PEG-polyaminoacid based micelles for controlled release of doxorubicin: Rational design, safety and efficacy study

Article abstract of DOI:10.1016/j.jconrel.2021.05.010

A library of amphiphilic monomethoxypolyethylene glycol (mPEG) terminating polyaminoacid co-polymers able to self-assemble into colloidal systems was screened for the delivery and controlled release of doxorubicin (Doxo). mPEG-Glu/Leu random co-polymers were generated by Ring Opening Polymerization from 5 kDa mPEG-NH₂ macroinitiator using 16:0:1, 8:8:1, 6:10:1, 4:12:1 γ-benzyl glutamic acid carboxy anhydride monomer/leucine N-carboxy anhydride monomer/PEG molar ratios. Glutamic acid was selected for chemical conjugation of Doxo, while leucine units were introduced in the composition of the polyaminoacid block as spacer between adjacent glutamic repeating units to minimize the steric hindrance that could impede the Doxo conjugation and to promote the polymer self-assembly by virtue of the aminoacid hydrophobicity. The benzyl ester protecting the γ-carboxyl group of glutamic acid was quantitatively displaced with hydrazine to yield mPEG_5kDa-b-(hydGlu_m-r-Leu_n). Doxo was conjugated to the diblock co-polymers through pH-sensitive hydrazone bond. The Doxo derivatized co-polymers obtained with a 16:0:1, 8:8:1, 6:10:1 Glu/Leu/PEG ratios self-assembled into 30–40 nm spherical nanoparticles with neutral zeta-potential and CMC in the range of 4-7 μM. At pH 5.5, mimicking endosome environment, the carriers containing leucine showed a faster Doxo release than at pH 7.4, mimicking the blood conditions. Doxo-loaded colloidal formulations showed a dose dependent cytotoxicity on two cancer cell lines, CT26 murine colorectal carcinoma and 4T1 murine mammary carcinoma with IC₅₀ slightly higher than those of free Doxo. The carrier assembled with the polymer containing 6:10:1 hydGlu/Leu/PEG molar ratio {mPEG_5kDa-b-[(Doxo-hydGlu)₆-r-Leu₁₀]} was selected for subsequent in vitro and in vivo investigations. Confocal imaging on CT26 cell line showed that intracellular fate of the carrier involves a lysosomal trafficking pathway. The intratumor or intravenous injection to CT26 and 4T1 subcutaneous tumor bearing mice yielded higher antitumor activity compared to free Doxo. Furthermore, mPEG_5kDa-b-[(Doxo-hydGlu)₆-r-Leu₁₀] displayed a better safety profile when compared to commercially available Caelyx.

Full text of DOI:10.1016/j.jconrel.2021.05.010

Journal of Controlled Release 335 (2021) 21–37
Contents lists available at ScienceDirect
Journal of Controlled Release
journal homepage: www.elsevier.com/locate/jconrel
PEG-polyaminoacid based micelles for controlled release of doxorubicin:
Rational design, safety and efficacy study
Silvia Brunatoa, Francesca Mastrottoa, Federica Bellatoa, Chiara Bastiancichb,
Alessandra Travanutc, Mariangela Garofaloa, Giuseppe Mantovanic, Cameron Alexanderc,
,
Veronique Preatb, Stefano Salmasoa *, Paolo Calicetia
a Department of Pharmaceutical and Pharmacological Sciences, University of Padova, via F. Marzolo 5, 35131 Padova, Italy
b
´
Universite catholique de Louvain, Louvain Drug Research Institute, Advanced Drug Delivery and Biomaterials, Avenue Mounier 73, 1200 Brussels, Belgium
c Molecular Therapeutics and Formulations Division, School of Pharmacy, University of Nottingham, University Park, Nottingham NG7 2RD, UK
A R T I C L E I N F O
A B S T R A C T
Keywords:
A library of amphiphilic monomethoxypolyethylene glycol (mPEG) terminating polyaminoacid co-polymers able
to self-assemble into colloidal systems was screened for the delivery and controlled release of doxorubicin
(Doxo). mPEG-Glu/Leu random co-polymers were generated by Ring Opening Polymerization from 5 kDa mPEG-
NH2 macroinitiator using 16:0:1, 8:8:1, 6:10:1, 4:12:1 γ-benzyl glutamic acid carboxy anhydride monomer/
leucine N-carboxy anhydride monomer/PEG molar ratios. Glutamic acid was selected for chemical conjugation
of Doxo, while leucine units were introduced in the composition of the polyaminoacid block as spacer between
adjacent glutamic repeating units to minimize the steric hindrance that could impede the Doxo conjugation and
to promote the polymer self-assembly by virtue of the aminoacid hydrophobicity.
Polyaminoacids
Doxorubicin
pH-sensitive drug carriers
Polymeric micelles
Controlled release
The benzyl ester protecting the γ-carboxyl group of glutamic acid was quantitatively displaced with hydrazine
to yield mPEG5kDa-b-(hydGlum-r-Leun). Doxo was conjugated to the diblock co-polymers through pH-sensitive
hydrazone bond. The Doxo derivatized co-polymers obtained with a 16:0:1, 8:8:1, 6:10:1 Glu/Leu/PEG ratios
self-assembled into 30–40 nm spherical nanoparticles with neutral zeta-potential and CMC in the range of 4-7
μ
M. At pH 5.5, mimicking endosome environment, the carriers containing leucine showed a faster Doxo release
than at pH 7.4, mimicking the blood conditions. Doxo-loaded colloidal formulations showed a dose dependent
cytotoxicity on two cancer cell lines, CT26 murine colorectal carcinoma and 4T1 murine mammary carcinoma
with IC50 slightly higher than those of free Doxo. The carrier assembled with the polymer containing 6:10:1
hydGlu/Leu/PEG molar ratio {mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10]} was selected for subsequent in vitro and in
vivo investigations. Confocal imaging on CT26 cell line showed that intracellular fate of the carrier involves a
lysosomal trafficking pathway. The intratumor or intravenous injection to CT26 and 4T1 subcutaneous tumor
bearing mice yielded higher antitumor activity compared to free Doxo. Furthermore, mPEG5kDa-b-[(Doxo-
hydGlu)6-r-Leu10] displayed a better safety profile when compared to commercially available Caelyx®.
1. Introduction
example, liposomal formulation of doxorubicin (Doxo) Doxil®,
approved in the US in 1995 for the treatment of ovarian cancer and
Nanosized drug delivery systems have emerged as a strategy to
enhance the therapeutic performance of anticancer drugs as they
improve their biopharmaceutical features and provide for accumulation
and release within tumor tissue, thus reducing systemic side effects due
to unspecific drug biodistribution [1]. Accordingly, successful results
have been obtained with several anticancer drugs loaded in nanosized
formulations including liposomes, micelles and nanoparticles [2,3]. For
AIDS-related Kaposi's sarcoma, provided a number of advantages mostly
correlated to safety improvement (reduced cardiotoxicity, nausea and
vomiting, as well as less myelosuppression), while general consensus has
not been reached concerning the increase survival in patients treated
with liposomal Doxo compared to the free drug [4]. Therefore, there is
still need of systems that can selectively target the disease site and
enhance the therapeutic outcome. For such a reason, nanovectors must
* Corresponding author.
E-mail address: stefano.salmaso@unipd.it (S. Salmaso).
https://doi.org/10.1016/j.jconrel.2021.05.010
Received 27 July 2020; Received in revised form 7 May 2021; Accepted 9 May 2021
Available online 11 May 2021
0168-3659/© 2021 Elsevier B.V. All rights reserved.

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
be sharply engineered to yield tailored structures that selectively deliver
anticancer drugs to the disease site by active or passive process and
release the drug according to controlled mechanisms.
Penicillin G sodium and Streptomycin sulfate. 4′,6-diamidino-2-phe-
nylindole dihydrochloride (DAPI), 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT), Corning® T-75 culture flasks, Rat
Serum and Triton® X-100 were purchased from Sigma-Aldrich (St.
Louis, MO, USA). Rat anti-mouse CD107a (LAMP-1) and chicken anti-rat
IgG(H + L) cross-adsorbed secondary antibody Alexa Fluor 488 conju-
gated were obtained by Thermoscientific (Waltham, MA, USA).
CT26 colon carcinoma cell line and 4T1 breast carcinoma cell line
were purchased from ATCC cell bank (Manassas, VA, USA). Six-weeks
old BALB/c female mice were provided by Janvier Labs (Le Genest-St-
Isle, France).
Polyaminoacid-based vectors are emerging as promising drug vehi-
cles due to their high biocompatibility and biodegradability [5]. Indeed,
the clearance of the components of drug carriers from the body is a
requisite for approval by regulatory agencies and clinical translation of
this class of therapeutic systems. Thus, polyaminoacid-based carriers
may represent a safer option to develop drug nanocarriers. In addition,
depending on their composition, the reactive pendant groups of the side-
chain of polyaminoacids, such as aspartic or glutamic acids, can be
exploited for the conjugation of drugs or diagnostic agents, or they can
be functionalized with stimuli-responsive linkers to endow drug release
by tissue and intracellular microenvironmental triggers e.g. enzymatic
pools, pH, redox potential [6]. Systems exploiting enzymes such as ca-
thepsins, which can cleave polyaminoacid based drug conjugates and
are overexpressed in tumor tissues, are under clinical development for
local controlled drug release [7]. Additionally, biopharmaceutical fea-
tures of this class of nanovectors can be properly tailored by tuning the
polymer composition. For example, the conjugation of hydrophilic
polymers (such as PEG) to hydrophobic polyaminoacids can yield
macromolecules that self-assemble into micelles, which provide for high
drug loading, prolonged blood circulation, and enhanced passive tumor
accumulation by enhanced permeation and retention (EPR) effect
[8–12]. However, the hydrophobic/hydrophilic balance of PEG-
polyaminoacid block-co-polymers must be sharply optimized, as it dic-
tates the stability of the micelles in biological environments, as well as
the release of physically encapsulated or chemically conjugated drug
molecules. The chemical and physical drug loading in the polymeric
micelles may offer the opportunity for combination therapy further
expanding the versatility and therapeutic potential of these systems
[13].
2.1. Synthesis of mPEG5kDa-NH2
2.1.1. Synthesis of mPEG5kDa-allyl carbamate
mPEG5kDa-OH (20.0 g, 4.02 mmol) was dissolved in chloroform (15
mL) and added of allyl isocyanate (1.30 g, 15.6 mmol) and triethylamine
(2.80 mL, 2.02 g, 20.0 mmol). The mixture was maintained under stir-
ring at room temperature for 18 h. The polymer was isolated by pre-
cipitation in diethyl ether (40 mL), centrifuged at 5000 rpm for 5 min,
rinsed three times with diethylether (40 mL) and finally dried under
reduced pressure. mPEG5kDa-allyl carbamate (19.5 g, 3.82 mmol) was
obtained as a white solid in 95% yield. The product was analyzed by 1H
NMR and MALDI-ToF mass spectroscopy using SuperDHB as the matrix.
MALDI-ToF spectrum showed the expected set of signals for the Na+
and K+ adducts of mPEG5kDa allyl carbamate. Using the chains with 110
oxyethylene repeating units as
a
representative example
-
C
225H449NO1⋅ 12Na+: theor. 4981.0, found 4980.8; C225H449NO1⋅ 12K+:
theor. 4996.9, found 4996.9. Complete spectrum is shown in Fig. SI-1
(Supplementary Information).
1
–
H NMR (400 MHz, CDCl ): δ 5.82 (m, 1H, CH CH ), 5.21 (dq, J =
–
3
2
–
–
–
–
In this work, we generated a library of amphiphilic di-block co-
polymers composed by a hydrophilic block of methoxypolyethylene
glycol (mPEG) and a polyaminoacid block with different ratios of glu-
tamic acid-γ-hydrazide (hydGlu) and leucine (Leu), to investigate the
polymer structure/drug loading relationship and efficiency of drug de-
livery to the tumor while reducing systemic drug exposure [14].
While drug delivery systems reported in the literature derive from
rather sophisticated conceptualization in response to the complexity of
tumor, we explored here a rational approach to achieve high drug
loading in the carrier, efficient drug delivery to the tumor with reduced
systemic Doxo exposure using biodegradable and non-toxic materials.
The antitumor efficacy and safety profile of selected micellar formula-
tions was estimated comparatively using two solid subcutaneous tumor
models, CT26 colorectal carcinoma and 4T1 breast tumor, with that of a
commercially available liposomal formulation of Doxo.
17.1, 1.6 Hz, 1H, CH CHH, 5.13 (dq, J = 10.2, 1.4 Hz, 1H, CH CHH,
4.97 (bs, 1H, (C=O)NH), 4.22 (m, 2H, -CH2OC(O)), 3.63 (s, 476H,
ꢀ (OCH2CH2)n), 3.36 (s, 3H, CH3O-).
2.1.2. Synthesis of mPEG5kDa-NH2
mPEG5kDa-allyl carbamate (10.1 g, 1.99 mmol) was solubilized in
methanol (10 mL). Cysteamine hydrochloride (1.13 g, 9.95 mmol) and
2,2′-dimethoxy-2-phenylacetophenone (DPAP) (0.102 g, 0.400 mmol)
were dissolved in 2 mL and 1 mL of methanol, respectively, and then
sequentially added to the mPEG5kDa-allyl carbamate solution. The final
mixture was aliquoted in vials and irradiated with UV light (λ = 350 nm,
1
36 W) for 3 h. When allyl signals could no longer be detected by H
NMR, the aliquots were pulled together, and methanol was removed
under vacuum. The resulting residue was solubilized in water (100 mL)
and extracted 3 times with ethyl acetate (100 mL) to remove organic
species deriving from the decomposition of DPAP. Next, the aqueous
phase was extracted 2 times with a 3:1 vol:vol dichloromethane: iso-
propanol (100 mL) and the organic phase was dried under reduced
pressure to obtain the mPEG5kDa-cysteamine (mPEG5kDa-NH2) as white
powder (9.95 g, 1.92 mmol), (98% yield). MALDI-TOF analysis of
mPEG5kDa-NH2 was performed using SuperDHB as a matrix.
2. Materials and methods
5.0 kDa methoxy-polyethyleneglycol (mPEG5kDa-OH) was purchased
from Sigma-Aldrich (St. Louis, USA). All the other chemical reagents
including salts and analytical grade solvents were purchased from
Sigma-Aldrich (St. Louis, MO, USA), VWR (Milan, Italy) or Carlo Erba
(Milan, Italy). mPEG5kDa-OH was azeotropically dried with toluene
under reduced pressure prior to use, while all other products were used
without further purification. Doxo hydrochloride was purchased from
LC laboratories (Woburn, MA, USA). The water used for the preparation
MALDI-ToF spectrum showed the expected set of signals for the H+,
Na+ and K+ adducts of this polymer. Using the chains with 110 oxy-
ethylene repeating units as
a
representative example
-
C227H456N2O112S⋅H+:
theor.
theor.
5036.0,
5058.0,
found
found
5035.8;
5057.9;
C
C
227H456N2O112S⋅Na+:
of solutions was “ultrapure” water (milliQ-grade, 0.06
μ
S cmꢀ 1), pro-
227H456N2O112S⋅K+: theor. 5073.9, found 5073.8. Complete spectrum
duced with a Millipore Milli-Q purification system (Burlington, MA,
USA). Dialysis membranes with a molecular weight cut off (MWCO) of
3.5–5 kDa and Flot-a-Lyzers with MWCO 3.5–5 kDa were purchased
from Prodotti Gianni (Milan, Italy).
is shown in Fig. SI-1.
SEC analysis (DMF + 0.1% LiBr): Mn,SEC = 7.3 kDa; Đ = 1.09.
1H NMR (400 MHz, CDCl3): δ 4.22 (m, 2H, -CH2OC(O)), 3.63 (m,
476H, ꢀ (OCH2CH2)n), 3.37 (s, 3H, CH3O), 3.29 (q, J = 5.9 Hz, 2H, (O)
CNHCH2), 3.17 (bs, 2H, CH2NH2), 2.94 (t, J = 7.3 Hz, 2H,
SCH2CH2NH2), 2.63 (t, J = 7.1 Hz, 2H, CH2 CH2CH2S), 1.90 (m, 2H,
CH2CH2S).
All the media and materials used for cell culture were obtained from
Gibco (Life Technologies, Waltham, MA, USA), including phosphate
buffered saline (PBS), RPMI 1640 Medium, Fetal Bovine Serum (FBS),
22

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
2.2. Synthesis of γ-benzyl-glutamic acid N-carboxyanhydride (BnOGlu-
2.4.1. Representative example: synthesis of mPEG5kDa-b-(BnOGlu8-r-Leu8)
mPEG5kDa-NH2 (0.500 g, 0.0970 mmol) was dissolved in toluene (10
mL), and the solvent was removed under reduced pressure to azeo-
tropically eliminate traces of moisture in the initial mPEG5kDa-NH2
starting material. Anhydrous DMF (12 mL) was added to mPEG5kDa-NH2
under argon atmosphere, the flask was sonicated for five minutes to
completely solubilize the macroinitiator, then the resulting solution was
degassed by argon bubbling for 30 min. For the synthesis of mPEG5kDa-b-
(BnOGlu8-r-Leu8), Glu-NCA (0.255 g, 0.871 mmol) and Leu-NCA (0.183
g, 1.16 mmol) were poured into a round-bottomed flask under argon
atmosphere, then the mPEG5kDa-NH2 initiator solution was added via
cannula, to give a mPEG5kDa-NH2:BnOGlu:Leu feed ratio of 1:9:12, and
the resulting mixture was stirred at room temperature. The reaction
solution was bubbled with argon twice a day for 30 min. After 7 days the
mixture was added dropwise under stirring to 100 mL of Et2O, and the
resulting suspension centrifuged for five minutes at 5000 rpm. The su-
pernatant was discarded, the precipitate was dissolved in CH2Cl2 (5 mL)
and then the side products, consisting in poly(BnOGlum-r-Leun) oligo-
mers were precipitated in methanol. After centrifugation for 5 min at
5000 rpm, the supernatant was recovered and dropwise added to 100
mL of Et2O to isolate, after filtration, mPEG5kDa-b-(BnOGlu8-r-Leu8) as a
white solid. The precipitate was dried for 24 h under reduced pressure to
remove traces of solvents, then characterized by 1H NMR and SEC.
mPEG5kDa-b-(BnOGlu8-r-Leu8) 1H NMR (400 MHz, CDCl3): δ 7.28 (m,
40H, CH aromatic), 5.02 (s, 16H, CH2Ph), 4.01 (m, 16H, C(O)CH, C(O)
CHNH2), 3.64 (s, 476H, ꢀ (OCH2CH2)n), 3.38 (s, 3H, CH3O), 2.70–2.17
(m, 58H, NH(CH2)3S(CH2)2, CH2CH2C(O), CH2CH(CH3)2), 0.86 (m,
48H, CH(CH3)2). SEC analysis (DMF + 0.1% LiBr): Mn = 10.5 kDa; Đ =
1.11.
NCA)
The protocol for γ-benzyl-glutamic acid NCA synthesis was adapted
from Markland et al. [15] and Williams et al. [16] γ-benzyl-glutamic
acid (19.9 g, 83.8 mmol) was suspended in tetrahydrofuran (THF, 200
mL). α-pinene (33.0 mL; 20.9 mmol) and solid triphosgene (9.14 g, 30.8
mmol) were sequentially added to the suspension under nitrogen at-
mosphere. The resulting mixture was heated under reflux at 56 ◦C until
the suspension turned yellow and clear (90 min). The solution was then
cooled down to room temperature and bubbled with nitrogen for 4 h to
remove gaseous co-product and potential traces of unreacted phosgene.
An aqueous 1 M NaOH trap was connected to the system during this
procedure to neutralize the gaseous species removed from the reaction
mixture. The solution was then concentrated to ~70 mL under reduced
pressure, and petroleum ether was added (50 mL). The resulting solution
was transferred into a conical flask, sealed, and left to crystallize at
ꢀ 18 ◦C overnight. γ-benzyl-glutamic acid NCA (16.3 g, 61.7 mmol) was
isolated by filtration in 82% yield.
1H NMR (400 MHz, CDCl3): δ 7.35 (m, 5H, CH aromatic), 6.64 (s, 1H,
NHCH), 5.14 (s, 2H, OCH2Ph), 4.38 (t, J = 6.1 Hz, 1H, NHCH), 2.59 (t, J
= 6.9 Hz, 2H, CH2C(O)), 2.26 (m, 1H, CHCHH), 2.13 (m, 1H, CHCHH).
13C NMR (101 MHz, CDCl3): δ 172.52 (1C, C(O)CH), 169.48 (1C, C
(O)O), 151.97 (1C, C(O)NH), 135.35 (1C, C aromatic), 128.85 (2C, CH
aromatic), 128.73 (2C, CH aromatic), 128.50 (1C, CH aromatic), 67.25
(1C, OCH2Ph), 57.07 (1C, CHCH2), 29.98 (1C, CH2CH2), 27.05 (1C,
CH2CH2).
FT-IR:
υ 3247, 2320, 1862, 1772, 1718, 1493, 1396, 1251, 1183,
1110, 796, 961 cmꢀ 1
.
Yield of purification process 75% (mol/mol).
2.3. Synthesis of leucine N-carboxyanhydride (Leu-NCA)
2.4.2. mPEG5kDa-BnOGlu16
mPEG5kDa-NH2 (2.00 g, 0.387 mmol), BnOGlu-NCA (1.78 g, 6.77
mmol, 17.5 eq), DMF (41 mL). After 5 days additional BnOGlu-NCA
(1.12 g, 4.25 mmol, 11 eq) were added, and the reaction stopped at
day 7. 1H NMR (400 MHz, CDCl3): δ 7.43–7.14 (m, 80H, CH aromatic),
5.02 (s, 32H, CH2Ph), 4.19–3.92 (m, 16H, C(O)CH), 3.64 (s, 476H,
(OCH2CH2)n), 3.38 (s, 3H, CH3O), 2.62–2.26 (m, 74H, NH(CH2)3S
(CH2)2, CH2CH2C(O). SEC analysis (DMF + 0.1% LiBr): Mn = 12.2 kDa;
Đ = 1.07. Yield of purification process 75% (mol/mol).
This NCA monomer was prepared as described by Smeets et al. [17].
Leucine (8.02 g, 61.0 mmol) was suspended in anhydrous THF (100 mL).
α
-pinene (25.0 mL, 157 mmol) and solid triphosgene (6.70 g, 22.2
mmol) were sequentially added under nitrogen atmosphere. The sus-
pension was heated under reflux at 64 ◦C until it turned yellow and clear
(~ 30 min). The solution was then cooled down to room temperature
and bubbled with nitrogen following the same procedure described
above for the synthesis of γ-benzyl-glutamic acid NCA. The solvent was
then removed under reduced pressure to yield a yellow oily residue.
Petroleum ether (300 mL) was added to induce the precipitation of Leu-
NCA as a white solid. This precipitate was filtered, washed with ice-cold
petroleum ether (3 × 25 mL), and dried under reduced pressure, to give
leucine N-carboxyanhydride (Leu-NCA) as a white solid (7.65 g, 48.6
mmol) with a recovery yield of 95% (mol/mol).
2.4.3. mPEG5kDa-b-(BnOGlu6-r-Leu10
)
mPEG5kDa-NH2 (2.00 g, 0.387 mmol), BnOGlu-NCA (1.02 g, 4.26
mmol, 10 eq), Leu-NCA (1.27 g, 8.13 mmol, 21 eq), DMF (73 mL). After
5 days additional BnOGlu-NCA (1.32 g, 13 eq) and Leu-NCA (1.03 g, 17
eq) were added, and the reaction stopped at day 7. 1H NMR (400 MHz,
CDCl3): δ 7.32 (m, 30H, CH aromatic), 5.03 (s, 18H, CH2Ph), 4.02 (m,
16H, C(O)CH, C(O)CHNH2), 3.64 (s, 476H, ꢀ (OCH2CH2)n), 3.38 (s, 3H,
CH3O), 2.71–2.17 (m, 54H, NH(CH2)3S(CH2)2, CH2CH2C(O), CH2CH
(CH3)2), 0.88 (m, 60H, CH(CH3)2). SEC analysis (DMF + 0.1% LiBr): Mn
= 10.4 kDa; Đ = 1.09. Yield of purification process 72% (mol/mol).
1H NMR (400 MHz, CDCl3): δ 6.43 (bs, 1H, NHCH) 4.33 (dd, J = 8.9,
4.0 Hz, 1H, NHCH), 1.82 (m, 2H, CH2CH), 1.68 (m, 1H, CH(CH3)2), 0.99
(m, 6H, CH(CH3)2).
13C NMR (101 MHz, CDCl3): δ 170.25 (1C, C(O)CH), 153.38 (1C, C
(O)NH), 56.33 (1C, CHCH2), 40.88 (1C, CH2CH), 25.06 (1C, CH(CH3)2),
22.79 and 21.59 (2C, CH(CH3)2).
2960, 1798, 1750, 1468, 1117, 1079 cmꢀ 1
.
2.4.4. mPEG5kDa-b-(BnOGlu4-r-Leu12
)
FT-IR:
υ
mPEG5kDa-NH2 (2.00 g, 0.387 mmol), BnOGlu-NCA (0.713 g, 2.71
mmol, 7 eq), Leu-NCA (1.58 g, 10.1 mmol, 26 eq), DMF (75 mL). After 5
days additional BnOGlu-NCA (0.682 g, 7 eq) and Leu-NCA (1.09 g, 18
eq) were added, and the reaction stopped at day 7. 1H NMR (400 MHz,
CDCl3): δ 7.32 (m, 20H, CH aromatic), 5.03 (s, 8H, CH2Ph), 4.02 (m,
16H, C (O)CH, C(O)CHNH2), 3.64 (s, 476H, (OCH2CH2)n), 3.38 (s, 3H,
CH3O), 2.72–2.17 (m, 50H, NH(CH2)3S(CH2)2, CH2CH2C (O), CH2CH
(CH3)2), 0.89 (m, 72H, CH(CH3)2). SEC analysis (DMF + 0.1% LiBr): Mn
= 9.9 kDa; Đ = 1.08. Yield of purification process 77% (mol/mol).
2.4. Synthesis of mPEG5kDa-b-(BnOGlum-r-Leun) random co-polymers
This procedure was adapted from the protocol reported by Zhao et al.
[18]. The mPEG5kDa-b-(BnOGlum-r-Leun) co-polymer library was syn-
thetized by Ring Opening Polymerization (ROP) using mPEG5kDa-NH2 as
the initiator, and Glu-NCA and Leu-NCA as the monomers at different
relative molar ratios. For the synthesis of the co-polymer library, the
amount of monomers to be used was calculated considering targeted
monomer conversions of 90% and 70% for Glu-NCA and Leu-NCA (Leu-
NCA was found to react more slowly than Glu-NCA in these co-
polymerization reactions), respectively, based on preliminary poly-
merization studies (Data not shown).
23

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
2.5. Synthesis mPEG5kDa-b-[(γ-hydrazide-glutamic acid)m-r-leucinen]
unreacted Doxo, the obtained red powder was dissolved in 1:1 v/v
DMSO:[10 mM Phosphate buffer, pH 7.8 (PB)] mixture, transferred into
a 3.5–5 kDa MWCO dialysis bag and dialyzed against the same solvent
mixture for 24 h, replacing the solvent four times. The presence of the
unreacted drug was monitored by thin layer chromatography (TLC) run
with MeOH after slides preconditioning with 5% v/v triethylamine
(TEA) in MeOH. Dialysis was carried out until complete disappearance
of free Doxo in the dialysis bag solution. Afterwards, the composition of
the dialysis medium was gradually varied from a 1:1 to 0:1 v/v DMSO:
PB ratio in six hours to remove the organic solvent and to induce the self-
assembling of the polymer-drug conjugates. Finally, the dialysis medium
was changed to deionized water (DI water) adjusted to pH 7.8 with
ammonia and dialyzed for further three hours to remove salts. As a
control of the assembling behavior, the same dialysis procedure was
performed on drug free polymers, in order to evaluate the assembly of
micelles in absence of the linked drug. The polymeric self-assembled
micelle suspension was then freeze-dried and stored at ꢀ 20 ◦C until
reconstitution, which was performed by dissolving micelle powders in
the required solvent and stirred for three hours on an orbital shaker.
The conjugation yield was assessed by calculating the number of
Doxo molecules per polymer chain. The drug/polymer chain molar ratio
was estimated according to the spectrophotometric absorbance of the
conjugated drug at λ = 488 nm in PB (calibration line: y = 10.416× +
0.0196, R2 = 0.9991), and the iodine assay for PEG detection [20] (λ =
535 nm, calibration curve: y = 0.0204× – 0.0052, R2 = 0.9918). The
[mPEG5kDa-b-(hydGlum-r-Leun)]
The protocol of polymer deprotection was adapted from Bae et al.
[11] mPEG5kDa-b-(BnOGlu8-r-Leu8) (0.500 g, 0.0639 mmol, 0.511 mmol
of benzyl ester units) was dissolved in toluene (10 mL) and the solvent
was removed under reduced pressure to azeotropically remove mois-
ture. Anhydrous DMF (3.5 mL) was then added to the solution under
argon atmosphere, followed by dropwise addition of hydrazine hydrate
(0.631 g, 19.7 mmol). The reaction was carried out at 40 ◦C under
stirring for 48 h. The resulting gel-like solution was diluted with addi-
tional 3.5 mL of DMF to decrease viscosity. The resulting solution was
precipitated in Et2O (100 mL) and the white solid was isolated by
centrifugation. Then, the product was re-dissolved in milliQ water (50
mL) and purified by dialysis (MWCO 3.5 kDa) for two days using 5 L of
deionized water as receiving medium. The polymer containing solution
was freeze-dried to obtain mPEG5kDa-b-(hydGlu8-r-Leu8) as a white solid
(0.389 g, 0.0470 mmol, 74% mol/mol yield). The γ-hydrazide Glu:Leu
monomer ratio was estimated by 1H NMR. The number of hydrazide
groups per polymer chain was calculated according to the modified
Snyder's assay for the evaluation of hydrazides [19] (λ = 500 nm, cali-
bration curve: y = 11.997× – 0.0318. R2 = 0.9982), and the iodine assay
for PEG detection [20] (λ = 535 nm, calibration curve: y = 0.0229× +
0.0207. R2 = 0.9933).
1H NMR (400 MHz, DMSO‑d6): δ 4.16 (m, 16H, C(O)CH, C(O)
CHNH2), 3.51 (s, 476H, ꢀ (OCH2CH2)n), 3.24 (s, 3H, CH3O), 2.05–1.45
(m, 58H, NH(CH2)3S(CH2)2, CH2CH2C(O), CH2CH(CH3)2), 0.86 (m,
48H, CH(CH3)2). SEC analysis (DMF + 0.1% LiBr): Mn = 8.9 kDa; Đ =
1.25. Number of hydrazides per polymer chain: 8.3.
same protocol was used to conjugate Doxo to mPEG5kDa-hydGlu16
,
mPEG5kDa-b-(hydGlu6-r-Leu10), mPEG5kDa-b-(hydGlu4-r-Leu12).
mPEG5kDa-(Doxo-hydGlu)16
.
Conjugation yield 44% (mol/mol), 7.0 Doxo molecules/polymer
The same procedure was applied to deprotect the other polymers of
the library and generate mPEG5kDa-hydGlu16, mPEG5kDa-b-(hydGlu6-r-
Leu10), mPEG5kDa-b-(hydGlu4-r-Leu12).
chain.
Doxo content: 32% (w/w).
mPEG5kDa-b-[(Doxo-hydGlu)8-r-Leu8].
mPEG5kDa-hydGlu16: 1H NMR (400 MHz, D2O): δ 4.34 (m, 16H, C(O)
CH), 3.74 (s, 476H, (OCH2CH2)n), 3.42 (s, 3H, CH3O-), 2.75–1.81 (m,
74H, NH(CH2)3S(CH2)2, CH2CH2C(O). SEC analysis (DMF + 0.1% LiBr):
Mn = 12.2 kDa; Đ = 1.03.
Conjugation yield 70% (mol/mol), 5.6 Doxo molecules/polymer
chain.
Doxo content: 29% (w/w).
mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10].
Number of hydrazide per polymer chain: 16.1. Yield of purification
process 68% (mol/mol) (0.326 g, 0.0424 mmol).
Conjugation yield 88% (mol/mol), 5.3 Doxo molecules/polymer
chain.
mPEG5kDa-b-(hydGlu6-r-Leu10): 1H NMR (400 MHz, DMSO‑d6) δ 4.16
(m, 16H, (C=O)CH, (C=O)CHNH2), 3.51 (s, 476H, (OCH2CH2)n), 3.24
(s, 3H, CH3O-), 2.05–1.45 (m, 54H, NH(CH2)3S(CH2)2, CH2CH2C(O),
CH2CH(CH3)2), 0.87 (m, 60H, CH(CH3)2). SEC analysis (DMF + 0.1%
LiBr): Mn = 8.8 kDa; Đ = 1.15. Number of hydrazide per polymer chain:
6.3. Yield of purification process 75% (mol/mol) (0.354 g, 0.0495
mmol).
Doxo content: 28% (w/w).
mPEG5kDa-b-[(Doxo-hydGlu)4-r-Leu12].
Conjugation yield 100% (mol/mol), 4.0 Doxo molecules/polymer
chain.
Doxo content: 23% (w/w).
2.7. Critical micelle concentration (CMC)
mPEG5kDa-b-(hydGlu4-r-Leu12): 1H NMR (400 MHz, DMSO‑d6) δ 4.07
(m, 16H, C(O)CH, C(O)CHNH2), 3.51 (s, 476H, (OCH2CH2)n), 3.24 (s,
3H, CH3O), 2.02–1.43 (m, 50H, NH(CH2)3S(CH2)2, CH2CH2C(O), CH2CH
(CH3)2), 0.87 (m, 72H, CH(CH3)2). SEC analysis (DMF + 0.1% LiBr): Mn
= 9.8 kDa; Đ = 1.20. Number of hydrazide per polymer chain: 4.2. Yield
of purification process 70% (mol/mol) (0.336 g, 0.0474 mmol).
2.7.1. CMC by pyrene assay for mPEG5kDa-b-(hydGlum-r-Leun)
The protocol by Ambrosio et al. [21] was adapted to assess the CMC
of the drug-free co-polymers mPEG5kDa-b-(hydGlum-r-Leun). 20 μL ali-
quots of a 5 mg/mL pyrene solution in acetone were transferred in vials
and the organic solvent was removed under reduced pressure. Then, 1
mL of mPEG5kDa-hydGlum-r-Leun in 10 mM phosphate buffer, 0.15 M
2.6. Synthesis and self-assembly of mPEG5kDa-b-[(Doxo-hydGlu)n-r-
Leum]
NaCl, pH 7.4 in the 0–300 μM range was added to each vial and stirred
overnight on an orbital shaker, in the dark at room temperature. Each
sample was prepared in triplicate. Afterwards, the samples were
centrifuged three times at 13,000 rpm for 3 min, and the supernatants
were analyzed by fluorescence spectroscopy with an excitation wave-
length of 335 nm and recording the emission spectra from 350 to 450
nm. The ratio between the emission intensities at 384 nm (I3) and 373
nm (I1), I3/I1, was plotted vs the polymer logarithmic concentration. The
CMC value was derived from the cross-point of the two straight lines
fitting through the points [22,23].
Conjugation of Doxo to mPEG5kDa-b-(hydGlu8-r-Leu8) was performed
using a protocol adapted from Bae et al. [11] mPEG5kDa-b-(hydGlu8-r-
Leu8) (0.300 g; 0.0416 mmol) was dissolved in an anhydrous 1:2 MeOH:
DMSO mixture (4.5 mL). Doxo hydrochloride (Doxo-HCl) (0.386 g,
0.666 mmol) was solubilized in 1.5 mL of anhydrous DMSO. The two
solutions were mixed and trifluoroacetic acid (TFA) (51 μL, 0.666 mmol)
was added as a catalyst. The reaction mixture was left under stirring in
the dark at room temperature for three days. The solution was then
dropwise added to Et2O (40 mL) and the red pellet was recovered by
centrifugation (5000 rpm, 5 min). This process was repeated 4 times and
the product was then desiccated under reduced pressure. To remove the
2.7.2. CMC by spectrofluorimetric analysis and DLS for mPEG5kDa-b-
[(Doxo-hydGlu)m-r-Leun]
1 mL of 0–50 μM of mPEG5kDa-b-[(Doxo-hydGlu)m-r-Leun] in 10 mM
24

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
phosphate buffer, 0.15 M NaCl, pH 7.4 was prepared and stirred over-
night on an orbital shaker in the dark at room temperature. Each sample
was prepared in triplicate. Then, the samples were analyzed by fluo-
rescence spectroscopy using an excitation wavelength of 490 nm and
recording the emission spectra in the λ = 510–600 nm range. The
emission intensities at λ = 555 nm was plotted vs the polymer concen-
tration and the CMC was derived from the cross-point of the two lines
fitting the points of the increasing fluorescence intensity with that fitting
the points of decreasing fluorescent intensity.
system was assessed by sampling 20 μL of the solution and tested by
spectrophotometric analysis (λmax 488 nm) using a calibration line (y =
10.416× + 0.0196, R2 = 0.9991) and molar extinction coefficient re-
ported in the literature (εM = 11,500 Mꢀ 1 cmꢀ 1 at 25 ◦C) [24]. The
percentage of released Doxo was calculated on the basis of the residual
concentration of Doxo inside the Float-A-Lyzer® G2 system with respect
to the drug concentration at time 0. The analysis was performed in
triplicate. The procedure was preliminarily validated by assessing the
release of free Doxo.
The same samples were analyzed by Malvern Dynamic Light Scat-
tering Zetasizer Ultra (Malvern, UK) equipped with optical fluorescence
filter wheel. The derived mean count rate was plotted vs the polymer
logarithmic concentrations. The CMC value were obtained from the
cross-point of the two straight lines fitting through the points.
2.11. In vitro biological studies
2.11.1. Cell cultures
CT26 murine colorectal carcinoma and 4T1 murine mammary car-
cinoma cell lines were cultured using RPMI 1640 Medium supplemented
with 10% Fetal Bovine Serum (FBS), 100 U/mL penicillin G sodium and
2.8. Size and zeta potential analysis
100 μg/mL streptomycin sulphate (complete medium). Cells were sub-
2
◦
The particle size, size distribution and zeta potential were measured
by using a Malvern Dynamic Light Scattering Zetasizer Nano (Malvern,
UK) equipped with a red laser (λ = 633 nm) at a fixed angle of 173◦ at
cultured in 75 cm culture flasks and incubated at 37 C and 5% CO2
atmosphere.
◦
25 C. DTS applications 6.12 software was used for the data analysis.
2.11.2. In vitro cytotoxicity
The particle size and size distribution were determined by using 100 μM
CT26 and 4T1 cell viability after incubation with Doxo or Doxo-
conjugated polymeric micelles was investigated by MTT (Thiazolyl
Blue Tetrazolium Bromide) assay [25]. Cells were seeded in a 96-well
mPEG5kDa-b-(hydGlum-r-Leun) and mPEG5kDa-b-[(Doxo-hydGlu)m-r-
Leun] colloidal suspensions in 10 mM phosphate, 150 mM NaCl, pH 7.4
(PBS). Analyses were performed soon after micelle preparation and after
3 min centrifugation at 3000 rpm. The size values were reported as
volume, number and intensity. The zeta-potential was obtained by using
plate at a density of 6 × 103 cells/well (200
μL/well) and grown at
37 ◦C and 5% CO2 for 24 h. Then, the medium was replaced with 100
μL
of free Doxo or Doxo-conjugated polymeric micelle suspension, at
100
μ
M
mPEG5kDa-b-(hydGlum-r-Leun) and mPEG5kDa-b-[(Doxo-
0.1–100 μM Doxo equivalent concentration range, in complete medium.
hydGlu)m-r-Leun] colloidal suspensions in 10 mM HEPES. For each
sample, DLS measurements were performed in triplicate with 10 runs
per 10-s measurement.
Cells incubated with complete medium only were used as a control.
After 48 or 72 h of incubation, the medium was discharged and cells
were gently rinsed with PBS (200 μL) and incubated with 200 μL of MTT
solution in complete medium (0.5 mg/mL) for 3 h at 37 ◦C. Afterwards,
the medium was removed and DMSO (200 L) was added to each well to
Stability of micelles over time was assessed by DLS analysis.
Colloidal dispersions of mPEG5kDa-b-(hydGlum-r-Leun) and mPEG5kDa-b-
[(Doxo-hydGlu)m-r-Leun] in PBS, pH 7.4 or 10 mM sodium acetate, 150
mM NaCl, pH 5.5 were prepared as reported above and added of 2.5% v/
μ
dissolve the formazan crystals. The absorbance at λ = 560 nm was
measured by a Thermo Fisher Scientific MultiSkan EX plate reader
(Waltham, MA, USA). Cell viability was expressed as relative percentage
of living cells with respect to the untreated ones, considered as negative
control (N = 3, n = 17).
v FBS to a final micelle concentration of 100 μM. The samples were
incubated at 37 ◦C and analyzed by DLS at scheduled time.
2.9. Transmission electron microscopy
2.11.3. Haemolysis
10 μL drop of a 100 μM polymeric micelles in MilliQ water were
Erythrocytes (RBCs) were isolated from heparinized mouse blood by
centrifugation at 500 xg for 5 min at 4 ◦C. The pellet was washed thrice
with sterile saline solution (0.9% w/v NaCl in Milli-Q water) and
resuspended in PBS, pH 7.4, to yield a 5% w/v haematocrit concentra-
placed on a homemade carbon coated copper grid, and the solvent was
allowed to dry at room temperature. Then, the samples were treated
with 1% uranyl acetate in distilled water for 5 min at room temperature
to provide for negative staining. Transmission electron microscopy an-
alyses were carried out using a FEI Tecnai G2 microscope (Hillsboro, OR,
USA). Particle size analysis was performed with ImageJ Software
(developed at the National Institute of Health, Bethesda, MD, USA). The
average size of the polymeric colloidal systems was calculated by
measuring 50 individual particles with ImageJ software version 1.51j8
(National Institutes of Health software package by Wayne Rasband;
Bethesda, MD, USA) and particle size distribution histograms were
generated.
tion. Afterward, volumes of 40
160
μL of RBC suspensions were mixed with
μ
L of mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] colloidal suspension in
PBS, pH 7.4, or Caelyx® at equivalent Doxo concentrations in the 0–600
μ
M range. Samples were prepared in quadruplicate and incubated at
◦
37 C for 1 h and then centrifuged at 500 xg for 5 min. 100
μ
L of su-
pernatants were plated in a 96-well plate and the released haemoglobin
was quantified spectrophotometrically at 405 nm using a VICTOR X3
plate reader (PerkinElmer, Waltham, MA-USA). Absorbance of
mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] colloidal suspensions in PBS at
the tested concentrations were subtracted from each sample. The hae-
molytic activity of mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] was plotted as
percentage of haemoglobin released with respect to RBC samples lysed
with 1% w/v Triton X-100. RBCs incubated with PBS were used as
negative control.
2.10. Release studies
Solutions of free Doxo and Doxo-conjugated polymeric micelles in
PBS were diluted at a Doxo equivalent concentration of 400 μM in PBS,
pH 7.4, or 10 mM sodium acetate, 150 mM NaCl, pH 5.5. The Doxo
equivalent concentration was selected to yield a co-polymer concen-
tration about 10 times higher than that of the conjugate with the highest
2.11.4. Confocal microscopy imaging
CT26 cells were seeded in 24-well plates containing glass dishes at a
CMC of the conjugates (6.7
μ
M for mPEG5kDa-(Doxo-hydGlu)16. Aliquots
density of 5 × 104 cells/well (500
μL/well) in complete medium and
(1.5 mL) of these solutions were transferred into a Float-A-Lyzer® G2
system, 3.5–5 kDa MWCO, and dialyzed against the same buffer (500
allowed to grow for 24 h under culture conditions. Afterwards, the
medium was discharged and replaced with 500
celles assembled with mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] at 5
Doxo equivalent concentration in complete medium. Cells were
μL of free Doxo or mi-
◦
mL) thermostatted with a water bath at 37 C under stirring. At fixed
μM
time points, the concentration of Doxo inside the Float-A-Lyzer® G2
25

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
incubated at 37 ◦C for 2 h or, after removal of the medium and cell wash
were administered by local intratumor injection or by tail vein intra-
venous injection. Eight groups were set: Group 1: control group (un-
with 400 μL of PBS, further incubated with complete medium for other 4
h. Afterwards, cells were gently rinsed twice with PBS (400
20 min at room temperature in the dark with 400 L of 4% w/v para-
formaldehyde (PFA) solution in PBS and rinsed with PBS (3 × 400 L).
μ
L), fixed for
treated, n = 9); Group 2: intratumor injection of Doxo (30 μL, n = 11);
μ
Group 3: intratumor injection of micelles assembled with mPEG5kDa-b-
[(Doxo-hydGlu)6-r-Leu10] (30 L, n = 10); Group 4: intravenous injec-
tion of Doxo (150 L, n = 8); Group 5: single intravenous injection of
micelles assembled with mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] (150 L,
n = 8); Group 6: single intravenous injection of Caelyx® (150 L, n = 8);
Group 7: three intravenous injections (once per week) of micelles
assembled with mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] (150 L, n = 8);
Group 8: three intravenous injections (once per week) of Caelyx® (150
μ
μ
Then, cells were incubated for 45 min at room temperature with 400
μL
μ
of PBS containing 5% v/v Rat Serum and 0.25% v/v Triton® X-100 to
permeabilize the cell membrane. Afterwards, the solution was dis-
charged, and the cells were stained for lysosomes detection. Cells were
μ
μ
incubated with 200
μ
L of rat anti-mouse anti-lysosomal-associated
μ
membrane protein 1 (anti-LAMP-1) antibody solution (1:100 dilution in
PBS added of 5% rat serum). After 1 h, cells were rinsed three times with
μ
L, n = 7). The dose of drug intratumorally injected was 3 mg/kg, while
PBS and incubated for 1 h with a 2
μ
g/mL 4′,6-diamidino-2-phenyl-
it was 15 mg/kg for each intravenous injection. Tumor growth was
assessed by caliper measurement; body weight and tumor volume were
checked every three days. The experimental endpoints were set at tumor
volume of 800 mm3, 20% body weight loss or metastasis-related side
effects. When either of these conditions appeared, mice were sacrificed.
indole (DAPI) and chicken anti-rat IgG(H + L) cross-adsorbed secondary
antibody Alexa Fluor 488 conjugated (1:500 dilution) solution in PBS
added of 5% fetal bovine serum (200
staining, respectively.
μL), for nuclei and lysosome
Finally, the wells were gently rinsed three times with PBS and once
with MilliQ water before being mounted on microscope slides using
Mowiol as mounting media prepared with 10 w/v% of Mowiol® 4–88
and 0.1% w/v 1,4-diazabicyclo[2.2.2]octane (DABCO; Sigma-Aldrich;
St Louis, MO, USA) in a 1:3.8 v/v glycerol/66 mM Tris-HCl buffer pH
8.5 mixture. Cells were imaged with a LSM 800 series Zeiss™ confocal
laser-scanning microscope (Jena, Germany) equipped with a 63× oil
immersion objective lens. Laser irradiation at 408, 488, 561 nm was
used to detect DAPI for nuclei imaging (blue), anti-LAMP-antibody for
lysosome imaging (green), Doxo (red), respectively. The images were
then processed with ZEN 2 (blue edition) from Zeiss™ Software and
elaboration was performed using ImageJ Software version 1.51j8.
2.12.3. Toxicity studies
Six-weeks old BALB/c female mice were randomly assigned into
three groups (n = 6) and each group received one intravenous injection
of mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10 or Caelyx® at equivalent dose
(15 mg/kg DOX) or control (PBS). Collection of blood in Sarstedt
Microvette 500 tubes were performed at day 3 after injection.
Aspartate aminotransferase (AST), alanine aminotransferase (ALT),
blood urea nitrogen (BUN), creatine phosphokinase (CPK), and lactate
dehydrogenase (LDH) levels were assayed using a Fujifilm DRI-CHEM
NX500i Analyzer (Fujifilm, Tokyo, Japan). After collection of blood,
animals were sacrificed and organs (lungs, liver, heart, spleen) were
collected in formol and underwent histological analysis.
Tissues from organs of mice underwent routine paraffin processing
2.12. In vivo studies
followed by sectioning at 4 μm and staining with Hematoxylin and Eosin
(HE). Hearts were sectioned longitudinally to allow visualization of all
four chambers and scored semi-quantitatively for severity of cardiac
injury [27–30]. The histologic variables under investigations were:
cytoplasmic vacuolation, infiltration by mononuclear cells, and inter-
stitial fibrosis with myofiber disorganization/atrophy. Each criterion
was scored on a 4-point scale (Table SI-1). Uncommon histologic find-
ings (mineralization, atrial thrombosis, perivascular lymphocytes) were
recorded by region as absent (0) or present (1). The scores for each
anatomical region were summed to obtain a total cardiac score for each
animal. For bilateral lung sections, semi-quantitative lung scoring
(Table SI-1) was achieved by recording the approximate percentage per
section occupied by a series of interstitial and parenchymal lesions.
Scores for each change were summed to obtain a composite score. A
similar approach was used to score liver and spleen pathology (Table SI-
1). The pathologist was blinded as to treatment status and mouse strain.
The animal experiments were performed according to the Belgian
national regulation guidelines and in agreement with EU Directive
1010/63/EU concerning the use of animals for experimental purposes.
The experiments were approved by the ethical committee for animal
´
care of the Faculty of Medicine of the Universite Catholique de Louvain
(2017/UCL/MD/34). Mice had free access to water and food and the
animal body weight was constantly monitored.
2.12.1. Anticancer activity on subcutaneous colorectal tumor model
CT26 cells (5 × 105 cells/mouse) were subcutaneously injected into
the right flank of the mice to allow reproducible tumor volume mea-
surements using an electronic caliper. Tumor volume was calculated
according to the formula: volume =
π
/6 x length x width2 [26]. Mice
were randomly assigned to treatment group when the tumor reached the
volume of 25 ± 2 mm3. Treatments were administered by local intra-
tumor injection or by tail vein intravenous injection. Five groups were
set: Group 1: control group (untreated, n = 9); Group 2: intratumor in-
2.13. Statistical analysis
jection of Doxo, (30
μ
L, n = 7); Group 3: intratumor injection of micelles
L, n = 6);
L, n = 7); Group 5:
intravenous injection of micelles assembled with mPEG5kDa-b-[(Doxo-
hydGlu)6-r-Leu10] (150 L, n = 8). The injected drug doses were 2.4 mg/
assembled with mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] (30
Group 4: intravenous injection of Doxo (150
μ
Statistical analysis was performed using GraphPad Prism, version
7.0a (GraphPad Software, USA). All results are expressed as mean ±
standard deviation (SD), except for results arising from cellular in vitro
and in vivo studies which are expressed as mean ± standard error of the
mean (SEM). The half maximal inhibitory concentration (IC50) values
were calculated using a nonlinear regression log(inhibitor) vs response,
variable slope. Statistical significance was attained for values of p < 0.05
and determined using two-way ANOVA for the in vitro and in vivo studies
(mouse body weight vs time; tumor growth vs time). Survival curves
were compared using a Mantel-Cox (log-rank) test. Outliers were
calculated using GraphPad software (significance level 0.01, two-sided)
and removed from the study.
μ
μ
kg and 12 mg/kg, for the intratumor and intravenous administration,
respectively. The therapeutic effect was evaluated by measuring the
tumor volume every other day. Body weights were assessed prior to each
tumor volume measurement. The experimental end point was set when
the tumor volume reached 800 mm3 or at 20% body weight loss. When
either of the last two conditions appeared, mice were sacrificed.
2.12.2. Anticancer activity on subcutaneous breast tumor model
4T1 cells (1 × 106 cells/mouse) were subcutaneously injected into
the BALB/c mice right flank, and the tumor growth was monitored with
the procedure described above. Mice were randomly assigned to treat-
ment group when the tumor volume reached 26 ± 1 mm3. Treatments
3. Results and discussion
The amphiphilic co-polymers [mPEG5kDa-b-(hydGlum-r-Leun)]
26

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
described in this study were designed to yield self-assembling doxoru-
bicin (Doxo) conjugates for tumor drug delivery. The co-polymers were
synthetized by Ring-Opening Polymerization [31] of γ-benzyl-glutamic
acid N-carboxyanhydride (BnOGlu-NCA) and leucine N-carboxyanhy-
dride (Leu-NCA), using mPEG5kDa-NH2 as initiator.
mPEG5kDa-NH2, with a number-average molecular mass of 5.4 kDa, was
confirmed by 1H NMR and MALDI ToF mass spectrometry (Fig. SI-1-SI-
3).
The PEG/polyaminoacid block molecular weight ratio was selected
according to the literature data for analogous block co-polymers, where
12 kDa mPEG was used as initiator of a 37 aspartic acid monomer block
[32]. In order to investigate the effect of the co-polymer composition on
the physicochemical properties of these materials and in turn the bio-
pharmaceutical performance of the drug conjugates, four mPEG5kDa-b-
(hydGlum-r-Leun) were produced by using 1:16:0, 1:8:8, 1:6:10, 1:4:12
mPEG/BnOGlu/Leu ratio ratios (Table 1, Fig. SI-4-SI-9).
Glutamic acid (Glu) was chosen as pendant group because its
γ-carboxyl group can be switched to hydrazide (hydGlu) and conjugate
Doxo through a pH-cleavable hydrazone bond, which intracellularly
releases the drug in an acidic lysosomal microenvironment. Leucine
(Leu) was included because its hydrophobic sidechain favors the self-
assembly of these amphiphilic block co-polymers and enhances the
stability of the resulting colloidal systems. Furthermore, Leu residues
can space the Glu monomers engaged for Doxo conjugation, limiting
hindrance constraints. Studies reported in the literature show, in fact,
that the conjugation of Doxo to poly-aspartic acid through hydrazide
bond yields only partial aspartate functionalization with the drug. This
results in a heterogeneous product, with non-controlled composition
and a number of unreacted pendant hydrazide groups [32] that can
affect the self-assembling behavior of these polymer conjugates, the
stability of resulting nanovehicles and, ultimately, the drug release.
Therefore, we aimed at identifying strategies that allow for both com-
plete derivatization of the pendant carboxylic groups of polyaminoacid-
based materials, and formation of stable self-assembled delivery nano-
vectors. Accordingly, we designed a library of mPEG5kDa-b-(hydGlum-r-
Leun) diblock co-polymers with different Glu/Leu composition (Fig. 1).
The 5 kDa mPEG-NH2 initiator was selected because it has been
approved for systemic pharmaceutical application [33], and has already
been successfully utilized to generate polyaminoacid-based self-assem-
bling systems [34]. Indeed, PEGylation contributes to the amphiphilicity
of materials including supramolecular systems and colloidal surfaces,
providing a flexible, hydrated coating to the self-assembled systems,
which is a requisite for “stealth” features and long circulation time in
blood [35].
The content of BnOGlu and Leu in the polymers was estimated by 1H
NMR by comparing the integral of the signal of terminal -OCH3 of mPEG
at 3.3–3.4 ppm with those of benzylic -OCH2- at 5.0–5.2 ppm of BnOGlu,
and of the -CH3 groups at 0.6–1.0 ppm of Leu (Fig. SI-10-SI-13). SEC
analysis showed that all polymers had low molecular weight dispersity.
The observed Mn,SEC values were slightly higher than the ones estimated
by 1H NMR, which can be ascribed to the difference in hydrodynamic
volume between the mPEG5kDa-polyaminoacid materials synthesized in
this study and the PMMA narrow standards utilized to calibrate the SEC.
The pendant γ-benzyl ester functionalities were converted into hy-
drazide moieties by treatment with hydrazine. 1H NMR analysis of the
purified co-polymers (Fig. SI-14-SI-17) showed complete removal of the
benzyl ester protecting group. A modified Snyder's assay [19] showed
that the number of γ-hydrazide groups per polymer chain was in
agreement with the theoretically expected. SEC analysis confirmed that
all final mPEG5kDa-b-(hydGlum-r-Leun) co-polymers retained low mo-
lecular dispersity.
Doxo conjugation to the hydrazide pendant groups of the co-polymer
was estimated by UV–Vis at λ = 488 nm in relation of the mPEG5kDa
concentration assessed by iodine assay. The conjugation efficiency
(number of Doxo derivatized hydrazide/total hydrazide) was found to
be inversely correlated to the hydGlu:Leu ratio being 44, 70, 88, and
γ-benzyl-glutamic acid N-carboxyanhydride (BnOGlu-NCA) and
leucine N-carboxyanhydride (Leu-NCA) monomers were synthesized by
γ-benzyl glutamate and leucine reaction with triphosgene in the pres-
100%
for
mPEG5kDa-hydGlu16
,
mPEG5kDa-b-(hydGlu8-r-Leu8),
mPEG5kDa-b-(hydGlu6-r-Leu10), and mPEG5kDa-b-(hydGlu4-r-Leu12),
respectively. This result is in good agreement with our initial hypothesis
represented in Fig. 1, which postulated that hydrazide functionalities of
the γ-hydGlu monomers must be sufficiently spaced along the polymer
chain to minimize steric hindrance provided by grafted drug molecules.
ence of α-pinene to trap HCl generated in the process (Scheme 1) [17].
Aminoacid benzyl ester protection was used to avoid possible side-
reactions of γ-carboxyl group of Glu during NCA monomer synthesis
and subsequent polymerization reactions. Furthermore, after polymer-
ization the γ-benzyl ester can be straightforward converted into the
corresponding γ-hydrazide moiety required for Doxo conjugation. Both
NCA monomers were isolated in high purity by crystallization (BnOGlu-
NCA) or precipitation (Leu-NCA).
3.1. Characterization of polymeric micelles
The critical micelle concentration (CMC) of drug-free mPEG5kDa-b-
(hydGlum-r-Leun) co-polymers was assessed using pyrene as a fluores-
cent probe. The CMC was calculated by elaboration of results of Fig. 2A-
D that reports the plots of the emission intensity ratio of the third (I3)
and first (I1) spectrum signals of the pyrene spectrum vs the logarithmic
polymer concentration [37].
The macroinitiator mPEG5kDa-NH2 was synthesized through a two-
step procedure which involves the introduction at the polymer chain-
end of a 1-alkene moiety, which can then be used to introduce a pri-
mary amine functionality by thiol-ene reaction [36] in the presence of
cysteamine. Firstly, mPEG5kDa-OH was converted into the mPEG5kDa
-
In agreement with results reported by Vega et al. for PEG-polyGlu
based materials [38], mPEG5kDa-hydGlu16 did not self-assemble under
allyl carbamate by treatment with allyl isocyanate in the presence of
triethylamine. The 1H NMR analysis confirmed the formation of carba-
mate derivative. Then, thiol-ene reaction [36] was carried out using
cysteamine in the presence of DPAP as the photo-initiator that was
activated at λ = 350 nm. The conversion of mPEG5kDa-OH into
the experimental conditions used for the CMC assay (CMC > 300 μM).
On the contrary, mPEG5kDa-b-(hydGlu8-r-Leu8), mPEG5kDa-b-(hydGlu6-
r-Leu10) and mPEG5kDa-b-(hydGlu4-r-Leu12) were found to assemble
with CMC of 92.2 ± 13.1 μM, 78.5 ± 8.6 μM, and 17.8 ± 13.1 μM,
Fig. 1. Schematic representation of mPEG5kDa-b-[(Doxo-hydGlu)m-r-Leun] with high (A) and low (B) ratio of hydGlu monomers with respect to Leu. Light green
sections: hydGlu monomer; dark green sections: Leu monomer; light blue chain: PEG; red stars: Doxo. (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)
27

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
Scheme 1. Synthesis of mPEG5kDa-b-[(Doxo-hydGlu)m-r-Leun]. i. triphosgene,
α
-pinene, THF, 56 ◦C, 90 min; ii. allyl isocyanate, Et3N, CHCl3, room temperature, 18
h; iii. Cysteamine hydrochloride, DPAP, MeOH, irradiation at λ = 350 nm, 3 h; iv. BnOGlu-NCA, Leu-NCA, DMF, room temperature, 7 days; v. hydrazine hydrate,
DMF, 40 ◦C, 48 h; vi. Doxo hydrochloride, TFA 0.2% (v/v), DMSO, 3 days.
respectively. These results show that, as expected, the increase of Leu/
fluorescence emission intensity at λ = 555 nm initially increased and
then decreased as consequence of the fluorophore self-quenching when
the polymers self-assemble (Fig. SI-19).
Glu molar ratio favors the co-polymer assembling, which is ascribable to
the hydrophobic isobutyl side moiety (hydrophobic constant
π = 1.64.
[39]), while in the case of mPEG5kDa-hydGlu16, that does not contain
hydrophobic moieties, assembling does not occur.
In this work, we observed that the high leucine content combined
with the presence of four Doxo molecules in mPEG5kDa-b-[(Doxo-
hydGlu)4-r-Leu12] resulted in conjugate precipitation into large aggre-
gates. Therefore, this Doxo-conjugate block co-polymer could not be
characterized and further investigated in vitro and in vivo.
Due to fluorescence interferences between Doxo and pyrene, a novel
analytical procedure was set-up to calculate the CMC of Doxo conju-
gated mPEG5kDa-b-(hydGlum-r-Leun) co-polymers, which exploits the
Doxo emission intensities at 555 nm. This analytical approach relies on
the quenching of the Doxo fluorescence when drug molecules are in
close proximity such as in the core of micelles [40,41].
The different fluorescence profiles reported in Fig. 2 E-G reflect the
co-polymer composition. Indeed, as the Leu content increases a lower
Doxo packing can be achieved that reflects on lower quenching rate as
the micelles assemble. Therefore, the CMC values were calculated at the
intercept of the regression lines of increasing and decreasing fluores-
cence. The CMC for mPEG5kDa-(Doxo-hydGlu)16, mPEG5kDa-b-[(Doxo-
hydGlu)8-r-Leu8] and mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] were 6.7
Fig. 2E-G reports the plots of emission intensities of Doxo at 555 nm
vs the concentration of the mPEG5kDa-b-[(Doxo-hydGlu)m-r-Leun] co-
polymers that were used to calculate the CMC. The increase of drug-
polymer concentration resulted in a linear increase of the spectropho-
tometric absorbance of Doxo at λ = 488 nm (Fig. SI-18) while the
± 0.9 μM, 4.9 ± 0.2 μM, and 4.6 ± 0.2 μM, respectively. The CMC of the
28

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
those obtained by fluorescence quenching analysis (Fig. SI-20). Inter-
estingly, the CMC values of copolymers containing Doxo are about 20-
fold lower than the ones obtained with the Doxo-free counterparts
indicating that Doxo plays a relevant role in copolymer association and
micelle packing. To note that the low CMC of mPEG5kDa-b-[(Doxo-
Glu)m-r-Leun] is a requisite for therapeutic applications of these delivery
systems. Indeed, low CMC prevents the dissociation of the nanocarrier
upon dilution in the blood.
Table 1
Composition and molecular weight features of co-polymers synthesized in this
work.
Polymersa
b
Glu:
Leua
Mn, NMR
(kDa)a
Mn, SEC
(kDa)
Ðb
mPEG5kDa-BnOGlu16
mPEG5kDa-b-(BnOGlu8-r-
Leu8)
16:0
8:8
9.3
8.2
12.2
10.5
1.07
1.11
mPEG5kDa-b-(BnOGlu6-r-
6:10
4:12
8.0
7.8
10.4
9.9
1.09
1.08
The size of the colloidal systems is relevant in dictating the fate of a
nanocarrier after injection in the bloodstream. Indeed, the diameter of
the nanosized therapeutics suitable for in vivo application should lay in
the range from 5 to 200 nm, which minimizes the renal clearance
through the glomerular capillaries occurring for colloids with size below
5 nm, and the RES removal for colloids with size above 200 nm [42].
Table 2 reports the size, polydispersity index and charge of the as-
semblies obtained with drug-free and Doxo-conjugated co-polymers.
Leu10
)
mPEG5kDa-b-(BnOGlu4-r-
Leu12
)
mPEG5kDa-hydGlu16
mPEG5kDa-b-(hydGlu8-r-
Leu8)
16:0
8:8
8.1
7.6
12.2
8.9
1.03
1.25
mPEG5kDa-b-(hydGlu6-r-
6:10
4:12
7.5
7.4
8.8
9.8
1.15
1.20
Leu10
)
mPEG5kDa-b-(hydGlu4-r-
Leu12
)
mPEG5kDa-b-(hydGlu8-r-Leu8) and mPEG5kDa-b-(hydGlu6-r-Leu10
)
)
a
b
Co-polymers composition was estimated by 1H NMR in CDCl3.
micelles had similar size, while the mPEG5kDa-b-(hydGlu4-r-Leu12
nanoparticles showed slightly smaller diameter with no statistically
significant differences although a decreasing size trend was observed
with the increase of Leu content, suggesting that Leu may have a little
influence on the size of the micelles.
Determined by SEC using DMF + 0.1% LiBr as the mobile phase, in a system
calibrated with PMMA standards.
mPEG5kDa-b-[(Doxo-Glu)m-r-Leun] was also assessed by light scattering
using a detection system equipped with fluorescent filter for Doxo
fluorescence removal, which provided CMC values in agreement with
The results reported in Table 2 show that the mPEG5kDa-(Doxo-
hydGlu)16
, mPEG5kDa-b-[(Doxo-hydGlu)8-r-Leu8] and mPEG5kDa-b-
Fig. 2. Self-assembly behavior of the polymer library. I3/I1 spectrum signal ratio of pyrene vs logarithmic concentration of (A) mPEG5kDa-hydGlu16, (B) mPEG5kDa-b-
(hydGlu8-r-Leu8), (C) mPEG5kDa-b-(hydGlu6-r-Leu10), and (D) mPEG5kDa-b-(hydGlu4-r-Leu12). Fluorescence intensity at 555 nm (arbitrary units: A.U.) of (E)
mPEG5kDa-(Doxo-hydGlu)16, (F) mPEG5kDa-b-[(Doxo-hydGlu)8-r-Leu8], (G) mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] vs conjugate concentration.
29

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
nanostructures with round shaped morphology typical of micelles
formed by block co-polymer where hydrophilic block is longer than the
hydrophobic portion of an amphiphilic co-polymer [43]. The size was in
agreement with the results obtained by DLS analyses (Fig. 3, Fig. SI-22,
Table SI-2). The non-spherical shape of the micelles observed by TEM
may explain the relatively high PDI obtained by DLS analysis. However,
it should be noted that high PDI have been already reported in the
literature in the case of self-assembling amphiphilic co-polymers [44].
Importantly, micelles were found to be fairly stable even in the
presence of proteins over 72 h of incubation at 37 ◦C either at pH 7.4 or
5.5. Indeed, the DLS results (Fig. SI-23) showed that the micelles did not
undergo aggregation, even though the standard deviation was higher
than that obtained in buffer, which is ascribable to detection in-
terferences due to the presence of proteins.
Table 2
Volume weighted hydrodynamic diameter (dH), PDI and zeta potential (ZP)
obtained by DLS analysis; diameter assessed by TEM analysis (dTEM) of
mPEG5kDa-b-(hydGlum-r-Leun) and mPEG5kDa-b-[(Doxo-hydGlu)m-r-Leun] as-
semblies. (ND: non-detectable; NA non-assessable).
dH (nm)
PDI
dTEM
(nm)
ZP (mV)
mPEG5kDa-hydGlu16
ND
ND
ND
ND
mPEG5kDa-b-(hydGlu8-r-Leu8)
51.0 ±
4.2
0.37 ±
0.07
47.9 ±
3.9
0.46 ±
0.17
mPEG5kDa-b-(hydGlu6-r-
50.2 ±
4.1
0.39 ±
0.05
45.5 ±
4.3
0.23 ±
0.02
Leu10
)
mPEG5kDa-b-(hydGlu4-r-
39.2 ±
5.6
0.30 ±
0.04
32.3 ±
3.4
0.03 ±
0.32
Leu12
)
mPEG5kDa-(Doxo-hydGlu)16
41.6 ±
4.9
0.34 ±
0.06
38.4 ±
4.5
2.31 ±
0.25
Altogether, the TEM and DLS results confirmed that the micellar
nanosystems obtained with the new block co-polymers loaded with
Doxo possess a suitable size for the passive tumor accumulation through
the EPR effect [45–47].
mPEG5kDa-b-[(Doxo-hydGlu)8-
r-Leu8]
33.3 ±
5.3
0.37 ±
0.05
26.5 ±
3.0
3.36 ±
0.25
mPEG5kDa-b-[(Doxo-hydGlu)6-
29.2 ±
1.1
0.36 ±
0.07
23.8 ±
3.0
3.61 ±
0.28
r-Leu10
]
mPEG5kDa-b-[(Doxo-hydGlu)4-
r-Leu12
NA
NA
NA
NA
The zeta potential analysis revealed that mPEG5kDa-b-(hydGlum-r-
Leun) micelles are close to neutrality, which is in agreement with the
mostly non-protonated state of the hydrazide groups at pH 7.4 being the
pKa typically in the range of 4–5 [48]. The mPEG5kDa-b-[(Doxo-
hydGlu)m-r-Leun] micelles possess a slightly positive zeta potential,
which can be ascribed to exposure of the amino group of the gluconic
moiety of few Doxo units.
]
[(Doxo-hydGlu)6-r-Leu10] assemble into colloidal systems with smaller
hydrodynamic diameter with respect to the drug-free counterparts. The
smaller size of Doxo-loaded micelles compared to the Doxo-free coun-
terparts can be ascribed to structural arrangements. The role played by
Doxo in the formation of micelles is also supported by the evidence that
mPEG5kDa-(Doxo-hydGlu)16 can self-assemble into nanoparticles while
mPEG5kDa-b-[(Doxo-hydGlu)4-r-Leu12] could not be analyzed by DLS
due to formation of large aggregates that underwent precipitation. To
note that, similarly to the Doxo free conjugates, the Doxo containing
counterparts showed a micelle size decrease trend with the increase of
Leu content, although no statistical significance was calculated along the
copolymer series.
3.2. Release studies
Doxo release from the micelles was investigated at pH 7.4 and 5.5,
which correspond to the bloodstream and lysosomal pH, respectively.
The release profiles reported in Fig. 4 show that the drug loaded
formulations
{mPEG5kDa-(Doxo-hydGlu)16
,
mPEG5kDa-b-[(Doxo-
hydGlu)8-r-Leu8], and mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10]} possess
similar release profile at the two pH tested, and the drug release is
remarkably faster under acid conditions. The formulations displayed
The TEM images of Doxo conjugates reported in Fig. 3 show that all
block co-polymers containing Leu formed similar homogeneous
Fig. 3. Size profiles by DLS (A) and TEM images of mPEG5kDa-(Doxo-hydGlu)16 (B), mPEG5kDa-b-[(Doxo-hydGlu)8-r-Leu8] (C), and mPEG5kDa-b-[(Doxo-hydGlu)6-r-
Leu10] (D). The TEM images underwent ImageJ analysis and histograms of the particle size distributions have been reported in Fig. SI-21.
30

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
strongly affected by the microenvironment that is influenced by the
structure of the supramolecular system. Interestingly, the profiles re-
ported in Fig. 4 show that the Doxo release rate increases with the Leu
content in the conjugate. In particular, statistically significant differ-
ences were found after 100 h of drug release between mPEG5kDa-(Doxo-
hydGlu)16 and mPEG5kDa-b-[(Dox-hydGlu)6-r-Leu10]. Indeed, consid-
ering that the drug is conjugated to the backbone in the same way in all
copolymers, it seems reasonable to think that differences in drug release
are due to the different access of water in the micelle core as reported
above which suggests that Leu and the resulting different intermolecular
interactions within the micelle core play a role in the drug release
profile.
Finally, despite the slow drug release observed in buffer, it is worth
mentioning that the drug release rate in the acid intracellular com-
partments may be facilitated by the contribution of lysosomal enzymatic
fragmentation as reported in the literature with other Doxo conjugated
poly(L-glutamic acid) based therapeutics [51].
3.3. Cell viability studies
As Doxo has demonstrated its therapeutic efficacy against several
solid
tumors,
mPEG5kDa-(Doxo-hydGlu)16
,
mPEG5kDa-b-[(Doxo-
hydGlu)8-r-Leu8], and mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] micelles
were tested on two tumor cell lines, murine colorectal carcinoma CT26
[52,53] and murine mammary carcinoma 4T1 [54]. Their anticancer
effect was investigated by MTT assay following cells treatment with
equivalent Doxo concentrations in the range of 0.1–100 μM for 48 or 72
h. A preliminary study was undertaken to evaluate the biocompatibility
of the drug free block co-polymers at concentrations corresponding to
the maximum equivalent concentrations used with the Doxo loaded
bioconjugates, namely 13
μM for mPEG5kDa-hydGlu16, mPEG5kDa-b-
(hydGlu8-r-Leu8), and 18
μ
M for mPEG5kDa-b-(hydGlu6-r-Leu10) for 72 h.
In all cases, the CT26 and 4 T1 cell viability resulted to be above 80%
(Fig. SI-24) indicating that the co-polymers have high biocompatibility,
which is a requisite for therapeutic applications.
Fig. 5A reports the cell viability profiles of both CT26 and 4 T1 cell
lines incubated with the drug-conjugates. Pharmacologically active
Doxo is released after the hydrazone bond cleavage, which is a requisite
for its localization into the nucleus of cancer cells and DNA intercalation
[11].
In the case of CT26 cells, the IC50 obtained with the two bio-
conjugates containing Leu were higher than free Doxo and decreased
with the incubation time. The mPEG5kDa-(Doxo-hydGlu)16 had similar
IC50 after 48 h and 72 h incubation. After 72 h cell incubation all bio-
conjugates showed similar IC50 (Fig. 5-B1).
The cytotoxicity towards 4T1 cancer cells was similar for all bio-
conjugates and at both tested incubation times (Fig. 5-B2). The similar
IC50 at 48 and 72 h might be ascribed to the rapid proliferation and
aggressiveness of these cells that may develop drug resistance. The lower
IC50 observed with 4T1 cells with respect to CT26 may be due to several
factors among which cell line sensitivity to the drug, rate of bioconjugate
access to the cytosolic compartment and drug availability resulting from
trafficking pathways and intracellular release rate.
Fig. 4. Doxo release profile from (A) mPEG5kDa-(Doxo-hydGlu)16
, (B)
mPEG5kDa-b-[(Doxo-hydGlu)8-r-Leu8],
(C)
mPEG5kDa-b-[(Dox-hydGlu)6-r-
○
Leu10], at pH 7.4 (●) and pH 5.5 ( ).
The cell viability profiles reported in Fig. 5-A and the IC50 data re-
ported in Fig. 5-B show that under the experimental conditions the
mPEG5kDa-b-[(Doxo-hydGlu)m-r-Leun] bioconjugates have biological
activity comparable to Doxo, which makes these systems very promising
for therapeutic applications. Indeed, usually Doxo bioconjugates possess
significantly lower cytotoxicity compared to the free drug [55,56],
which implies the use of high drug doses.
high stability at pH 7.4 with negligible drug release over 9 days, which is
crucial in preventing undesired drug release in the bloodstream (pH 7.4)
upon administration, thus reducing drug-related systemic toxicity as-
cribable to off-target release and disposition [5,49].
At pH 5.5, Doxo was slowly released to yield 17, 23 and 29% released
drug in 9 days from mPEG5kDa-(Doxo-hydGlu)16, mPEG5kDa-b-[(Doxo-
hydGlu)8-r-Leu8] and mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10], respec-
tively. In general, the release rate through the hydrazone cleavage was
much slower than that expected according to our experience with
polymer bioconjugates and studies reported in the literature [50].
However, it should be noticed that the hydrolysis of chemical bonds is
mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] showed the lowest IC50 value
in both cell lines, which can be ascribed to better cell uptake and/or to
the more efficient Doxo release reported in Fig. 4.
mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] was the co-polymer with
highest performance in terms of derivatization efficiency of the hydGlu
monomers with Doxo, drug release and biological activity in vitro.
31

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
Fig. 5. A. Viability profiles of CT26 cells (left) and 4T1 cell line (right) after 48 (top panels) and 72 h (bottom panels) incubation with Doxo, mPEG5kDa-(Doxo-
hydGlu)16, mPEG5kDa-b-[(Doxo-hydGlu)8-r-Leu8], mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10], at drug equivalent concentration; B. IC50 of Doxo and micellar formulations
referred as drug concentration (mean ± SEM, N = 3, n = 17) ***p < 0.001, **p < 0.01 *p < 0.05.
Accordingly, this bioconjugate was selected for the further in vitro and in
vivo investigations.
resulting from the overlapping of micelles red spots co-localized with the
lysosome green staining) or in the cytosol. This confirmed the intracel-
lular delivery of the drug by the nanocarrier and its trafficking through
the lysosomal compartment where the presence of the acidic environ-
ment can trigger the cleavage of the pH sensible hydrazone bond [57],
resulting in Doxo release. mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] per-
sisted into lysosomes over 4 h cell incubation resulting in the released
Doxo disposition in the nuclei. It is worth noting that at this time point
the fluorescence in the nuclei of cells incubated with free Doxo was
negligible indicating that the drug diffuses from the nucleus during in-
cubation with medium.
3.4. Confocal analyses
The intracellular fate of free Doxo and mPEG5kDa-b-[(Doxo-
hydGlu)6-r-Leu10] in the CT26 cells was assessed by immunofluores-
cence confocal microscopy. The co-localization of free or conjugated
Doxo (red fluorescence) within lysosomes (green fluorescence) was
investigated by labeling lysosomes with primary antibody against
Lysosomal-Associated Membrane Protein-1 (LAMP-1). Fig. 6 shows
representative images of cells incubated with free Doxo or mPEG5kDa-b-
[(Doxo-hydGlu)6-r-Leu10] and for 2 h or for 2 h plus 4 h incubation with
fresh complete medium.
3.5. In vivo studies
Upon 2 h cell incubation with free Doxo, the drug fluorescence was
mostly associated to the nuclei, while mPEG5kDa-b-[(Doxo-hydGlu)6-r-
Leu10] was mostly disposed in the lysosomal compartment (yellow spots
Based on the in vitro results, the anticancer activity of mPEG5kDa-b-
[(Doxo-hydGlu)6-r-Leu10] was tested on CT26 and 4T1 subcutaneous
syngeneic tumor models [58] following intratumor (IT) and intravenous
32

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
Fig. 6. Confocal microscopy images of CT26 cell incubated with: Doxo or mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] for 2 h; and with Doxo or mPEG5kDa-b-[(Doxo-
hydGlu)6-r-Leu10] for 2 h followed by 4 h incubation with fresh medium. Scale bar: 20 μm.
(IV) administrations. The Doxo dose was selected according to the
studies reported by Bae et al. [40] and Lv et al. [59] showing that pol-
yaminoacidic based micellar systems display anticancer activity in mice
with doses in the 10–20 mg/kg range. As the Doxo-loaded PEGylated
liposomal trademark formulation Caelyx® is a nanomedicine used for
breast cancer treatment, this formulation was also used as reference in
our 4T1 in vivo studies.
23 days after intratumoral and intravenous administration, respectively.
The therapeutic activity of the polymeric nanocarrier was also
investigated in vivo with 4T1-bearing mice. The animals were randomly
divided into groups and treated either locally with intratumor injections
of 3 mg/kg Doxo equivalent dose, or intravenously with single or mul-
tiple injections of 15 mg/kg Doxo equivalent dose (three administra-
tions, once per week).
A preliminary study was carried out to assess the in vivo toxicity of
mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] by intravenous administration to
healthy mice of 12 mg/kg Doxo equivalent bioconjugate dose and
monitoring the animal body weight and wellbeing for 14 days. No
weight loss, distress or behavior changes were observed during this time,
confirming the good tolerability of the bioconjugates at the selected dose
(Data not shown). In the first antitumor study, free Doxo or Doxo
equivalent dose of mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] was adminis-
tered either intratumorally or intravenously to CT26 tumor bearing mice
at 2.4 mg/kg and 12 mg/kg, respectively. The intratumor administration
allows for achieving higher drug dose at the tumor site while reducing
the risk of side effects compared to the systemic route [60].
Fig. 8B shows that after intratumor administration, the Kaplan-Meier
survival curves of Doxo and mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10
]
present similar median survival, which is higher compared to the un-
treated group (30 and 31 days for the Doxo and mPEG5kDa-b-[(Doxo-
hydGlu)6-r-Leu10] treated groups vs 27 days for the control group).
Spontaneous metastases spreading from the primary tumor were
observed in all the groups, which can be ascribed to the highly tumor-
igenic and invasive behavior of 4T1 tumor model [61]. This prompted
the sacrifice of the mice for metastasis-related side effects and not for the
tumor size.
Fig. 8A shows that after intratumor administration, a slight decrease
of body weight was noticed for treated groups {Doxo and mPEG5kDa-b-
[(Doxo-hydGlu)6-r-Leu10]} starting from day 18. Similar body weight
loss was observed for untreated mice suggesting that this was due to the
metastasis onset, rather than to the toxicity of the treatment. The pres-
ence of metastasis was confirmed by autopsies, demonstrating that lungs
and spleen were the organs mainly colonized by metastatic cells. Based
on the tumor volume values before metastasis onset and until the
moment of sacrifice, a significant difference between Doxo treated group
and mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] treated mice was detected
(Fig. 8B; refer to supporting information for statistical comparison be-
tween different groups).
The intratumoral injection of mice did not induce weight loss in the
week following the treatment (Fig. 7A). The intravenous administration
of mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] did not impact the mouse
weight curve while free Doxo induced weight loss in most of the animals
starting from the first few days following treatment (Fig. 7B).
mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] treated mice survived longer
than the untreated and free Doxo treated groups, showing a higher
median survival time of 30 days and 23 days after intratumoral and
intravenous administration, respectively.
Fig. 7B and E show a significant tumor growth delay after intratumor
and intravenous injections of the bioconjugate with respect to the un-
treated group starting one week following treatment (day 15) refer to
supporting information for statistical comparison between different
groups, proving the therapeutic efficacy of the bioconjugate. However,
no significant difference in tumor growth was observed compared to
Doxo (except between day 19 and 23). mPEG5kDa-b-[(Doxo-hydGlu)6-r-
Leu10] treated mice survived longer than the untreated and free Doxo
treated groups, showing a higher median survival time of 30 days and
The intravenous administration to mice bearing 4T1 subcutaneous
tumors was performed by single injection and weekly multiple admin-
istrations. This treatment schedule was planned on the basis of the fast
tumor growth rate observed during previous studies. Free Doxo and
Caelyx®, a trademark Doxo liposomal formulation approved for the
treatment of several cancers including breast, ovarian and multiple
myeloma [62,63], were administered as references. The high toxicity of
free Doxo observed after a single injection to the CT26 cancer bearing
33

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
Fig. 7. In vivo antitumor efficacy of mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] intratumorally (left panels) and intravenously (right panels) injected in subcutaneous
CT26 colorectal cancer bearing mice. Doxo equivalent doses of 2.4 mg/kg and 12 mg/kg were administered intratumorally and intravenously, respectively. A) and
D): Animal body weight in the first three weeks following treatment (refer to supporting information for statistical comparison between different groups); B) and E):
Tumor growth curves (refer to supporting information for statistical comparison between different groups); C) and F): Kaplan-Meier Survival curve. Results are
expressed in mean ± SEM, n = 6–9; Statistical analysis of the treatments is referred to the untreated group: ***p < 0.001, **p < 0.01 *p < 0.05.
mouse model, also documented by the literature [40,59], did not allow
for its use as a reference for the multiple administrations schedule.
Interestingly, 4T1 tumor bearing mice treated by IV with free Doxo
exhibited significant body weight loss during the first two weeks
compared to animals treated by single or multiple administration of
mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] (Fig. 8D; refer to supporting in-
formation for statistical comparison between different groups).
Furthermore, significant lower body weight decrease was found after
Caelyx® formulation injected by a multiple administration regimen,
caused a severe and irreversible body weight loss, highlighting its
toxicity at this regimen and dose. Therefore, mPEG5kDa-b-[(Doxo-
hydGlu)6-r-Leu10] possesses a comparable therapeutic performance and
a safer profile with respect to Caelyx®.
The median survival time increased after single or multiple
mPEG5kDa-(Doxo-hydGlu)6-r-Leu10 administration and single Caelyx®
administration with respect to the control untreated group (p < 0.001).
The median survival was 35 and 37 days for single administration of
single and multiple injection of mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10
]
compared to the reference liposomal formulation Caelyx® administered
at equivalent Doxo dose in single or multiple administrations,
respectively.
mPEG5kDa-(Doxo-hydGlu)6-r-Leu10
and
Caelyx®,
respectively,
compared to 27 days for untreated mice (Fig. 8E). On the contrary,
single free Doxo administration and multiple Caelyx® administration
dramatically reduced the survival rate. Overall, no significant survival
difference was observed between single and weekly injections of
mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10], which can be ascribed to the
appearance of metastasis-related side effects that required the mouse
sacrifice regardless of the tumor volumes. Indeed, the aggressiveness of
this tumor model does not allow for long-term endpoint.
Single or multiple intravenous administration of mPEG5kDa-b-
[(Doxo-hydGlu)6-r-Leu10] slowed down tumor growth compared to
untreated mice (refer to supporting information for statistical compar-
ison between different groups). Moreover, nearly overlapped tumor
volume profiles found between mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10
]
(single or multiple injections) and Caelyx® multiple injections. Multiple
mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] injections show that the bio-
conjugate can be safely administered to achieve enhanced anticancer
activity while limiting the toxicity of the drug. On the contrary, the
With the aim to investigate the tolerability of the polymer-based
drug carrier a safety study was undertaken. In vitro studies showed
that mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] polymeric micelles induced
34

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
Fig. 8. In vivo antitumor efficacy of mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] intratumorally (A, B, C) and intravenously (D, E, F) injected in subcutaneous 4T1 breast
cancer bearing mice. Doxo equivalent doses of 3 mg/kg and 15 mg/kg were administered intratumorally and intravenously, respectively. Animal body weight in the
first three weeks following treatment (A and D; refer to supporting information for statistical comparison between different groups); tumor growth curves (B and E;
refer to supporting information for statistical comparison between different groups); Kaplan-Meier Survival curve (C and F) after administration of single (A, B, C) or
single and weekly (D, E, F) (see arrows on “time” axis) administration of Doxo and mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10]. Caelyx® was also injected intravenously as
○
control (D, E, F). Results are expressed in mean ± SEM, n = 8–11. A, B, C: For panels C and F, statistical analysis is referred to untreated group (*), Doxo ( ), Caelyx®
multiple injections (●), mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] single injection (∝), mPEG5kDa-b-[(Doxo-hydGlu)6-r-Leu10] multiple injections (■); triple marks: p <
0.001, double marks: p < 0.01, single mark: p < 0.05.
less than 4.5%, haemolysis (Fig. SI-25), which is the threshold for a
haemolytic reaction [64]. After i.v. administration to mice, the blood
analysis (Table SI-3) showed that mPEG5kDa-b-[(Doxo-hydGlu)6-r-
Leu10] did not significantly increase the levels of blood markers for
organ toxicity, AST, BUN and CPK, while a slight increase of ALT and
LDH was observed in comparison to untreated mice. On the contrary,
Caelyx® significantly increased all marker levels except BUN, which
suggests that the polymeric micelles are significantly safer to liver and
heart than Caelyx®. Histologic analyses (Fig. SI-26) showed that the
polymeric carrier has lower cardiac toxicity with respect to the reference
Caelyx®, without induction of mononuclear cell infiltration and fibrosis.
Hepatocytic cytoplasmic hydropic lesions were also less frequent and
severe in mice treated with polymeric micelles than Caelyx®. Overall,
the polymeric micelles showed a safer profile with respect to Caelyx® of
which toxicity is extensively documented in the literature [65–69].
It should be noted that the main objective of this study was to
evaluate the proof of concept of mPEG5kDa-b-[(Doxo-hydGlu)m-r-Leun]
co-polymers as Doxo delivery systems for cancer treatment. For this
reason, we chose to perform an anti-tumor efficacy study on two
different subcutaneous tumor models. Considering the preliminary re-
sults obtained in this work, further studies will be carried out to evaluate
the biodistribution and anticancer efficacy on orthotopic models.
4. Conclusions
The results obtained in this work show that the identification of the
requisites allow for the rational design of PEG-polyaminoacid compo-
sition yielding co-polymers with optimized anticancer drug conjugation
that self-assemble into colloidal nanocarriers and release the drug under
controlled conditions. The introduction of leucine as spacer among the
glutamic monomers minimizes steric hindrance that prevents complete
co-polymer derivatization with drugs, and modulates the amphiphilicity
35

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
[12] Y. Bae, N. Nishiyama, K. Kataoka, In vivo antitumor activity of the folate-
conjugated pH-sensitive polymeric micelle selectively releasing adriamycin in the
intracellular acidic compartments, Bioconjug. Chem. 18 (2007) 1131–1139.
[13] X. Yang, J.J. Grailer, I.J. Rowland, A. Javadi, S.A. Hurley, V.Z. Matson, D.
A. Steeber, S. Gong, Multifunctional stable and pH-responsive polymer vesicles
formed by heterofunctional triblock copolymer for targeted anticancer drug
delivery and ultrasensitive MR imaging, ACS Nano 4 (2010) 6805–6817.
[14] R.B. Weiss, G. Sarosy, K. Clagett-Carr, M. Russo, B. Leyland-Jones, Anthracycline
analogs the past, present, and future, Cancer Chemother. Pharmacol. 18 (1986)
185–197.
of the conjugate, which contributes to its self-assembly. Therefore, ac-
cording to the physicochemical properties of the conjugated drug,
suitable Glu/Leu molar ratio must be selected to compromise a high
derivatization efficiency of the Glu monomers with drug and the
increased hydrophobicity of the polyaminoacid block due to Leu, which
contributes to the colloidal stability of the assembled materials.
In the case of Doxo (543.5 Da and 1.27 logP), the co-polymer
including 6 hydGlu and 10 Leu [PEG5kDa-b-(hydGlu6-r-Leu10)] was
found to yield a drug conjugate with the best physicochemical and
biopharmaceutical properties, drug loading, particle size, self-
association, stability, and site selective controlled drug release. The
resulting conjugate displayed excellent pharmaceutical performance by
enhancing the selective anticancer activity of Doxo and significantly
limiting its toxic side-effects, which provides for remarkably better
therapeutic outcome in term of survival with respect to the parent drug
Doxo. Furthermore, it may also offer a higher safety with respect to
commercial colloidal systems designed to deliver Doxo.
[15] P. Markland, G.L. Amidon, V.C. Yang, Modified polypeptides containing γ-benzyl
glutamic acid as drug delivery platforms, Int. J. Pharm. 178 (1999) 183–192.
[16] A.J. Williams, V.K. Gupta, Self-assembly of a rodlike polypeptide on solid surfaces:
role of solvent, molecular weight, and time of assembly, J. Phys. Chem. B 105
(2001) 5223–5230.
[17] N.M.B. Smeets, P.L.J. Van Der Weide, J. Meuldijk, J. Vekemans, L.A. Hulshof,
A scalable synthesis of l-Leucine-N-carboxyanhydride, Org. Process. Res. Dev. 9
(2005) 757–763.
[18] W. Zhao, Y. Gnanou, N. Hadjichristidis, Fast and living ring-opening
polymerization of α-amino acid N-carboxyanhydrides triggered by an “alliance” of
primary and secondary amines at room temperature, Biomacromolecules 16
(2015) 1352–1357.
[19] S.L. Snyder, P.Z. Sobocinski, An improved 2, 4, 6-trinitrobenzenesulfonic acid
method for the determination of amines, Anal. Biochem. 64 (1975) 284–288.
[20] G.E.C. Sims, T.J. Snape, A method for the estimation of polyethylene glycol in
plasma protein fractions, Anal. Biochem. 107 (1980) 60–63.
The outcome of the systematic investigation we performed to select
the most promising co-polymer can be exploited as background to
generate other materials based on mixed polyaminoacid by introducing
semisynthetic hydrophobic aminoacids as alternative to leucine that can
modulate the self-assembling behavior of the amphiphilic co-polymer,
the stability of the resulting colloidal carrier, the monomer conjuga-
tion efficiency with drug for pH-controlled release and that can, in
addition, be exploited for physical encapsulation of anticancer drugs
with poor aqueous solubility thus offering new opportunities for com-
bination therapies.
[21] E. Ambrosio, M. Barattin, S. Bersani, S. Shubber, S. Uddin, C.F. van der Walle,
P. Caliceti, S. Salmaso, A novel combined strategy for the physical PEGylation of
polypeptides, J. Control. Release 226 (2016) 35–46.
˜
[22] L. Pineiro, M. Novo, W. Al-Soufi, Fluorescence emission of pyrene in surfactant
solutions, Adv. Colloid Interf. Sci. 215 (2015) 1–12.
[23] J. Aguiar, P. Carpena, J.A. Molina-Bolıvar, C.C. Ruiz, On the determination of the
critical micelle concentration by the pyrene 1: 3 ratio method, J. Colloid Interface
Sci. 258 (2003) 116–122.
[24] H.T.T. Duong, F. Hughes, S. Sagnella, M. Kavallaris, A. Macmillan, R. Whan,
J. Hook, T.P. Davis, C. Boyer, Functionalizing biodegradable dextran scaffolds
using living radical polymerization: new versatile nanoparticles for the delivery of
therapeutic molecules, Mol. Pharm. 9 (2012) 3046–3061.
Acknowledgement
[25] M.B. Hansen, S.E. Nielsen, K. Berg, Re-examination and further development of a
precise and rapid dye method for measuring cell growth/cell kill, J. Immunol.
Methods 119 (1989) 203–210.
This work was supported by the Italian Ministry of Education, Uni-
versity and Research (MIUR) through the “Progetti di ricerca di Rile-
vante Interesse Nazionale” (PRIN) scheme (MIUR grant code:
20175XBSX4).
[26] C. Bastiancich, K. Vanvarenberg, B. Ucakar, M. Pitorre, G. Bastiat, F. Lagarce,
´
V. Preat, F. Danhier, Lauroyl-gemcitabine-loaded lipid nanocapsule hydrogel for
the treatment of glioblastoma, J. Control. Release 225 (2016) 283–293.
[27] C.J. Zeiss, D.M. Gatti, O. Toro-Salazar, C. Davis, C.M. Lutz, F. Spinale, T. Stearns,
M.B. Furtado, G.A. Churchill, Doxorubicin-induced cardiotoxicity in Collaborative
Cross (CC) mice recapitulates individual cardiotoxicity in humans, G3 (Bethesda,
Md.) 9 (2019) 2637–2646.
Appendix A. Supplementary data
[28] V.G. Desai, E.H. Herman, C.L. Moland, W.S. Branham, S.M. Lewis, K.J. Davis, N.
I. George, T. Lee, S. Kerr, J.C. Fuscoe, Development of doxorubicin-induced chronic
cardiotoxicity in the B6C3F1 mouse model, Toxicol. Appl. Pharmacol. 266 (2013)
109–121.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.jconrel.2021.05.010.
[29] V.G. Desai, C.K. Joshua, V. Vijay, C.L. Moland, E.H. Herman, T. Lee, T. Han, S.
M. Lewis, K.J. Davis, L. Muskhelishvili, S. Kerr, J.C. Fuscoe, Early biomarkers of
doxorubicin-induced heart injury in a mouse model, Toxicol. Appl. Pharmacol. 281
(2014) 221–229.
References
[1] W.-l. Ye, J.-b. Du, R. Na, Y.-f. Song, Q.-b. Mei, M.-g. Zhao, S.-y. Zhou, Cellular
uptake and antitumor activity of DOX-hyd-PEG-FA nanoparticles, PLoS One 9
(2014), e97358.
[30] W.J. Reagan, M. York, B. Berridge, E. Schultze, D. Walker, S. Pettit, Comparison of
cardiac troponin I and T, including the evaluation of an ultrasensitive assay, as
indicators of doxorubicin-induced cardiotoxicity, Toxicol. Pathol. 41 (2013)
1146–1158.
[2] C. LoPresti, H. Lomas, M. Massignani, T. Smart, G. Battaglia, Polymersomes: nature
inspired nanometer sized compartments, J. Mater. Chem. 19 (2009) 3576–3590.
[3] K. Miyata, R.J. Christie, K. Kataoka, Polymeric micelles for nano-scale drug
delivery, React. Funct. Polym. 71 (2011) 227–234.
[31] C. Deng, J. Wu, R. Cheng, F. Meng, H.-A. Klok, Z. Zhong, Functional polypeptide
and hybrid materials: precision synthesis via α-amino acid N-carboxyanhydride
polymerization and emerging biomedical applications, Prog. Polym. Sci. 39 (2014)
330–364.
[4] N. Zhao, M.C. Woodle, A.J. Mixson, Advances in delivery systems for doxorubicin,
J. Nanomed. Nanotechnol. 9 (2018).
[32] Y. Bae, S. Fukushima, A. Harada, K. Kataoka, Design of environment-sensitive
supramolecular assemblies for intracellular drug delivery: polymeric micelles that
are responsive to intracellular pH change, Angew. Chem. 115 (2003) 4788–4791.
[33] S.S. Banerjee, N. Aher, R. Patil, J. Khandare, Poly (ethylene glycol)-prodrug
conjugates: concept, design, and applications, J. Drug Deliv. 2012 (2012) 103973.
[34] A.M. Eckman, E. Tsakalozou, N.Y. Kang, A. Ponta, Y. Bae, Drug release patterns
and cytotoxicity of PEG-poly (aspartate) block copolymer micelles in cancer cells,
Pharm. Res. 29 (2012) 1755–1767.
[5] H. Xu, Q. Yao, C. Cai, J. Gou, Y. Zhang, H. Zhong, X. Tang, Amphiphilic poly
(amino acid) based micelles applied to drug delivery: the in vitro and in vivo
challenges and the corresponding potential strategies, J. Control. Release 199
(2015) 84–97.
[6] A. Ponta, Y. Bae, PEG-poly (amino acid) block copolymer micelles for tunable drug
release, Pharm. Res. 27 (2010) 2330–2342.
[7] E. Bernabeu, M. Cagel, E. Lagomarsino, M. Moretton, D.A. Chiappetta, Paclitaxel:
what has been done and the challenges remain ahead, Int. J. Pharm. 526 (2017)
474–495.
[35] Z. Amoozgar, Y. Yeo, Recent advances in stealth coating of nanoparticle drug
delivery systems, in: Wiley Interdisciplinary Reviews: Nanomedicine and
Nanobiotechnology 4, 2012, pp. 219–233.
[8] H. Maeda, J. Wu, T. Sawa, Y. Matsumura, K. Hori, Tumor vascular permeability
and the EPR effect in macromolecular therapeutics: a review, J. Control. Release 65
(2000) 271–284.
[36] A. Dondoni, The emergence of thiol–ene coupling as a click process for materials
and bioorganic chemistry, Angew. Chem. Int. Ed. 47 (2008) 8995–8997.
[37] X. Zhang, W. Tang, Z. Yang, X. Luo, H. Luo, D. Gao, Y. Chen, Q. Jiang, J. Liu,
Z. Jiang, PEGylated poly (amine-co-ester) micelles as biodegradable non-viral gene
vectors with enhanced stability, reduced toxicity and higher in vivo transfection
efficacy, J. Mater. Chem. B 2 (2014) 4034–4044.
[9] K.N. Sill, B. Sullivan, A. Carie, J.E. Semple, Synthesis and characterization of
micelle-forming PEG-poly (amino acid) copolymers with iron-hydroxamate cross-
linkable blocks for encapsulation and release of hydrophobic drugs,
Biomacromolecules 18 (2017) 1874–1884.
[10] H. Cabral, K. Miyata, K. Osada, K. Kataoka, Block copolymer micelles in
nanomedicine applications, Chem. Rev. 118 (2018) 6844–6892.
[11] Y. Bae, W.-D. Jang, N. Nishiyama, S. Fukushima, K. Kataoka, Multifunctional
polymeric micelles with folate-mediated cancer cell targeting and pH-triggered
drug releasing properties for active intracellular drug delivery, Mol. BioSyst. 1
(2005) 242–250.
[38] J. Vega, S. Ke, Z. Fan, S. Wallace, C. Charsangavej, C. Li, Targeting doxorubicin to
epidermal growth factor receptors by site-specific conjugation of C225 to poly (L-
glutamic acid) through a polyethylene glycol spacer, Pharm. Res. 20 (2003)
826–832.
36

S. Brunato et al.
Journal of Controlled Release 335 (2021) 21–37
ˇ
`
[39] V. Pliska, M. Schmidt, J.-L. Fauchere, Partition coefficients of amino acids and
[55] E. Markovsky, H. Baabur-Cohen, R. Satchi-Fainaro, Anticancer polymeric
nanomedicine bearing synergistic drug combination is superior to a mixture of
individually-conjugated drugs, J. Control. Release 187 (2014) 145–157.
[56] A. Balasso, S. Salmaso, P. Pontisso, A. Rosato, S. Quarta, A. Malfanti, F. Mastrotto,
P. Caliceti, Re-programming pullulan for targeting and controlled release of
doxorubicin to the hepatocellular carcinoma cells, Eur. J. Pharm. Sci. 103 (2017)
104–115.
hydrophobic parameters
π of their side-chains as measured by thin-layer
chromatography, J. Chromatogr. A 216 (1981) 79–92.
[40] Y. Bae, N. Nishiyama, S. Fukushima, H. Koyama, M. Yasuhiro, K. Kataoka,
Preparation and biological characterization of polymeric micelle drug carriers with
intracellular pH-triggered drug release property: tumor permeability, controlled
subcellular drug distribution, and enhanced in vivo antitumor efficacy, Bioconjug.
Chem. 16 (2005) 122–130.
[57] U. Beyer, T. Roth, P. Schumacher, G. Maier, A. Unold, A.W. Frahm, H.H. Fiebig,
C. Unger, F. Kratz, Synthesis and in vitro efficacy of transferrin conjugates of the
anticancer drug chlorambucil, J. Med. Chem. 41 (1998) 2701–2708.
[58] L.P. Belnap, P.H. Cleveland, M.E.M. Colmerauer, R.M. Barone, Y.H. Pilch,
Immunogenicity of chemically induced murine colon cancers, Cancer Res. 39
(1979) 1174–1179.
[41] Y. Bae, K. Kataoka, Intelligent polymeric micelles from functional poly (ethylene
glycol)-poly (amino acid) block copolymers, Adv. Drug Deliv. Rev. 61 (2009)
768–784.
[42] O. Zagorodko, J.J. Arroyo-Crespo, V.J. Nebot, M.J. Vicent, Polypeptide-based
conjugates as therapeutics: opportunities and challenges, Macromol. Biosci. 17
(2017) 1600316.
[59] S. Lv, M. Li, Z. Tang, W. Song, H. Sun, H. Liu, X. Chen, Doxorubicin-loaded
amphiphilic polypeptide-based nanoparticles as an efficient drug delivery system
for cancer therapy, Acta Biomater. 9 (2013) 9330–9342.
[43] S.C. Owen, D.P.Y. Chan, M.S. Shoichet, Polymeric micelle stability, Nano Today 7
(2012) 53–65.
[44] H. Phan, R.I. Minut, P. McCrorie, C. Vasey, R.R. Larder, E. Krumins, M. Marlow,
R. Rahman, C. Alexander, V. Taresco, A.K. Pearce, Role of Self-Assembly
Conditions and Amphiphilic Balance on Nanoparticle Formation of PEG-PDLLA
Copolymers in Aqueous Environments 57, 2019, pp. 1801–1810.
[60] C.Y.X. Chua, J. Ho, S. Demaria, M. Ferrari, A. Grattoni, Emerging Technologies for
Local Cancer Treatment 3, 2020, p. 2000027.
[61] B.A. Pulaski, S. Ostrand-Rosenberg, Mouse 4T1 breast tumor model, Curr. Protoc.
Immunol. 39 (2000) 20–22.
´
[45] F. Danhier, O. Feron, V. Preat, To exploit the tumor microenvironment: passive and
[62] T. Tejada-Berges, C.O. Granai, M. Gordinier, W. Gajewski, Caelyx/Doxil for the
treatment of metastatic ovarian and breast cancer, Expert. Rev. Anticancer. Ther. 2
(2002) 143–150.
active tumor targeting of nanocarriers for anti-cancer drug delivery, J. Control.
Release 148 (2010) 135–146.
[46] Y. Matsumura, H. Maeda, A new concept for macromolecular therapeutics in
cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and
the antitumor agent smancs, Cancer Res. 46 (1986) 6387–6392.
[47] V. Torchilin, Tumor delivery of macromolecular drugs based on the EPR effect,
Adv. Drug Deliv. Rev. 63 (2011) 131–135.
[63] M.E.R. O’Brien, N. Wigler, M. Inbar, R. Rosso, E. Grischke, A. Santoro, R. Catane,
D.G. Kieback, P. Tomczak, S.P. Ackland, Reduced cardiotoxicity and comparable
efficacy in a phase III trial of pegylated liposomal doxorubicin HCl (CAELYXTM/
Doxil®) versus conventional doxorubicin for first-line treatment of metastatic
breast cancer, Ann. Oncol. 15 (2004) 440–449.
¨
[48] S. Raddatz, J. Mueller-Ibeler, J. Kluge, L. Waß, G. Burdinski, J.R. Havens, T.
[64] M. Rahimi, K.D. Safa, E. Alizadeh, R. Salehi, Dendritic chitosan as a magnetic and
biocompatible nanocarrier for the simultaneous delivery of doxorubicin and
methotrexate to MCF-7 cell line, New J. Chem. 41 (2017) 3177–3189.
[65] A. Maksimenko, F. Dosio, J. Mougin, A. Ferrero, S. Wack, L.H. Reddy, A.A. Weyn,
E. Lepeltier, C. Bourgaux, B. Stella, L. Cattel, P. Couvreur, A unique squalenoylated
and nonpegylated doxorubicin nanomedicine with systemic long-circulating
properties and anticancer activity, Proc. Natl. Acad. Sci. U. S. A. 111 (2014)
E217–E226.
J. Onofrey, D. Wang, M. Schweitzer, Hydrazide oligonucleotides: new chemical
modification for chip array attachment and conjugation, Nucleic Acids Res. 30
(2002) 4793–4802.
[49] M. Li, W. Song, Z. Tang, S. Lv, L. Lin, H. Sun, Q. Li, Y. Yang, H. Hong, X. Chen,
Nanoscaled poly (L-glutamic acid)/doxorubicin-amphiphile complex as pH-
responsive drug delivery system for effective treatment of nonsmall cell lung
cancer, ACS Appl. Mater. Interfaces 5 (2013) 1781–1792.
[50] S. Salmaso, A. Semenzato, S. Bersani, F. Mastrotto, A. Scomparin, P. Caliceti, Site-
selective protein glycation and PEGylation, Eur. Polym. J. 44 (2008) 1378–1389.
[66] F.A. Boratto, M.S. Franco, A.L.B. Barros, G.D. Cassali, A. Malachias, L.A.
M. Ferreira, E.A. Leite, Alpha-tocopheryl succinate improves encapsulation, pH-
sensitivity, antitumor activity and reduces toxicity of doxorubicin-loaded
liposomes, Eur. J. Pharm. Sci. 144 (2020) 105205.
˜´
[51] J.J. Arroyo-Crespo, A. Arminan, D. Charbonnier, L. Balzano-Nogueira, F. Huertas-
´
Lopez, C. Martí, S. Tarazona, J. Forteza, A. Conesa, M.J. Vicent, Tumor
microenvironment-targeted poly-L-glutamic acid-based combination conjugate for
[67] C. Stark, P. Taimen, T. Savunen, J. Koskenvuo, Pegylated and liposomal
doxorubicin is associated with high mortality and causes limited cardiotoxicity in
mice, BMC Res. Notes 11 (2018) 148.
enhanced triple negative breast cancer treatment, Biomaterials 186 (2018) 8–21.
´
[52] F. Danhier, T.T.B. Kouhe, N. Duhem, B. Ucakar, A. Staub, N. Draoui, O. Feron,
´
V. Preat, Vitamin E-based micelles enhance the anticancer activity of doxorubicin,
[68] R. Zhang, Y. Zhang, Y. Zhang, X. Wang, X. Gao, Y. Liu, X. Zhang, Z. He, D. Wang,
Y. Wang, Ratiometric delivery of doxorubicin and berberine by liposome enables
superior therapeutic index than DoxilⓇ, Asian J. Pharm. Sci. 15 (2020) 385–396.
[69] T. Safra, F. Muggia, S. Jeffers, D.D. Tsao-Wei, S. Groshen, O. Lyass, R. Henderson,
G. Berry, A. Gabizon, Pegylated liposomal doxorubicin (doxil): reduced clinical
cardiotoxicity in patients reaching or exceeding cumulative doses of 500 mg/m2,
Ann. Oncol. 11 (2000) 1029–1033.
Int. J. Pharm. 476 (2014) 9–15.
[53] X. Gao, B. Wang, X. Wei, W. Rao, F. Ai, F. Zhao, K. Men, B. Yang, X. Liu, M. Huang,
Preparation, characterization and application of star-shaped PCL/PEG micelles for
the delivery of doxorubicin in the treatment of colon cancer, Int. J. Nanomedicine
8 (2013) 971–982.
[54] C.J. Lovitt, T.B. Shelper, V.M. Avery, Doxorubicin resistance in breast cancer cells
is mediated by extracellular matrix proteins, BMC Cancer 18 (2018) 41.
37