Accepted Manuscript

Factorial design formulation optimization and in vitro characterization of curcumin-

loaded PLGA nanoparticles for colon delivery

Mohamed A. Akl, Alma Kartal-Hodzic, Timo Oksanen, Hatem R. Ismael, Mohsen M.

Afouna, Marjo Yliperttula, Ahmed M. Samy, Tapani Viitala

PII: S1773-2247(16)30006-5
DOI: 10.1016/j.jddst.2016.01.007
Reference: JDDST 161

To appear in: Journal of Drug Delivery Science and Technology

Received Date: 30 October 2015

Revised Date: 5 December 2015
Accepted Date: 12 January 2016

Please cite this article as: M.A. Akl, A. Kartal-Hodzic, T. Oksanen, H.R. Ismael, M.M. Afouna, M.
Yliperttula, A.M. Samy, T. Viitala, Factorial design formulation optimization and in vitro characterization
of curcumin-loaded PLGA nanoparticles for colon delivery, Journal of Drug Delivery Science and
Technology (2016), doi: 10.1016/j.jddst.2016.01.007.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

1 Factorial design formulation optimization and in vitro characterization of

2 curcumin-loaded PLGA nanoparticles for colon delivery

PT
4 Mohamed A. Akla,b, Alma Kartal-Hodzica, Timo Oksanena, Hatem R. Ismaelb, Mohsen M. Afounab,

RI
5 Marjo Yliperttulaa, Ahmed M. Samyb and Tapani Viitalaa,*

SC
6

a
7 Centre for Drug Research, Division of Pharmaceutical Biosciences, Faculty of Pharmacy, University

U
8 of Helsinki, Finland
AN
b
9 Department of Pharmaceutics and Ind. Pharmacy, Faculty of Pharmacy (Boys), Al-Azhar University,
M

10 Nasr City, Cairo, Egypt

11
D
TE

12 * Corresponding author:

13 Dr. Tapani Viitala, Centre for Drug Research, Division of Pharmaceutical Biosciences, Faculty of
EP

14 Pharmacy, University of Helsinki, P.O. Box 56, 00014 Helsinki, Finland

C

15 Phone: +358504154529
AC

16 E-mail: [email protected]

1
 ACCEPTED MANUSCRIPT

17 Abstract

18 We have employed a 23 factorial design approach for optimizing curcumin-loaded PLGA nanoparticles

19 with a specific interest in colon treatment. Curcumin-loaded PLGA nanoparticles were prepared by

PT
20 using the single emulsion solvent evaporation technique. In vitro characterizations of the curcumin-

21 loaded nanoparticles revealed that the mean particle sizes of the nanoparticles ranged from 181.5 nm to

RI
22 206.9 nm, the zeta potential values were in the range of -30.6 to -41.7 mV, the encapsulation

efficiencies were between 58.1 - 83.2 % and the drug release from the formulations was in the range of

SC
23

24 34.4 - 62.8 %. The properties of the optimized curcumin-loaded PLGA nanoparticles predicted by the

U
25 23 factorial design approach correlated very well with the experimentally determined particle size of
AN
26 219.6 nm, zeta potential of -36.8 mV, encapsulation efficiency of 74.4 % and a 56.2 % cumulative drug

27 release after 24 h. In vitro cellular uptake studies with HT-29 cells showed that the optimized
M

28 curcumin-loaded PLGA nanoparticle exhibited a much higher cellular uptake of curcumin (i.e. 7.01 ±

29 0.33 µg / 106 cells) than a native curcumin solution (3.74 ± 0.56 µg / 106 cells). Stability studies also
D

30 showed that all investigated formulations were stable at least for 2 months.
TE

31
EP

32
C

33
AC

34 Key words: Curcumin; PLGA nanoparticles; Poly vinyl alcohol; Factorial design; colon targeting

35

2
 ACCEPTED MANUSCRIPT

36 1. Introduction

37 Colon cancer is the third leading cause of cancer deaths in the United States [1]. The American Cancer

38 Society has estimated that there will be about 93 090 new cases of colon cancer and 39 610 new cases

PT
39 of rectal cancer in the United States in 2015, and these are expected to cause about 49 700 deaths

40 during 2015. The lifetime risk of developing a colorectal cancer is about 1 in 20 [1]. The most common

RI
41 treatment for cancer is chemotherapy. However, chemotherapy is often associated with a number of

drawbacks such as nonselective distribution of drugs, multidrug resistance, enhanced drug toxicity,

SC
42

43 undesirable side effects on normal tissue and inherent lack of beneficial response of cytotoxic

U
44 anticancer drugs [2,3]. Therefore, there is an urgent need to develop therapeutic modalities with no or
AN
45 minimal side effects on healthy tissues or organs.

46 Natural herbal extracts, such as curcumin, hold high potential as chemotherapeutic agents because they
M

47 are often safe and do not show side effects on healthy tissues or organs. They also show chemo
D

48 preventive activities against malignancy. In recent years curcumin has drawn the attention of research
TE

49 because it sensitizes cancer cells for chemotherapy by inducing programmed cell death [4,5]. Curcumin

50 is a polyphenol of turmeric derived from the roots (rhizomes) of Curcuma longa. Several studies have
EP

51 shown that curcumin has anti-inflammatory [6], antioxidant [7] and antimicrobial effects [8]. The most

52 important effect is, however, its potential use against cancer due to its ability to suppress the
C

53 proliferation of a wide variety of tumor cells [4,9]. Curcumin has been reported to be a potent inhibitor
AC

54 of the nuclear factor-kappa B (NF-kappa B) activation in various human cancer cell lines [10,11,12]. It

55 has also been reported to decrease multidrug resistance by down regulating the P-glycoprotein (PGP)

56 expression in resistant cells. Unfortunately, the high potential of curcumin for treatment of cancer and

57 chronic inflammation is hampered by its low aqueous solubility at physiological pH conditions, rapid

58 decomposition in alkaline media and instability towards light. For example, curcumin is practically

3
 ACCEPTED MANUSCRIPT

59 insoluble in neutral aqueous solutions (i.e. ~ 1 µM), but dissolves readily in organic solvents such as

60 alcohols, ketones, esters, and organic acids (i.e. in the range between 1 – 30 mM) [13]. The solubility

61 of curcumin in dichloromethane has also been reported to be at least 3 mM [14]. The hydrophobic

62 character of curcumin results in pharmacokinetic restrictions such as low absorption, poor

PT
63 bioavailability by oral route, extensive metabolism and rapid elimination [15]. Numerous approaches

RI
64 have been made to ameliorate the bioavailability of curcumin. These include the use of adjuvants,

65 which can block metabolic pathways of curcumin (for example Piperine, a known inhibitor of hepatic

SC
66 and intestinal glucuronidation) [15]. The main strategies used to overcome the physicochemical

67 limitations of curcumin and to increase its bioavailability are based on loading the compound in

68
U
nanocarriers, such as liposomes [16,17], polymeric micelles [18], phospholipid complexes [19] and
AN
69 polymeric nanoparticles [15,20]. Nanocarriers have been shown to increase the solubility, provide

70 longer circulation times, enhance permeability, and increase resistance towards metabolic processes of
M

71 curcumin. Some formulations have been designed to prolong the release of curcumin, while others have
D

72 additional mechanisms to improve the cellular delivery or intracellular release.

TE

73 Biodegradable polymeric nanoparticles (PNPs) have widely been developed as nanoscale drug delivery

74 vehicles for cancer treatment due to their excellent endocytosis efficiency, passive tumor-targeting,
EP

75 high encapsulation efficiency, and ability to deliver a wide range of therapeutic agents [21,22,23].
C

76 Nanoencapsulation protects the molecules from premature degradation, improves their solubility, and
AC

77 promotes controlled drug release and drug targeting. Nanoparticles (NPs) can at best represent a low

78 risk of toxicity and they can often enhance the drug efficacy, i.e. specificity, tolerability and therapeutic

79 index [24]. Poly (lactic-co-glycolic acid) (PLGA) is an approved biodegradable polymer with good

80 biocompatibility. Therefore, PLGA and its various derivatives have widely been explored as carriers

81 for controlled delivery of a large number of both small molecular drugs and macromolecular

4
 ACCEPTED MANUSCRIPT

82 therapeutics [25,26]. The majority of PLGA-based drug delivery carriers are in the form of injectables,

83 but PLGA-based nanoparticles are also continuously investigated for oral delivery [27,28,29].

84 The main objective of this study was to preliminary investigate the use of PLGA for encapsulating

PT
85 curcumin into PNPs for oral colon delivery purposes. A prerequisite for this is to confirm that

86 curcumin-loaded PLGA nanoparticles withstand the harsh conditions in the stomach and small intestine

RI
87 before they reach the colon. Therefore, we have focused on the systematic optimization of the basic

composition of curcumin-loaded PLGA nanoparticles with the aim to improve the aqueous

SC
88

89 dispersability of curcumin, to protect it from degradation in simulated gastrointestinal fluids and to

U
90 increase curcumin uptake in colon cancer cells. Curcumin-loaded PNPs (C-PNPs) were prepared with
AN
91 an emulsion solvent evaporation technique by using a 23 factorial design approach to optimize the

92 composition of the C-PNP. The physicochemical properties and in vitro characterization of C-PNPs
M

93 were determined and performed through particle size, zeta potential, drug encapsulation, drug release

94 kinetics, Fourier-Transform infrared spectroscopy and differential scanning calorimetry studies.

D

95 Furthermore, the stability of the C-PNPs at room temperature and 4 °C, the stability of the optimized
TE

96 C-PNP against enzyme and bile salt, and the in vitro cellular uptake of the optimized C-PNP

97 formulation in colon cancer cells were also investigated.

EP

98
C

99 2. Materials and Methods

AC

100 2.1 Materials

101 Curcumin was purchased from Cayman chemical company (China). Acid terminated Poly (D, L-

102 lactide-co-glycolide) (PLGA, 50:50, Mw = 7.000-17.000, Resomer® RG 502H), Polyvinyl alcohol

103 (PVA, mw 31,000–50,000 – 87-89 % hydrolyzed), Dichloromethane (DCM) and Phosphate Buffer

5
 ACCEPTED MANUSCRIPT

104 Saline (PBS) tablet were purchased from Sigma-Aldrich (Germany). Potassium dihydrogen phosphate

105 was purchased from Riedel-de-Haen (Germany). Acetonitrile, Phosphoric acid and Methanol were

106 purchased from Fluka (United Kingdom). OPTI-MEM I, McCoy's 5a Medium, and TrypLE Express

107 were obtained from Gibco, Life Technologies (USA).

PT
108

RI
109 2.2. Preparation and characterization of the formulations

SC
110 2.2.1. Experimental design

U
111 Statgraphics® plus software version 4 was used to adapt the 23 factorial design approach to optimize
AN
112 the curcumin-loaded PLGA nanoparticle (C-PNP). The independent variables for optimization were 1)

113 the amount of PLGA in the organic phase (X1 = c PLGA), 2) the PVA concentration in the aqueous
M

114 phase (X2 = c PVA) and 3) the amount of curcumin in the organic phase (X3 = c CUR). The selected

115 responses were 1) the mean particle size of the C-PNPs (Y1), 2) the zeta potential of the C-PNPs (Y2),
D

116 3) the encapsulation efficiency (Y3) and 4) the percentage (%) of the cumulative drug released after 24
TE

117 h (Y4). Each independent variable was given a high and low level value as shown in Table 1.
EP

118

119 Table 1. Independent and dependent variables used for the 23 factorial design approach.
C
AC

Independent variables Level Dependent variables

Low High

Amount of PLGA 100 mg 200 mg Particle size

6
 ACCEPTED MANUSCRIPT

Concentration of PVA 2% 4% Zeta potential

Amount of curcumin 10 mg 20 mg Encapsulation efficiency

In vitro release over 24 h.

PT
120

RI
121

SC
122 2.2.2 Preparation of curcumin-loaded PLGA nanoparticles (C-PNPs)

U
123 C-PNPs were prepared by using the single emulsion-solvent evaporation technique reported by Khalil
AN
124 et al. [24] with slight modifications. Briefly, 100-200 mg of PLGA polymer was dissolved in 5 ml of

125 dichloromethane (DCM) in a glass tube. Then 10 or 20 mg of curcumin powder was added to the
M

126 polymer/solvent mixture and allowed to dissolve for 30 minutes, with intermittent vortexing. The

127 organic phase containing PLGA and curcumin was rapidly added drop wise into a glass tube containing
D

128 10 mL of PVA in an aqueous solution, while vortexing. Once all the drug/polymer mixture was added
TE

129 to the PVA solution, the contents were vortexed for an additional 10 s at a high setting. The tube

130 contents were then emulsified by using sonication for 7 min at 40% amplitude in an ice - water bath by
EP

131 using a probe sonicator (Vibra-Cell VCX 750 sonicator, Sonics & Materials Inc., Newtown, CT, USA).

The resulting fine (O/W) emulsion was immediately poured into 30 ml of an aqueous PVA 0.5% w/v
C

132

solution under rapid stirring with a magnetic stirrer. Dichloromethane was then evaporated from the
AC

133

134 droplets under magnetic stirring at 800 rpm for 3 h. The nanoparticles were collected by centrifugation

135 at 20,000 rpm for 15 minutes and washed 3 times with Milli-Q water. Supernatants were collected to

136 evaluate the encapsulation efficiency of curcumin. Finally, the pellet of the nanoparticles was re-

137 suspended in 5 ml of Milli-Q water.

7
 ACCEPTED MANUSCRIPT

138

139 2.2.3. Physicochemical characterization and drug encapsulation efficiency

140 2.2.3.1. Particle size determination

PT
141 The average particle size and PDI of the C-PNPs were measured with dynamic light scattering by using

RI
142 ZetaSizer Nano ZS instrument (Malvern Instruments Ltd, Worcestershire, UK). Samples for particle

143 size measurements were prepared by taking 300 µl of each nanoparticle suspension and diluting these

SC
144 with 3 ml of Milli-Q water and sonicating for 30 s. Particle size measurements were performed at RT

145 with a detection angle of 90о. Each sample was measured in triplicate.

U
AN
146

147
M

148
D

149
TE

150 2.2.3.2. Zeta potential measurements

EP

151 Samples for zeta potential measurements were prepared in the same way as those for particle size

measurements. Zeta potentials for three replicates of each sample were measured at RT with a
C

152
AC

153 ZetaSizer Nano ZS instrument (Malvern Instruments Ltd, Worcestershire, UK).

154

155 2.2.3.3. Determination of drug encapsulation efficiency

8
 ACCEPTED MANUSCRIPT

156 The encapsulation efficiency of curcumin in the C-PNPs was determined by a direct technique reported

157 by Tsai et al. [20] with minor modifications. Briefly, 1 ml of each C-PNP suspension was centrifuged

158 at 15,000 rpm for 30 min. The supernatant was removed after centrifugation and 1 ml of acetonitrile (a

159 common solvent for PLGA and the drug) was added to the C-PNP pellets. Then the samples in

PT
160 acetonitrile were sonicated for 15 min to break the C-PNP structure and to release the curcumin. The

RI
161 amounts of curcumin in the C-PNPs were analyzed by UPLC at 426 nm after dilution with the mobile

162 phase composed of 38% acetonitrile and 62% of 0.05% phosphoric acid. The drug encapsulation

SC
163 efficiency (EE %) for each C-PNP was then calculated by using the following equation:

U
164
AN
165 % EE = [(Curcumin encapsulated/Curcumin total) × 100]. (1)
M

166

167
D
TE

168

169 2.2.4. In vitro drug release and release kinetics

EP

170 2.2.4.1. In-vitro curcumin release

C

171 The in vitro release of curcumin from the C-PNPs was determined by the dialysis bag method reported
AC

172 earlier by Zanotto-Filho et al. [30] with slight modifications. Briefly, 1 ml of each C-PNP solution was

173 transferred in dialysis bags (Sigma, St. Louis, MO, USA) with a molecular cut-off 12-14 kDa. The bags

174 were suspended in 150 ml of release medium containing 2% SDS. SDS was employed in the buffer

175 solution not only to maintain sink condition but also to provide solubility for curcumin in the aqueous

176 phase. The whole set-up was placed in a shaking water bath (Grant OLS200, Cambridge Ltd, England)

9
 ACCEPTED MANUSCRIPT

177 adjusted to a constant speed of 100 rpm at 37 ± 0.5°C. Samples were withdrawn at appropriate time

178 intervals from the outer solution to estimate the percentage of the drug released. In order to mimic the

179 in vivo release characteristics of the C-PNP formulations, a three-stage approach with three-different

180 pH release media was used following the recommendations on methods for dosage forms testing by US

PT
181 Pharmacopeia 36 [31]. The time interval and medium used for the three different stages were the

RI
182 following:

Stage 1. Time interval: 1 – 2 h. Simulated gastric fluid (SGF); 0.2 g sodium chloride, 0.32 g pepsin and

SC
183

184 0.7 ml concentrated HCl in 100 ml water, pH 1.2.

U
185 Stage 2. Time interval: 3 – 5 h. Simulated intestinal fluid (SIF); 0.68 mg KH2PO4 and 7.7 ml of 0.2 M
AN
186 NaOH in 100 ml water, pH 6.8.
M

187 Stage3. Time interval: 6 – 24 h. Simulated colonic fluid SCF; PBS, pH 7.4.

188 One ml of sample was taken from the outer solution at appropriate time intervals, i.e. 1, 2, 3, 4, 5, 6, 7,
D

189 8, 12 and 24 h. The sample was then filtered by using a 0.45 µm filter membrane. The curcumin
TE

190 concentration was determined by diluting the filtrate with the mobile phase mixture composed of 38%

191 of acetonitrile and 62% of 0.05 % phosphoric acid and measured with UPLC at λmax 426 nm. After the
EP

192 sample was taken, 1 ml of fresh medium was always added back to the outer media.
C

193
AC

194 2.2.4.2. Drug release kinetics

195 The in vitro drug release data were fitted to three different kinetic models which are often used to

196 describe the behavior of the drug release from nanospheres, i.e. zero-order, first-order and Higuchi

197 diffusion models. The proper mode of release was determined on the basis of the correlation coefficient

10
 ACCEPTED MANUSCRIPT

198 (r) for the linear regression fit of the parameters involved, where the highest correlation coefficient

199 represents the actual mode of the release.

200

PT
201 2.2.5. Characterization of the optimized C-PNP formulation

RI
202 2.2.5.1. Morphology of the polymeric nanoparticles

SC
203 Transmission electron microscopy (TEM) was used for morphological analysis of the optimized C-PNP

204 formulation. The samples of freshly made C-PNP suspensions in aqueous media were dropped onto a

U
205 copper grid with a carbon film support and extra solution was removed with a filter paper. Samples
AN
206 were observed directly with TEM (JEOL 1200 EX II, JEOL Ltd., Tokyo, Japan) after they were dried

207 below room temperature without further staining.

M

208
D

209
TE

210 2.2.5.2. Fourier- Transform infrared spectroscopy (FT-IR)

EP

211 The FT-IR spectra (range 4000–650 cm−1) were recorded on freeze dried samples using a Vertex 70

212 (Bruker Optics GmbH, Germany) spectrophotometer (spectral resolution of 4 cm−1 and 32 co-added
C

213 scans) equipped with a MIRacleTM ATR device (Pike Optics, Germany) with a single reflection
AC

214 diamond crystal (1.8 mm spot size). The raw materials, the blank optimized PNP without curcumin and

215 the optimized C-PNP formulations were analyzed with attenuated total reflectance (ATR). The samples

216 were deposited on top of a diamond crystal and secured with a high-pressure clamp. The average of

217 characteristic peaks of IR transmission spectra was recorded from triplicate samples.

11
 ACCEPTED MANUSCRIPT

218

219 2.2.5.3. Differential scanning calorimetry (DSC)

220 DSC studies were performed in order to characterize the physical state of curcumin in the optimized C-

PT
221 PNP formulation. Thermograms of the raw materials, the blank optimized PNP without curcumin and

222 the optimized C-PNP formulations were obtained by using a DSC 823 instrument (Mettler Toledo,

RI
223 Switzerland). About 4 – 5 mg of each sample was weighed, crimped into an aluminum pan along with

SC
224 the standard reference aluminum pan and analyzed at a scanning temperature range from 50 to 300 ◦C

225 at a heating rate of 10 ◦C/min. The baseline was optimized before each run.

226
U
AN
227
M

228
D

229
TE

230 2.2.6. In vitro cellular drug uptake

EP

231 2.2.6.1. Cell culture and cell preparation

Human Caucasian colon adenocarcinoma (HT29) cells, obtained from the European Collection of Cell
C

232
AC

233 Cultures (ECACC), were cultured in a McCoy's 5a medium supplemented with 10% FBS at a density

234 of about 2 x 106 cells/mL in 75 cm2 culture flasks. The cells were maintained at 37 °C, 95% relative

235 humidity and 5% CO2. The cells were harvested weekly from plastic flasks (75 cm2) with TrypLE

236 Express and the medium was changed twice a week.

237

12
 ACCEPTED MANUSCRIPT

238 2.2.6.2. Cellular drug uptake assay

239 HT29 cells were cultured in a 12-well plate with a cell density of 1× 104 cells/per well and incubated in

240 a 5% CO2 incubator at 37 °C for 24 h before cellular drug uptake studies. The cell association assay

PT
241 was carried out at 37 °C in OPTI-MEM I reduced serum medium which contained either enough pure

242 curcumin or the optimized C-PNP formulation. Each well contained the same absolute amount of 200

RI
243 µg curcumin. Washing the wells with ice-cold PBS three times terminated the incubation of both the

SC
244 cells with pure curcumin and the optimized C-PNP formulation. The washed cells were lysed in PBS

245 containing 0.5% Triton X-100, after which they were vortexed for 3 min. Then the cell lysate was

U
246 centrifuged to remove the cellular debris. The amount of curcumin in the cell lysate was determined by
AN
247 diluting the cell lysate with the mobile phase mixture composed of 38% of acetonitrile and 62% of 0.05

248 % phosphoric acid and measured with UPLC at λmax = 426 nm.
M

249
D

250
TE

251 2.2.7. In vitro stability of C-PNP formulations

EP

252 The evaluation of the in vitro stability of the C-PNP formulations was performed with two different

253 tests: 1) the effect of storage temperature on particle size and 2) the effect of SGF, SIF and SCF on
C

254 particle size. The particle size of each C-PNP stored for up to two months at 4°C and RT was regularly
AC

255 measured at RT with a ZetaSizer Nano ZS instrument (Malvern Instruments Ltd, Worcestershire,

256 UK).The pH dependency of the stability of the C-PNPs was evaluated at three pHs (i.e. pH 1.2, 6.8,

257 7.4) in order to evaluate the stability of the C-PNPs in the gastrointestinal tract. The C-PNPs were

258 incubated at 37 °C and samples were collected at predetermined time points of 0, 0.5, 2, 6 and 24 h.

259 Simulated gastric fluid (SGF) with and without pepsin and simulated intestinal fluid (SIF) with and

13
 ACCEPTED MANUSCRIPT

260 without pancreatin were prepared according to the US Pharmacopeia 36 [31]. SGF was composed of

261 0.2% sodium chloride (NaCl), 0.2% pepsin, and 0.7% hydrochloric acid (HCl) in water, and the pH

262 was adjusted to 1.2. SIF was composed of 0.68% monobasic potassium phosphate (KH2PO4) (pH

263 adjusted to 6.8 using NaOH) and 3 mM sodium taurocholate to simulate the effect of bile salts. For

PT
264 each C-PNP formulation, 5 ml of the suspension was added to 15 ml of simulated fluid followed by

RI
265 incubation for 2 h in SGF and 6 h in SIF at 37 °C.

SC
266

267 3. Results

268
U
3.1. Curcumin-loaded PLGA nanoparticles for optimizing the C-PNP composition
AN
269 The single-emulsion solvent-evaporation method was successfully utilized for preparing a range of C-
M

270 PNP formulations. All together eight different C-PNP formulations with different ratios of curcumin,

271 PLGA and PVA (C-PNP1 to C-PNP8, Table 2) were prepared and characterized according to the 23
D

272 factorial design approach. The amounts of curcumin (X1), PLGA (X2) and PVA (X3) were chosen as
TE

273 independent parameters, whereas measurements of particle size, polydispersity index (PDI), zeta

274 potential (ZP), encapsulation efficiency (EE%) and percentage of drug released after 24 h (DR%) were
EP

275 used as input of the dependent parameters in the 23 factorial design model in order to find the most
C

276 optimal C-PNP composition.

AC

277

278 3.2. Experimentally determined input parameters of C-PNPs for optimization purposes

279 Table 2 shows the measured values of particle size, PDI and ZP, EE% and DR% for the eight C-PNP

280 formulations prepared during the optimization stage. It is clear that the particle sizes of all the C-PNPs

14
 ACCEPTED MANUSCRIPT

281 were in the nano-size range. The smallest particle size of 181.5 nm was observed for C1-PNP and the

282 largest particle size of 206.9 nm was observed for C4-PNP. The PDI values of the C-PNPs were also

283 very small (i.e. between 0.010 – 0.060), which shows that the particle size populations of the C-PNPs

284 were very homogeneous. A small PDI value (i.e. ~ 0.2) is often taken as an indication of a homogenous

PT
285 particle size population, while a larger PDI value (i.e. > 0.3) reflects a heterogeneous particle size

RI
286 population.

The ZP of all of the C-PNP formulations displayed a high negative value between -30.5 mV to -41.7

SC
287

288 mV. A ZP value for nanoparticles in this range usually predicts that the nanoparticles are very well

U
289 dispersed in the aqueous media and that the nanosuspensions will have a very good stability and
AN
290 tolerance against aggregation.

291
M

292
D

293
TE

294 Table 2. Composition of curcumin-loaded PLGA nanoparticles and experimentally measured input
EP

295 responses.

Run Factor Response

C
AC

(Code value)

X1 X2 X3 PDI Particle size Zeta potential Encapsulation Cumulative drug

(nm) Y1 (mV) Y2 efficiency (%) Y3 release (DR%)

Y4
C-PNP1 100 4 10 0.060 ± 0.011 181.5 ± 2.2 -41.7 ± 0.7 59.1 ± 1.9 34.4

C-PNP2 200 4 20 0.040 ± 0.041 204.6 ± 2.8 -37.5 ± 1.4 77.9 ± 2.0 39.056

15
 ACCEPTED MANUSCRIPT

C-PNP3 100 4 20 0.020 ± 0.015 192.2 ± 0.5 -33.5 ± 1.2 61.8 ± 1.7 40.28

C-PNP4 200 2 20 0.010 ±0.014 206.9 ± 2.1 -40.8 ± 0.3 83.2 ± 2.1 62.822

C-PNP5 100 2 10 0.030 ± 0.008 185.0 ± 2.1 -30.6 ± 0.8 58.1 ± 1.3 36.822

C-PNP6 200 4 10 0.035 ± 0.017 191.1 ± 1.9 -36.2 ± 1.2 66.3 ± 0.4 46.48

PT
C-PNP7 100 2 20 0.034 ± 0.035 196.7 ± 1.4 -37.5 ± 0.8 63.9 ± 3.5 36.67

RI
C-PNP8 200 2 10 0.042 ± 0.034 200.9 ± 2.2 -40.0 ± 0.9 75.5 ± 2.3 38.58

296

SC
297 The EE%s of the C-PNPs were also very promising. A high EE% is usually desirable for NPs because

U
298 it increases the probability that the NPs can deliver and release a sufficient amount of drug to the target
AN
299 site, especially if the NPs are designed so that they have a proper triggered release mechanism and drug

300 release rate. The highest and the lowest EE% of 83.2% and 55.1% were observed for C4-PNP and C1-
M

301 PNP, respectively (Table 2). The EE%s of the C-PNPs in this study clearly correlated with the particle

302 size, i.e. the larger the particle size of the C-PNP the higher the EE% and vice versa.
D
TE

303 The in vitro drug release characteristics of the C-PNPs were determined by using a three-different pH

304 release media (SGF pH 1.2, SIF pH 6.8 and SCF pH 7.4) approach in order to mimic the GI tract. The
EP

305 in vitro release profiles of curcumin were obtained by graphing the cumulative percentage of the drug

306 released with respect to the total amount of curcumin encapsulated in the C-PNPs as a function of the
C

307 time (Figure 1). Negligible amounts of curcumin were released in the buffer medium at pH 1.2 which
AC

308 indicates that curcumin is not released from any of the C-PNPs at gastric and ileal pH (acidic pH).

309 Only a slight release of curcumin took place at pH 6.8 (i.e. < 5 %), whereas a substantial amount of

310 curcumin was released at the higher colonic pH values of pH = 7.4. All the C-PNPs also displayed a

311 sustained release of curcumin up to 24 h at colonic pH values (Figure 1). The highest release of

312 curcumin after 24 h, 62.822 %, was observed for the C-PNP4 formulation (Figure 1A, Table 2),

16
 ACCEPTED MANUSCRIPT

313 whereas the lowest release of curcumin, 34.4 %, was observed for the C-PNP1 formulation (Figure 1A,

314 Table 2).

315 The release kinetic parameters obtained by fitting kinetic models to the drug release data of the C-PNP

PT
316 formulations are shown in Table S1 (supporting information). The release of curcumin from the C-

317 PNPs followed mainly the Higuchi diffusion model, except for the C-PNP7 formulation which followed

RI
318 best the zero order model. This indicates that the release of curcumin from the C-PNP formulations was

mainly taking place by diffusion from the matrix and not due to the degradation of PLGA.

SC
319

320

321
U
AN
322
M

323
D
TE
C EP
AC

324 A)

325

17
 ACCEPTED MANUSCRIPT

PT
RI
SC
326 B)

U
AN
327 Figure 1. In-vitro release of curcumin from A) C-PNP1 – C-PNP4 and B) C-PNP5 – C-PNP8

328 formulations.
M

329
D

330 3.3. Optimization of C-PNP composition by using the 23 factorial design model
TE

331 The criteria for optimizing the C-PNP formulation were smaller particle size, higher zeta potential,

332 higher EE % and higher DR%. The experimentally measured input parameters were analyzed by using
EP

333 the Statgraphics® plus software. The ANOVA analysis for the dependent parameters where the
C

334 statistical significance of each effect was tested by comparing the mean square against an estimate of
AC

335 the experimental error is shown in Table S2 (supporting information). The analysis provided

336 mathematical relationships between the independent variables and the dependent responses in the form

337 of polynomial equations (Table S3, supporting information). These polynomial equations represent the

338 quantitative effect of the process variables (X1, X2 and X3) and their interaction on the designated

339 responses and can be graphically visualized through Pareto charts, main effects and surface plots

18
 ACCEPTED MANUSCRIPT

340 (Figures S1-S3, supporting information). Table S2 and Figures S1-S3 show that the independent

341 variables selected for our study, i.e. the amount of PLGA (X1), concentration of PVA (X2) and amount

342 of curcumin (X3), significantly influenced the dependent parameters of particle size (Y1), zeta potential

343 (Y2), EE % (Y3) and drug release % (Y4).

PT
344 The final optimal composition of the C-PNP9 formulation obtained from the analysis by using the 23

RI
345 factorial design model contained 176.8 mg of PLGA, 2% of PVA and 16.6 mg of curcumin (Table 3).

This optimized formulation was further characterized by particle size, PDI, ZP, EE%, DR%, FT-IR

SC
346

347 spectroscopy, DSC and finally evaluated with in vitro cell uptake studies.

U
348
AN
349
M

350
D

351 Table 3. Composition of factors and responses for the optimized C-PNP9 formulation.
TE

Independent Variable Low High Optimum

X1 : Amount of PLGA in organic Phase 100 mg 200 mg 176.8 mg
EP

X2 : Concentration of PVA in Aqueous Phase 2% 4% 2%

X3 : Amount of Curcumin in organic phase 10 mg 20 mg 16.6 mg
C

Response Predicted Observed Residual

AC

Y1 : Particle size 202.3 nm 219.6 nm -17.3

Y2 : Zeta potential -39.7 mV -36.8 mV 2.9
Y3 : Entrapment Efficiency 76.4 % 74.4% 2.0
Y4 : Cumulative drug release after 24 h. 50.4% 56.2% -5.8
352

353

19
 ACCEPTED MANUSCRIPT

354 3.4. Physico-chemical properties of optimized PNPs

355 The experimentally measured particle size, PDI, ZP and EE% of the optimized C-PNP9 formulation

356 were 219.6 ± 1.6 nm, 0.186 ± 0.023, -36.8 and 74.4 ± 8.2%, respectively (Table 3). The average

PT
357 diameter and PDI demonstrate that the optimized C-PNP9 formulation is composed of nanoparticles

358 with a narrow size distribution. The size, PDI and ZP of a blank PNP with the same composition as the

RI
359 optimized C-PNP but without curcumin were found to be 198.3 ± 4.1 nm, 0.270 ± 0.021 and –41.5 ±

0.1 mV, respectively.

SC
360

361 In order to assess the protection of the content under acidic conditions and to verify an effective

U
362 delivery at neutral pH the in vitro release profile of curcumin from the optimized C-PNP9 formulation
AN
363 was investigated in buffers with different pH values. As was previously seen for the C-PNP1 – C-PNP8

364 formulations, negligible amounts of curcumin were released in the buffer medium at pH 1.2 meaning
M

365 that curcumin was not released from the optimized C-PNP at gastric and ileal pH (acidic pH). This
D

366 indicates that the optimized C-PNP9 formulation was able to protect curcumin under highly acidic
TE

367 conditions. Similarly, only a slight release of curcumin took place at pH 6.8 (< 5 %), whereas a

368 substantial amount of curcumin was released from the optimized C-PNP9 at higher colonic pH values
EP

369 of pH = 7.4 (i.e. 56.16 % over 24 h and 72.65 % over 48 h). The in vitro release data of curcumin for

370 the optimized C-PNP9 formulation were best fitted by a Higuchi diffusion model with r2 = 0.974369
C

371 (Table S1, supporting information).

AC

372

373 3.5. FT-IR analysis

374 FT-IR spectroscopy was used to identify the chemical composition of the materials used for the

375 optimized C-PNP9 formulation, and for revealing any significant interactions between the materials in

20
 ACCEPTED MANUSCRIPT

376 the formulation mixtures. Figure 2A shows the characteristic FT-IR spectra of acid terminated PLGA

377 powder, curcumin powder, the blank optimized nanoparticle without curcumin (B-PNP9) and the

378 optimized curcumin-loaded nanoparticle (C-PNP9).

379 The PLGA spectrum shows a strong peak at 1750 cm-1 which is attributed to carbonyl stretching (C=O)

PT
380 in the PLGA backbone [32]. The strongest peak, seen at 1087 cm-1, is a characteristic peak for PLGA

RI
381 and originates from the C‐O‐C stretching. The shoulder at 1050 cm-1 is attributed to C‐CH3 vibrations

382 from the polymer backbone, while the peak at 1454 cm-1 is due to the C‐H stretching in the methyl

SC
383 groups. The bands at 1385 cm-1 and 877 cm-1 are assigned to O–H in-plane and out-of-plane bending

384 vibrations, respectively. The band at 757 cm-1 is assigned to out-of-plane bending vibration of C–H.

385
U
Other peaks such as 1184 cm-1 were assigned to bonded hydroxyl end groups of the PLGA chains, and
AN
386 1268 cm-1 and 1385 cm-1 to the PLGA ester groups. The blank optimized nanoparticle without

curcumin (B-PNP9) showed the same characteristic peaks in the FT-IR spectrum as the pure PLGA
M

387

388 powder which indicates that PLGA remains intact during the processing steps to form nanoparticles.
D

389 The FT-IR spectrum of curcumin (Figure 2A) shows a characteristic absorption band at 3,490 cm-1,
TE

390 which can be attributed to the phenolic O–H stretching vibration [33]. The bands at 1506 and 1429 cm-1

391 are due to the stretching vibrations of C–C of the benzene ring and olefinic bending vibration of the
EP

392 C=C group bound to the benzene ring, respectively [34]. These peaks were shifted to 1515 and 1453

393 cm-1 in the optimized C-PNP9 formulation which confirmed that curcumin was encapsulated by the
C

394 PLGA polymer. The peak at 1629 cm-1 is attributed to the carbonyl (–C=O) stretching of the
AC

395 conjugated ketone. This was shifted to 1753 cm-1 in the optimized C-PNP9 formulation. Other peaks in

396 the curcumin spectrum were the sharp absorption peaks at 1602 cm-1 corresponding to the benzene ring

397 stretching [35], the peak at 810 cm-1 originating from the stretching vibration of C–O in –C–OCH3 and

398 the peak for the C–O–C stretching of ether that appeared at 1029 cm-1 [36]. The peak for the phenyl

399 ring was observed at 867 cm-1 in the optimized C-PNP9 formulation. All major bands of both curcumin
21
 ACCEPTED MANUSCRIPT

400 and PLGA shift to higher frequencies in the optimized C-PNP9 formulation indicating that curcumin

401 and PLGA were bound together to form a more stable nanoparticle. The optimized C-PNP9 formulation

402 showed characteristic peaks which were close to the principal IR peaks of the drug, confirming the

403 presence of curcumin in the nanospheres.

PT
404

RI
405

406

SC
407

408

409
U
AN
M
D
TE
C EP

410 A)
AC

411

22
 ACCEPTED MANUSCRIPT

PT
RI
SC
412 B)

413
U
Figure 2. A) FT-IR spectra of curcumin powder, acid terminated PLGA powder, blank optimized
AN
414 nanoparticle without curcumin (B-PNP9) and optimized curcumin-loaded nanoparticle (C-PNP9). B)
M

415 DSC Thermograms of curcumin powder, acid terminated PLGA powder, blank optimized nanoparticle

416 (B-PNP9) and optimized C-PNP9 formulation.

D

417
TE

418 3.6. DSC analysis

419 DSC measurements were employed to study the crystal transformation of the optimized C-PNP9
EP

420 formulation. The DSC curves of native curcumin powder, acid terminated PLGA powder, B-PNP9 and
C

421 optimized C-PNP9 are shown in Figure 2B. Native curcumin powder displayed a sharp peak at 181.9
AC

422 °C corresponding to the melting point of crystalline regions. The PLGA polymer showed a small peak

423 at around 52 °C. This peak is referred to the relaxation peak that follows the glass transition and is

424 consistent with earlier studies [37,38]. No notable melting point was observed for PLGA due to its

425 amorphous nature. The DSC curves of pure PLGA show that the polymer is thermally stable up to

426 220˚C. Moreover, it can also be seen from Figure 2B that the PLGA polymer had the same melting

23
 ACCEPTED MANUSCRIPT

427 peak as the optimized C-PNP9 formulation. This indicates that the encapsulation process did not affect

428 the polymer structure. However, no crystalline drug material in the optimized C-PNP9 formulation was

429 detected as the sharp peak of curcumin was absent.

430

PT
431 3.7. Morphology of optimized PNPs

RI
432 The morphological appearance of the optimized C-PNP9 and blank optimized B-PNP9 without

SC
433 curcumin were visualized with a Transmission Electron Microscope (TEM) (Figure 3A and 3B,

434 respectively). The TEM images show that both samples present nanostructures with a spherical

U
435 morphology with no evidence of particle aggregation. The sizes of the optimum C-PNP9 and B-PNP9
AN
436 are 189.064 ± 20.336 nm and 195.94 ± 25.201 nm, respectively. The sizes obtained from the TEM

437 measurements are in sound accordance with the results obtained from the particle size measurements
M

438 by dynamic light scattering (Table 2).

D

439
TE

440 3.8. In vitro cellular uptake of curcumin

EP

441 HT-29 human colorectal adenocarcinoma cells were chosen as target cells to demonstrate the success

442 of using the optimized C-PNP9 formulation as a formula for improving the cell uptake and cell
C

443 internalization of curcumin. Since curcumin was practically insoluble in water it was necessary to
AC

444 prepare the pure curcumin reference solution by dissolving it in an aqueous solution of DMSO (1.4 %

445 V/V). Figure 3C shows the curcumin cell uptake in HT-29 cells when the same absolute amount of

446 curcumin (200 µg) was incubated with the cells in the form of the pure reference solution and the

447 optimized C-PNP9 formulation. The curcumin uptake by HT-29 cells in case of native curcumin

448 solution and the optimized C-PNP9 formulation after 2 h was 3.7 ± 0.6 and 7.0 ± 0.3 µg per 106 cells,

24
 ACCEPTED MANUSCRIPT

449 respectively. Thus, the optimized C-PNP9 formulation displayed a higher curcumin uptake with a

450 significant difference when compared with the pure curcumin reference solution (P = 0.00059 at a

451 significance level of 0.001).

PT
452

453

RI
454

SC
455

456

U
AN
457 A) B)
M
D
TE
EP

458
C
AC

459 C

25
 ACCEPTED MANUSCRIPT

460 Figure 3. A TEM image of A) blank optimum nanoparticles without curcumin (B-PNP9) and B)

461 optimized curcumin loaded PLGA nanoparticles (C-PNP9). C) In vitro HT-29 cell uptake of curcumin

462 from pure curcumin solution and the optimized C-PNP9 formulation (n = 3).

PT
463

464 3.9. In vitro stability studies of C-PNP formulations

RI
465 The stability of all of the C-PNP formulations in this study was investigated by storing the samples for

SC
466 2 months at room temperature and at 4ºC. Table S4 summarizes the particle sizes and PDI values of the

467 C-PNP formulations measured after one month and two months of storage. The particle size and PDI of

468
U
the different C-PNP formulations have clearly remained within a good range, which indicates a good
AN
469 long term stability of all the C-PNP formulations prepared in this study.
M

470 In view of oral delivery we also examined the colloidal stability of the optimized C-PNP formulation in

471 a GI tract environment by using simulated gastric fluid (SGF) with and without enzyme and simulated
D

472 intestinal fluid (SIF) with bile salt. The optimized C-PNP9 formulation was incubated at 37 °C in
TE

473 different pH media (i.e. 1.2, 6.8 and 7.4) with and without enzyme in order to mimic the simulated GI

474 fluids: SGF, SIF and SCF. The colloidal stability of the optimized C-PNP9 was then monitored by
EP

475 measuring the particle size during incubation in 1) SGF without enzyme for one day, 2) SGF with
C

476 pepsin for 2 h and 3) SIF with bile salt for 6 h. As can be seen from Tables S5 and S6, no significant
AC

477 changes in particle size and PDI were observed for the optimized C-PNP9 formulation incubated at 37

478 °C either with SGF with or without enzyme, or with SIF containing bile salt.

479

480 4. Discussion

26
 ACCEPTED MANUSCRIPT

481 Nanoparticle-based drug delivery systems are of high importance because they can significantly affect

482 the stability and functionality and also the pharmacokinetics and pharmacodynamics of the drug of

483 interest. In a quest of developing an optimum oral formulation for colon treatment, we have prepared a

484 series of curcumin-loaded PLGA nanoparticles (C-PNPs) and used a 23 factorial design approach to

PT
485 achieve a small size, maximum entrapment efficiency and specific maximal release of curcumin in a

RI
486 colonic environment. The C-PNPs were successfully obtained with the single-emulsion solvent-

487 evaporation method. PVA was used as the emulsifier since it has widely been used as a stabilizer for

SC
488 PLGA polymers and other pharmaceutical formulations. PLGA was chosen as a carrier matrix due to

489 its many advantages and attractive properties. Namely, PLGA i) is a biodegradable polymer with good

490
U
biocompatibility, ii) is approved by FDA and EMEA for drug delivery systems, iii) has well described
AN
491 formulation methods which are easy to adapt to various types of drugs e.g. hydrophilic or hydrophobic

492 small molecules or macromolecules, iv) has the ability to protect the encapsulated drug from
M

493 degradation, v) enables the preparation of sustained release formulations, vi) provides the possibility to
D

494 easily modify surface properties to introduce stealth properties and/or better interaction with biological
TE

495 materials and vii) allows to develop targeted nanoparticles to specific organs or cells [26]. Moreover,

496 PLGA degrades into nontoxic lactic acid and glycolic acid in the body. PLGA has also been shown to
EP

497 be a suitable matrix for encapsulating lipophilic drugs [39]. Furthermore, biodegradable polymers, such

498 as PLGA, often keep their rigidity in physiological temperatures, thus preventing detergents and bile
C

499 salts to permeate into the polymer matrix. This in combination with the fact that PLGA can be tuned to
AC

500 degrade at a specific pH makes it a highly suitable matrix for oral drug delivery purposes where it is of

501 utmost importance to protect the drug of interest from the harsh environment in the stomach and the

502 intestine that contain a lot of enzymes and detergents.

27
 ACCEPTED MANUSCRIPT

503 Particle size and size distribution (PDI) are important parameters when developing suitable

504 nanomedicines for therapeutic purposes. They determine the in vivo distribution, toxicity and the

505 targeting ability of nanoparticle systems and consequently the fate of administered nanoparticles by

506 influencing their pharmacokinetics, including the time of circulation, absorption and distribution

PT
507 [40,41]. In addition, particle size and size distribution can also influence the drug loading, drug release

RI
508 and stability of drug inside nanoparticles [ 42 ]. Moreover, the size of the particles influences

509 substantially their ability to act as drug delivery vehicles. For example, 0.1 µm sized nanoparticles have

SC
510 shown to have about a 10-fold higher intracellular uptake in an in vitro cell culture model compared to

511 10 µm-sized microparticles [43]. Furthermore, nanoparticles which are smaller than 0.1 µm have been

512
U
reported to have about a 27-fold higher transfection efficacy than nanoparticles which are larger than
AN
513 0.1 µm [ 44 ]. A recent study has also demonstrated that nanoparticles smaller than 200 nm are

514 distributed more efficiently in tumors than larger ones [45]. Particle size governs the distribution of
M

515 nanoparticles in the body and smaller particles are considered to be more efficient for passive targeting
D

516 to tumor tissues through the enhanced permeability and retention effect [46]. Thus, by reducing the
TE

517 particle size of a drug delivery system and by using nanosized formulations the colonic residence time

518 is believed to increase, but they can also provide additional benefits for cancer therapy, such as a
EP

519 selective accumulation in inflamed tissues. Reducing the size of an oil-in-water emulsion has been

520 shown to be an efficient approach for preparing polymer-based nanoparticle drug delivery systems,
C

521 especially for lipophilic drugs such as curcumin. The particle size and PDI values for all of the C-PNPs
AC

522 in this study clearly show that they are all homogeneous nanoformulations in the size range 180 – 210

523 nm (Table 2). This was also confirmed by the TEM images of the blank and curcumin-loaded

524 optimized C-PNP formulations (Figure 3A and Figure 3B). These results are in good agreement with

525 other studies showing similar morphology of PLGA NPs with sizes between 100 - 163 nm

526 [15,20,37,47,48].
28
 ACCEPTED MANUSCRIPT

527 Zeta potential, on the other hand, is a measure of the surface charge of the particles, which in turn

528 determines the stability of colloidal suspensions, and mucoadhesion properties of particles [49]. The

529 surface charge of the particles controls their stability in nanoparticulate formulations through strong

530 electrostatic repulsions between the particles. Therefore, the zeta potential values, irrespective of

PT
531 charge type, either positive or negative, should preferentially be high (i.e. > 30 mV) in order to ensure

RI
532 stability and avoid aggregation of the particles. Moreover, surface charge plays a key role in the

533 cellular uptake and intra-cellular distribution of nanoparticles [46,50]. PLGA nanoparticles are in

SC
534 general negatively charged due to the dissociation of the free terminal carboxylic group of the acid-

535 terminated PLGA polymer. Table 2 shows that the C-PNPs in this study display a sufficiently high

536
U
negative zeta potential to ensure that the nanoparticles will disperse very well in the aqueous media and
AN
537 that the nanosuspensions will have a very good stability and tolerance against aggregation.
M

538 Determination of the encapsulation parameters, particularly the EE% for nanoparticle formulations is

539 of utmost importance for evaluating the therapeutic efficacy of a drug delivery system. It is generally
D

540 recognized that the encapsulation efficiency of a chemical compound in PLGA polymers correlates
TE

541 with the concentration, molecular weight, and volume ratio of PLGA [51,52]. The EE%s of curcumin

542 for the C-PNPs prepared in this study were between 58-83% (Table 2). A recent study has reported a
EP

543 curcumin EE% of up to 92% for PLGA-based nanoparticles for oral delivery [38]. The higher curcumin
C

544 EE% in that particular study compared to the C-PNPs prepared in the current study can be due to the
AC

545 use of PLGA and PVA of different molecular weights and/or a different preparation technique (i.e.

546 solid-in-oil-in-water solvent evaporation technique). However, other studies that have used the same

547 technique as the current study for preparing PLGA-based nanoparticles for oral delivery of curcumin

548 and another poorly soluble drug glipizide have reported EE%s of 76% and 16-57%, respectively

549 [27,28]. Thus, the EE%s of the C-PNPs in this study can be considered to be sufficiently high for oral

29
 ACCEPTED MANUSCRIPT

550 delivery. The high EE% can be attributed to several factors. Firstly, the hydrophobic nature of PLGA

551 molecules makes it relatively easy to entrap lipophilic curcumin into PLGA. Secondly, there is a

552 favorable interaction taking place between the phenolic hydroxyl of curcumin and the carboxylic

553 groups of PLGA. Thirdly, the lipophilic nature of curcumin and consequently its low solubility in

PT
554 aqueous solutions results in a minimal loss of the drug to the external aqueous phase during the

RI
555 formulation process.

A sustained and triggered release of the drug in a delivery system is an important property closely

SC
556

557 related to its pharmacokinetics and efficacy. The in vitro release characteristics of the C-PNPs in Figure

U
558 1 show that practically no curcumin was released from any of the C-PNP formulations in the SGF
AN
559 medium (~ zero), whereas a slight release of curcumin took place in the SIF (i.e. < 5% of curcumin

560 released). Once the pH reached neutral values (i.e. pH 7.4), curcumin was, however, rapidly released in
M

561 a biphasic manner reaching a sustained release of 34-63% after 24 h. This behavior differs is different

562 from the one presented in the study by Xie et al. [38] where a burst release of curcumin of up to 25%
D

563 and a sustained release of 48% at pH = 2 were reported for PLGA-based NPs for oral administration.
TE

564 Furthermore, Xie et al. [38] and Cartiera et al. [27] reported that curcumin-loaded PLGA-based NPs for

565 oral delivery released curcumin in a biphasic manner reaching a sustained release of ~55% and ~20%
EP

566 after 24 h. Thus, it can be concluded that the sustained release of curcumin from the C-PNPs in this
C

567 study is high and in good agreement with other similar studies. The fact that negligible amounts of
AC

568 curcumin were released in the pH range 2-6.8 clearly shows that a triggered release of curcumin from

569 the C-PNPs in an environment mimicking the colon is taking place, and that the amount of curcumin

570 released is substantially higher in a colonic environment than in an environment mimicking the

571 stomach or small intestine. This means that the main absorption of curcumin from the C-PNP

572 formulations should take place in the large intestine, and not in the stomach or small intestine, unless

30
 ACCEPTED MANUSCRIPT

573 the C-PNPs are taken up earlier in the villi of the small intestine. However, the uptake of the C-PNPs

574 before they reach the colon could possibly be minimized by a proper surface functionalization of the

575 NPs. Thus, the optimized C-PNP9 formulation developed here should enable a further development of

576 an oral formulation for triggered, sustained, targeted and effective colon treatment with curcumin.

PT
577 The kinetic constants and mode of release obtained from the release kinetic fits to the in vitro release

RI
578 profiles of the C-PNPs (Figure 1) are shown in Table S1. The controlled release pattern of curcumin

from the C-PNPs followed mainly the Higuchi diffusion model with a half-life of 7-29 h. This indicates

SC
579

580 that a sustained release of curcumin from the C-PNPs mainly takes place by diffusion from the PLGA

U
581 matrix and not through extensive degradation/erosion processes of the PLGA nanoparticles. This is in
AN
582 sound accordance with the fact that degradation of PLGA can last from weeks to months [52], and if

583 the diffusion of the drug is faster than the matrix degradation, then the mechanism of drug release
M

584 should occur mainly by diffusion [23,53]. The release mechanism of the C-PNPs is in good agreement

585 with recent studies that have reported that drug-loaded PLGA NPs exhibit a controlled release pattern
D

586 following the Higuchi diffusion model [17,20,54].

TE

587 The factorial design analysis approach is a helpful tool during formulation optimization as it allows
EP

588 elucidating the relationship between the dependent and independent variables during the optimization

589 process. The main effects, interaction profiles, response surface and contour plots obtained from the 23
C

590 factorial design analysis of the C-PNPs in this study are shown in Figures S1-S3. Table S3 shows the
AC

591 corresponding regression equations for the responses obtained by analyzing the experimentally

592 determined input parameters. The coefficients in the regression equations with more than one factor

593 term represent an interaction term. A positive value for a response in the regression equation represents

594 an effect that favors the optimization (synergistic effect), while a negative value indicates an inverse

595 relationship (antagonistic effect) between the factors and the response [55]. Thus, the regressions

31
 ACCEPTED MANUSCRIPT

596 equations in Table S3, and Figures S1-S3 clearly show that X1 and X3 have the main effects on the

597 dependent parameters while X2 has a similar effect as X3 but to a lesser extent. X1 and X3 have a

598 positive effect on Y1 and Y3, while X2 has a negative effect on Y1, Y3 and Y4. On the other hand, X1

599 has a positive effect on Y2, while X2 and X3 have a negligible effect on it. The regression equation for

PT
600 particle size in Table S3 also clearly shows a negative sign for X2 (PVA), which indicates that the size

RI
601 of the nanoparticles decreases when the concentration of the stabilizer is increased, whereas a positive

602 sign for X1 (PLGA) and X3 (curcumin) indicates a positive effect on the particle size, i.e. an increasing

SC
603 particle size with increasing concentration. A more detailed discussion concerning the effects of the

604 independent variables is summarized in the supporting information.

U
AN
605 The observed and predicted physicochemical characteristics, i.e. values of the dependent responses

606 under investigation for the optimized C-PNP9 formulation, are shown in Table 3, along with the
M

607 residual values. It can clearly be seen that the experimentally measured and the predicted values are

608 well in line. The fact that the observed values for particle size, zeta potential, EE % and DR% after 24
D

609 h are close to the predicted values validates the successful use of the 23 factorial design and thus also
TE

610 the equations derived for the dependent variables.

EP

611 FT-IR spectroscopy is a technique for identifying the chemical composition of a material, but can also

612 be used for revealing interactions between materials in formulation mixtures. The optimized C-PNP9
C

613 formulation showed characteristic peaks which were close to the principal IR peaks of the drug (Figure
AC

614 2A), confirming the presence of curcumin in the nanospheres, but also indicating that there are no

615 strong interactions between the drug and the polymers used for the nanoencapsulation process. Similar

616 results were also obtained by Sampath et al. [56] during an FT-IR analysis of curcumin-loaded PLGA

617 NPs. The DSC measurements (Figure 2B) did not detect any crystalline drug material in the optimized

618 C-PNP9 sample as the sharp peak of curcumin was absent. Thus, the thermograms suggest that

32
 ACCEPTED MANUSCRIPT

619 curcumin is in an amorphous or disordered-crystalline phase of molecular dispersion form or in a solid

620 solution state inside the PLGA polymer matrix [48]. The DSC results further support the FT-IR results

621 that there are no significant interactions between the drug and any other components in the optimized

622 C-PNP9 formulation. The indication that curcumin was molecularly dispersed in the PLGA polymer

PT
623 matrix is also supported by the fact that the release mechanism of curcumin from the optimized C-PNP9

RI
624 formulation followed the Higuchi diffusion model. Polakovic et al. [57] have shown this with lidocaine

625 loaded PNPs. They concluded that when lidocaine is molecularly dispersed in a polymer matrix its

SC
626 release kinetics fit better to a diffusion model, whereas if lidocaine is in crystal form within the

627 polymer matrix its release kinetics could be described by a dissolution model.

U
AN
628 The in vitro cell uptake studies in Figure 3C clearly show an enhanced uptake of curcumin in HT-29

629 human colorectal adenocarcinoma cells for the optimized C-PNP9 formulation compared to the native
M

630 curcumin solution. The enhanced uptake can be attributed to the small particle size of the optimized C-

631 PNP9 formulation and the effect of PVA on the fluidity of the cell membrane. In addition, PLGA-
D

632 nanoparticles are internalized by cells partly via fluid-phase pinocytosis as well as via clathrin-
TE

633 mediated endocytosis. This allows PLGA NPs to rapidly escape the endolysosomes and enter the

634 cytoplasm even within 10 min of incubation. This facilitates the interactions of PLGA NPs with the
EP

635 vesicular membranes which leads to transient and localized destabilization of the membrane. This
C

636 consequently results in the escape of the PLGA NPs into the cytosol [26,58].
AC

637 The overall stability of the C-PNPs prepared in this study can be attributed to their high zeta potential

638 values as well as the protective effect of PLGA. In general, bile acid affects the GI stability of

639 polymeric NPs. The optimized C-PNP9 formulation prepared in this study was, however, expected to

640 be more stable against bile acid because it consists of PLGA with a transition temperature (Tc) of 52

641 °C, which is much higher than the temperature of the gastrointestinal tract. Polymer matrices still keep

33
 ACCEPTED MANUSCRIPT

642 their rigidity below the Tc, which means that detergents and bile salts have a harder job to permeate

643 into the polymer matrix. An increase of PDI of the optimized C-PNP9 formulation incubated in SIF

644 with bile acid for one hour may still be attributed to the effect of bile salt. In conclusion, the stability

645 studies revealed a good colloidal stability of the optimized curcumin-loaded PLGA formulation in

PT
646 synthetic gastrointestinal fluids, even though a slight increase in size distribution was observed in

RI
647 presence of bile salt.

SC
648

649 5. Conclusions

650
U
Curcumin-loaded PLGA nanoparticles for colon delivery were successfully prepared with the emulsion
AN
651 solvent evaporation technique. The use of the 23 factorial design model enabled development of an
M

652 optimized curcumin-loaded PLGA-based nanoformulation using minimum amount of raw materials

653 and minimum time. On the basis of the optimization criteria it was found that the composition of the
D

654 optimized formulation should contain 176.758 mg PLGA, 2 % PVA and 16.636 mg curcumin. The
TE

655 observed responses of particle size (~ 220 nm), zeta potential (> -30 mV), EE% (> 70%) and drug

656 release after 48 h (> 70%) correlated well with the predicted values obtained from the 23 factorial
EP

657 design model. DSC studies confirmed the presence of curcumin in an amorphous or disordered-

658 crystalline phase of molecular dispersion form or in a solid solution state in the polymer matrix. FT-IR
C

659 studies confirmed that there are no substantial interactions between curcumin and PLGA. TEM images
AC

660 of the optimized blank (B-PNP9) and curcumin-loaded PLGA NPs (C-PNP9) revealed that the NPs

661 were discrete, non-aggregated, spherical in shape, and displayed a good size distribution. The results of

662 small size, sustained release, good colloidal stability in synthetic gastrointestinal fluids, long term

663 stability and a clearly higher cellular uptake in HT-29 cells compared to pure curcumin solution show

34
 ACCEPTED MANUSCRIPT

664 that the optimized curcumin-loaded PLGA nanoparticle formulation (C-PNP9) has high potential as an

665 initial platform for a further development of an efficient oral-targeted drug delivery system to the

666 colon, especially if it is further functionalized with a mucoadhesive polymer (e.g. chitosan) or a

667 specific targeting ligand (e.g. wheat germ agglutinin or specially designed peptides for colon targeting).

PT
668

RI
669 Acknowledgements

670 This research was funded by Center for International Mobility CIMO, Egypt (Grant No. KM-14-9066)

SC
671 and Academy of Finland (Grant Nos. 137 053 and 263861). We also want to address our special thanks

U
672 to Tiina Lipiäinen and Sara Fraser for their kind help with the experimental work of the DSC and FTIR
AN
673 studies, respectively. The Electron Microscopy Unit of University of Helsinki is acknowledged for help

674 with acquiring the TEM images.

M

675
D

676
TE

677
EP

678

679
C
AC

680

681

682

683 References

35
 ACCEPTED MANUSCRIPT

1 American Cancer Society,

2015.http://www.cancer.org/cancer/colonandrectumcancer/detailedguide/colorectal-cancer-key-

statistics. Last accessed 21.10.2015.

PT
2 M. Das, C. Mohanty, S.K. Sahoo, Ligand-based targeted therapy for cancer tissue, Expert Opin.

Drug. Deliv. 6 (2009) 285-304.

RI
3 S.K. Sahoo, S. Parveen, J. Panda, The present and future of nanotechnology in human health care,

SC
Nanomedicine 3 (2007) 20-31.

4 B.B. Aggarwal, K.B. Harikumar, Potential therapeutic effects of curcumin, the anti-inflammatory

U
agent, against neurodegenerative, cardiovascular, pulmonary, metabolic, autoimmune and
AN
neoplastic diseases, Int. J. Biochem. Cell Biol. 41 (2009) 40-59.

5 R.A. Fryer, C. Galustian, A.G. Dalgelish, Recent advances and developments in treatment
M

strategies against pancreatic cancer, Curr. Clin. Pharmacol. 4 (2009) 102-12.

6 X. Gao, J. Kuo, H. Jiang, D. Deeb, Y. Liu, G. Divine, R.A. Chapman, S.A. Dulchavsky, S.C.
D

Gautam, Immunomodulatory activity of curcumin: suppression of lymphocyte proliferation,

TE

development of cell-mediated cytotoxicity, and cytokine production in vitro, Biochem. Pharmacol.

68 (2004) 51-61.
EP

7 A.P. Lakshmanan, K. Watanabe, R.A. Thandavarayan, F.R. Sari, H. Meilei, V. Soetikno, S.

C

Arumugam, V.V. Giridharan, K. Suzuki, M. Kodama, Curcumin attenuates hyperglycaemia-

AC

mediated AMPK activation and oxidative stress in cerebrum of streptozotocin-induced diabetic rat,

Free Radic. Res. 45 (2011) 788-795.

8 R. De, P. Kundu, S. Swarnakar, T. Ramamurthy, A. Chowdhury, G.B. Nair, A.K. Mukhopadhyay,

Antimicrobial activity of curcumin against Helicobacter pylori isolates from India and during

infections in mice, Antimicrob. Agents Chemother. 53 (2009) 1592-1597.

36
 ACCEPTED MANUSCRIPT

9 Z.M. Shao, Z.Z. Shen, C.H. Liu, M.R. Sartippour, V.L. Go, D. Heber, M. Nguyen, Curcumin

exerts multiple suppressive effects on human breast carcinoma cells, Int. J. Cancer. 98 (2002) 234-

240.

PT
10 S. Bisht, G. Feldmann, S. Soni, R. Ravi, C. Karikar, A. Maitra, A. Maitra, Polymeric nanoparticle-

encapsulated curcumin (‘nanocurcumin’): a novel strategy for human cancer therapy, J.

RI
Nanobiotech. 5 (2007) 3.

SC
11 R.K. Maheshwari, A.K. Singh, J. Gaddipati, R.C. Srimal, Multiple biological activities of

curcumin: a short review, Life Sci. 78 (2006) 2081-2087.

U
12 S. Lev-Ari, L. Strier, D. Kazanov, L. Madar-Shapiro, H. Dvory-Sobol, I. Pinchuk, B. Marian, D.
AN
Lichtenberg, N. Arber, Celecoxib and curcumin synergistically inhibit the growth of colorectal

cancer cells, Clin. Cancer Res. 11 (2005) 6738-6744.

M

13 W.W. Quitschke, Ch 5, Bioavailability and metabolism of curcuminoids in Natural Compounds as

Inducers of Cell Death, vol 1, Eds M. Diederich and K. Noworyta, Springer, 2012.
D

14 H. Kharkwal, K. Bala, D.D. Joshi, D. Pande Katare, Bioavailability enhancement of curcuminoids

TE

using natural polymer, Der Pharm. Lettre 4 (2012) 1698-1711.

15 P. Anand, H.B. Nair, B. Sung, A.B. Kunnumakkara, V.R. Yadav, R.R. Tekmal, B.B. Aggarwal,
EP

Design of curcumin-loaded PLGA nanoparticles formulation with enhanced cellular uptake, and
C

increased bioactivity in vitro and superior bioavailability in vivo, Biochem. Pharmacol. 79 (2010)
AC

330-338.

16 A. Lazar, S. Mourtas, I. Youssef, C. Parizot, A. Dauphin, B. Delatour, S.G. Antimisiaris, C.

Duyckaerts, Curcumin-conjugated nanoliposomes with high affinity for Ab deposits: possible

applications to Alzheimer disease, Nanomed. Nanotechnol. 9 (2013) 712-721.

37
 ACCEPTED MANUSCRIPT

17 M.M. Yallapu, M. Jaggi, S.C. Chauhan, Curcumin nanoformulations: a future Nanomedicine for

cancer, Drug Discov. Today 17 (2012) 71-80.

18 M. Gou, K. Men, H. Shi, M. Xiang, J. Zhang, J. Song, J. Long, Y. Wan, F. Luo, X. Zhao, Z. Qian,

PT
Curcumin-loaded biodegradable polymeric micelles for colon cancer therapy in vitro and in vivo,

Nanoscale 3 (2011) 1558-1567.

RI
19 A.O. Elzoghby, W.M. Samy, N.A. Elgindy, Albumin-based nanoparticles as potential controlled

SC
release drug delivery systems, J. Control. Rel. 157 (2012) 168-182.

20 Y.M. Tsai, C.F. Chien, L.C. Lina, T.H. Tsai, Curcumin and its nano-formulation: the kinetics of

U
tissue distribution and blood-brain barrier penetration, Int. J. Pharm. 416 (2011) 331-338.
AN
21 K. Ganguly, T.M. Aminabhavi, A.R. Kulkarni, Colon targeting of 5-Fluorouracil using

polyethylene glycol cross-linked chitosan microspheres enteric coated with cellulose acetate
M

phthalate, Ind. Eng. Chem. Res. 50 (2011) 11797–11807.

22 D. Peer, J.M. Karp, S. Hong, O.C. Farokhzad, R. Margalit, R. Langer, Nanocarriers as an emerging
D

platform for cancer therapy, Nat. Nanotech. 2 (2007) 751-760.

TE

23 K.S. Soppimath, T.M. Aminabhavi, A.R. Kulkarni, W.E. Rudzinski, Biodegradable polymeric

nanoparticles as drug delivery devices, J. Control. Rel. 70 (2001) 1–20.

EP

24 N.M. Khalil, T.C. do Nascimento, D.M. Casa, L.F. Dalmolin, A.C. de Mattos, I. Hoss, MA.
C

Romano, R.M. Mainardes, Pharmacokinetics of curcumin-loaded PLGA and PLGA–PEG blend

AC

nanoparticles after oral administration in rats, Coll. Surf. B Biointerf. 101 (2013) 353-360.

25 R.C. Mundargi, V.R. Babu, V. Rangaswamy, P. Patel, T.M. Aminabhavi, Nano/micro technologies

for delivering macromolecular therapeutics using poly(D, L-lactide-co-glycolide) and its

derivatives, J. Control. Rel. 125 (2008) 193-209.

38
 ACCEPTED MANUSCRIPT

26 F. Danhier, E. Ansorena, J.M. Silva, R. Coco, A. Le Breton, V. Préat, PLGA-based nanoparticles:

An overview of biomedical applications, J. Control. Rel. 161 (2012) 505-522.

27 M.S. Cartiera, E.C. Ferreira, C. Caputo, M.E. Egan, M.J. Caplan, W.M. Saltzman, Partial

PT
correction of cystic fibrosis defects with PLGA nanoparticles encapsulating curcumin, Mol.

Pharm. 7 (2009) 86-93.

RI
28 S. Jain, S. Saraf, Influence of processing variables and in vitro characterization of glipizide loaded

SC
biodegradable nanoparticles, Diab. Metabol. Syndr; Clin. Res. & Rev. 3 (2009) 113-117.

29 T. Ma, L. Wang, T. Yang, G. Ma, S. Wang, Homogeneous PLGA-lipid nanoparticles as a

U
promising oral vaccine delivery system for ovalbumin, 9 (2014) 129-136.
AN
30 A. Zanotto-Filho, K. Coradini, E. Braganhol, R. Schröder, C.M. de Oliveiram, A. Simões-Pires,

A.M. Battastini, A.R. Pohlmann, S.S. Guterres, C.M. Forcelini, R.C. Beck, J.C. Moreira,
M

Curcumin-loaded lipid-core nanocapsules as a strategy to improve pharmacological efficacy of

curcumin in glioma treatment, Eur. J. Pharm. Biopharm. 83 (2013) 156-167.

D

31 USP 36 NF 31. United State Pharmacopeia and National Formulary, 2013.

TE

32 N.T. Paragkumar, D. Edith, J.-L. Six, Surface characteristics of PLA and PLGA films, Appl. Surf.

Sci. 253 (2006) 2758-2764.

EP

33 R. Benassi, E. Ferrari, S. Lazzari, F. Spagnolo, M. Saladini, Theoretical study on Curcumin: A

C

comparison of calculated spectroscopic properties with NMR, UV–vis and IR experimental data, J.
AC

Mol. Struct. 892 (2008) 168–176.

34 T.M. Kolev, E.A. Velcheva, B.A. Stamboliyska, M. Spiteller, DFT and experimental studies of the

structure and vibrational spectra of curcumin, Int. J. Quantum Chem.102 (2005) 1069-1079.

35 A. Anitha, M. Sreeranganathan, K.P. Chennazhi, V.K. Lakshmanan, R. Jayakumar, In vitro

combinatorial anticancer effects of 5-fluorouracil and curcumin loaded N, O-carboxymethyl

39
 ACCEPTED MANUSCRIPT

chitosan nanoparticles toward colon cancer and in vivo pharmacokinetic studies, Eur. J. Pharm.

Biopharm. 88 (2014) 238-251.

36 K. Renuka, M. Ramakant, V. Ashwni, D. Pankaj, K. Vivek, G. Varsha, K.P. Sarvesh, R.M.

PT
Prabhat, K.D. Anil, Colon-specific delivery of curcumin by exploiting Eudragit-decorated chitosan

nanoparticles in vitro and in vivo, J. Nanopart. Res. 15 (2013) 1893.

RI
37 A. Mukerjee, J.K. Vishwanatha, Formulation, characterization and evaluation of curcumin-loaded

SC
PLGA nanospheres for cancer therapy, Anticancer Res. 29 (2009) 3867-3875.

38 X. Xie, Q. Tao, Y. Zou, F. Zhang, M. Guo, Y. Wang, H. Wang, Q. Zhou, S. Yu, PLGA

U
nanoparticles improve the oral bioavailability of curcumin in rats: characterizations and
AN
mechanisms, J. Agric. Food Chem. 59 (2011) 9280-9289.

39 C. Wang, P.C. Ho, L.Y. Lim, Wheat germ agglutinin-conjugated PLGA nanoparticles for
M

enhanced intracellular delivery of paclitaxel to colon cancer cells, Int. J. Pharm. 400 (2010) 201-

210.
D

40 D.P. Lankveld, A.G. Oomen, P. Krystek, A. Neigh, A. Troost-de Jong, C.W. Noorlander, J.C. Van
TE

Eijkeren, R.E. Geertsma, W.H. De Jong, The kinetics of the tissue distribution of silver

nanoparticles of different sizes, Biomaterials 31 (2010) 8350-8361.

EP

41 S.A. Kulkarni, S.-S. Feng, Effects of particle size and surface modification on cellular uptake and
C

biodistribution of polymeric nanoparticles for drug delivery, Pharm. Res. 30 (2013) 2512-2522.
AC

42 V.J. Mohanraj, Y. Chen, Nanoparticles: a review, Trop. J. Pharm. Res. 5 (2006) 561-573.

43 M.P. Desai, V. Labhasetwar, E. Walter, R.J. Levy, G.L. Amidon, The mechanism of uptake of

biodegradable microparticles in Caco-2 cells is size dependent, Pharm. Res. 14 (1997) 1568-1573.

44 S. Prabha, W.-Z. Zhou, J. Panyam, V. Labhasetwar, Size-dependency of nanoparticle-mediated

gene transfection: studies with fractionated nanoparticles, Int. J. Pharm. 244 (2002) 105-115.

40
 ACCEPTED MANUSCRIPT

45 C. He, Y. Hu, L. Yin, C. Tang, C. Yin, Effects of particle size and surface charge on cellular

uptake and biodistribution of polymeric nanoparticles, Biomaterials 31 (2010) 3657–3666.

46 F. Akhtar, M.A. Rizvi, S.K. Kar, Oral delivery of curcumin bound to chitosan nanoparticles cured

PT
Plasmodium yoelii infected mice, Biotech. Adv. 30 (2012) 310-320.

47 S. Abbas, E. Karangwa, M. Bashari, K. Hayat, X. Hong, H.R. Sharif, X. Zhang, Fabrication of

RI
polymeric nanocapsules from curcumin-loaded nanoemulsion templates by self-assembly,

SC
Ultrason. Sonochem. 23 (2014) 81-92.

48 K.L. Nair, A.K. Thulasidasan, G. Deepa, R.J. Anto, G.S. Kumar, Purely aqueous PLGA

U
nanoparticulate formulations of curcumin exhibit enhanced anticancer activity with dependence on
AN
the combination of the carrier, Int. J. Pharm. 425 (2012) 44-52.

49 J. Vandervoort, A. Ludwig, Biocompatible stabilizers in the preparation of PLGA nanoparticles: a

M

factorial design study, Int. J. Pharm. 238 (2002) 77-92.

50 J. Panyam, W.Z. Zhou, S. Prabha, S.K. Sahoo, V. Labhasetwar, Rapid endo-lysosomal escape of
D

poly (D, L-lactide-co-glycolide) nanoparticles: implications for drug and gene delivery. FASEB J.
TE

16 (2002) 1217-1226.

51 C.E. Astete, C.M. Sabliov, Synthesis and characterization of PLGA nanoparticles. J. Biomater. Sci.
EP

Polym. Ed. 17 (2006) 247–289.

C

52 C. Wischke, S.P. Schwendeman, Principles of encapsulating hydrophobic drugs in PLA/PLGA

AC

microparticles, Int. J. Pharm. 364 (2008) 298-327.

53 T. Niwa, H. Takeuchi, T. Hino, N. Kunou, Y. Kawashima, Preparations of biodegradable

nanospheres of water-soluble and insoluble drugs with D, L-lactide/glycolide copolymer by a

novel spontaneous emulsification solvent diffusion method, and the drug release behaviour, J.

Control. Rel. 25 (1993) 89-98.

41
 ACCEPTED MANUSCRIPT

54 K.K. Chereddy, R. Coco, P.B. Memvanga, B. Ucakar, A. des Rieux, G. Vandermeulen, V. Préat,

Combined effect of PLGA and curcumin on wound healing activity, J. Control. Rel. 171 (2013)

208-215.

PT
55 Y. Duan, S. Xu, Q. Wang, J. Liu, Z. Zhang, Optimization of preparation of DHAQ-loaded PEG-

PLGA-PEG nonaparticles using central composite design, J. Mater. Sci. Mater. Med. 17 (2006)

RI
559-563.

SC
56 M. Sampath, R. Lakra, P. Korrapati, B. Sengottuvelan, Curcumin loaded poly (lactic-co-glycolic)

acid nanofiber for the treatment of carcinoma, Coll. Surf. B: Biointerf. 117 (2014) 128-134.

U
57 M. Polakovc, T. Görner, R. Gref, E. Dellacherie, Lidocaine loaded biodegradable nanospheres II.
AN
Modelling of drug release, J. Control. Rel. 60 (1999) 169–177.

58 J.K. Vasir, V. Labhasetwar, Biodegradable nanoparticles for cytosolic delivery of therapeutics,

M

Adv. Drug. Deliv. Rev. 59 (2007) 718-28.

D
TE
C EP
AC

42