PEG, PTX-PEG-[cRADfK], and
PTX-PEG-E-[c(RGDfK)2] conjugates or with PGA-PTX, PGA-PTX-[cRADfK], and PGA-
PTX-E-[c(RGDfK)2] conjugates at PTX-equivalent concentrations, and were seeded on
coated plates in the presence of complete EGM-2 medium. After 8 h of incubation (37°C; 5%
CO2), wells were imaged using Nikon TE2000E inverted microscope integrated with Nikon
DS5 cooled CCD camera by 4X objective, brightfield technique.
Endothelial cell adhesion assay
The ability of free and PGA or PEG conjugated E-[c(RGDfK)2] to bind v3 integrin on the
cell surface was evaluated by HUVEC adherence to fibrinogen following incubation with the
conjugates, as previously described [48]. Shortly, flat bottom 96-well culture plates were
coated with 0.5 µg/well fibrinogen (overnight, 4ºC). Following three washes with PBS, the
wells were blocked with 1 % (w/v) bovine serum albumin (BSA) for 1 h at 37ºC. HUVEC
were harvested in PBS with 2.5 mM EDTA, re-suspended in EBM-2 serum-free media and
were incubated with the different treatments for 30 min at RT. Treated HUVEC were then
plated at 5 x 104 cells/well and allowed to attach to fibrinogen-coated plates for 1 h at 37ºC.

Following incubation, unattached cells were removed by rinsing the wells with PBS.
Attached cells were fixed with 3.7% (v/v) formaldehyde, stained with 0.5% (w/v) crystal
violet, and imaged using a Nikon TE2000E inverted microscope integrated with a Nikon DS5
cooled CCD camera by 6X objective, bright field technique. The number of attached cells
was quantified with NIH ImageJ processing and analysis software. Non-specific binding was
determined by adhesion to BSA-coated plates.
Migration assay
Cell migration assay was performed as previously described [49]. HUVEC (1.5x105) were
exposed to PTX, PTX prodrug, PTX-PEG, PTX-PEG-[cRADfK], and PTX-PEG-E-
[c(RGDfK)2] conjugates or to PGA-PTX, PGA-PTX-[cRADfK], and PGA-PTX-E-
[c(RGDfK)2] conjugates at PTX-equivalent concentrations, that were added to the upper
chamber of transwells for 2 h of incubation prior to migration towards vascular endothelial
growth factor (VEGF). Migration was normalized so that 100% represents migration to
VEGF alone without exposure to any compound.
Statistical methods
Data is expressed as mean ± s.d. Statistical significance was determined using an unpaired t-
test. P<0.05 was considered statistically significant. All statistical tests were two-sided.
RESULTS
Multiple polymers, each with their own advantages, have been employed as integral
parts of targeted polymer therapeutics for efficient drug delivery. However, we lack a
detailed comparison between different polymers in a similar system. Therefore, we evaluated
drug release and cytotoxicity of three polymeric carriers (PGA, PEG, and HPMA-copolymer)

following conjugation to PTX in two mammary adenocarcinoma cell lines (human MDA-
MB-231 and murine 4T1). To address the fundamental question regarding the influence of
polymeric scaffold multivalency on the anti-angiogenic activity of PTX conjugates, we also
decorated PGA-PTX and PEG-PTX with E-c(RGDfK)2, an αvβ3 integrin targeting moiety.
We assessed the anti-angiogenic behavior (i.e. cell proliferation, capillary-like tubular-
structure formation, cell migration, and cell adhesion to fibrinogen) of these conjugates in an
integrin-overexpressing cell line (HUVEC), as well as each conjugate’s cytotoxic effect on
integrin-overexpressing cancer cell lines.
Synthesis, characterization, and drug release profiles for PGA, PEG, and HPMA
bearing PTX
a) Synthesis of PGA-PTX conjugate (1)
We conjugated PGA-PTX, as previously described [25], via PTX introduction by
esterification of the 2’-hydroxyl with DIC/HOBT activated carboxylic residues of PGA
(Figure 1). We determined average molecular weight (Mw) and dispersity (Mw/Mn, Đ) by
SEC as 17.7 kDa (Đ = 1.3) for PGA and 25.3 kDa (Đ = 1.3) for PGA-PTX conjugate. Drug
loading determination via UV-Vis and HPLC demonstrated a range of 4.9 mol% (around 6-7
molecules of PTX per chain) with a free-drug content of less than 2% of the total payload.
b) Synthesis of PEG-EMCH-PTX conjugate (10)
In order to conjugate PTX to PEG (Mw = 9.7 kDa), we used 6-maleimidocaproyl hydrazine
(EMCH) as a pH-responsive linker and achieved efficient conjugation via a Thiol-Maleimide
reaction. Therefore, the resulting conjugate (Mw = 10.9 kDa) contains a hydrazone bond that
is cleaved at the acidic pH of endosomes and lysosomes resulting in intracellular drug release
[46] (Figure 2).

c) Synthesis of HPMA copolymer-GFLG-FK-PTX conjugate (18)
Figure 3 displays the synthetic pathway and the chemical structure of HPMA copolymer-
GFLG-FK-PTX conjugate 3. We conjugated PTX to the HPMA copolymer through a
cathepsin B-hydrolyzable Gly-Phe-Leu-Gly (GFLG) peptidic sequence followed by a Phe-
Lys-PABC carbonate linker (referred in the text as FK). The conjugation to the HPMA
copolymer used a three-step procedure: 1) PTX attachment to the FK–PABC linker, 2) PTX-
linker conjugation to HPMA copolymer–GFLG–ONp (Mw = 31.6 kDa, Đ = 1.66), and
finally, 3) removal of the alloc protecting group. The resulting conjugate 18 (Mw
(theoretical) = 34.1 kDa,) was water-soluble with a PTX loading (determined by HPLC) of
1.12 mol% (around 2 PTX molecules per polymeric chain).
c) Hydrodynamic diameter of the conjugates
We characterized the hydrodynamic diameter and size distribution of our conjugates, finding
a diameter of ~6.8 nm for PGA-PTX and 4.9 nm for PEG-PTX in physiological PBS,
indicating a behavior as monomolecular micelles. However, we found a larger diameter for
the HPMA copolymer-GFLG-FK-PTX conjugate (~19.51 nm), possibly due to some degree
of aggregation. Irrespective of this small aggregation effect, all conjugates demonstrated the
appropriate size to take full advantage of the EPR effect and enhanced tumor accumulation.
Table 1 summarizes the physico-chemical features of the different conjugates.
Table 1. Physico-chemical characteristics of the different PTX-bearing conjugates
Mw†
Size§
Payload¶
PGA-PTX
PEG-PTX
HPMA-PTX
25.3
10.9
34.1
6.8
4.9
19.51
4.9
1‡
1.12

‡ 1 molecule of PTX per chain. † Mw = molecular weight (in kDa). Measured by GPC in PBS
§
¶
(20 mM, pH= 7). Diameter, determined by DLS, in nm. Distribution by volume.
Determined by HPLC against PTX calibration curve in acetonitrile-water gradient and by UV
( = 227 nm) in mol%.
d) Drug release from the conjugates
In order to assess linker degradation, we performed a preliminary drug release assay using an
incubation with acetate buffer (pH 5.5) and cathepsin B for the PGA and HPMA conjugates.
As expected, both conjugates exhibited defined time-dependent PTX release (Figure 4A and
4C respectively). Ester hydrolysis in the case of PGA and a 1,6-elimination followed by a
decarboxylation after the hydrolysis of the FK linker for HPMA yielded PTX as the main
metabolite released from the polymer, as determined by liquid chromatography–mass
spectrometry (LC-MS) experiments using a MALDI-TOF as the MS detector. An incubation
of 48 h led to a 16.5% PTX release from PGA-PTX conjugate, and similarly, a 13.2% release
from the HPMA-GFLG-FK-PTX conjugate. In the absence of cathepsin B, we observed a
negligible amount of PTX release from the PGA conjugate and only 3% of the total payload
for the HPMA conjugate. This observation confirmed that cathepsin B effectively hydrolyzed
the peptide-base backbone or GFLG linker in PGA and HPMA copolymer, respectively.
The non-peptidic EMCH linker in PEG-PTX conjugate contains an acid labile
hydrazone bond and is insensitive to cathepsin B hydrolysis. Hence, we employed a drug
release assay upon incubation at pH 5.5 in this case (Figure 4B). Similar to the previous
conjugates, PTX represented the main released metabolite from the polymer, as determined
by analytical HPLC. Following 48 h incubation at pH 5.5, the conjugate released 19.7% PTX.
PTX was not released at pH 7.5 and remained under 2%.

Cell growth inhibition of the non-targeted PTX-bearing conjugates
To evaluate whether PTX retained its cytotoxic activity following polymer conjugation, we
determined potential inhibitory activity on the proliferation of human MDA-MB-231 and
murine 4T1 mammary adenocarcinoma cell lines after 72 h of incubation.
a) Inhibition of MDA-MB-231 cells
Initial results suggested that all PTX conjugates mediated a decrease in the proliferation of
human MDA-MB-231 cells. PGA-PTX conjugate exhibited an IC50 value of ~80 nM (Figure
5A), PEG-PTX exhibited a lower IC50 value of ~8 nM PTX-equivalent concentrations
(Figure 5B), while HPMA-PTX-FK exhibited higher IC50 of ~150 nM (Figure 5C). We found
that the IC50 of free PTX to be ~10 nM.
b) Inhibition of 4T1 cells
Interestingly, the conjugates displayed considerable differences concerning the proliferation
of murine 4T1 cells. Free PTX exhibited an IC50 of ~50 nM and PGA-PTX exhibited a
similar level of inhibition with an IC50 of ~150 nM (Figure 5D), while PEG-PTX exhibited a
less effective IC50 of ~9000 nM (Figure 5E) and HPMA-PTX-FK exhibited an IC50 of ~1000
nM (Figure 5F). Overall, this suggests a clear advantage for PGA conjugation.
Conjugates for anti-angiogenic and cytotoxic combination therapy: αvβ3 integrin-
targeted PGA-PTX and PEG-PTX conjugates
We wanted to address the fundamental question of whether the conjugation of an integrin-
targeting moiety (E-[c(RGDfK)2]) to the polymeric backbone could influence drug release
and cytotoxicity of PTX conjugates. In addition, we also aimed to assess whether the
introduction of c(RGDfK), whose structure is very similar to the reported anti-angiogenic

molecule Cilengitide (c(RGDfV)), can confer an additional anti-angiogenic activity on the
conjugate.
Based on our initial results, we chose PGA and PEG conjugates for further
conjugation with E-[c(RGDfK)2]. PGA-PTX and PEG-PTX displayed similar sizes (6.8 and
4.9 nm) and both exhibited time-dependent release of PTX after 48 h. In addition, while the
PEG conjugate displayed the greatest inhibitory effect on human MDA-MB-231 mammary
cancer cells, the PGA conjugate was the best on murine 4T1 mammary cancer cells.
Another unresolved issue in polymer therapeutics is the influence of the number of
targeting moieties attached to one single backbone. Although the ratio between PTX and
RGD in both conjugates was equal to 1, PGA contains more than 6 molecules per side-chain
while PEG enables the conjugation of only one moiety at one end of its chain.
Synthesis and characterization of RGD and PTX containing conjugates
a) Synthesis of PGA-PTX-E-[c(RGDfK)2] and control conjugates
We synthesized PGA-PTX-E-[c(RGDfK)2] (4) carrying PTX and E-[c(RGDfK)2] at similar
loadings, 4.9 and 5.5 mol% respectively (and also a conjugate with c(RADfK) (3) to be used
as a control in all the experiments) following reported protocols [25] (Figure 1). In order to
avoid crosslinking reactions with the carboxylate present in E-[c(RGDfK)2], we separately
activated the remaining glutamic carboxylates in PGA-PTX (1) with N-hydroxysuccinimide
and coupled via amide bond to the E-[c(RGDfK)2] α-free amino group in its linker glutamic
amino acid (E).
b) Synthesis of PTX-PEG-E-[c(RGDfK)2] and control conjugates
Similarly to PGA-PTX-E-[c(RGDfK)2], we conjugated the cyclic peptides to the NHS
activated carboxylated of Trt-S-PEG-CO-NHS forming an amide (8 and 9 for RGD and RAD

respectively, Figure 2). We deprotected the Thiol groups of these conjugates and immediately
conjugated them to the maleimide moiety of EMCH-PTX prodrug (11 and 12) [46].
c) Hydrodynamic diameter of E-[c(RGDfK)2] targeted conjugates
In order to evaluate the change in size of our conjugates after the introduction of the targeting
moiety, we determined hydrodynamic diameter by DLS. PGA-PTX-E-[c(RGDfK)2]
conjugate exhibited a diameter of ~8.6 nm, while PTX-PEG-E-[c(RGDfK)2] conjugate
exhibited a diameter of ~10.44 nm (Figure 6).
d) PTX release from integrin-targeted PGA and PEG conjugates
We next evaluated drug release kinetics upon incubation with either cathepsin B or acidic
pH for PGA-PTX-E-[c(RGDfK)2]. Cathepsin B was responsible for the observed time-
dependent release profile, with ~60% of the total PTX detached after 72 h (Figure 7A).
Moreover, we found that a pH of 5.5 did not hydrolyze the ester bond during the same time-
scale.
The synthetic linker in the PEG-PTX conjugate was not expected to be cleaved by
cathepsin B; therefore, we focused on pH-dependent drug release. As expected, we found a
progressive degradation of the hydrazone linker at acidic pH (Figure 7B). Following 72 h of
incubation at pH 5.5, the conjugate released ~25% of PTX, while we observed no significant
release at neutral pH.
Cell growth inhibition of v3 integrin-expressing cell lines by RGD-targeted PTX-
bearing conjugates
In order to evaluate whether conjugation of RGD peptides to the polymeric backbone
has any effect on the internalization and cytotoxic activity of PTX, we determined cell-

growth inhibitory efficiency on the proliferation on two αvβ3-expressing cell lines, namely
MDA-MB-231 [50] and HUVEC [51].
Inhibition of MDA-MB-231 cell growth
We incubated αvβ3 integrin-expressing MDA-MB-231 (Figure 8A) with different
conjugates and used free PTX as a reference. PGA-PTX conjugates inhibited the proliferation
of MDA-MB-231 cells and exhibited IC50 values of ~300 nM PTX-equivalent concentrations
for PGA-PTX-E-[c(RGDfK)2] and PGA-PTX-c(RADfK) and ~1000 nM for PGA-PTX
(Figure 8B). However, PTX-PEG-E-[c(RGDfK)2], PTX-PEG-c(RADfK), and PTX-PEG
exhibited an IC50 in the range of 10-20 nM PTX-equivalent concentrations, which was very
similar to the powerful activity of free PTX. Surprisingly, the PTX-EMCH prodrug showed a
higher IC50 of ~150 nM (Figure 8C).
Anti-angiogenic cascade inhibition by PGA-PTX-E-[c(RGDfK)2] and PTX-PEG-E-
[cRGDfK)2]
With the initial results at hand, we evaluated the anti-angiogenic properties of PGA-PTX-
E-[c(RGDfK)2] and PTX-PEG-E-[c(RGDfK)2] compared to free PTX. To that end, we
carried out endothelial cell proliferation, capillary-like tube formation, cell adhesion, and
migration assays using HUVECs.
a) Inhibition of the Endothelial cell proliferation
We incubated αvβ3 integrin-expressing HUVECs (Figure 8D) with different
conjugates and used free PTX as a reference. After 72 h of exposure, free PTX exhibited an
IC50 value of ~2 nM. In the same range, PGA-PTX-E-[c(RGDfK)2] conjugate inhibited
HUVEC proliferation at an IC50 of ~8 nM (Figure 8E). PTX-PEG and PTX-PEG-E-
[c(RGDfK)2] presented an IC50 of ~20 nM, and ~25 nM, respectively, while PTX-EMCH

prodrug exhibited a higher IC50 of ~50 nM (Figure 8F). However, a shorter time exposure (15
min) of the cells to the conjugates revealed a significant difference between the two targeted
polymer conjugates. While PGA-PTX-E-[c(RGDfK)2] exhibited an IC50 of ~2000 nM
(Figure 8G), we could not establish an IC50 value for PTX-PEG-E-[c(RGDfK)2], probably
due to its slow release (Figure 8H). Table 4 summarizes the different IC50 values.
b) Capillary-like tubular structures inhibition
Next, we examined the effect of the conjugates on capillary-like tubular structure-forming
ability of HUVEC in Matrigel, an additional crucial step in the angiogenic cascade of events
(Figure 9A). Free PTX and PGA-PTX-E-[c(RGDfK)2] conjugate inhibited the formation of
tubular structures of HUVEC by ~25%.
Interestingly, while PTX-PEG-E-[c(RGDfK)2] displayed a similar activity to free PTX at
the same PTX-equivalent (~20%), neither the PTX-PEG conjugate or the PTX prodrug
(PTX- EMCH) inhibited the formation of the tubular structures under the same conditions
(Figure 9C) [25, 52] ].
c) Cell adhesion assays
We also determined the ability of the conjugates to adhere to fibrinogen in order to evaluate
the targeting specificity of conjugated E-[c(RGDfK)2] binding to v3 integrin. As expected,
a solution of 50 µM of free E-[c(RGDfK)2] abrogated HUVEC adhesion by ~60%, while at
the same concentration, the free c(RADfK) peptide displayed no significant effect on the
adhesion of endothelial cells to fibrinogen. RGD-equivalent doses of PGA-PTX-E-
[c(RGDfK)2] or PTX-PEG-E-[c(RGDfK)2] inhibited the adhesion by 70% and 60%,
respectively (Figure 9B). Free PTX, PTX-PEG, and PTX-PEG-c(RADfK) control conjugates

(at PTX-equivalent concentration) displayed negligible effects on endothelial cell adhesion.
Unexpectedly, PTX- prodrug also inhibited HUVEC adhesion by ~30% (Figure 9D).
d) Cell migration assays
To complete the evaluation of the anti-angiogenic activity, we finally assessed the effect of
conjugates on HUVEC migratory activity. We carried out all assays in non-cytotoxic PTX-
equivalent concentrations while using short incubation times. Using these conditions, we
could evaluate the anti-angiogenic effect of PTX and the conjugates while eliminating PTX
cytotoxic activity towards HUVECs.
Under these conditions, PGA-PTX-E-[c(RGDfK)2] inhibited HUVEC migration by
50% [25]; however, PTX-PEG-E-[c(RGDfK)2], PTX-PEG-[cRADfK], PTX-PEG conjugates,
and the PTX-EMCH prodrug did not significantly affect cell migratory ability (Data not
shown).
DISCUSSION
Emerging strategies for the improvement of cancer treatment include the conjugation of
potent cytotoxic chemotherapeutic drugs, such as paclitaxel (PTX) to a hydrophilic polymeric
backbone to improve pharmacokinetics and reduce harmful side effects and the targeting of
αvβ3 integrin receptors. Encouragingly, both of these strategies are separately under
evaluation in clinical trials. Additionally, the conjugation of both anti-angiogenic and
cytotoxic agents to the same polymer chain to gain a synergistic effect has also proven to be a
successful strategy. This is exemplified by our previous studies of an anti-αvβ3 and PTX bi-
specific macromolecule targeting tumor cells and their endothelial microenvironment using a
PGA polymer [25], the promising in vivo results of Doxorubicin-PEG-E-[c(RGDfK)2] [38],

and the potent anti-cancer activity of an HPMA copolymer conjugated with RGDfK and
docetaxel [39].
Our aim in this manuscript was to study the influence of the polymeric backbone and
the linker in the anti-angiogenic and cytotoxic performance of conjugates. In addition to the
above-mentioned PGA-PTX-E-[c(RGDfK)2] conjugate, other conjugates using the same
rational design and bearing different backbones have been reported [53]. However, to the best
of our knowledge, there are no published studies that directly compare conjugates containing
these particular polymeric chains under the same conditions.
In this manuscript, we have synthesized, characterized, and compared the cytotoxic
activity of PGA-PTX, PEG-PTX, and HPMA copolymer-PTX (Figures 1, 2, and 3,
respectively). The choice of linker between the polymer and the drug also represents an
important decision in the total design of a nanomedicine/polymer therapeutic. While PGA is
biodegradable and displays adequate drug release kinetics in the presence of cathepsin B,
without the need of any specific linker other than a simple ester bond, the methacrylic
backbone of HMPA copolymer is not degradable. Therefore, HPMA copolymerized during
synthesis included a GFLG sequence within the side chains. Studies have reported that the
GFLG tetrapeptide is recognized and hydrolyzed by cathepsin B to ensure sustainable and
time-dependent drug release within the tumor cell (Figure 3). PEG is also non-degradable,
and for PEG, we chose an established hydrolyzable linker, 6-maleimidocaproyl hydrazine,
which reacts with PTX to yield a PTX-EMCH prodrug (6). This prodrug contains an acid-pH
labile hydrazone bond in its structure and can be linked to thio-modified PEG by a thio-
maleimide reaction (Figure 2). We next compared the three conjugates in terms of drug
loading, size, drug release, and anti-cancer in vitro activity (Table 2).

Table 2. Comparison of PEG, PGA and HPMA copolymer- Drug conjugates
Conjugate
Linker
Trigger
BD†
Size§
Max. Drug Release¶
IC50 (nM)*
MDA-MB-231
4T1
150
9000
1000
PGA-PTX
PEG-PTX
HPMA-PTX
Amide/Ester
Hydrazide
GFLG
Cath B / pH
pH
Cath B
Y
N
N
6.8
4.9
19.51
16
20
13
80
8
150
† stands for Biodegradability of the polymeric backbone. § Hydrodynamic diameter
determined by dynamic light scattering, in nm. ¶ Percentage of drug released after 48 h. * IC50
values were calculated at PTX-equivalent concentrations.
The first requirement for a biocompatible polymer-drug conjugate is solubility in
water. Free PTX only dissolves in organic solvents and it is co-administered in a formulation
together with CremophorEL, which can cause irritation and hypersensitivity [32].
Conjugation of PTX to our polymers yielded water-soluble nanoconstructs up to, at least, 10
mg/mL.
PGA-PTX (25 kDa) and PEG-PTX (11 kDa) conjugates demonstrated a similar
hydrodynamic diameter corresponding to monomolecular micelles (5-7 nm). HMPA
copolymer-PTX (34.1 kDa) displayed a degree of aggregation (19 nm), which slows down
target cell internalization.
The multivalence of PGA and HPMA enables different drug loading and in this study,
PGA conjugate loading was 4.9 mol%, which ensures 6-7 molecules of PTX/conjugate, while
HPMA copolymer loading was lower, at 1.12%, containing around 2 molecules of
PTX/conjugate. Both nano-constructs demonstrated cathepsin B-dependent drug-release
(Figure 4A and 4C) with a release after 48 h of 16% and 13% of total drug, respectively, with
minimal release at pH 5.5. In contrast, as every chain of PEG contains a PTX prodrug, the
release at 48 h in an acid lysosomal-mimicking pH was ~20% (Figure 4B).
We tested the cell proliferation inhibition properties of the three conjugates in two
breast cancer cell lines: human MDA-MB-231 and murine 4T1 adenocarcinomas. Curiously,
at equivalent doses of PTX, PEG-PTX displayed 10-fold greater potency than the PGA-PTX
conjugate in inhibiting MDA-MB-231 proliferation (IC50 = 8 versus IC50 = 80 nM).

Conversely, we observed opposite findings with 4T1 (IC50 = 150 versus IC50 = 9000 nM for
PGA and PEG, respectively). This observation might be explained by the difference in
cathepsin B expression in different cells [51] and pH levels, which affect the balance between
the kinetic constants of enzymatic degradation and pH-dependent hydrolysis. Our findings
with the HPMA copolymer-PTX suggested lower potency when compared to PGA-PTX in
both cell lines. It is possible that both the PGA and HPMA backbone may wrap PTX in a
very efficient fashion and, therefore, access of cathepsin B to the conjugate core may be very
restricted. However, while HPMA is a non-degradable methacrylate, PGA can be
progressively degraded by cathepsin B and this way, PTX may be released in a more rapid
manner (Figure 4).
The main conclusion of this part of our work was that the ability of the conjugates to
inhibit cell growth, at PTX-equivalent concentrations, depends amongst other characteristics,
on the cell type and the cathepsin B level of expression in every type of cell and not only on
the architecture or drug loading of the conjugates. Therefore, we recommend cathepsin B
level quantification of neoplastic cells for each patient in order to inform on the
administration of the most relevant polymer therapeutic construct. These conclusions are in
line with the European Medicines Agency (EMA) recommendations towards the need of
personalized medicine [54].
We next focused our interest on the introduction of RGD-containing molecules to
actively drive our conjugates towards endothelial cells as part of an anti-angiogenic strategy.
We conjugated E-[c(RGDfK)2] to PGA at an equimolar ratio with respect to PTX (around 5
mol% each, Figure 1) to enable a direct comparison between the multivalent PGA-PTX-E-
[c(RGDfK)2] and the monovalent PTX-PEG-E-[c(RGDfK)2] (Figure 2). Although the
introduction of the bulky and charged E-[c(RGDfK)2] molecule can double the size of the