You are on page 1of 10

International Journal of Pharmaceutics 485 (2015) 31–40

Contents lists available at ScienceDirect

International Journal of Pharmaceutics


journal homepage: www.elsevier.com/locate/ijpharm

Injectable nanoparticle-loaded hydrogel system for local delivery of


sodium alendronate
Urszula Posadowska a , Martin Parizek b , Elena Filova b , Malgorzata Wlodarczyk-Biegun c ,
Marleen Kamperman c, Lucie Bacakova b , Elzbieta Pamula a, *
a
AGH University of Science and Technology, Faculty of Materials Science and Ceramics, Department of Biomaterials, al. A. Mickiewicza 30, 30-059 Krakow,
Poland
b
Academy of Sciences of the Czech Republic, Institute of Physiology, Department of Biomaterials and Tissue Engineering, Videnska 1083, 14220 Prague 4-Krc,
Czech Republic
c
Wageningen University, Laboratory of Physical Chemistry and Colloid Science, Dreijenplein 6, 6703HB, Wageningen, The Netherlands

A R T I C L E I N F O A B S T R A C T

Article history: Systemic administration of bisphosphonates, e.g. sodium alendronate (Aln) is characterized by extremely
Received 14 January 2015 low bioavailability and high toxicity. To omit aforementioned drawbacks an injectable system for the
Received in revised form 28 February 2015 intra-bone delivery of Aln based on Aln-loaded nanoparticles (NPs-Aln) suspended in a hydrogel matrix
Accepted 2 March 2015
(gellan gum, GG) was developed. Aln was encapsulated in poly(lactide-co-glycolide) (PLGA 85:15) by
Available online 4 March 2015
solid–oil–water emulsification. Drug release tests showed that within 25 days all the encapsulated drug
was released from NPs-Aln and the release rate was highest at the beginning and decreased with time. In
Keywords:
contrast, by suspending NPs-Aln in a GG matrix, the release rate was significantly lower and more
Sodium alendronate
Injectability
constant in time. The GG–NPs-Aln system was engineered to be easily injectable and was able to
Gellan gum reassemble its structure after extrusion as shown by rheological measurements. In vitro studies showed
PLGA that the GG–NPs-Aln was cytocompatible with MG-63 osteoblast-like cells and it inhibited RANKL-
Nanoparticles mediated osteoclastic differentiation of RAW 264.7 cells. The injectability, the sustained local delivery of
Osteoporosis small doses of Aln and the biological activity render the GG–NPs-Aln system promising for the local
treatment of osteoporosis and other bone tissue disorders.
ã 2015 Published by Elsevier B.V.

1. Introduction cancer pain symptoms (Mitra and Jones, 2012) and as a bone
antitumour agent (Cornelis et al., 2014). However, its systemic
Bisphosphonates, e.g. sodium alendronate (Aln), are one of the application beside low bioavailability (<1% when administered
most important classes of drugs for the treatment of bone tissue orally) evokes numerous side effects such as fever, esophageal
pathologies. It is known that the Aln prevents hydroxyapatite loss erosions, ulcers and problems with gastrointestinal tract (Marshall
and inhibits osteoclast-mediated bone resorption (Hayden et al., et al., 2000).
2014). This drug is mainly used for the treatment of osteoporosis, The side effects and limitations of oral administration may be
bone metastasis and Paget’s disease (Ezzati Nazhad Dolatabadi overcome by local delivery of Aln. So far, there have been few
et al., 2014). More recently it has been proposed as adjuvant reports in the literature describing systems for local bisphosph-
analgesic in combination with opioids in the management of onates administration; they were based on hydroxyapatite micro-

Abbreviations: Aln, sodium alendronate; NPs-Aln, sodium alendronate loaded nanoparticles; NPs, nanoparticles; GG, gellan gum; GG–NPs- Aln, injectable nanoparticle-
loaded hydrogel system; PLGA, poly(lactide-co-glycolide); FDA, Food and Drug Administration; PVA, polyvinyl alcohol; OPA, O-phthaldialdehyde; PBS, phosphate buffered
saline; DMEM, Dulbecco’s modified Eagle’s Minimum Essential Medium; RPMI, Roswell Park Memorial Institute (Medium); FBS, fetal bovine serum; MG-63, osteoblast-like
cells; RAW 264.7, murine monocyte-like cells (macrophages); RANKL, receptor activator of nuclear factor kappa-B ligand; TRAP, tartrate-resistant acid phosphatase; RTCA,
real-time cell analyzer; UHQ-water, ultra high quality water; Zr(Acac)4, zirconium acetylacetonate; S/O/W, solid–oil–water emulsification; DLS, dynamic light scattering; %EE,
encapsulation efficiency; %LE, loading efficiency; AFM, atomic force microscopy; FMAX, maximal extrusion force; MWCO, molecular weight cut-off; G0 , storage modulus; G00 ,
loss modulus.
* Corresponding author. Tel.: +48 12 617 44 48; fax: +48 12 617 33 71.
E-mail addresses: uposadow@agh.edu.pl (U. Posadowska), parizek@biomed.cas.cz (M. Parizek), filova@biomed.cas.cz (E. Filova), gosia.wlodarczyk-biegun@wur.nl
(M. Wlodarczyk-Biegun), marleen.kamperman@wur.nl (M. Kamperman), lucy@biomed.cas.cz (L. Bacakova), epamula@agh.edu.pl (E. Pamula).

http://dx.doi.org/10.1016/j.ijpharm.2015.03.003
0378-5173/ ã 2015 Published by Elsevier B.V.
32 U. Posadowska et al. / International Journal of Pharmaceutics 485 (2015) 31–40

spheres (Lee et al., 2013), calcium phosphate microspheres (Kim Ultra high quality water (UHQ-water) was produced in an UHQ-PS
et al., 2010), poly(b-hydroxybutyrate-co-b-hydroxyvalerate)–hy- purification system (Elga, UK). Poly(lactide-co-glycolide) (PLGA,
droxyapatite composites (Huang et al., 2009) or carbopol gels 85:15, Mn = 100 kDa, d = 1.9) was synthesized by a ring-opening
(Reddy et al., 2005). Attempts to coat titanium implants with polymerization in bulk at 100  C using a biocompatible Zr(Acac)4
bisphosphonates to improve osteointegration have also been initiator (Dobrzynski et al., 2001) and kindly provided by the
described (Harmankaya et al., 2013). Nevertheless the approach of Centre of Polymer and Carbon Materials of the Polish Academy of
local delivery of Aln is still poorly explored. Sciences in Zabrze, Poland.
There is a need to develop a local delivery system for Aln which
will be easy to handle by the surgeon, and will remain at the 2.2. Nanoparticle fabrication
targeted area of the affected tissue. We hypothesize that such a
system will assure local controlled delivery of small doses of the PLGA nanoparticles were fabricated using a double emulsifi-
drug, so systemic side effects typical for oral Aln administration cation/solvent evaporation technique. A portion of 10 mg of Aln
will be reduced. Moreover, the system will be administered via powder (solid – S) was emulsified in a 1.67% w/v PLGA solution in
injection so it will be less-invasive and traumatic than typical dichloromethane (6 ml, oil – O), forming the primary emulsion (S/
surgical intervention performed to fill in the defect or treat affected O). The primary emulsion was then introduced dropwise into an
area with bone graft. To this end, we developed an injectable aqueous solution with a stabilizer, PVA (4% w/v, 20 ml) on ice
material consisting of Aln-loaded nanoparticles suspended in a during ultrasonication (3 min, 40% of the cycle, Sonics Vibra
hydrogel matrix. CellTM, USA) forming the secondary emulsion (S/O/water).
A resorbable polymer, poly(lactide-co-glycolide) (PLGA), was Afterwards, the emulsion was stirred at 500 rpm overnight in
chosen to encapsulate Aln because this FDA-approved copolymer order to evaporate dichloromethane. The dispersion of nano-
has a tunable degradation rate (i.e., by molar ratio of lactide to particles was centrifuged (14,000 rpm, 4  C, 20 min) and washed
glycolide, chain structure and molecular weight) and can be easily three times with UHQ-water. The obtained nanoparticles were
processed into nanocarriers using emulsification (Shive and freeze-dried for 48 h and then stored at 4  C. Before further
Anderson, 1997). As a hydrogel matrix gellan gum (GG) was used. analyses the particles were redispersed in UHQ-water using
GG is an anionic polysaccharide produced by bacteria ultrasound.
(Sphingomonas elodea) during aerobic fermentation (Doner and
Douds, 1995). GG is widely applied in the food and pharmaceutical 2.3. Size and zeta potential of nanoparticles
industries (Babu et al., 2010) and more recently it has been
proposed for cartilage (Oliveira et al., 2010) and bone (Douglas The size as well as the polydispersity index of empty and Aln-
et al., 2014) regeneration. Further advantages of GG are low cost, loaded nanoparticles (NPs, NPs-Aln) were determined by dynamic
biodegradability and low toxicity of its degradation products, as light scattering (DLS) using a Zetasizer Nano ZS (Malvern Instru-
well as the fact that it is not animal-derived, avoiding regulatory ments). The size of the nanoparticles was determined according to
concerns (Gantar et al., 2014). the Stokes–Einstein equation. The zeta potential was measured by
The aim of this study was first to manufacture and characterize a laser Doppler electrophoresis method using the same apparatus
PLGA nanoparticles loaded with Aln. In the second step, drug and calculated using the Smoluchowski equation based on
delivery system based on Aln-loaded nanoparticles and GG electrophoretic mobility data.
hydrogel was produced and characterized in terms of its
mechanical properties (rheology), surgical handiness (injectability, 2.4. Drug encapsulation and loading efficiencies
swelling) and Aln release kinetics. Finally, to confirm the feasibility
of the developed system in the treatment of bone tissue defects In order to assess the amount of encapsulated drug, the
and diseases, in vitro tests in contact with osteoblast-like cells and supernatant obtained after particles centrifugation was collected.
osteoclast-like cells were conducted. Aln content was examined spectrophotometrically (UV-VIS Cecil
CE 2502, Corston UK) at 332 nm with the use of an OPA test (Anhalt,
2. Materials and methods 1977). Encapsulation efficiency (%EE) and loading efficiency (%LE)
were calculated according to formulae (1) and (2), respectively:
2.1. Materials and chemicals
mass of Aln in nanoparticles
%EE ¼  100% (1)
initial mass of Aln in the system
Sodium alendronate (Aln) was a gift from Polpharma S.A.
Poland (batch no.: 504041227). Polyvinyl alcohol (PVA, Mowiol1
4-88, Mn = 31,000 Da) and gellan gum (GelzanTM CM, classified as
mass of Aln in nanoparticles
low-acyl gellan by the manufacturer) were bought from Sigma– %LE ¼  100% (2)
mass of nanoparticles
Aldrich, Poland. Dichloromethane, methanol, 2-mercaptoethanol,
isopropanol, orthophthaldialdehyde (OPA), phosphate buffered
saline (PBS) buffer (concentrate), calcium chloride, potassium
hydroxide and borate acid were all of analytical grade and were 2.5. Morphology of nanoparticles
purchased from POCh, Poland. Chemicals used for biological tests
included acid phosphatase, TRAP (leukocyte tartrate resistant acid Morphology evaluation of the unloaded NPs and NPs-Aln was
phosphatase kit, Sigma–Aldrich), LIVE/DEAD1 viability/cytotoxic- performed by atomic force microscopy (AFM). A drop of
ity assay for mammalian cells (calcein-AM and ethidium homo- nanoparticles suspension (0.1% w/v) was placed on a micro-
dimer-1, Invitrogen, Molecular Probes) and RANKL (receptor scopic glass slide (Thermoscientific, Menzel-Glaser, Germany)
activator of nuclear factor kappa-B ligand, Sigma–Aldrich, USA). and the solvent was allowed to evaporate for 12 h at room
Cell culture media (DMEM, Dulbecco’s modified Eagle’s Minimum temperature. Topographic images were recorded in contact
Essential Medium; RPMI-1640) were purchased from Sigma– mode AFM (Explorer, Thermomicroscopes, spring constant
Aldrich, USA. Osteoblast-like MG-63 cells were obtained from the k = 0.02 N/m, image size 5 mm  5 mm). All the images were
European Collection of Cell Cultures, Salisbury, UK; and RAW processed using SPMLab6.02 software provided by the AFM
264.7 murine monocytes were obtained from ATCC, TIB-71, USA. manufacturer.
U. Posadowska et al. / International Journal of Pharmaceutics 485 (2015) 31–40 33

2.6. Fabrication of nanoparticle-loaded hydrogel system bags (ZelluTransRoth, MWCO 12 kDa) were filled with 1 ml of
sample, end sealed and immediately immersed in the vials with
GG powder was dissolved in UHQ-water at an initial 20 ml of PBS at pH 7.2 under continuous stirring (50 rpm) at 37  C.
concentration of 1.4% w/v at 90  C. After the temperature of the At predetermined time intervals, 2 ml of buffer was collected and
GG hydrogel was decreased to 60  C, a portion of the NPs-Aln the amount of drug was quantified using an OPA assay as described
suspension (10% w/v) was added. To adjust the injectability, the elsewhere (Anhalt, 1977).
matrices were cross-linked by addition of 6% w/v CaCl2 solution.
The final GG–NPs-Aln samples contained 0.7% w/v GG, 1% w/v 2.11. In vitro tests with osteoblast-like and osteoclast- like cells
nanoparticles and 0.3% w/v CaCl2. As a reference, GG cross-linked
hydrogels containing 0.7% w/v GG and 0.3% w/v CaCl2 were also Cellular response of osteoblast-like and osteoclast-like cells
prepared. was studied in contact with extracts from GG–NPs-Aln system,
which were prepared in two culture media (DMEM and RPMI) at
37  C for 72 h, in conditions similar to drug release studies.
2.7. Injectability of nanoparticle-loaded hydrogel system
Osteoblast-like MG-63 cells were cultured in DMEM (supple-
mented with 10% fetal bovine serum (FBS), gentamicin 40 mg/ml)
GG–NP-Aln samples were characterized in terms of injectability
and RAW 264.7 monocyte-like cells were cultured in RPMI
by a method described elsewhere (Ghadiri et al., 2013). The
(supplemented with 10% FBS, gentamicin 40 mg/ml) at 37  C in a
maximal force needed to extrude 1 ml of sample from a 2 ml
humidified air atmosphere containing 5% of CO2. A real-time cell
syringe equipped with a standard needle (18 G, 0.85 mm and
analyzer (RTCA, xCELLigence, Roche Applied Science, Mannheim,
1.45 mm of inner and outer diameter respectively, Becton Dick-
Germany and ACEA Biosciences, San Diego, California) was used
inson) was measured in compression mode on a universal testing
to evaluate the survival and growth of cells in the prepared
machine (Zwick 1435, Germany). The crosshead speed during the
extracts continuously, during a 7 day time span. RTCA is an
test was 50 mm/min.
impedance-based system used for dynamic and label-free
monitoring of adherent cells attachment, spreading and prolifer-
2.8. Swelling experiments
ation. As adherent cells spread and proliferate on the bottom of
the e-plate covered with gold electrodes, cell membranes act as
To analyze the swelling properties of the hydrogel-based
an insulator blocking detection of electric current by the
systems, samples of known weight were immersed in a sealed
electrodes (Peper et al., 2014). First, 162 ml of cell culture medium
glass container filled with phosphate buffer (PBS; pH 7.4) (Kim
as control samples, i.e. pure DMEM or RPMI, and the extracts from
et al., 2007) for 24 h at 37  C. Then the samples were taken out
GG–NPs-Aln system in DMEM or RPMI were poured into the wells
from PBS and their mass was determined again. The swelling
of a 96-well sensory E plate (E-Plate 96, BioTech a.s., Prague, CR,
degree was calculated using a gravimetric method according to
Cat. No. 05232368001), and background impedance was measured
the formula (3) :
in each well. After that, 18 ml FBS were added into each well.
ðmass after swelling  initial massÞ Subsequently, a dense suspension of cells (5  103 in 20 ml of
Swelling degree ¼ 100%
initial mass DMEM for MG-63 or of RPMI for RAW 264.7) was added to the
(3) wells in triplicates. To elicit osteoclastic differentiation of RAW
264.7 monocyte-like cells, 100 ng/ml of RANKL was used. RAW
264.7 cells in RPMI without RANKL were also cultured. The
2.9. Rheological measurements
impedance was measured every 30 min. Cell index values
(reflecting cell attachment, spreading and proliferation) were
To characterize the rheological properties, tests were performed
calculated automatically by the instrument according to the
on GG–NPs-Aln system and GG hydrogel (Physica MCR 501
formula (4):

ðimpedance at individual time interval  background impedanceÞ


Cell index ¼ (4)
15V

Rheometer, Anton Paar, Graz, Austria; Couette CC10/T200 coaxial To determine the viability of cells on day 1 and 7 cells were
geometry with a bob diameter 10.002 mm, cup diameter stained using LIVE/DEAD1 viability/cytotoxicity assay according to
10.845 mm). For each sample a layer of paraffin oil was applied to manufacturer’s protocol. Briefly, cells were incubated with calcein-
avoid evaporation. Gelation was analyzed by measuring storage and AM and ethidium homodimer-1 (2 and 4 mM in PBS, respectively)
loss moduli (G0 and G00 ) as a function of time during application of for 10 min at 37  C. Calcein-AM, metabolized by esterase activity of
oscillatory sinusoidal deformation (0–30 min, frequency f = 1 Hz, living cells, emitted green fluorescence, and ethidium homodimer-
strain c = 0.01%). During the second time interval (30–60 min) gels 1, which penetrated into dead cells through their damaged
were broken by increasing the strain (c = 0.01–100%, f = 1 Hz). In the membrane, produced red fluorescence. After incubation of cell
third time interval (60–90 min), the strain was reduced and with these dyes, microphotographs were taken under an
recovery of the gel was analyzed (frequency f = 1 Hz, strain epifluorescence microscope (Olympus IX-51 with DP-70 digital
c = 0.01%). All measurements were performed at 37  C. camera, Japan).
To determine the differentiation of RAW 264.7 cells towards
2.10. Drug release studies osteoclasts, the cells after 7 days of culture were stained by using
a TRAP staining kit according to the manufacturer’s protocol.
The release profiles of Aln from NPs-Aln suspended in PBS Briefly, the cells were fixed in citrate/acetone solution and
buffer as well as from GG–NPs-Aln system were analyzed. Dialysis incubated in solutions of acetate, naphtol AS-BI phosphoric acid
34 U. Posadowska et al. / International Journal of Pharmaceutics 485 (2015) 31–40

and Fast Garnet GBC Salt with or without tartrate for 1 h at 37  C 3.3. Drug release studies
in dark. After rinsing with deionized water, the cells were
counterstained with acid hematoxylin. Afterwards microphoto- An initial burst release phase measured after day 1 was 17% and
graphs were taken under IX 51 Olympus microscope equipped 2% for NPs-Aln and GG–NPs-Aln respectively (Fig. 1A and B). For
with DP70 digital camera. NPs-Aln the release of Aln was the highest within the first 5 days of
the experiment (200 mg of Aln), it decreased to 120 mg in the 2nd
2.12. Statistics 5-day time interval, followed by 80 mg for the 3rd and 4th 5-day
time intervals and finally it diminished to 10 mg in the last analyzed
The results were presented as a mean  standard error of the 5-day time interval (i.e., days 20–25 of the experiment) (Fig. 4A).
mean (S.E.M.) from 3 individual measurements (n = 3). The For GG–NPs- Aln the release rate was slower and constant in time
statistical analyses were performed using t-test and significant (linearity of zero-order kinetics); the system delivered 16  3 mg of
differences were assumed at ***p < 0.001, **p < 0.01 and *p < 0.05. Aln within each 5-day interval (Fig. 4B). In total, after 25 days, 95%
and 8% of the drug was released from NPs-Aln and GG–NPs-Aln,
3. Results respectively. After 30 days, the release curves of NPs-Aln leveled off
while the injectable GG–NPs-Aln system still showed the ability to
3.1. Properties of nanoparticles deliver drug with constant doses.

Nanoparticles were formulated with an average diameter of ca. 3.4. In vitro studies
270 nm and ca. 230 nm for empty and Aln-loaded batches,
respectively (Table 1). The polydispersity was low (0.2 and MG-63 osteoblasts-like cells had normal spindle-shaped
0.3 for Aln-loaded and empty nanoparticles, respectively), which morphology in both pure DMEM and extracts of GG–NPs-Aln
corresponded to a narrow size distribution (Fig. 1A and B). Laser (DMEM/GG–NPs-Aln) (Fig. 5A). The LIVE/DEAD1 viability/cyto-
electrophoresis demonstrated a negative zeta potential in the toxicity assay showed that the initial attachment of MG-63 cells
range of 28 and 30 mV (Table 1). The solubilization efficiency and cell densities achieved in both media were similar on days 1
and loading efficiencies of Aln in the nanoparticles were ca. 70% and 7. A waste majority of cells was viable, i.e. stained in green, and
and 5%, respectively. AFM pictures (Fig. 1C and D) revealed that the only few cells were dead, i.e. stained in red, in both DMEM and
nanoparticles were spherical and had smooth morphology. DMEM/GG–NPs-Aln extract (Fig. 5A). In accordance with these
results, the continuous impedance- based measurement showed
3.2. Properties of nanoparticle-loaded hydrogel system overlapping curves of MG-63 cells in both DMEM and DMEM/GG–
NPs-Aln extract during a 7-day culture (Fig. 5B).
Injectability tests of the systems (Fig. 2A) showed that the RAW 264.7 macrophages cultured in pure RPMI displayed
measured extrusion force, FMAX, using the standard 18 G needle for mostly a round morphology, they partially and weakly adhered to
both GG and GG–NPs-Aln was lower than 10 N. For GG hydrogels the surface of culture dish (Fig. 6A). The microscopic observations
the force was 8.1 1.1 N and it increased to 9.5  1.0 N upon after RANKL addition (Fig. 6B) showed that the cells formed
nanoparticles addition (no significant difference was found aggregates, which can be considered as a tendency to fuse and to
according to t-test). It is worth noting that for non-cross-linked form multinucleated cells, and some of them were spread having
GG hydrogels (without Ca2+) the extrusion force was significantly polygonal or spindle-shaped morphology, which are indicators of
lower, i.e. 0.81  0.06 N (data not shown). For GG–NPs-Aln systems their osteoclastic differentiation (Steinberg and Hiken, 2007;
swelling was significantly lower (42.0  3.5%) than for GG hydro- Araujo et al., 2009). In tested conditions, i.e. when culture was
gels (105  2.3%) (Fig. 2B). performed in GG–NPs-Aln extracts (Fig. 6C), the cells were mostly
The elastic as well as the viscous response of GG and GG–NPs- round, the spread cells were less numerous and the cell
Aln were analyzed by measuring storage (G0 ) and loss (G00 ) moduli, morphology was similar to that of cells cultured in pure RPMI.
respectively (Fig. 3). In the first time interval (0–30 min) the elastic The presence and activity of TRAP, i.e. an important marker of
response dominated the viscous behavior for both GG and GG– osteoclastic differentiation, manifested with brown-reddish color,
NPs-Aln. For GG a G0 of 8.5 kPa was reached (Fig. 3A) whereas G00 was most apparent in RAW 264.7 cells cultured in the medium
reached 0.5 kPa (Fig. 3B). For GG–NPs-Aln G0 and G00 were found with RANKL and least apparent in cells in GG–NPs-Aln extracts. In
7.4 kPa (Fig. 3A) and 0.5 kPa (Fig. 3B), respectively. During breaking addition, in RPMI + RANKL medium, RAW 264.7 macrophages
of the hydrogels by strain sweep deformation (s = 0.01–100%) G0 reached higher cell population density in comparison with cells in
and G00 decreased continuously; G0 decreased to 10 Pa, while G00 pure RPMI and GG–NPs-Aln extracts.
decreased to 80 Pa. Fig. 3C and D shows G0 and G00 moduli as a Microscopic observation went along with impedance-based
function of strain during the breaking test up to 10%. It is apparent curves (Fig. 6D). For RAW 264.7 cells cultured in RPMI with
that both the drop in G0 and in G00 was faster for GG–NPs-Aln than addition of RANKL cell index was the highest suggesting RAW
for GG hydrogel. During the last interval (60–90 min) both GG and 264.7 osteoclastic differentiation (i.e., generation of impedance by
GG–NPs-Aln restored the initial values of G0 and G00 ; they reached cell aggregation and spreading), and also proliferation (i.e.,
6.8 kPa and 0.4 kPa for GG hydrogel and 7.1 kPa and 0.5 kPa for the generation of impedance by occupancy of the well bottom by a
GG–NPs-Aln injectable system, respectively. higher number of cells). In the case of RAW 264.7 cells cultured in

Table 1
Physico-chemical analysis of empty and Aln-loaded PLGA nanoparticles.

Nanoparticles DH Pdl z %EE %LE


(nm) (mV) (%) (%)
NPs (empty) 267  7 0.290  0.029 30.5  0.6 – –
NPs-Aln 234  4 0.195  0.006 27.7  0.2 70.3  4.9 5.1  0.6

DH – hydrodynamic diameter, Pdl – polydispersity index, z – zeta potential, %EE – encapsulation efficiency, %LE – loading efficiency.
U. Posadowska et al. / International Journal of Pharmaceutics 485 (2015) 31–40 35

Fig. 1. Histograms presenting size distribution of empty (A) and Aln-loaded nanoparticles (B) obtained using Malvern Zetasizer NanoZS. AFM visualization of empty (C) and
Aln-loaded nanoparticles (D) (Explorer, Thermomicroscopes).

RPMI + RANKL/GG–NPs- Aln extracts the cell index reached lower tissue as a minimally invasive and handy injection, was based on
values which were similar as for RAW cells cultured in RPMI. hydrogel containing Aln encapsulated within PLGA nanoparticles.
Our hypothesis was that the system would assure local and
4. Discussion sustained release of drug, which is known to inhibit osteoclasts
maturation. In the study NPs-Aln were obtained and their
We aimed at manufacturing a new type of Aln delivery system physicochemical properties were assessed. Then, NPs-Aln were
for the local treatment of osteoporosis and other bone diseases. suspended in GG hydrogels and the hydrogel matrix was cross-
The system, intended to be administered inside affected bone linked with calcium ions to provide structural integrity suitable for

Fig. 2. Extrusion force (A) and swelling (B) of the cross-linked hydrogel (GG) and nanoparticles loaded system (GG–NPs-Aln). Significant difference ***p < 0.001 according to
t-test.
36 U. Posadowska et al. / International Journal of Pharmaceutics 485 (2015) 31–40

Fig. 3. Rheological characteristics of the cross-linked hydrogel (GG) and nanoparticles loaded system (GG–NPs-Aln). Storage modulus, G0 , (A) and loss modulus, G00 , (B) during
assembling (0–30 min, f = 1 Hz, strain, s = 0.01%), breaking (30–60 min, f = 1 Hz, s = 0.01–100%) and reassembling (60–90 min, f = 1 Hz, s = 0.01%). G0 (C) and G00 (D) as a function
of strain (f = 1 Hz, s = 0.01–10%) plotted on a logarithmic scale.

injection. Afterwards, the kinetics of the drug release was encapsulation of 70% of the drug introduced in the system, and in
measured. Lastly, to assess biological activity of developed system the final nanoparticles the drug loading efficiency was equal to
cellular response of bone-forming and bone-resorbing cells was 5%.
evaluated. The average diameter of NPs-Aln was 230 nm with a narrow size
distribution as reflected by the polydispersity index (below 0.2).
4.1. Properties of Aln-loaded PLGA nanoparticles NPs-Aln were spherical as shown by AFM observations and
demonstrated a negative zeta potential (ca. 28 mV), which can be
Aln-loaded PLGA nanoparticles were successively formulated attributed to the presence of the ionized carboxyl groups of PLGA
according to a solid/oil/water (S/O/W) method, which enabled on the particles’ surface (Le Broc-Ryckewaert et al., 2013).

Fig. 4. Aln release curves from drug-loaded nanoparticles (A) and gellan gum based injectable system (B). Cumulative curves (black squares and lines) as well as doses
released from 1 ml of NPs-Aln suspension (1% w/v) or 1 ml of GG–NPs-Aln system during 5-day intervals (boxes) are shown.
U. Posadowska et al. / International Journal of Pharmaceutics 485 (2015) 31–40 37

Fig. 5. Morphology of osteoblast-like MG-63 cells cultured in DMEM/GG–NPs-Aln extracts and in DMEM on day 1 and 7 (A). The living cells were stained with LIVE/DEAD1
viability/cytotoxicity assay and observed under Olympus IX51 epifluorescence microscope equipped with DP70 digital camera; objective 10, scale bar = 200 mm. Cell index
of MG-63 cells cultured in DMEM or in DMEM/GG–NPs-Aln extracts during a 7-day culture, measured using xCELLigence system real-time cell analyzer (B). (For interpretation
of the references to color in text, the reader is referred to the web version of this article.)

Interestingly, unloaded nanoparticles were slightly bigger (ca. higher than the GG matrix itself (0.7% w/v) was found not to
270 nm) and less uniform (polydispersity ca. 0.3) than NPs-Aln. The significantly affect its injectability (see Fig. 2A). A slight but not
dispersion of Aln powder in the oily phase in the first step of the S/ statistically significant increase in extrusion force (FMAX 1 N)
O/W emulsification may have resulted in solid Aln nuclei on which, possibly resulted from an increase in viscosity of the system due to
during the second step, polymer chains precipitated more easily friction arising between nanoparticles and GG chains. Still, the
thus creating smaller and more uniform nanoparticles. As a result, obtained values (9.5  1.0 N) were lower than the injectability of
the process of nanoparticle formation became more controlled commercially available medical products (e.g. Radiesse1 – a
when Aln powder was introduced to the system. dermal filler, FMAX 27 N) (Ji et al., 2012) and were in the range of the
injectability of tricalcium phosphates pastes proposed for bone-
4.2. Properties of injectable GG–NPs-Aln system grafting applications (5.4  1.1 N–18.5  1.7 N) (Montufar et al.,
2013).
Although there are many evidences that PLGA nanoparticles are Swelling of the GG–NPs-Aln system was found to be lower (42%)
potent drug delivery vehicles it was also reported that they are than that of GG (105%) (see Fig. 2B). Presumably this can be related
often instable, agglomerate and are prone to phagocytosis in vivo either to the hydrophobic character of the PLGA nanoparticles that
(Li et al., 2014; Kohane, 2007). Therefore it may be advantageous to possibly repelled water molecules or to the location of the NPs-Aln
process nanoparticles into implantable or preferably injectable within the matrix. It is known that swelling occurs when water
forms (for example by suspending them in hydrogels). In this molecules fill the pores within the hydrogel matrix (Gantar et al.,
study, GG hydrogel was used as a matrix to suspend NPs-Aln. GG 2014). Par analogy it can be speculated that NPs-Aln are located
forms a double helix out of two left-handed chains. Three pairs of within the pores of the hydrogel and occupy available space.
such double helices build an associate that constitutes a gel. The Nevertheless, from the application point of view and hydrogel
structure exposes carboxylic groups on the periphery that are structural integrity, a low extent of swelling is beneficial.
prone to interact with bivalent positive ions (e.g. Ca2+), which In order to characterize the GG–NPs-Aln system prior, during and
physically crosslink the hydrogel and thus increase its stiffness after injection the rheological tests were divided in three stages:
(García et al., 2011). When enriched with drug loaded nano- gelation (assembling), breaking (injection) and reassembling (self-
particles (e.g. NPs-Aln), functionality of the GG hydrogel expands healing). It was found that both the GG–NPs-Aln system and the
towards therapeutic injections. reference cross-linked GG hydrogel behave elastically predominant-
In order to determine the handiness of such hydrogel-based ly (G0 > G00 ) (see Fig. 3A and B, time 0–30 min). Addition of NPs-Aln
injectable system, the extrusion force from the syringe was into GG resulted in a very small decrease in G0 but it did not influence
determined. Suspending NPs-Aln at a concentration of 1% w/v, i.e. G00 . This may be ascribed to the larger distance between the GG
38 U. Posadowska et al. / International Journal of Pharmaceutics 485 (2015) 31–40

Fig. 6. Morphology of RAW 264.7 cells (non-differentiated macrophages) in RPMI medium (A), morphology of cells differentiating towards osteoclasts in RPMI + RANKL; red
arrows indicate cell fusion (B), morphology of cells cultured in RPMI + RANKL/GG–NPs- Aln extracts (C) after staining cells for TRAP (objective 20, scale bar = 100 mm). Cell
index of RAW 264.7 cells cultured in RPMI or in RPMI + RANKL or RPMI + RANKL/GG–NPs- Aln extracts during a 7-day culture, measured using xCELLigence system real-time
cell analyzer (B). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

chains due to the presence of NPs-Aln which possibly hindered the of different types of drugs (Fredenberg et al., 2011; Mundargi et al.,
formation of calcium mediated cross-links. 2008). Thus by using PLGA 85:15 of relatively high molecular
To simulate the injection process, the GG–NPs-Aln gel was weight (Mn = 100 kDa, Mw = 190 kDa) we intended to slow down
broken by applying a high shear rate (see Fig. 3A and B, time hydrolytic degradation of the polymer matrix, reduce the burst
30–60 min), which transformed GG–NPs-Aln into a flowing sol release and obtain retarded and sustained delivery of the
(both G0 and G00 decreased drastically and G00 became higher than encapsulated drug. Our approach resulted in an initial burst
G0 ). The drop in both moduli (G0 , G00 ) was due to structural changes release of 17% after 1 day and all the encapsulated drug was
within the gel network: possibly cross-links destruction or an released during 25 days. It is probable that the initial burst release
alignment of hydrogel chains occurred as shown in other studies from NPs-Aln was associated with Aln adsorbed on the nanoparti-
(Morais et al., 2013). The G' of the GG–NPs-Aln system dropped cle surfaces, while in the following stage Aln was released upon
much faster than the G0 of GG when strain deformations were hydrolytic degradation of PLGA. It was found, however, that NPs-
applied (see Fig. 3C and D), which suggests that nanoparticle Aln did not assure a constant drug release rate: within the first 5-
addition caused the system to become weaker than GG. In general, day time interval 200 mg of Aln was released, while in the fifth
the viscosity drop upon shearing of the GG–NPs-Aln showed that analyzed 5-day time interval only 10 mg of Aln was released from a
the system proposed would be suitable and handy to extrude. 1 ml of NPs-Aln suspension (see Fig. 4A).
The structural integrity of both GG and GG–NPs-Aln was In contrast, by using GG–Aln-NPs system controlled delivery of
immediately recovered after decreasing the strain back to 0.01% Aln was obtained: the system delivered 16  3 mg of Aln within
(see Fig. 3A and B, time 60–90 min). Both G0 and G00 were restored to each 5-day interval up to day 25, which cumulatively accounted to
their initial values, i.e. G0 = 7.1 kPa, G00 = 0.5 kPa, which indicates gel 8% of total drug entrapped within the system (see Fig. 4B). In our
reformation. The self-healing behavior may be ascribed to system drug release was retarded due to the suspension of the NPs
attractive electrostatic interactions that can be reformed after in a GG hydrogel matrix. It is well known that the diffusion rate of
being broken. A similar behavior, i.e. electrostatically-driven molecules through a hydrogel depends mainly on: the amount of
immediate structure restoration after elimination of the shear, liquid in the hydrogel, the distance between polymer chains (the
was previously described by Wang et al. for colloidal gels obtained mesh size), the structural flexibility (Bertz et al., 2013) and their
from oppositely charged gelatin nanospheres (Wang et al., 2011). size and charge (Golmohamadi and Wilkinson, 2013). We
The rheological tests showed that GG–NPs-Aln system behaved hypothesize that for the GG–NPs-Aln system degradation of
predominantly elastically and displayed self-healing behavior, i.e. nanoparticles and thus drug release was retarded mainly due to
the ability to restore its structure upon destruction for example lower availability of free water molecules within the hydrogel as
during injection. compared with the situation when the nanoparticles were freely
suspended in aqueous media (e.g. PBS). As a result hydrolytic
4.3. Drug release from NPs-Aln and injectable GG–NPs-Aln system degradation of PLGA nanoparticles was hindered and less drug was
released from NPs-Aln. However the hydrogel was not an obstacle
In our studies the controlled release of small doses of Aln was to drug release because both Aln and GG are negatively charged
considered a key criterion in selecting the compositions for intra- thus drug transport through the gel structure might be even
bone drug administration. The nanoparticles with encapsulated enhanced. The diffusion of a negatively charged probe (Oregon 2C)
Aln were produced from PLGA 85:15 which is known to degrade through a negatively charged hydrogel (alginate) was already
more slowly than PLGA containing equal molar ratio of lactide and described by Golmohamadi and Wilkinson (2013), who found that
glycolide, i.e. 50:50. The latter is most often used for encapsulation there is depletion of anionic probes within the polyanion structure.
U. Posadowska et al. / International Journal of Pharmaceutics 485 (2015) 31–40 39

4.4. Biological effects of injectable GG–NPs-Aln References

Real time cell analysis together with viability tests and Anhalt, J.P., 1977. Assay of gentamicin in serum by high-pressure liquid
chromatography. Antimicrob. Agents Chemother. 11 (4), 651–655.
microscopic observations were used to assess the behavior of Araujo, J.C., Poblenz, A., Corn, P., Parikh, N.U., Starbuck, M.W., Thompson, J.T., Lee, F.,
osteoblast-like cells and the efficiency of osteoclastic differen- Logothetis, C.J., Darnay, B.G., 2009. Dosatinib inhibits both osteoclast activation
tiation of macrophages in contact with extracts from GG–NPs- and prostate cancer PC-3-cell induced osteoclast formation. Cancer Biol. Ther. 8
(22), 2153–2159.
Aln system. In general, differentiation of macrophages occurs in Babu, R.J., Sathigari, S., Kumar, M.T., Pandit, J.K., 2010. Formulation of controlled
the presence of the recombinant RANKL factor, which causes release gellan gum macro beads of amoxicillin. Curr. Drug Deliv. 7 (1), 36–43.
cell fusion and transformation into multinucleated osteoclast- Bertz, A., Wöhl-Bruhn, S., Miethe, S., Tiersch, B., Koetz, J., Hust, M., Bunjes, H.,
Menzel, H., 2013. Encapsulation of proteins in hydrogel carrier systems for
like cells (Steinberg and Hiken, 2007). We found that RANKL- controlled drug delivery: influence of network structure and drug size on
stimulated RAW 264.7 cells tended to fuse, exhibited a flattened release rate. J. Biotechnol. 163 (2), 243–249.
morphology, were positively stained for TRAP and their Cornelis, F., Truchetet, M.E., Amoretti, N., Verdier, D., Fournier, C., Pillet, O., Gille, O.,
Hauger, O., 2014. Bisphosphonate therapy for unresectable symptomatic benign
normalized cell index was the highest (see Fig. 6), which are
bone tumors: a long-term prospective study of tolerance and efficacy. Bone 58
indicators of osteoclastic differentiation (Steinberg and Hiken, (2), 11–16.
2007; Araujo et al., 2009). When such cells were cultured in Dobrzynski, P., Kasperczyk, J., Janeczek, H., Bero, M., 2001. Synthesis of
GG–NPs-Aln extracts they were round, poorly spread and biodegradable copolymers with the use of low toxic zirconium compounds. 1.
Copolymerization of glycolide with L- lactide initiated by Zr(Acac)4.
behaved as non-differentiated macrophages. The curves plotting Macromolecules 34, 5090–5099.
cell index of RANKL- stimulated RAW 264.7 in contact with Doner, L.W., Douds, D.D., 1995. Purification of commercial gellan to mono gellan
GG–NPs-Aln extracts and cells cultured under control con- cation salts results in acute modification of solution and gel forming properties.
Carbohydr. Res. 273 (2), 225–233.
ditions (i.e. in pure RPMI medium) overlapped. Evidently, Douglas, T., Wlodarczyk, M., Pamula, E., Declercq, H., de Mulder, E., Bucko, M.,
alendronate-induced blockage of RANKL-stimulated cytoskeleton Balcaen, L., Vanhaecke, F., Cornelissen, R., Dubruel, P., Jansen, J., Leeuwenburgh,
organization of monocytes for osteoclasts (Vaananen, 2005) took S.J., 2014. Enzymatic mineralization of gellan gum hydrogel for bone tissue-
engineering applications and its enhancement by polydopamine. J. Tissue Eng.
place, causing an impedance drop for cultures in GG–NPs-Aln Regen. Med. 8 (11), 906–918.
extracts. The results obtained are in accordance with the study Ezzati Nazhad Dolatabadi, J., Hamishehkar, H., Eskandani, M., Valizadeh, H., 2014.
performed by Kim et al. (2010) who showed that Aln-loaded Formulation, characterization and cytotoxicity studies of alendronate sodium-
loaded solid lipid nanoparticles. Colloid Surf. B 117, 21–28.
drug delivery systems reduced osteoclastic differentiation. So, it Fredenberg, S., Wahlgren, M., Reslow, M., Axelsson, A., 2011. The mechanisms of
can be concluded that injectable GG–NPs-Aln systems down- drug release in poly(lactic-co-glycolic acid)-based drug delivery systems – a
regulated osteoclasts formation. review. Int. J. Pharm. 415 (1–2), 34–52.
Gantar, A., da Silva, L.P., Oliveira, J.M., Marques, A.P., Correlo, V.M., Novak, S., Reis, R.
The real time cell analysis of MG-63 cells show that no change
L., 2014. Nanoparticulate bioactive-glass-reinforced gellan-gum hydrogels for
in cell index was found between reference conditions and GG–NPs- bone-tissue engineering. Mater. Sci. Eng. C 43, 27–36.
Aln extracts. Moreover, the morphology as well as the viability of García, M.C., Alfaro, M.C., Calero, N., Munoz, J., 2011. Influence of gellan gum
the cells did not differ (see Fig. 5). It altogether indicates no concentration on the dynamic viscoelasticity and transient flow of fluid gels.
Biochem. Eng. J. 55, 73–81.
negative impact of GG–NPs-Aln on model bone-forming cells. Thus Ghadiri, M., Chrzanowski, W., Lee, W.H., Dehghani, F., Fathi, A., Rohanizedeh, R.,
injectable GG–NPs-Aln system was found to inhibit osteoclasts 2013. Alginate/clay (Laponite1) composite with enhanced properties for
differentiation and at the same time not to affect osteoblasts biomedical applications. Appl. Clay Sci. 85, 64–73.
Golmohamadi, M., Wilkinson, K.J., 2013. Diffusion of ions in a calcium alginate
functions. This is a proof that drug released from the developed hydrogel-structure is the primary factor controlling diffusion. Carbohydr.
system retained its biological activity. Polym. 94 (1), 82–87.
Harmankaya, N., Karlsson, J., Palmquist, A., Halvarsson, M., Igawa, K., Andersson, M.,
Tengvall, P., 2013. Raloxifene and alendronate containing thin mesoporous
titanium oxide films improve implant fixation to bone. Acta Biomater. 9 (6),
5. Conclusion 7064–7073.
Hayden, R.S., Vollrath, M., Kaplan, D.L., 2014. Effects of clodronate and alendronate
on osteoclast and osteoblast co-cultures on silk-hydroxyapatite films. Acta
Our system based on Aln-loaded PLGA nanoparticles suspended
Biomater. 10, 486–493.
in gellan gum hydrogel matrix: (i) was found to be injectable and Huang, W., Wang, Y., Ren, L., Du, C., Shi, X., 2009. A novel PHBV/HA microsphere
restored its elastic structure after extrusion, (ii) assured local and releasing system loaded with alendronate. Mater. Sci. Eng. C 29, 2221–2225.
uniform drug delivery, and (iii) inhibited osteoclasts differentia- Ji, D.-Y., Kuo, T.-F., Wu, H.-D., Yang, J.-C., Lee, S.-Y., 2012. A novel injectable chitosan/
polyglutamate polyelectrolyte complex hydrogel with hydroxyapatite for soft-
tion without affecting osteoblasts functions. The results show the tissue augmentation. Carbohydr. Polym. 89, 1123–1130.
potential of our approach for the local treatment of osteoporosis Kim, J., Kim, I.S., Cho, T.H., Lee, K.B., Hwang, S.J., Tae, G., Noh, I., Lee, S.H., Park, Y., Sun,
and other bone tissue pathologies. K., 2007. Bone regeneration using hyaluronic acid-based hydrogel with bone
morphogenic protein-2 and human mesenchymal stem cells. Biomaterials 28
(10), 1830–1837.
Kim, C.W., Yun, Y.P., Lee, H.J., Hwang, Y.S., Kwon, I.K., Lee, S.C., 2010. In situ
Conflict of interest fabrication of alendronate-loaded calcium phosphate microspheres: controlled
release for inhibition of osteoclastogenesis. J. Control. Release 147 (1),
45–53.
The authors have no conflict of interest to declare. Kohane, D.S., 2007. Microparticles and nanoparticles for drug delivery. Biotechnol.
Bioeng. 96 (2), 203–209.
Le Broc-Ryckewaert, D., Carpentier, R., Lipka, E., Daher, S., Vaccher, C., Betbeder, D.,
Acknowledgments
Furman, C., 2013. Development of innovative paclitaxel-loaded small PLGA
nanoparticles: study of their antiproliferative activity and their molecular
The authors would like to thank to Professor Piotr Dobrzynski interactions on prostatic cancer cells. Int. J. Pharm. 454 (2), 712–719.
(Center of Polymer and Carbon Materials, Polish Academy of Lee, J.H., Koa, I.H., Jeon, S., Chae, J., Chang, J., 2013. Micro-structured hydroxyapatite
microspheres for local delivery of alendronate and BMP-2 carriers. Mater. Lett.
Sciences) for providing us with degradable copolymer, Polpharma 105, 136–139.
S.A. for providing sodium alendronate and Dr. Lukasz Zych Li, Y., Kroger, M., Liu, W.K., 2014. Endocytosis of PEGylated nanoparticles
(Department of Ceramics and Refractories, Faculty of Materials accompanied by structural and free energy changes of the grafted polyethylene
glycol. Biomaterials 35 (3), 8467–8478.
Science and Ceramics) for the access to the Zetasizer. Polish Marshall, J.K., Rainsford, K.D., James, C., Hunt, R.H., 2000. A randomized controlled
National Science Centre (Grant no: 2012/05/B/ST8/00129) and trial to assess alendronate-associated injury of the upper gastrointestinal tract.
Ministry of Health of the Czech Republic (Grant No. NT13297-4/ Aliment. Pharmacol. Ther. 14 (11), 1451–1457.
Mitra, R., Jones, S., 2012. Adjuvant analgesics in cancer pain: a review. Am. J. Hosp.
2012) provided financial support to this project. Palliat. Care 29 (1), 70–79.
40 U. Posadowska et al. / International Journal of Pharmaceutics 485 (2015) 31–40

Montufar, E.B., Maazouz, Y., Ginebra, M.P., 2013. Relevance of the setting reaction to label-free assessment of T-cell-mediated killing of adherent cells. J. Immunol.
the injectability of tricalcium phosphate pastes. Acta Biomater. 9 (4), Methods 405, 192–198.
6188–6198. Reddy, G.T., Kumar, T.M., Veena, K.M., 2005. Formulation and evaluation of
Morais, D.S., Rodrigues, M.A., Silva, T.I., Lopes, M.A., Santos, M., Santos, J.D., Botelho, alendronate sodium gel for the treatment of bone resorptive lesions in
C.M., 2013. Development and characterization of novel alginate-based periodontitis. Drug Deliv. 12 (4), 217–222.
hydrogels as vehicles for bone substitutes. Carbohydr. Polym. 95 (1), 134–142. Shive, M.S., Anderson, J.M., 1997. Biodegradation and biocompatibility of PLA and
Mundargi, R.C., Babu, V.R., Rangaswamy, V., Patel, P., Aminabhavi, T.M., 2008. Nano/ PLGA microspheres. Adv. Drug Deliv. Rev. 28 (1), 5–24.
micro technologies for delivering macromolecular therapeutics using poly(D,L- Steinberg, T.H., Hiken, J.F., 2007. P2 receptors in macrophage fusion and osteoclast
lactide-co-glycolide) and its derivatives. J. Control. Release 125 (3), 193–209. formation. Purinergic Signal. 3 (1–2), 53–57.
Oliveira, J.T., Santos, T.C., Martins, L., Picciochi, R., Marques, A.P., Castro, A.G., Neves, Vaananen, K., 2005. Mechanism of osteoclast mediated bone resorption—rationale
N.M., Mano, J.F., Reis, R.L., 2010. Gellan gum injectable hydrogels for cartilage for the design of new therapeutics. Adv. Drug Deliv. Rev. 57 (7), 959–971.
tissue engineering applications: in vitro studies and preliminary in vivo Wang, H., Hansen, M.B., Lowik, D.W.P.M., van Hest, J.C.M., Li, Y., Jansen, J.A.,
evaluations. Tissue Eng. A 16 (1), 343–353. Leeuwenburgh, S.C.G., 2011. Oppositely charged gelatin nanospheres as building
Peper, J.K., Schuster, H., Löffler, M.W., Schmid-Horch, B., Rammensee, H.G., blocks for injectable and biodegradable gels. Adv. Mater. 23 (12), 119–124.
Stevanovi c, S., 2014. An impedance-based cytotoxicity assay for real-time and

You might also like