Dual cancer targeting using estrogen functionalized chitosan nanoparticles loaded with doxorubicin-estrone conjugate: A quality by design approach

Article abstract of DOI:10.1016/j.ijbiomac.2020.08.172

In this study, estrone was used as targeting functionality in chitosan nanoparticles (DoxEs-CSEsNPs) carrying doxorubicin-estrone conjugate for dual targeted intracellular delivery to breast cancer cells. Estrone was conjugated with Dox and CS and characterized by FTIR and FT-NMR spectroscopy. Dox/DoxEs containing CSEsNPs were prepared with ionic gelation method and for the effect of formulation variables a 3-factor, 3-level Box-Behnken design (BBD) was explored, which predict the responses like particle size (Y₁) and percent entrapment efficiency (%EE) (Y₂) when CSEs: TPP ratio (X₁), sonication time (X₂) and stirring speed (X₃) were selected as independent variables. The Dox-CSEsNPs and DoxEs-CSEsNPs were characterized for size, shape, PDI, surface charge and thermal analysis. The drug entrapment efficiency was 66.33 ± 2.82% and 62.25 ± 2.63% for Dox-CSEsNPs and DoxEs-CSEsNPs formulation respectively. The in vitro release, haemolytic toxicity, and fluorescent microscopy studies were also assessed. Anticancer activity on the MCF-7 cell line indicated the higher potency of DoxEs-CSEsNPs as compared to Dox-CSEsNPs, DoxEs, and Dox solution. The findings are decisive for selective targeting of antineoplastic agents to the ERs, which indicate that the DoxEs loaded CSEsNPs were able to significantly improve the efficacy of Dox.

Full text of DOI:10.1016/j.ijbiomac.2020.08.172

International Journal of Biological Macromolecules 164 (2020) 2881–2894
Contents lists available at ScienceDirect
International Journal of Biological Macromolecules
journal homepage: http://www.elsevier.com/locate/ijbiomac
Dual cancer targeting using estrogen functionalized chitosan
nanoparticles loaded with doxorubicin-estrone conjugate: A quality by
design approach
Balak Das Kurmi a, Rishi Paliwal b, Shivani Rai Paliwal a,
⁎
a
Pharmaceutics Research Laboratory, SLT Institute of Pharmaceutical Sciences, Guru Ghasidas Vishwavidyalaya, Bilaspur 495009, CG, India
Nanomedicine and Bioengineering Research Laboratory, Department of Pharmacy, Indira Gandhi National Tribal University, Amarkantak 4848487, MP, India
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 26 March 2020
Received in revised form 14 August 2020
Accepted 21 August 2020
Available online 24 August 2020
In this study, estrone was used as targeting functionality in chitosan nanoparticles (DoxEs-CSEsNPs) carrying
doxorubicin-estrone conjugate for dual targeted intracellular delivery to breast cancer cells. Estrone was conju-
gated with Dox and CS and characterized by FTIR and FT-NMR spectroscopy. Dox/DoxEs containing CSEsNPs
were prepared with ionic gelation method and for the effect of formulation variables a 3-factor, 3-level Box-
Behnken design (BBD) was explored, which predict the responses like particle size (Y1) and percent entrapment
efficiency (%EE) (Y2) when CSEs: TPP ratio (X1), sonication time (X2) and stirring speed (X3) were selected as in-
dependent variables. The Dox-CSEsNPs and DoxEs-CSEsNPs were characterized for size, shape, PDI, surface
charge and thermal analysis. The drug entrapment efficiency was 66.33 2.82% and 62.25 2.63% for Dox-
CSEsNPs and DoxEs-CSEsNPs formulation respectively. The in vitro release, haemolytic toxicity, and fluorescent
microscopy studies were also assessed. Anticancer activity on the MCF-7 cell line indicated the higher potency
of DoxEs-CSEsNPs as compared to Dox-CSEsNPs, DoxEs, and Dox solution. The findings are decisive for selective
targeting of antineoplastic agents to the ERs, which indicate that the DoxEs loaded CSEsNPs were able to signif-
icantly improve the efficacy of Dox.
Keywords:
Nanoparticles
Chitosan
Doxorubicin
Estrogen receptor
Dual targeting
Quality by design
Box–Behnken design
© 2020 Published by Elsevier B.V.
1. Introduction
estrogen receptors are members of nuclear receptor super family; how-
ever, in cancerous conditions, they remain frequently present on cell
Breast cancer is a frequently occurring cancer among women and
presents significant challenges for its treatment. Conventional breast
cancer therapy has several obstacles such as non-specific
biodistribution of the drug, low intracellular drug concentration, and oc-
currence of multidrug resistance [1,2]. Tumor cell targeting could be an
option for breast cancer treatment that would evade the limitations of
conventional chemotherapy. A number of molecular targets have been
identified on the basis of pathophysiology of cancer cells that can be uti-
lized for targeted breast carcinoma therapy [3,4].
Breast cancer is classified on the basis of expression of hormone re-
ceptors or target protein such as hormone receptor positive cancer (i.e.
estrogen receptor or progesterone receptor), epidermal growth recep-
tor (EGFR or HER2) over expressed cancer and triple negative breast
cancer (TNBC) that do not express any of these receptors means ER (-
ve), PR (-ve) and HER 2 (-ve) as well [5]. The hormone responsive breast
cancers are estrogen receptor positive (ER+), progesterone receptor
positive (PR+) or both and among them majority of hormone-
responsive breast cancer are ER+ (approximately 75%). Though,
surface as well [6]. These cell surface estrogen receptors are associated
with the growth and proliferation of breast cancer. Such high level of ex-
pression of estrogen receptors has been widely exploited for targeted
drug delivery as drug conjugates [7,8], as targeted nanocarriers [9,10]
and also for gene delivery [11] to ER+ breast cancers. However, dual
targeting of both nuclear ER and cell surface ER using specific drug deliv-
ery design is not reported till date.
Targeted nanoparticles have been widely explored for site specific
drug delivery of chemotherapeutic agents with improved pharmacoki-
netics and lower systemic toxicity [12]. It is noteworthy that the proper-
ties of selected polymeric material for the preparation of such targeted
nanoparticle play an important role as it affect not only the clinical per-
formance of the delivery system but also affect the regulatory clearance
at the time of translation of the product from bench to bedside. Chito-
san, a natural biopolymer, is biocompatible, biodegradable, and safe ma-
terial for the construction of nanoparticles. It offers modification
opportunities due to the presence of several amino (−NH2) group in
its chemical structure that are resourceful for the tethering of various li-
gands and transforming it into a suitable carrier for biomedical applica-
tion such as drug/gene targeting [13]. Doxorubicin (Dox) is one of the
most popular and effective anticancer molecules commonly used in
⁎
Corresponding author.
E-mail address: srai2k@gmail.com (S.R. Paliwal).
https://doi.org/10.1016/j.ijbiomac.2020.08.172
0141-8130/© 2020 Published by Elsevier B.V.

2882
B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
breast cancer chemotherapy. However, its clinical use is associated with
dose-dependent toxicities such as myelosuppression and cardiotoxicity.
This may be turned up in to a clinically appreciable formulation based
product using nanotechnological transformation altering improved effi-
cacy and limiting side-effects [14].This is why a large number of drug-
conjugates and Dox encapsulated carriers have been developed till
date to achieve targeted Dox delivery and to circumvent its clinical
side-effects [15].
Conventional methods of experimentation and optimization con-
sider only one variable at a time and hence factor interactions cannot
be determined. Recently, quality by design (QbD) has been recom-
mended by ICH (Q8) guidelines to optimize the critical process param-
eters for getting the desired target quality in the product. Design of
experiment (DoE) procedure commences with predetermined objec-
tives and focuses on a better understanding of process parameters and
product design. Further, DoE is used to correlate the relationship be-
tween independent variables (factors) and dependent variables (re-
sponses). DoE produces more reliable results with a lesser number of
experiments with an added advantage of extrapolation of data by plot-
ting the results. A response surface methodology has been explored to
identify the response of independent variables for the optimization of
pharmaceutical formulations. Among several design options, Box-
Benhken design is the most commonly applied design with a minimum
run of experiments [16].
filtered. The obtained filtrate was further concentrated and precipitated
with cold diethyl ether and air-dried to yield carboxylic acid derivative
of Es (i.e., Es-SA). Further, Dox containing -NH2 group and COOH-
estrone was conjugated with corbodimide chemistry using N,N′-DCC
and N-NHS. Briefly, the Es-SA (40 mg, 0.1 mmol) was reacted with
Dox (87 mg, 0.15 mmol) in the presence of DCC (61.8 mg, 0.3 mmol),
NHS (34.5 mg, 0.3 mmol) and TEA (42 μL, 0.3 mmol) in DMSO at RT
under nitrogen (N2) atmosphere for 24 h. The product was filtered to re-
move N,N-dicyclohexylurea (DCU) and then lyophilized (Labconco 4.5
L, Freezone Plus cascade benchtop freez dryer, USA) to remove DMSO.
2.2.2. Conjugation of estrone with chitosan
To obtain Es conjugated CS (CSEs) polymer, an amidation reaction
was applied as described by Guo et al., 2014 with slight modification
[19]. Functionalized Es-SA as obtained earlier was conjugated with pri-
mary amine groups of CS by using NHS/EDC mediated carbodimide cou-
pling reaction. The Es-SA (80 mg, 0.2 mmol), NHS (34.5 mg, 0.3 mmol)
and EDC (57.5 mg, 0.3 mmol) were taken in anhydrous DCM for 24 h at
room temperature under constant stirring (Remi 2-MLH). After evapo-
ration of the solvent, the product was added into CS acetic acid solution
(1% v/v, pH = 4). After 24 h the reaction was completed and the conju-
gate was dialyzed against distilled water for 72 h in order to remove free
reactants and lyophilized (Labconco 4.5 L, Freezone Plus cascade bench-
top freeze dryer).
Thus, considering all these aspects, we hypothesized to construct Es
conjugated Dox incorporated chitosan-estrone nanoparticles (CSEsNPs)
formulation further called as DoxEs-CSEsNPs in order to explore dual ER
targeting at the nuclear and cellular level both. BBD design was applied
to optimize the process parameters for the preparation of nanoparticles.
Here, Dox-Es conjugate may play a dual role. First, when the conjugate is
reached in the intracellular compartment of cancer cells, it will bind to
ERs which carries Dox-Es towards the nucleus i.e., the site of doxorubi-
cin action and hence may limit efflux of Dox. Second, Dox-Es, if leached
from nanoparticles during transit in the blood pool; conjugated Dox-Es
may navigate the Dox to the target site by protecting it from non-
specific distribution to non-target sites. The Dox/DoxEs containing
CSEsNPs preparations were investigated for different characterization
parameters. The qualitative localization study was performed using
fluorescence microscopy. Cytotoxic potential and targetability was
assessed on MCF-7 cell lines. The developed formulations were also
evaluated for in vivo performance on tumor-bearing rat model where
various pharmacokinetic parameters were also determined.
2.2.3. Characterization of DoxES and CSEs conjugates
The conjugation of CSEs was confirmed by FT-IR (8400S, Shimadzu)
and FT-NMR (Bruker's AVANCE-III, 500 MHz) spectroscopy for various
shifts and peaks and interpreted for different groups.
2.3. Preparation of chitosan nanoparticles
Dox and DoxEs loaded chitosan nanoparticle was prepared by
ionic gelation of CSEs with TPP with slight modification on our previ-
ously reported method Fig. 1B [20]. Firstly, CSEs was dissolved in
acidic solution containing 1.0% (v/v) glacial acetic acid. Then pH of
the solution was increased to 4.7 by adding 0.1 N NaOH. The TPP
was added to the above solution with different CSEs/TPP ratio
under constant stirring (Remi 2-MLH). After 1 h stirring, the solution
was sonicated (PCiTM 3.5 L 100) and centrifuged (Remi, C-24) at
15000 rpm for 30 min on glycerol bed. The CSEs nanoparticles were
collected at the bottom of the centrifuge tube as a transparent gel
pellet. The pellet of chitosan nanoparticles was dried by a freeze
dryer (Labconco 4.5 L, Freezone Plus cascade benchtop freez dryer,
USA) before characterization. In order to get drug entrapment, the
drug (Dox/DoxEs) was added in to the CSEs solution in glacial acetic
acid before the addition of TPP solution.
2. Materials and methods
2.1. Materials
Chitosan (CS), estrone (Es), sodium tripolyphosphate (TPP), succinic
anhydride (SA), dimethylaminopyridine (DMAP), N hydroxysuccinimide
(NHS), dicyclohexylcarbidiimide (DCC), triethylamine (TEA), N-(3-
dimethylaminopropyl)-N-ethylcarbidiimide (EDC) and dialysis mem-
brane (MWCO 12–14 kDa) were purchased from Himedia, India. Doxoru-
bicin was provided by Sun Pharma (Vadodara, India). All other reagents
and solvent were either of analytical or HPLC grade.
2.3.1. Experimental design
For the optimization of formulation parameters, a 3-factor and 3-
level Box-Behnken (BBD) design was employed using Design-Expert
Software (Stat-Ease Inc., Minneapolis, MN). This design was selected be-
cause it requires only three levels of each independent variables and ex-
ploring quadratic response surfaces with second-order polynomial
models suggesting the minimum number of experimental runs capable
to indicate the variable interaction. It does not contain the extreme level
of variables which may leads to difficulties with respect to formulation
development and unsatisfactory results.
The independent variables were, (X1) CSEs to TPP ratio, (X2)
sonication time and (X3) the stirring speed, with 3 level viz., low
level (−1), mid-level (0), high level (+1) respectively, while depen-
dent variables were (Y1) particle size and (Y2) percent entrapment
efficiency. Table S1 represents the level of these dependent and inde-
pendent variables. Design matrix consisting of 17 experimental runs
including five centre points was constructed. The computer-
2.2. Methods
2.2.1. Conjugation of Es with Dox
DoxEs conjugate was synthesized using the method described by Rai
et al (2008) and Cao et al (2008) with some modifications [17,18]
(Fig. 1A). Briefly, Es was activated using SA, for that Es (270 mg,
1 mmol), SA (150 mg, 1.5 mmol), DMAP (122 mg, 1 mmol), and TEA
(139 μL, 1 mmol) were dissolved in dioxane, and the resulting solution
was stirred (Remi 2-MLH) overnight at room temperature (RT). The di-
oxane was evaporated under vacuum (MAC New Delhi), completely
and the residue was dissolved in a minimum quantity of DCM then

B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
2883
Fig. 1. [A] Schematic diagram demonstrating conjugation of estrone with Dox (DoxEs) and chitosan(CSEs); [B] Scheme of method of preparation of chitosan nanoparticles.
generated nonlinear quadratic model equation of the design is as
follows:
(R2), percent coefficient of variance (% CV), predicted residual sum of
squares (PRSS) etc. with the level of significance at p < 0.05. Response
surface analysis applied for measuring the effect of the independent var-
iables on formulation attributes. The resulted perturbation and 3D plots
were critically observed for showing the direct effect and interactions
which was considered for the selection of the experimental model.
The effect of factors on responses was analyzed by BBD and optimiza-
tion of the formulation was done to attain a lower particle size of NPs
and higher percent entrapment efficiency using overlay graph.
2
Y ¼ A0 þ A1X1 þ A2X2 þ A3X3 þ A4X1X2 þ A5X2X3 þ A6X1X3 þ A7X1
2
2
þ A8X2 þ A9X3
where, A0: the intercept representing the arithmetic average of all of 17
runs; A1-A9: the regression coefficient obtained from the results of ex-
perimental response to dependent variable Y; X1, X2, and X3: are the in-
dependent variables; X1X2, X2X3, and X1X3 are the interaction terms
which represents the effect on the response on a variation of two factors
simultaneously; X12, X22, and X23: quadratic terms.
Seventeen batches of CSEsNPs were formulated as suggested by de-
sign expert and the effect of factors on responses were evaluated
(Table 1). The best-fitting model or the quality of the model was evalu-
ated after considering different statistically parameters obtained by
analysis of variance (ANOVA) like p value, coefficient of determination
2.3.2. Differential scanning calorimetry
Differential scanning calorimetry characterizations of CS, Es, CSEs,
Dox, and DoxEs-CSEsNPs were carried out with differential scanning
calorimeter (DSC 4000, PerkinElmer). Each sample was scanned be-
tween 40 °C and 350 °C and blank aluminium pans were employed as
reference. The temperature of maximal excess heat capacity was de-
fined as the phase transition temperature.

2884
B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
Table 1
2.4. Haemolysis studies
Observed response obtained after performing the factorial runs given by BBD.
Fresh human blood sample (1.0 mL) was collected into vials
(HiAnticlot Vials Flat Bottom) and immediately centrifuged at 2000 ×g
for 5 min leading to separate RBCs. Obtained RBC pellet was washed
3–4 times with normal saline solution and re-suspended into saline so-
lution. Further, about 1.0 mL of the RBCs suspension was added to
1.0 mL of either of 0.01, 0.02, 0.03, 0.04, and 0.05%, w/v solution of differ-
ent formulations (i.e., Dox, DoxEs, Dox-CSEsNPs, DoxEs-CSEsNPs), dou-
ble distilled water (100% haemolysis) and normal saline (blank
samples). These tubes were incubated for 1 h at 37 °C with gentle
hand shaking. On completion of incubation, centrifugation was done
for 5 min at 2000 ×g. The obtained supernatant was analyzed by UV–
Vis spectrophotometer (UV-1800, Shimadzu) at ʎmax415 nm as re-
ported elsewhere [21].
Runs
Factor 1
Factor 2
B:ST
Factor 3
C:SS
Response 1
Response 2
A:CSEs:TPP
% wt/wt
PS
EE
%
Sec
/Sec
nm
1
2
3
4
5
6
7
8
−1
0
0
−1
0
1
1
0
0
−1
−1
0
1
1
0
0
0
0
−1
1
0
0
1
0
0
1
−1
1
−1
0
−1
0
0
1
1
1
−1
0
0
1
0
−1
0
0
−1
−1
0
0
0
366.42
257.66
213.23
387.12
263.58
134.41
131.78
268.26
196.32
394.54
357.34
292.12
169.98
173.56
246.27
237.61
240.18
66.24
69.32
60.87
68.45
62.32
52.64
53.91
66.37
61.12
69.35
67.21
62.38
57.82
58.64
61.37
59.43
60.12
9
10
11
12
13
14
15
16
17
2.5. Cytotoxicity studies
The cytotoxicities of the developed formulations were estimated
using the MTT assay. The MCF-7 cells were treated with free Dox,
DoxEs, Dox-CSEsNPs and DoxEs-CSEsNPs formulations at the equivalent
concentration of Dox (1, 5 and 10 μg/mL). After 48 h, cell lines were ex-
posed to MTT solution (20 μL, 5 mg/mL in PBS). These cells were washed
after 4 h followed by addition of dimethyl sulfoxide to ensure complete
dissolution of formed formazan crystals. The absorbance at 570 nm was
measured using microplate reader (BIO-RAD, model 680, UK) compar-
ing with untreated control cells. Percent cell viability was than deter-
mined at different concentrations. To study the role of estrogen
receptor-mediated uptake a competitive binding in presence of free es-
trone was also performed. The cells were pre-incubated with estrone for
1 h before treatment with the formulations.
0
0
2.3.3. Determination of particle size, polydispersity index (PDI), zeta-
potential and morphology of chitosan nanoparticles
The average particle size, PDI, and zeta potential of nanoparticles
was determined on the basis of dynamic light scattering principle
using Zetasizer (ZS90 zetasizer, Malvern Instrument, UK). All measure-
ments were performed in triplicate at 25 °C.The morphology of CSEsNPs
were examined by a high-performance digital imaging TEM (TECNAI
200 Kv TEM, Fei, Electron Optics). One drop of the suspended solution
was spread onto a carbon-coated copper grid and stained with 2% (w/
v) phosphotungstic acid. After drying at RT the samples were placed
for TEM analysis using an accelerating voltage of 200 kV.
2.6. In vivo studies
The studies were performed in compliances with committee for the
purpose of control and supervision of experiments on animal, Ministry
of culture, Government of India and the study protocols were also ap-
proved by Institutional Animal Ethics Committee (Ref. No. 224/IAEC/
Pharmacy/2018) Guru Ghasidas Vishwavidyalaya, Bilaspur, CG, India.
2.3.4. Determination of drug entrapment efficiency
Percent entrapment efficiency (EE) of Dox in CSEsNPs was estimated
by an indirect estimation of free drug. The free unentrapped drug was
removed after centrifugation by supernatant with the washing of nano-
particles at least four times for complete removal of free drug and esti-
mated using UV spectrophotometer at λmax 480 nm (UV-1800,
Shimadzu).
2.6.1. Blood level and biodistribution of Dox
Female rats (Spargue-Dawley, 40–45 days) were taken and kept
under standard conditions and fed with laboratory food. After 10 days,
DMBA in soyabean oil (65 mg/kg) was administered orally to induce
breast cancer. Approximately after 3 months, tumor had developed in
the animals. In vivo blood levels and biodistribution patterns were esti-
mated on these tumor bearing animals. Briefly, tumor-induced animals
were randomly distributed into five groups. The first group received sa-
line and served as control. While second, third, fourth, and fifth group
animals received free Dox, DoxEs, Dox-CSEsNPs, and DoxEs-CSEsNPs
formulations respectively intravenously at the equivalent dose of Dox
at 5 mg/kg body weight via tail vein injection. Animals were sacrificed
at different time intervals for the collection of visceral organs such as
heart, lungs, liver, kidney, and tumor immediately. The blood samples
were also collected every time. Dox was then extracted from the col-
lected tissues using methanol after homogenization. Amount of Dox
was quantified in blood and organs/tissue by using a previously re-
ported HPLC method (YL9100 HPLC) [22]. The data were normalized
to tissue weight.
2.3.5. In vitro drug release profile
The dialysis tube-based diffusion technique was used for the assess-
ment of in vitro drug release behaviour of the formulations. Two-
milliliter solution of the selected formulation was placed in the dialysis
sac of benzoylated dialysis membrane (MWCO, 12–14 kDa, Himedia),
hermetically tied, and immediately suspended in 20 mL of aqueous
phosphate buffer saline (PBS pH 7.4) receptor medium. The sink condi-
tion was maintained in the receptor compartment through constant
stirring using magnetic stirrer and withdrawing large samples (Remi
2-MLH) at 37 0.5 °C. The Dox release was determined in withdrawn
aliquot, after appropriate dilution, using UV spectrophotometer (UV-
1800, Shimadzu).
2.3.6. Effect of pH on pharmaceutical characteristics
The pH of the biological mediums may affect the formulation charac-
teristics. To evidence any such change, the effect of pH on the formula-
tions was determined by measuring the particle size, zeta potential and
PDI at various pH values using Zetasizer. The nanoparticles dispersion
was incubated in 20 mM acetate buffer (pH 5.8 and 7.4) at 37 1 °C
up to 4 h and then observed at least three times for above parameters.
2.6.2. Qualitative estimation using fluorescence microscopy
The fluorescence microscopy was performed in female rats to con-
firm the accumulation of different Dox formulations. The selected nano-
particle formulations were injected in the tail vein of the rats. The
treated rats were then sacrificed after 6 h and major organs tissue like
heart, liver, lungs kidney, and breast were excised. These tissues were

B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
2885
treated, sectioned, and observed under inverted fluorescent microscope
3.2. Preparation of chitosan nanoparticles
(Leica DM IL LED, Germany).
Chitosan nanoparticles can protect the trapped drug from proteo-
lytic degradation and offers a selective intracellular targeting for the
loaded drug in a passive/active manner. CSNPs were prepared by a
well-established ionic gelation method, which consists the ionic inter-
2.6.2.1. Histopathology analysis. Histopathology analysis was done to as-
sess the effect of Dox on heart tissues. The different formulations were
administered to the rats. The heart was collected, washed with saline
and fixed with 10% formalin. After fixation, it was embedded in paraffin,
sectioned and mounted on glass slides. The samples were stained with
hematoxylin and eosin and observed under light microscope [23].
action between positively charged amino groups (-NH+3 ) of CSEs and
5−
the negatively charged phosphate groups (-P3O ) of TPP at acidic pH
10
[26].
2.7. Statistical analysis
3.3. Optimization and statistical analysis of experimental data
All data were expressed as mean
SD and Analysis of variance
(ANOVA) was applied for the comparisons among three or more groups
followed by post hoc Tukey-Kramer test. Student's t-test was performed
for comparison between two groups. For statistical significance p < 0.05
was considered for all comparisons.
Box-Behnken Design (BBD) was used for the optimization of process
parameters. It provided generous-sized valuable information and
proclaimed the value of statistical design for carrying out experiments.
The selected independent variables like the ratio of CSEs to TPP, sonica-
tion time, and stirring speed played a critical role in the development of
NPs with influence on the observed responses for particle size (PS) and
% entrapment efficiency (EE). The three factors, three levels BBD with 17
runs experimental design were performed to assess the effect of inde-
pendent variables on the observed responses. Different variables practi-
cal data were analyzed with Design-Expert software based on the
constrained condition of desirability (Table 1) and quadratic polynomial
equations generated by ANOVA were statistically validated to set-out
the fitness of model and of the different variables. The selected indepen-
dent variables like the ratio of CSEs to TPP, ST, and SS played a critical
role in the development of NPs with influence on the observed re-
sponses for PS and EE. Polynomial equations point out the main and in-
teraction effects which were determined by the assessment of statistical
parameters. The positive or negative sign for the values in polynomial
equations coefficient describe the positive effect or negative effect re-
spectively of independent variables on the responses in quadratic equa-
tion. The quadratic model considered best fitted and highly significant
on the basis of F value (>0.001). The p-value <0.05 was considered sta-
tistically significant for this model. The insignificant lack of fit value
(>0.05) applied for the prediction of the model. The ratio of the ex-
plained variation and total variation for the responses known as the co-
efficient of determination (R2) was used for the determination of the
degree of the fit model (≈1) along with the maximum adjR2 which
also explains the suitability of design and increase with the introduction
of new variables only. The difference between R2 andadjR2 should be re-
sided in less the 0.2 for the conclusion of the goodness of the model. The
reproducibility of the model expressed as percent coefficient of variance
(% CV) is a ratio of the standard error of the estimate and observed re-
sponse mean. The % CV deliberate for the purpose of reproducibility
and below 10% of CV can be considered for reproducible model
[27,28]. ANOVA fit statics presents another parameter called as ade-
quate precision used to measures the range in predicted response and
its associated error or called signal to noise ratio (> 4 considered as de-
sirable value). Additionally, the sum of squares (SS) for each indepen-
dent variable ascribed the model to analyze the percentage
contributions of the same. Perturbation plots and response surface 3D
plots were used for observation of the effects of the predetermined fac-
tors on the response i.e. PS and EE based on the model polynomial func-
tions, to determine the change in the response surface (Figs. S1 & S2).
The steep and curvature slope structure of the graph is known for the
sensitive response change with that particular factor, whereas compar-
atively, flat line indicated the insensitivity of independent variables over
the response. The perturbation graph is useful to observe the most affec-
tive variables for that response. These plots infer the role of each vari-
able on each response which can be perceived by the developed
formulations. Additionally, the check point optimized formulations
were also used to confirm the magnitude of independent variables pre-
dicted by the generated equation for validity evaluation of model.
3. Result and discussion
3.1. Conjugation and characterization of DoxEs and CSEs
Conjugation of Dox and Es was carried via primary amine group to
the carboxylic group. An excess amount of Dox was introduced to
achieve complete reaction with Es-SA. DoxEs was synthesized under
two steps namely i) Activation of estrone by succinic anhydride to con-
vert estrone into estrone succinate presenting -COOH group and ii) Con-
jugation of Dox and activated estrone (Es-Succinate) by amide bond
formation with -NH2 group of Dox and -COOH group of estrone succi-
nate by carbodimide chemistry using DCC and NHS as coupling agents
in DMSO. Chitosan and Es conjugate polymer (CSEs) was also synthe-
sized by conjugating amino groups of CS and carboxylic group of Es-
SA by using NHS/EDC as coupling agents. Estrone conjugated Dox and
CS were characterized by IR and NMR spectroscopy. Different IR bands
at 1712.79 cm−1C=O (Ester), 3053.32 cm−1CH2 (Methylene),
1249.87 cm−1 C-O-C (Ester), 3018.6 cm−1 CH3 (Methyl),
3585.67 cm−1 -OH (carboxylic acid), 1614.42 cm−1 Primary amide
band; 3346.50 cm−1, (N\\H stretch); and 1583.56 cm−1, (C\\N stretch)
are observed in the spectra. The peak at 1678.07 cm−1 (C_O stretch of
CO-NH) confirmed the conjugation of Es-SA with Dox via amide linkage
[18].
The IR spectrum of CSEs showed the stretching of -OH groups at
3344.57 cm−1 with broad peak because of the hydrogen bonds. The
-OH band may be overlapping the stretching band of -NH. Two consec-
utive bands between 2900 and 2800 cm−1 represents the C\\H sym-
metric and asymmetric stretching respectively. The presence of bands
in between 1300 and 1200 cm−1 shows the ester (C-O-C) and amide
(CO-NH) peaks respectively. The disappearance of characteristic
stretching vibration of the amino group in conjugates and the presence
of characteristic C=O bands of estrone around 1700 cm−1 confirm the
CSEs synthesis [24,25].
The 1H NMR spectra of DoxEs conjugate exhibited typical peaks at
3.633 ppm (H3C-O-C-1), 5.625 ppm (HC-10), 6.887 ppm (HC-3) and
7.059 ppm (HC-2 and HC-4). The spectrum showed an intense peak at
2.984 ppm, which characterize the methylene protons of the Es. Shift
at 7.039 ppm helps to distinguish the -CONH amide group of conjugated
system. 1HNMR spectra of CSEs also showed all characteristic proton
peaks of CS and Es. There was a chemical shift in the CS and Es peaks
due to a steric effect. The shifted peak assignments were found at
5.625 for aromatic -OH and 8.006 for -NH sec amide of conjugates.
Chemical shift at 4.852 ppm for C-1 and at 3.976 ppm for C-2 observed
respectively. As suggested earlier, the spectrum of CSEs also showed an
intense peak at 2.983 ppm, which is the characteristic of methylene pro-
tons of the Es and shift at 7.054 ppm helps to distinguish the -CONH
amide group of the conjugated system [17,21].

2886
B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
3.3.1. Analysis of responses
effect than CSEs:TPP. It is evident that the concentration of TPP in-
creased the PS significantly possibly due to high anion interaction
with positively charged CSEs ions. On increasing concentration of
CSEs, the particle size of NPs increased but opposite to TPP effect, the
ZP of NPs also increased due to the positive charge of CSEs. During the
stirring, high shearing forces reduces the particle size but it also in-
creases the surface energy and aggregates forms to reduce this energy
by reducing the surface area. However, sonication energy breaks the ag-
gregates and the particle size further decreased. The strong negative co-
efficient of all three main factors indicates that particle size is inversely
proportional to all the studied variables. This was in agreement with our
earlier reported studies [20].
The mathematical relationship between the independent variables
and the responses was expressed using the response surface plots. The
interaction effect of X1 and X2wasanalyzed by keeping X3 constant;
the interactions effect of X1 and X3 was analyzed by keeping X2 at con-
stant level; and the effect of X2 and X3 and their interaction was studied
when X1 was kept at fixed level, on response Y1respectively (Fig. S2[R1-
R3]). When we increased the magnitude of X1 and X2 simultaneously, a
negative effect on the particle size was observed (coefficient of X1X2 =
−0.4875) as presented in Fig. S2 (R1). The X1 and X3, also showed the
same response with coefficient of X1 X3 = −4.375 (Fig. S2 (R2)). The in-
teraction effect of X2 and X3 variable on PS was less significant (coeffi-
cient of X2 X3 = 12.8425X2X3) on the PS as shown in Fig. S2(R3)
where X2 showed negative relationship while X3 has no major effect
over the Y1.Particle size of the NP is the deciding factor which influences
the drug release, bioavailability and efficacy of the formulations. Nano-
particles undergo cellular internalization by endocytosis; here PS has in-
verse relationship on cellular drug uptake and influence the drug
bioavailability.
Seventeen trials of CSEsNPs were formulated in accordance with
BBD and were fitted into the experimental design provided by the
Design-Expert software. Table 1 represents the influence of indepen-
dent variables on responses.
3.3.1.1. Response Y1 (particle size). Transformation of data assumed as an
important factor in statistical analysis. Further, the maximum to mini-
mum ratio of practical value for dependent variables Y1 was found
2.99 which suggested no power transformation required. The transfor-
mation of response plays a vital role in data analysis. Transformation
is essential if the error (residuals) is a function of the magnitude of re-
sponse (predicted values). The thumb rule of power transformation in
responses is necessary. The thumb rule of power transformation re-
sponses is when ratio of maximum to minimum response is greater
than 10 transformations is not required, whereas less than 3, transfor-
mation is required.
ANOVA calculations measure the effects of concentration of CSEs to
TPP ratio, ST, and SS on the response PS. The suitability of the model
for the analysis of the independent and dependent variables was done
on the basis of, lack of fit (F value 1.26) and the Prob > F value of
p < 0.0001, low SD, high R2, and lower PRESS value, which suggested
quadratic model for Y1 response (Table S2).The ANOVA statistical data
calculation confirms significant model on the basis of (Prob > F less
than 0.0001) along with F value for response (Y1 was 56.44). Lack of
fit F value PS suggested that lack of fit was not significant relative to
the pure error. The various multiple regression terms i.e. predR2 and
adjR2 values for PS were in good agreement i.e. 0.9689 and 0.8835, re-
spectively and indicated that the predicted responses are good and fit
to the model. Additionally, adequate precision (24.715) determined
the signal-to-noise ratio desirable (>4). The obtained quadratic model
is utilized to find the experimental design space.
3.3.1.2. Response Y2 (% entrapment efficiency). The maximum to mini-
mum ratio was found to be 1.31 for the Y2 response so power transfor-
mation is not required similar to Y1 response. Quadratic model was
suggested for the Y2 response for analyzing it on the basis of lack of fit
value, and other model summary statistics. The Prob > F value of
P < 0.0183, low standard deviation, high R2 and lower predicted resid-
ual error sum of square (PRESS) values. The data of ANOVA (Table S3)
suggested the model was found significant (Prob > F is less than
0.05). The Model F value for response Y2 was 5.41, which defines the
model was significant. Magnitude of CSEs to TPP ratio, were significant
model terms that influence the Y2 response (P < 0.05). Lack of fit F
value for Y2 was 0.49 which implies that lack of fit was not significant
relative to the pure error. The predR2 and adjR2 values for response Y1
were 0.6742 and 0.7637, respectively. This indicates that predR2 value
and adjR2 value are in good agreement indicating a good fit. Adequate
precision for response Y2 was 61.503 indicates that it is an adequate
signal.
Obtained quadratic equation for Y1 (PS)
PS ðY1Þ ¼ 251:18–111:961 ꢀ 1−27:0725 ꢀ 2−9:55625
ꢀ
3−0:4875 X1X2–4:375X1X3 þ 12:8425X2X3 þ 18:8875
ꢀ 12−5:105 ꢀ 22−6:2425 ꢀ 3
ð1Þ
2
The regression equation indicates the effect of all the three formula-
tion variables (X1, X2 and X3), where X1, X2, and X3 are the main effects
influencing the response Y1. The X1 X2, X1 X2, X2 X3, X12, X22, and X32 are
the interaction terms phrases with second-order factors which stand for
the nonlinear relationship between the dependent variables and the in-
dependent variable and showed that the PS was changing when two in-
dependent variables were altered simultaneously. The quadratic
equation positive and negative sign are magnitude of independent var-
iables synergistic and antagonistic effects, respectively on the response.
The quadratic equation for PA suggested the negative effect of CSEs to
TPP ratio, ST and SS. Apart from this, from perturbation graphs obtained,
the influence of individual independent factors on the PS was also ob-
served (Fig. S1). For response Y1, Factor A (CSEs:TPP) was showing ver-
tical slope line which implies that the CSEs to TPP ratio was most
influencing for PS. Additionally, factor B (ST) and factor C (SS) provided
noticeable slope and slight bend respectively suggesting somewhat less
The generated quadratic equation for Y2 (EE)
EE ¼ 61:922–6:03 ꢀ 1−2:23125 ꢀ 2 þ 0:07125
2
ꢀ
3−0:965X1X2−0:425X1X3−1:7975X2X3 0:88975 ꢀ 1
2 2
þ 0:92775 ꢀ 2 þ 0:57275 ꢀ 3
The regression equation of response Y2 showed negative relation-
ship with CSEs to TPP ratio and ST but positive relationship with SS for-
mulation variables. As evidenced by the negative regression coefficient
for X1 and X2 in quadratic equation, CSEs to TPP ratio and ST was the
major factor influenced the (Y2) EE. A possible explanation may be
that, ST may provide high energy for aggregation of NPs which leads
to increased size CSNPs and EE as well. The perturbation graphs for re-
sponse Y2 (EE), factor A and B show steep curvature, whereas factor C
showed slight curvature. It indicates that CSEs to TPP ratio and ST was
important for determining EE. The 3D response surface graphs (Fig. S2
R'1-R'3) were articulated the mathematical relationship b/w the factors
on the responses. It can be concluded from these plots that the two
Table 2
Optimized process parameters of Box-Behnken design.
Level
CSEs:TPP
(ratio)
ST(sec)
SS
−1
0
+1 −1
0
+1 −1
0
+1
Parameters level
Optimized magnitude
Optimized process
parameters
1:1 2:1 4:1 60 90 120 Slow Medium Fast
0.399
1.6:1
−0.842
75
0.999
Medium

B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
2887
Fig. 2. Particle size distribution and TEM Photomicrograph of (A) Upper left and right Dox-CSEsNPs and (B) Lowe left and right DoxEs-CSEsNPs respectively.
variables are negatively influencing the response Y2, and third one have
positive effect may arise due to stirring energy provide more loading of
drugs. The negative regression coefficient for simultaneous increase of
indicate crystalline nature of both the components. Polymer CS has
shown peak at a temperature of 88.73 °C; whereas CSEs has shown
peaks at 77.73 °C and 264.39 °C. This clearly suggested the conjugation
of Es with CS. However, DoxEs-CSEsNPs formulation showed no sharp
endotherm, suggesting decrease in crystallinity of Dox and Es which
can be attributed to the fact that drug can be molecularly dispersed
into the polymer matrix [30]. A minor peak at the area where Dox and
Es present also suggested the compatibility between the drug and the
polymers as well as reduction in sharpness of the peaks strengthen
the entrapment of drug [31]. TEM images also confirmed about the mor-
phological uniformity with almost homogenous shading of developed
nanoparticles suggesting the compact structure of NPs formulations
(Fig. 2).
X1 and X2 designated that the CSEs to TPP ratio and the ST had an in-
verse relationship with the EE.
3.3.2. Optimization and validation
The optimization of response i.e., PS and % EE with desirability as
functions was undergone simultaneously. The overlay plot was plotted
to acquire the optimized formulation conditions. The optimum formula-
tion developed according to the set criteria of minimum PS to maximum
%EE. Thereby, another batch of CSEsNPs with the predicted value of the
formulation factors was prepared to validate the optimization protocol.
The parameters of optimized formulation were 2:1 CSEs to TPP ratio,
30 s ST and medium SS which attains the requirements of optimization.
The predicted value for PS and EE was 200.49 nm and 64.22% as shown
in overlay plot respectively (Fig. S3). The optimized CSEsNPs formula-
3.5. Determination of pharmaceutical characteristics
As per Box-Behnken Design runs, selected formulation parameters
were used to prepare CSEsNPs as described above. The formulations
were characterized for different pharmaceutical parameters such as
PS, PDI, ZP and SM. The average particle size of Dox-CSEsNPs and
tion showed 199
3.25 nm PS and 65.5
1.95% EE, which was in
good harmony with the predicted values. The desirability was found
to be 0.911 (Table 2 and Fig. 3). The dependent variable (i.e. PS and
EE) predicted value was found 200.49 nm and 64.22% respectively as
shown in overlay plot. Conclusively the both obtained values were
found in good harmony and close to each other.
DoxEs-CSEsNPs were found to be 198.2
14.3 nm and 206.4
15.1 nm respectively (Fig. 2). The significant increase in the PS of
DoxEs-CSEsNPs was observed as compared to Dox-CSEsNPs, this
might be due to the entrapment of Es conjugated Dox which is lipophilic
and may be less cationic in nature as compared to Dox. Polydispersity
index (PDI) of Dox-CSEsNPs and DoxEs-CSEsNPs were found to
3.4. DSC and TEM characterization
Different thermograms of CS, Es, Dox, CSEs, and DoxEs-CSEsNPs
were obtained and are shown in Fig. S4. Dox has shown the solid-
liquid transition at 202–204 °C in addition partially superimposed by
an endotherm due to decomposition at 245 °C and this was continued
to higher temperature [29]. Pure Dox and Es showed the sharp melting
endothermic peaks at 202.35 °C and 266.73 °C respectively, which
0.159
0.012 and 0.166
0.017 with ZP +30.6
2.4 mV and
+28.3 2.8 mV respectively. ZP is essential to determine the stability
of the colloidal preparation. The high ZP values play important role in
stabilizing the formulations due to high repulsive forces, which hinders
aggregation nanoparticles [32]. Percent drug entrapment efficiency for
both formulations was found 66.33
2.82% and 62.25
2.63%
Table 3
Optimized pharmaceutical characteristics of the targeted nanoparticles.
Formulation code
Average particle size (nm)
PDI
Entrapment efficiency (%)
Zeta potential (mV)
Dox-CSEsNPs
DoxEs-CSEsNPs
198.2
206.4
14.3
15.1
0.159
0.166
0.012
0.017
66.33
62.25
2.82%
2.63%
+30.6
+28.3
2.4 mV
2.8 mV

2888
B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
Fig. 3. Effect of pH on characterization parameters of Dox-CSEsNPs and DoxEs-CSEsNPs (mean SD, n = 3).
respectively (Table 3). This association of Dox and chitosan nanoparti-
cles might occur due to possible hydrophilic interaction between NH2
group of chitosan and OH and NH2 group of Dox. However, the notable
difference in EE might be due to lipophilic nature and molecular size of
DoxEs conjugates different from plain DOX.
1.2% respectively. The existing biphasic drug release behaviour is also
supported by several research groups [33–36]. However, conjugation
of Dox with Es decreased drug release that could be owing to structural
veracity offered by the coupling of Es, and lipophilic nature which might
result into dual barrier effect for Dox diffusion. In vitro release results of
Dox-CSEsNPs and DoxEs-CSEsNPs formulations were further analyzed
after treating data in different kinetic models such as zero- order, first-
order, Korsmeyer-Peppas, Higuchi, and Hixson–Crowell. Both formula-
tions Dox-CSEsNPs and DoxEs-CSEsNPs best-fitted in Higuchi model
(R2 = 0.964 and 0.961) as compared to other models, on the basis of ob-
tained values of regression coefficients for various kinetics models
(Table S4). The developed formulations were best fitted to Higuchi
model that depict the diffusion controlled release behaviour [37].
3.5.1. Effect of pH on characterization parameters of Dox-CSEsNPs and
DoxEs-CSEsNPs
Fig. 3 shows effect of pH on various formulations. Both the formula-
tions get slightly reduced in particle size at acidic pH (i.e. pH 5.8) due to
CS solubility at acidic environment and this may be exploited for the de-
sired drug release at acidic tumor microenvironment. Simultaneously,
zeta potential and PDI were also increased due to positive charge and
solubilisation of chitosan at this pH.
3.7. Haemolysis studies
3.6. In vitro drug release profile
As expected among all the tested formulations, plain Dox was found
to possess maximum haemolytic activity while DoxEs-CSEsNPs has
shown negligible haemolytic activity. In all the cases, haemolytic activ-
ity was concentration-dependent. Plain Dox has high haemolytic activ-
ity (4.64 0.208) at a concentration of 0.05% w/v in comparison to final
formulation of DoxEs-CSEsNPs (1.3 0.1) at similar concentration of
Dox. These results of haemolytic toxicity of Dox matched with the ear-
lier reports available in literature [38,39]. Conjugation of Es to Dox
also reduced the haemolytic toxicity of Dox; this was further drastically
reduced via the incorporation of conjugate in CSEsNPs due to masking of
direct action on RBCs (Fig. 5).
The release profiles of Dox-CSEsNPs and DoxEs-CSEsNPs formula-
tions were performed at physiological pH 7.4 (37 0.5 °C) using dialy-
sis method. The obtained in vitro release pattern of both the
formulations is shown in Fig. 4. The obtained results showed that the
Dox release was higher approximately 30.87 1.4% and 26.5 0.98%
for both Dox-CSEsNPs and DoxEs-CSEsNPs formulations respectively
in the initial hours (3 to 5 h) reflecting a burst release of the drug. The
burst release may be due to some surface adsorbed drug or may be
due to entrapment of the drug in outer stratum of the nanoparticles.
After 8 h, the drug release rate was observed comparatively slower
and sustained for 72 h. The approximate drug release from Dox-
CSEsNPs and DoxEs-CSEsNPs after 72 h was 82.7
1.3% and 77.0
3.8. Cytotoxicity studies
Cytotoxic potential of Dox, DoxEs, Dox-CSEsNPs and DoxEs-CSEsNPs
was determined against MCF-7 cell lines. Results depicted that cell via-
bility in all the four tested formulations, i.e., DoxEs, Dox-CSEsNPs,
DoxEs-CSEsNPs and free Dox was concentration-dependent (Fig. 6).
IC50 value of free Dox solution was higher than DoxEs-CSEsNPs, Dox-
CSEsNPs and DoxEs. All nanoparticles based Dox formulations showed
comparatively low IC50 value (i.e., IC50 < 10 μg/mL), however, DoxEs-
CSEsNPs had significant higher cytotoxicity than DoxEs, Dox-CSEsNPs
and free Dox solution. This effect may be attributed to the incorporation
of Es in nanoparticles which enhanced cellular uptake, via receptor-
mediated internalization by the cells [40]. This was further confirmed
by performing competitive study after pre-incubation of cells with
free ligand i.e. Es. Higher cell viability (%) was observed when Es was in-
cubated prior to treatment with DoxEs-CSEsNPs, Dox-CSEsNPs and
DoxEs. From these results, it is confirmed that the cytotoxicity of these
Es tethered NPS was reduced due to saturation of receptors by free li-
gand and higher cytotoxic effect of DoxEs-CSEsNPs and Dox-CSEsNPs
were attributed to its receptor-mediated endocytosis. The estrone con-
jugation with Dox might have contributed to lower IC50 value. DoxEs-
Fig. 4. In vitro drug release profile of Dox-CSEsNPs and DoxEs-CSEsNPs formulations
(mean SD, n = 3).

B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
2889
Fig. 5. Comparative studies of percent haemolysis of various formulations (10×).
Fig. 6. In vitro cytotoxic activity of Dox, DoxEs, Dox-CSEsNPs and DoxEs-CSEsNPs using the MCF-7 cell at different concentrations after 48 h (mean SD, n = 3); Left; in absence of estrone;
Right; in presence of estrone.

2890
B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
Fig. 7. Plasma concentration profiles of Dox after tail vein intravenous administration of
free Dox, DoxEs, Dox-CSEsNPs and DoxEs-CSEsNPs formulations (mean SD, n = 3).
Fig. 8. Organ distribution profile of Dox, DoxEs, Dox-CSEsNPs and DoxEs-CSEsNPs
formulations (after 6 h), Statistical significance compared with DoxEs-CSEsNPs group:
*p < 0.05, **p < 0.01, ***p < 0.001 and ns = non-significant.
CSEsNPs showed higher uptake than Dox-CSEsNPs owing to receptor-
mediated endocytosis by the MCF-7 cell lines. This might explain the
higher toxicity of DoxEs-CSEsNPs than DoxEs, Dox-CSEsNPs and free
Dox solution.
in vivo. In vitro release study also supported these results. The half-life
(t1/2) of Dox-CSEsNPs and DoxEs-CSEsNPs were 12.036 0.126 h and
11.749 0.981 h respectively and were significantly greater than the
half-life of DoxEs(6.778 0.764 h) and free Dox(7.419 0.832 h) re-
spectively. This may be due to slow and sustained release of drug from
NPs formulation. The half-life was increased almost doubled using
Dox-CSEsNPs and DoxEs-CSEsNPs than DoxEs conjugate and free Dox.
Overall, increased t1/2 and AUC, and decreased elimination rate remain
constant in the case of NPs as compared with the free drug and therefore
would be in favour of better efficiency of the drug.
3.9. In vivo studies
3.9.1. Blood level studies
Blood level studies of Dox DoxEs, Dox-CSEsNPs and DoxEs-CSEsNPs
were assessed to determine the release and performance of formula-
tions in vivo. The formulations were centrifuged to remove
unentrapped drug and drug concentration equivalent to that of free
drug (5 mg/kg body weights) was administered to the female rats by
i.v. route. The blood plasma was used to determine the concentration
of Dox in blood samples at various time intervals.
The blood levels of Dox after administration of DoxEs-CSEsNPs, Dox-
CSEsNPs, DoxEs and free Dox after 1 h were found to be 20.7 0.34,
21.6 0.35, 10.3 0.40 and 13.5 0.23 μg/mL, and after 6 h, 12.1
3.9.2. Quantitative biodistribution of Dox in different organs/tissues
The quantitative estimation of drug concentration in various organs
like heart, liver, lungs, kidney, and tumor after i.v. administration of free
Dox at 6 h time interval showed maximum accumulation of the drug in
non-target organs. Dox concentration after 6 h of i.v. administration in
various tissues were 18.3
the liver, 33 2.0 ng/g in the lung, 49.6 2.0 ng/g in the kidney, and
55.6 3.5 ng/g in the tumor. The formulations DoxEs, Dox-CSEsNPs
3.0 ng/g in the heart, 65.1
4.0 ng/g in
0.70, 8.8
Dox and DoxEs were cleared rapidly from the circulation. About 1.3
0.11%and 1.7 0.09% of the injected dose was present in the blood
0.55, 2.9
0.30 and 3.3
0.45 μg/mL, respectively. Free
and DoxEs-CSEsNPs have shown different biodistribution profile in
comparison to free Dox. These formulations were more accumulated
in the breast tumor than other organs (Fig. 8). The drug recovered
from the tumor in case of DoxEs-CSEsNPs (163.3 6.4 ng/g) was signif-
icantly higher than free Dox (55.6 3.5 ng/g, p < 0.001), DoxEs (94
6.0 ng/g, p < 0.001) and Dox-CSEsNPs formulation (132.2 4.0 ng/g,
p < 0.01) in 6 h. The results clearly showed that there were approxi-
mately 3, 1.7- and 1.2-fold increase in tumor uptake in case of DoxEs-
CSEsNPs formulation as compared with free Dox, DoxEs and Dox-
CSEsNPs formulations respectively. The increased uptake of DoxEs-
CSEsNPs may be due to the quick ERs recognition and internalization
by the cell membrane of tumor tissue similar to our previous reports
[10,41].
after 24 h in respective cases (Fig. 7). Pharmacokinetic parameters of
Dox-CSEsNPs and DoxEs-CSEsNPs formulation showed clearance from
the blood with an elimination rate constant (Kele) of 0.0575
0.006 h−1and 0.0590 0.004 h−1 respectively (Table 4). However, in
the case of DoxEs and free Dox the Kele were obtained 0.1021
0.041 h−1 and 0.0933
0.012 h−1 respectively. AUC values for the
blood drug concentrations from Dox-CSEsNPs and DoxEs-CSEsNPs
were 280.585 9.17 μg-h/mL and 317.729 9.98 μg-h/mL compared
to 89.167
5.76 μg-h/mL and 107.044
6.58 μg-h/mL values for
DoxEs and free Dox, respectively. AUC results suggested the sustained
release of drug from Dox-CSEsNPs and DoxEs-CSEsNPs formulation
Table 4
Results for various pharmacokinetic parameters of different Dox formulations.
Parameters
Formulation code
Dox
DoxEs
Dox-CSEsNPs
DoxEs-CSEsNPs
Elimination rate constant Kel (h−1
AUC (μg-h/mL)
)
0.0933
107.044
0.901
0.084
7.419
0.012
6.58
0.034
0.0042
0.832
0.1021
89.167
0.988
0.100
6.778
0.041
5.76
0.041
0.0054
0.764
0.0575
280.585
0.557
0.032
12.036
0.006
9.17
0.023
0.0013
0.126
0.0590
317.729
0.480
0.028
11.749
0.004
9.98
0.0039
0.002
0.981
V
d (L)
Cl (L/h)
−1
t
h
1/2

B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
2891
Fig. 9. Fluorescent photomicrograph of biodistribution (450×) after 6 h of tail vein intravenous administration of Dox-CSEsNPs and DoxEs-CSEsNPs, respectively. (A & A′) Heart (B & B′)
Liver (C & C′) Lung, (D & D′) Kidney and (E& E′) Tumor.
3.9.3. Qualitative observation using fluorescence microscopy and cardiac
histopathology
insignificant accumulation in the heart and kidney compared with
other formulation. These results might be attributed to the efficient in-
teraction between estrone-conjugated Dox and cellular ERs [17] over-
expressed on breast cancer cells [41]. The cardiac histopathology was
also performed for the determination of Dox associated cardiac toxicity
and is shown in Fig. 10. These images clearly showed that significant
lower toxicity was shown by DoxEs, Dox-CSEsNPs and DoxEs-CSEsNPs
formulation in comparison to free Dox. The study showed that it is
Qualitative localization of Dox-CSEsNPs and DoxEs-CSEsNPs in dif-
ferent tissues was also observed in tumor bearing female rats. Dox pres-
ence was determined by their fluorescence emission from particular
organ tissue using fluorescence microscopy (at 45×) after 6 h tail vein
intravenous administration. As shown in Fig. 9, DoxEs-CSEsNPs formu-
lation showed significant accumulation in the tumor tissue and

2892
B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
Fig. 10. Histopathological analysis image demonstrating cardiotoxicity of free Dox, DoxEs, Dox-CSEsNPs and DoxEs-CSEsNPs formulations (10×).
possible to achieve reduced cardiotoxicity due to Es conjugation and si-
multaneously encapsulation of Dox/Dox-Es in CSNPs [42]. Dual
targeting could be useful in achieving the goal of magic bullet concept
of targeted drug delivery for cancer therapy; more specifically breast
cancer therapy using estrone as ligand for both intracellular and extra-
cellular receptor in ER+ cancer.
process parameters to obtained better product quality i.e. PS, ZP, %EE
and EE etc. The data obtained from in vitro studies and in vivo pharma-
cokinetic, and biodistribution studies also suggested that the Dox deliv-
ery with Es conjugation through estrone chitosan conjugates
nanoparticles can be a promising drug delivery strategy for anticancer
bioactive as a whole. The developed formulation could be useful in re-
ducing the dosing frequency as well as dose. Further, it has been ren-
dered least toxic against blood cells and cardiac tissues and this could
help additionally in the reduction of cardiotoxicity of Dox. Hence, we
propose that estrone-mediated biodisposition and cellular interaction
of CSEsNPs, especially at the target sites would be useful and helpful
for the upcoming research in the field of anticancer drug delivery.
4. Conclusion
The present work was aimed to develop targeted Dox delivery sys-
tems based on estrone chitosan conjugates nanoparticles. The use of
targeting ligand (Es) for over-expressed tumor cell receptors is widely
accepted as a concept and module, which may be used to selectively de-
liver therapeutic agents to these cells while sparing non-target healthy
cells. Controlled and targeted Dox delivery with Es conjugation and
uploading this conjugate in estrone chitosan conjugate-based nanopar-
ticle can be a promising drug delivery strategy for anticancer bioactive
to avoid the associated adverse effect as well as for potential therapeutic
enhancement. Quality by Design approach has been successfully
employed as unique tool for the optimization of nanoparticles formula-
tions. It provides efficient understanding of the connection between in-
CRediT authorship contribution statement
The role of contributing authors of this manuscript is as follows:
BDK: Data curation, Writing- Original draft preparation, Software.
RP: Software, Formal analysis, Writing- Reviewing and Editing.
SRP: Conceptualization, Methodology, Funding acquisition, Supervi-
sion, Investigation, Writing- Reviewing and Editing, Project
administration.

B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
2893
[2] P. Tekchandani, B.D. Kurmi, S.R. Paliwal, Nanomedicine to deal with cancer cell biol-
ogy in multi-drug resistance, Mini-Rev. Med. Chem. 17 (18) (2017) 1793–1810,
https://doi.org/10.2174/1389557516666160219123222.
List of abbreviations
[3] R. Munagala, F. Aqil, R.C. Gupta, Promising molecular targeted therapies in breast
cancer, Indian J. Pharm. 43 (3) (2011) 236–245, https://doi.org/10.4103/0253-
7613.81497.
[4] B.D. Kurmi, P. Patel, R. Paliwal, S.R. Paliwal, Technology, molecular approaches for
targeted drug delivery towards cancer: a concise review with respect to nanotech-
nology, J. Drug Deliv. Sci. Technol. (2020) 101682, https://doi.org/10.1016/j.jddst.
2020.101682.
[5] V. Masoud, G. Pages, Targeted therapies in breast cancer: new challenges to fight
against resistance, World J. Clin. Oncol. 8 (2) (2017) 120–134, https://doi.org/10.
5306/wjco.v8.i2.120.
[6] D.C. Marquez, R.J. Pietras, Membrane-associated binding sites for estrogen contrib-
ute to growth regulation of human breast cancer cells, Oncogene 20 (39) (2001)
5420–5430, https://doi.org/10.1038/sj.onc.1204729.
[7] F.A. Holmes, R.S. Walters, R.L. Theriault, A.D. Forman, L.K. Newton, M.N. Raber, et al.,
Phase II trial of taxol, an active drug in the treatment of metastatic breast cancer, J.
Natl. Cancer Inst. 83 (24) (1991) 1797–1805, https://doi.org/10.1093/jnci/83.24.
1797-a.
[8] R.W. Boyle, C.K. Johnson, D. Dolphin, Iodination and heck alkynylation of 5,15-
diphenylporphyrin. A convenient entry to asymmetrically meso-substituted por-
phyrins, J. Chem. Soc. Chem. Commun. (5) (1995) 527–528, https://doi.org/10.
1039/C39950000527.
[9] S.R. Paliwal, R. Paliwal, N. Mishra, A. Mehta, S.P. Vyas, A novel cancer targeting ap-
proach based on estrone anchored stealth liposome for site-specific breast cancer
therapy, Curr. Cancer Drug Targets 10 (3) (2010) 343–353, https://doi.org/10.
2174/156800910791190210.
[10] S.R. Paliwal, R. Paliwal, H.C. Pal, A.K. Saxena, P.R. Sharma, P.N. Gupta, et al., Estrogen-
anchored pH-sensitive liposomes as nanomodule designed for site-specific delivery
of doxorubicin in breast cancer therapy, Mol. Pharm. 9 (1) (2012) 176–186, https://
doi.org/10.1021/mp200439z.
BBD
CS
DCC
DCM
DCU
DMAP
DMSO
Dox
Box–Behnken design
Chitosan
Dicyclohexylcarbodiimide
Dichloromethane
N,N-dicyclohexylurea
Dimethylaminopyridine
Dimethyl sulfoxide
Doxorubicin
Dox-CSEsNPs
Dox containing estrone chitosan conjugates NPs
DoxEs
DoxEs-CSEsNPs
Dox and Es conjugate
DoxEs containing estrone chitosan conjugates NPs
DSC
EDC
ERs
Es
Differential scanning calorimetry
N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide
Estrogen receptors
Estrone
FT NMR Fourier-Transform NMR
FTIR
H
HPLC
IR
MCF-7
mL
Fourier-transform infrared spectroscopy
Hour
High-performance liquid chromatography
Infrared
Michigan Cancer Foundation-7
Milliliter
[11] B.S. Reddy, R. Banerjee, 17Beta-estradiol-associated stealth-liposomal delivery of
anticancer gene to breast cancer cells, Angew. Chem. Int. Ed. Eng. 44 (41) (2005)
6723–6727, https://doi.org/10.1002/anie.200501793.
[12] B.D. Kurmi, J. Kayat, V. Gajbhiye, R.K. Tekade, N.K. Jain, Micro- and nanocarrier-
mediated lung targeting, Expert Opin. Drug Deliv. 7 (7) (2010) 781–794, https://
doi.org/10.1517/17425247.2010.492212.
[13] F.Y. Samadi, Z. Mohammadi, M. Yousefi, S. Majdejabbari, Synthesis of raloxifene-
chitosan conjugate: a novel chitosan derivative as a potential targeting vehicle,
Int. J. Biol. Macromol. 82 (2016) 599–606, https://doi.org/10.1016/j.ijbiomac.2015.
10.041.
mmol
Millimoles
MWCO Molecular weight cut-off
NaOH
NH2
NHS
NMR
NPs
PDI
PS
QbD
RGD
SA
Sodium hydroxide
Amino
N hydroxysuccinimide
Nuclear magnetic resonance
Nanoparticles
[14] J. Prados, C. Melguizo, R. Ortiz, C. Velez, P.J. Alvarez, J.L. Arias, et al., Doxorubicin-
loaded nanoparticles: new advances in breast cancer therapy, Anti Cancer Agents
Med.
Chem.
12
(9)
(2012)
1058–1070,
https://doi.org/10.2174/
Polydispersity index
Particle size
187152012803529646.
[15] S. Rai, R. Paliwal, B. Vaidya, P.N. Gupta, S. Mahor, K. Khatri, et al., Estrogen(s) and an-
alogs as a non-immunogenic endogenous ligand in targeted drug/DNA delivery,
Curr. Med. Chem. 14 (19) (2007) 2095–2109, https://doi.org/10.2174/
092986707781368432.
[16] A. Kaur, B.S. Bhoop, S. Chhibber, G. Sharma, V.S. Gondil, O.P. Katare, Supramolecular
nano-engineered lipidic carriers based on diflunisal-phospholipid complex for
transdermal delivery: QbD based optimization, characterization and preclinical in-
vestigations for management of rheumatoid arthritis, Int. J. Pharm. 533 (1) (2017)
206–224, https://doi.org/10.1016/j.ijpharm.2017.09.041.
Quality by Design
Arginine-Glycine-Aspartic acid
Succinic anhydride
Surface morphology
Triethylamine
Sodium tripolyphosphate
Transferrin receptors
Zeta-potential
SM
TEA
TPP
TRs
ZP
[17] S. Rai, R. Paliwal, B. Vaidya, K. Khatri, A.K. Goyal, P.N. Gupta, et al., Targeted delivery
of doxorubicin via estrone-appended liposomes, J. Drug Target. 16 (6) (2008)
455–463, https://doi.org/10.1080/10611860802088481.
[18] N. Cao, S.S. Feng, Doxorubicin conjugated to D-alpha-tocopheryl polyethylene glycol
1000 succinate (TPGS): conjugation chemistry, characterization, in vitro and in vivo
evaluation, Biomaterials 29 (28) (2008) 3856–3865, https://doi.org/10.1016/j.
biomaterials.2008.05.016.
[19] Y. Guo, M. Chu, S. Tan, S. Zhao, H. Liu, B.O. Otieno, et al., Chitosan-g-TPGS nanopar-
ticles for anticancer drug delivery and overcoming multidrug resistance, Mol.
Pharm. 11 (1) (2014) 59–70, https://doi.org/10.1021/mp400514t.
Declaration of competing interest
The authors confirm that this article content has no conflict of
interest.
[20] S. Wadhwa, R. Paliwal, S.R. Paliwal, S.P. Vyas, Hyaluronic acid modified chitosan
nanoparticles for effective management of glaucoma: development, characteriza-
tion, and evaluation, J. Drug Target. 18 (4) (2010) 292–302, https://doi.org/10.
3109/10611860903450023.
[21] B.D. Kurmi, V. Gajbhiye, J. Kayat, N.K. Jain, Lactoferrin-conjugated dendritic
nanoconstructs for lung targeting of methotrexate, J. Pharm. Sci. 100 (6) (2011)
2311–2320, https://doi.org/10.1002/jps.22469.
[22] L.M. Rose, K.F. Tillery, S.M. el Dareer, D.L. Hill, High-performance liquid chromato-
graphic determination of doxorubicin and its metabolites in plasma and tissue, J.
Chromatogr. 425 (2) (1988) 419–423, https://doi.org/10.1016/0378-4347(88)
80049-5.
Acknowledgement
This work is supported by Science and Engineering Research Board,
Department of Science and Technology, India (SB/YS/LS-333/2013) to
SRP in the form of young scientist project grant. BDK is thankful for pro-
viding research fellowship to the agency.
Appendix A. Supplementary data
[23] M. Habib-ur-Rehman, M. Tahir, K.P. Lone, W. Sami, Ethanol induced hepatotoxicity
in albino rats, J. Coll. Physicians Surg. Pak. 21 (10) (2011) 642–643, https://doi.
org/10.2011/JCPSP.642643.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.ijbiomac.2020.08.172.
[24] D. Mishra, N. Jain, V. Rajoriya, A.K. Jain, Glycyrrhizin conjugated chitosan nanoparti-
cles for hepatocyte-targeted delivery of lamivudine, J. Pharm. Pharmacol. 66 (8)
(2014) 1082–1093, https://doi.org/10.1111/jphp.12235.
References
[25] R. Gao, X. Su, X. He, L. Chen, Y. Zhang, Preparation and characterisation of core-shell
CNTs@MIPs nanocomposites and selective removal of estrone from water samples,
Talanta 83 (3) (2011) 757–764, https://doi.org/10.1016/j.talanta.2010.10.034.
[1] R.L. Siegel, K.D. Miller, A. Jemal, Cancer statistics, 2017, CA Cancer J. Clin. 67 (1)
(2017) 7–30, https://doi.org/10.3322/caac.21387.

2894
B.D. Kurmi et al. / International Journal of Biological Macromolecules 164 (2020) 2881–2894
[26] R. Paliwal, S.R. Paliwal, K. Sulakhiya, B.D. Kurmi, R. Kenwat, A. Mamgain, Chitosan-
based nanocarriers for ophthalmic applications, in: S. Maiti, S. Jana (Eds.), Polysac-
charide Carriers for Drug Delivery, Elsevier 2019, pp. 79–104.
[27] A. Czyrski, J. Sznura, The application of Box-Behnken-design in the optimization of
HPLC separation of fluoroquinolones, Sci. Rep. 9 (1) (2019), 19458https://doi.org/
10.1038/s41598-019-55761-z.
[28] S.-N. Nam, H. Cho, J. Han, N. Her, J. Yoon, Photocatalytic degradation of acesulfame K:
optimization using the Box–Behnken design (BBD), Process. Saf. Environ. Prot. 113
(2018) 10–21, https://doi.org/10.1016/j.psep.2017.09.002.
[29] F. Haghiralsadat, G. Amoabediny, M.H. Sheikhha, B. Zandieh-Doulabi, S.
Naderinezhad, M.N. Helder, et al., New liposomal doxorubicin nanoformulation
for osteosarcoma: drug release kinetic study based on thermo and pH sensitivity,
Chem. Biol. Drug Des. 90 (3) (2017) 368–379, https://doi.org/10.1111/cbdd.12953.
[30] D.M. Benival, P.V. Devarajan, Lipomer of doxorubicin hydrochloride for enhanced
oral bioavailability, Int. J. Pharm. 423 (2) (2012) 554–561, https://doi.org/10.1016/
j.ijpharm.2011.11.035.
[31] S. Tummala, M.N. Satish Kumar, A. Prakash, Formulation and characterization of 5-
fluorouracil enteric coated nanoparticles for sustained and localized release in
treating colorectal cancer, Saudi Pharm. J. 23 (3) (2015) 308–314, https://doi.org/
10.1016/j.jsps.2014.11.010.
[32] R. Paliwal, S.R. Paliwal, G.P. Agrawal, S.P. Vyas, Chitosan nanoconstructs for im-
proved oral delivery of low molecular weight heparin: in vitro and in vivo evalua-
tion, Int. J. Pharm. 422 (1–2) (2012) 179–184, https://doi.org/10.1016/j.ijpharm.
2011.10.048.
[35] R.L. Ding, F. Xie, Y. Hu, S.Z. Fu, J.B. Wu, J. Fan, et al., Preparation of endostatin-loaded
chitosan nanoparticles and evaluation of the antitumor effect of such nanoparticles
on the Lewis lung cancer model, Drug Deliv. 24 (1) (2017) 300–308, https://doi.org/
10.1080/10717544.2016.1247927.
[36] R. Rukmangathen, I.M. Yallamalli, P.R. Yalavarthi, Biopharmaceutical potential of
selegiline loaded chitosan nanoparticles in the management of Parkinson’s disease,
Curr. Drug Discov. Technol. 16 (4) (2019) 417–425, https://doi.org/10.2174/
1570163815666180418144019.
[37] K. Rajpoot, S.K. Jain, Colorectal cancer-targeted delivery of oxaliplatin via folic acid-
grafted solid lipid nanoparticles: preparation, optimization, and in vitro evaluation,
Artif. Cells Nanomed. Biotechnol. 46 (6) (2018) 1236–1247, https://doi.org/10.
1080/21691401.2017.1366338.
[38] K. Jain, N.K. Jain, Surface engineered dendrimers as antiangiogenic agent and carrier
for anticancer drug: dual attack on cancer, J. Nanosci. Nanotechnol. 14 (7) (2014)
5075–5087, https://doi.org/10.1166/jnn.2014.8677.
[39] D. Guo, C. Shi, X. Wang, L. Wang, S. Zhang, J. Luo, Riboflavin-containing
telodendrimer nanocarriers for efficient doxorubicin delivery: high loading capac-
ity, increased stability, and improved anticancer efficacy, Biomaterials 141 (2017)
161–175, https://doi.org/10.1016/j.biomaterials.2017.06.041.
[40] C.S. Porto, G.L. Gunsalus, C.W. Bardin, D.M. Phillips, N.A. Musto, Receptor-mediated
endocytosis of an extracellular steroid-binding protein (TeBG) in MCF-7 human
breast cancer cells, Endocrinology 129 (1) (1991) 436–445, https://doi.org/10.
1210/endo-129-1-436.
[41] S.A. Fuqua, Y. Cui, Targeting the estrogen receptor in clinical breast cancer, Breast
[33] L. Zhao, G. Yang, Y. Shi, C. Su, J. Chang, Co-delivery of Gefitinib and chloroquine by
chitosan nanoparticles for overcoming the drug acquired resistance, J.
Nanobiotechnol. 13 (2015) 57, https://doi.org/10.1186/s12951-015-0121-5.
[34] V. Kamat, I. Marathe, V. Ghormade, D. Bodas, K. Paknikar, Synthesis of monodisperse
chitosan nanoparticles and in situ drug loading using active microreactor, ACS Appl.
Mater. Interfaces 7 (41) (2015) 22839–22847, https://doi.org/10.1021/acsami.
5b05100.
Dis. 15 (2002) 3–11, https://doi.org/10.3233/bd-2002-15102.
[42] L. Polgar, E. Lajko, P. Soos, O. Lang, M. Manea, B. Merkely, et al., Drug targeting to de-
crease cardiotoxicity - determination of the cytotoxic effect of GnRH-based conju-
gates containing doxorubicin, daunorubicin and methotrexate on human
cardiomyocytes and endothelial cells, Beilstein J. Org. Chem. 14 (2018)
1583–1594, https://doi.org/10.3762/bjoc.14.136.