Mechanistic profiling of the siRNA delivery dynamics

COREL-07492; No of Pages 10 Journal of Controlled Release xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Controlled Release journal homepage: www.elsevier.com/locate/jconrel

F

O

4

Stefano Colombo a,1, Dongmei Cun b,1,⁎, Katrien Remaut c, Matt Bunker d, Jianxin Zhang d, Birte Martin-Bertelsen a, Anan Yaghmur a, Kevin Braeckmans c,e, Hanne M. Nielsen a,⁎⁎, Camilla Foged a

5 6 7 8 9

a

1 0

a r t i c l e

11 12 13 14 15

Article history: Received 31 October 2014 Received in revised form 19 December 2014 Accepted 20 December 2014 Available online xxxx

16 17 18 19 20 21 22

Keywords: siRNA Drug delivery Nanomedicine PLGA DOTAP Sustained release

Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen O, Denmark Department of Pharmaceutical Sciences, Shenyang Pharmaceutical University, Shenyang, Wenhua Road 103, 110016, China Biophotonic Imaging Group, Lab of General Biochemistry and Physical Pharmacy, Ghent University, Harelbekestraat 72, 9000 Ghent, Belgium d Molecular Profiles Ltd, 8 Orchard Place, Nottingham Business Park, Nottingham NG8 6PX, UK e Centre for Nano- and Biophotonics, Ghent University, Harelbekestraat 72, 9000 Ghent, Belgium

R O

b c

i n f o

a b s t r a c t

Understanding the delivery dynamics of nucleic acid nanocarriers is fundamental to improve their design for therapeutic applications. We investigated the carrier structure–function relationship of lipid–polymer hybrid nanoparticles (LPNs) consisting of poly(DL-lactic-co-glycolic acid) (PLGA) nanocarriers modified with the cationic lipid dioleoyltrimethyl-ammoniumpropane (DOTAP). A library of siRNA-loaded LPNs was prepared by systematically varying the nitrogen-to-phosphate (N/P) ratio. Atomic force microscopy (AFM) and cryo-transmission electron microscopy (cryo-TEM) combined with small angle X-ray scattering (SAXS) and confocal laser scanning microscopy (CLSM) studies suggested that the siRNA-loaded LPNs are characterized by a core–shell structure consisting of a PLGA matrix core coated with lamellar DOTAP structures with siRNA localized both in the core and in the shell. Release studies in buffer and serum-containing medium combined with in vitro gene silencing and quantification of intracellular siRNA suggested that this self-assembling core–shell structure influences the siRNA release kinetics and the delivery dynamics. A main delivery mechanism appears to be mediated via the release of transfection-competent siRNA–DOTAP lipoplexes from the LPNs. Based on these results, we suggest a model for the nanostructural characteristics of the LPNs, in which the siRNA is organized in lamellar superficial assemblies and/or as complexes entrapped in the polymeric matrix. © 2014 Published by Elsevier B.V.

48 49 50 51 52 53 54

T

C

E

The opportunity for mediating a highly specific cellular gene silencing via the RNA interference (RNAi) machinery by introducing small interfering RNA (siRNA) molecules in the cytoplasm may open the way for the development of innovative therapeutic applications [1]. Although siRNA has the potential to be an effective therapeutic drug, efficient cytoplasmic delivery still remains a major hurdle [2]. A growing number of delivery vehicles have been reported to mediate siRNA delivery into the cytoplasm of the target cells [3]. Nanoparticles composed of biodegradable polymers, such as poly(DL-lactic-co-glycolic acid) (PLGA), are attractive siRNA carriers because of their low toxicity, good colloidal stability, and the possibility of obtaining sustained release of their payload [4]. However, the siRNA delivery efficiency of PLGA

N C O

46 47

1. Introduction

U

44 45

R

41 39 38

43

23 24 25 26 27 28 29 30 31 32 33 34 35 36

R

37 40 42

P

3Q2

D

2

Mechanistic profiling of the siRNA delivery dynamics of lipid–polymer hybrid nanoparticles

E

1Q1

⁎ Correspondence to: D. Cun, Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen O, Denmark. ⁎⁎ Corresponding author. E-mail addresses: [email protected] (D. Cun), [email protected] (H.M. Nielsen). 1 These authors contributed equally.

nanoparticles is generally poor, as compared to that observed with lipid-based carriers [5]. Therefore, incorporation of certain cationic excipients, such as poly(ethyleneimine) [6] or lipids [7,8] into PLGA nanoparticles is a widely used strategy to improve their transfection capability [3]. Cationic lipids, e.g. dioleoyltrimethylammoniumpropane (DOTAP), have been successfully combined with PLGA, formulating siRNA in lipid–polymer hybrid nanoparticles (LPNs) by using various preparation procedures [9–12]. Among these, LPNs prepared at a DOTAP:PLGA weight ratio of 15:85 by using a double emulsion solvent evaporation (DESE) method resulted in nano-sized carriers demonstrating i) high siRNA loading, ii) sustained release, iii) enhanced transfection efficiency in vitro, and iv) encouraging therapeutic effect in vivo [8,13,14]. Nevertheless, little is known about the siRNA delivery mechanism(s) and dynamics mediated by this carrier, even though such knowledge is of importance for optimizing further the performance and the design for therapeutic applications. In the presented work, the physicochemical properties of a library of LPNs were systematically investigated. On the basis of the experimental studies, we propose a model for the nanostructural characteristics of the particles, which influences the particle's siRNA release kinetics, and is thus decisive for the siRNA delivery mechanism.

http://dx.doi.org/10.1016/j.jconrel.2014.12.026 0168-3659/© 2014 Published by Elsevier B.V.

Please cite this article as: S. Colombo, et al., Mechanistic profiling of the siRNA delivery dynamics of lipid–polymer hybrid nanoparticles, J. Control. Release (2014), http://dx.doi.org/10.1016/j.jconrel.2014.12.026

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75

2

S. Colombo et al. / Journal of Controlled Release xxx (2014) xxx–xxx

2.2. Preparation and characterization of LPNs

105

127

The siRNA-loaded LPNs were prepared by applying the DESE method as previously reported [14] using N/P ratios between 1 and 62. For the labeled particles, 10 mol% of each component was replaced with fluorophore-labeled siRNA or PLGA, respectively. A volume of 125 μl of 8–495 μΜ siRNA solution was emulsified in 250 μl of a 2.25 mg DOTAP/12.75 mg PLGA binary mixture in CH2Cl2 by sonication to form the primary emulsion, which was dispersed in 2% (w/v) PVA aqueous solution, resulting in the formation of a water-in-oil-in-water (w1/o/ w2) double emulsion. Alternatively, the LNPs were prepared using the modified oil-in-water (o/w) single emulsion solvent evaporation (SESE) method [10,16]. Briefly, 125 μl siRNA solution (250 nM) was added to a 2.25 mg DOTAP/12.75 mg PLGA binary mixture in CH2Cl2, followed by 1 min vortexing, and 350 μl MeOH was added, resulting in the formation of a single phase by vigorously mixing for 1 min. After incubation at room temperature (rt) for 3 h, 300 μl water and 300 μl CHCl3 were added. Upon complete phase separation, the organic phase was transferred to 1.6 ml of a 2% (v/v) PVA solution, the phases were immediately vortexed for 1 min and sonicated for 90 s. With the evaporation of CH2Cl2, the emulsion droplets were gradually solidified as LPNs. The nanoparticles were washed three times, and freeze dried as previously described [14]. The mean particle diameter (Z-average), polydispersity index (PDI) and zeta-potential were measured as previously described [14].

128

2.3. Dissociation of lipoplexes and determination of encapsulation efficiency

129 130

The siRNA was extracted by dissolving the LPNs in 200 μl CHCl3 followed by extraction with 500 μl of TE buffer supplemented with 100 mg/ml heparin and 100 mM OG (referred to as HD solution below) to dissociate siRNA/DOTAP complexes. The mixture was vortexed for 5 min at rt and the phases were separated by centrifugation (18,000 ×g, 4 °C, 20 min). The aqueous phase was collected and incubated at 37 °C for 5 min to evaporate the residual CHCl3. The siRNA

99 100 101 102

106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126

131 132 133 134 135

C

97 98

E

95 96

R

93 94

R

91 92

O

89 90

C

87 88

N

85 86

U

83 84

138

The molecular interaction between siRNA and DOTAP during the nanoparticle formation process was visualized by imaging the localization of the molecules in the primary emulsion using Cy3-EGFP siRNA (8 μM, final concentration). For DOTAP, various amounts of unlabeled DOTAP were mixed with 2 μg of NBD-DOTAP resulting in N/P ratios between 0 and 62. These emulsions were visualized using a Zeiss LSM510 CLSM (Carl Zeiss, Jena, DE) equipped with an argon laser (458/488 nm) and a HeNe laser (543 nm).

139

2.5. Small-angle X-ray scattering (SAXS)

147

SAXS measurements were performed at the beamline I911-SAXS4 (MAX II storage ring, MAX-lab synchrotron facility, Lund University, SE) using a 3.5 T multipole-wiggler producing a high-flux photon beam with a wavelength of 0.91 Å. The scattering was recorded using a PILATUS 1 M pixel area detector (DECTRIS, Baden, CH). With the applied instrumental setup, the covered q-range of interest was 0.01 to 0.83 Å−1. Silver behenate (CH3–(CH2)20–COOAg) with a d-spacing of 58.38 Å was used to calibrate the angular scale of the measured intensity [17]. The samples were measured in quartz glass capillaries (diameter 1.5 mm) and thermostated in a custom-built sample holder block of brass, which was connected to a circulating water bath (Julabo, Seelbach, DE). The temperature was pre-set to 25 °C, and the sample exposure time was 300 s. The 2D scattering data were azimuthally averaged, normalized to the incident radiation intensity and the sample exposure time and corrected for background and detector inhomogeneities using the software BioXTAS RAW [18]. The radially averaged intensity I is given as a function of the scattering vector q (q = 4π sinθ / λ, where λ is the wavelength and 2θ is the scattering angle).

148

2.6. Cryo-transmission electron microscopy (cryo-TEM)

166

Morphological analysis was carried out by cryo-TEM using a Tecnai G2 20 TWIN transmission electron microscope (FEI, Hillsboro, OR, USA). Samples for cryo-TEM were prepared using a FEI Vitrobot Mark IV, under controlled temperature and humidity conditions within an environmental vitrification system. A small droplet (5 μl) was deposited onto a Pelco Lacey carbon-filmed grid and spread carefully. Excess liquid was removed resulting in the formation of a thin (10–500 nm) sample film. The samples were immediately plunged into liquid ethane and kept at −180 °C. The vitrified samples were subsequently transferred in liquid nitrogen to an Oxford CT3500 cryo holder connected to the electron microscope. The sample temperature was continuously kept below −180 °C. All observations were made in bright field mode at an acceleration voltage of 120 kV. Digital images were recorded with a Gatan Imaging Filter 100 CCD camera (Gatan, Pleasanton, CA, USA).

167 168

2.7. In vitro siRNA release

181

Freeze-dried LPNs were dispersed in TE buffer to 10 mg/ml and incubated at 37 °C with shaking (50 rpm). At each time interval, the release medium was collected after centrifugation for 12 min at 18,000 ×g, and replaced with fresh TE buffer. The siRNA was quantified by using the RiboGreen® RNA reagent, with or without the addition of HD solution during the siRNA isolation. The surface morphology of the nanoparticles during the release study was examined by atomic force microscopy (AFM) as previously described [19].

182 183

F

104

81 82

2.4. Confocal laser scanning microscopy (CLSM)

O

103

2′-O-methyl modified dicer substrate asymmetric siRNA duplexes directed against enhanced green fluorescent protein (EGFP-siRNA) [15] and Alexa488-labeled EGFP-siRNA were provided by Integrated DNA Technologies (IDT, Coralville, IA, USA) as dried, purified and desalted duplexes, and re-annealed as recommended by the supplier. Cy3-labeled siRNA (Cy3-EGFP) was purchased from Dharmacon Research (Lafayette, CO, USA) as dried 2′-hydroxyl, annealed, purified and desalted duplexes. The siRNA sequences and modification patterns are reported in Supplementary data Table S1. DNA oligos were purchased from TAGC (Copenhagen, DK) as de-salted and reverse phase column purified oligos (Supplementary data Table S2). PLGA (lactide: glycolide molar ratio 75:25, Mw: 20 kDa) was purchased from Wako Pure Chemical Industries (Osaka, JP). Fluorescent PLGA-FPR648 was acquired from PolySciTech (Akina, West Lafayette, IN, USA). Polyvinylalcohol (PVA) 403 with an 80.0% degree of hydrolysis was provided by Kuraray (Osaka, JP). DOTAP chloride salt and NBD-DOTAP (1-oleoyl-2-[6-[(7-nitro-2-1,3-benzoxadiazol-4-yl)amino]hexanoyl]3-trimethylammonium propane, chloride salt) were purchased from Avanti Polar Lipids (Alabaster, AL, USA). Heparin and octyl-β-Dglucopyranoside (OG) were acquired from Sigma-Aldrich (St. Louis, MO, USA). Quant-iT™ RiboGreen® RNA Reagent and Tris–EDTA buffer (10 mM Tris, 1 mM EDTA, pH 7.5) (TE buffer) were acquired from Molecular Probes, Invitrogen (Paisley, UK). RNase-free diethyl pyrocarbonate (DEPC)-treated Milli-Q water was used for all buffers and dilutions. Additional chemicals were obtained commercially at analytical grade (SigmaAldrich).

R O

78 79

P

2.1. Materials

D

77

80

concentration was measured using the RiboGreen® RNA reagent ac- 136 cording to the manufacturer's instructions. 137

E

2. Materials and methods

T

76


140 141 142 143 144 145 146

149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165

169 170 171 172 173 174 175 176 177 178 179 180

184 185 186 187 188 189


3

190

2.8. Cell transfection

2.11. Statistics

249

191 192

The human non-small lung carcinoma cell line H1299 stably expressing EGFP (EGFP-H1299) was used. The cells were maintained in RPMI 1640 medium containing 10% fetal bovine serum (FBS, v/v) (PAA Laboratories, Pasching, AT) as previously reported [14,20]. The EGFP-H1299 cells were seeded in 12-well tissue culture plates (Corning, NY, USA) at a density of 1 × 105 cells/well. After 24 h (50% confluence), the cells were transfected with 500 μl of LPNs dispersed in RPMI 1640 medium containing 10% FBS (v/v) and incubated for 24 h. The cells were washed twice with 1 ml of phosphate-buffered saline (PBS, Sigma-Aldrich), and the transfection medium was replaced with full culture medium. After 48 h of transfection (90–95% confluence), the cells were washed twice with 1 ml PBS and detached using TrypLE™ Express (Invitrogen). The EGFP silencing was evaluated by flow cytometry using a Gallios flow cytometer (Beckman Coulter, Brea, CA, USA). The cell viability was estimated by propidium iodide (PI, Invitrogen) staining applying 1 μg/ml PI. Data was analyzed using FlowJo 7.6.5 (Three Star, Ashland, OR, USA).

Statistical analysis was performed using SigmaPlot v. 12.0 (Systat Software, San Jose, CA, USA) and GraphPad (GraphPad, La Jolla, CA, USA). Statistically significant differences were assessed by analysis of variance (ANOVA).

250 251

3. Results

254

204 205 206 207

2.9. Reverse transcription (RT) and stem–loop quantitative polymerase chain reaction (qPCR)

210

After trypsinization, the cells were pelleted by centrifugation for 5 min at 1100 ×g and washed twice with 1 ml PBS. The pellets were immediately frozen on dry ice and stored at − 80 °C. RNA isolation and purification were performed as previously described [20]. The recently described method was followed to perform the stem– loop RT qPCR [20]. Deoxynucleotide mix, Transcriptor Reverse Transcriptase buffer, Transcriptor Reverse Transcriptase and Protector RNase Inhibitor were all obtained from Roche (Basel, CH). The qPCR was performed with a LightCycler® 480 (Roche) using the Sybr Green® Master mix (Roche). The housekeeping genes snoRNA U109 (GenBank ID: AM055742.1) and let7a (NCBI ID: NR_029476) were selected for normalization of data as reported elsewhere [20]. The PCR data analyses were performed using the qpcR package in R (R Foundation for Statistical Computing, Vienna, AT) and the Cp values were determined by fitting and calculating the maximum of the second derivative [21]. The ΔΔCP method was used to quantify the siRNA [22] and the results were expressed in arbitrary units (a.u.). The applied stem–loop RT and PCR primers are reported in Supplementary data Table S2.

223 224 225 226 227 228

229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248

C

221 222

E

219 220

R

217 218

R

215 216

N C O

213 214

2.10. Dual-color fluorescence fluctuation spectroscopy (dcFFS) and fluorescence single particle tracking (fSPT) Freeze-dried particles were re-suspended in TE buffer, eventually supplemented with 50% (v/v) FBS, to a final concentration of 10 mg/ml and incubated at 37 °C with shaking. Prior to the analysis, aliquots were withdrawn and diluted with MilliQ water to 0.1 mg/ml for dcFFS and 0.2 mg/ml for fSPT, respectively. The samples were analyzed in glassbottom 96-well plates (Greiner Bio-one, Frickenhausen, DE). For the dcFFS measurements, an Eclipse TE300 inverted microscope (Nikon, Tokyo, JP) equipped with a water immersion objective (Plan Apo 60×, NA 1.2, collar rim correction, Nikon) was used. The 488 and 647 nm laser beams of a krypton–argon laser (Biorad, Cheshire, UK) were used. The intensities of both excitation wavelengths were controlled independently, using an acousto optic tunable filter (Opto-Electronique, St. Remy Les Chevreuse, FR). The microscope was equipped with a Time-Correlated Single Photon Counting (TCSPC) Data Acquisition module (PicoQuant, Berlin, DE), and data were recorded using SymPhotime (PicoQuant) and NIS (Nikon) software. The dcFFS spectra were analyzed as reported in Supplementary data and fSPT was performed and analyzed as previously reported [23].

U

211 212

T

208 209

Different nanoparticle formulations (2a–2e) were prepared by using the DESE method and systematically varying the N/P ratio from 1–62 by varying the siRNA loading at a constant DOTAP:PLGA weight ratio (15:85, Table 1). In addition, two control formulations were prepared including a formulation without DOTAP (1) and a formulation prepared using the SESE method (3). The mean hydrodynamic diameter of the LPNs prepared using the DESE method ranged from 213 to 286 nm and the PDI values were ≤ 0.16 suggesting narrow size distributions. The zeta-potential was significantly affected by the N/P ratio, and an N/ P ratio dependent increase in the zeta-potential was observed ranging from −27.0 ± 7.9 mV for 2a to +42.6 ± 2.5 mV for 2e (p b 0.001). However, the zeta-potentials for the nanoparticles at N/P ratios of 5, 10 and 62 (2c–e) were all in the same range (approximately +40 mV). The siRNA encapsulation efficiencies exceeded 30% for 2a–e and 3, and were significantly higher compared to 1 (p b 0.001). The relatively low encapsulation efficiency of 1, as compared to the other formulations, is a result of the absence of the DOTAP component. The siRNA encapsulation efficiency was apparently negatively affected by increasing the N/P ratio, although the differences were of little statistically significance (p b 0.05). This might be explained by a relatively larger loss of siRNA during preparation for the formulations prepared at the higher N/P ratios, as compared to the formulations with lower N/P ratios. This is a result of adjusting the N/P ratio by varying the initial amount of siRNA added to the inner water phase. Therefore, the siRNA loading of the nanoparticles also decreased with increasing N/P ratios.

F

202 203

O

200 201

252 253

3.1. Preparation and physicochemical characterization of siRNA-loaded 255 LPNs 256

R O

198 199

P

196 197

D

195

E

193 194

257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281

3.2. The siRNA partitioning in the primary emulsion is dependent on the N/P 282 ratio 283 To further investigate the potential molecular interactions between siRNA and DOTAP during LPN preparation, the spatial localization of siRNA in the primary w1/o emulsion was evaluated as a function of the N/P ratio (Fig. 1). In the absence of DOTAP the siRNA was localized in the inner w1 phase (Fig. 1A–C). At N/P = 0.1–1 an amount of siRNA was still detectable in the inner w1 phase (Fig. 1D–I), but with increasing DOTAP concentration in the oil phase, the siRNA was translocated to the w1/o interface and the continuous oil phase, i.e. at N/P ratios of 1 and 10 (Fig. 1G–L), the siRNA concentrated at the w1/o interface and ringlike patterns appeared. When the N/P ratio was increased to 62, the siRNA completely co-localized with DOTAP in the continuous organic phase (Fig. 1M–O). For 3, the siRNA was driven into the organic phase upon DOTAP–siRNA precomplexation and phase extraction [10,11].

284 285 286 287 288 289 290 291 292 293 294 295 296

3.3. The in vitro release kinetics of siRNA/DOTAP complexes from LPNs is N/P 297 ratio dependent 298 The siRNA release kinetics in buffer during the first 48 h of incubation was characterized to determine the influence of the DOTAP modification and the N/P ratio on the release. To discriminate between siRNA present as lipoplexes and non-complexed siRNA, heparin and detergent (HD) was added to dissociate the lipoplexes. In the absence of HD, only non-complexed siRNA is measured, and the difference between the siRNA concentration measured in the presence and absence of HD represents the amount of siRNA released as lipoplexes. Identical amounts of


299 300 301 302 303 304 305 306

4

Table 1 Physicochemical properties of the LPNs. Results denote mean ± SD (n = 3). Results significantly different from N/P = 1 (2a) are indicated: *p b 0.05, **p b 0.01, and ***p b 0.001.

322 323 324 325 326 327 328 329

215.0 ± 10.1* 240.7 ± 3.5 286.4 ± 8.0*** 213.2 ± 6.7** 222.5 ± 2.8* 216.1 ± 6.8** 159.3 ± 6.6***

0.09 ± 0.04 0.16 ± 0.08 0.13 ± 0.02 0.09 ± 0.01 0.10 ± 0.01 0.07 ± 0.01 0.20 ± 0.02

−39.0 ± 1.2 −27.0 ± 7.9 −25.3 ± 7.2 +40.3 ± 1.9*** +43.5 ± 1.1*** +42.6 ± 2.5*** +35.3 ± 4.5***

9.3 ± 3.0*** 63.3 ± 5.0 60.8 ± 7.5 66.1 ± 7. 3 54.6 ± 2.0* 54.7 ± 3.3* 32.0 ± 12.8**

74.0 74.0 37.0 14.8 7.4 1.2 74.0

3.1 ± 1.0 46.9 ± 0.4 19.8 ± 2.5 9.8 ± 1.1 4.0 ± 0.2 0.7 ± 0.0 10.5 ± 4.2

O

3.4. The surface topography during release is influenced by the DOTAP 330 modification 331

R O

Based on the release data presented in Fig. 2, AFM topography images of 1 and 2e were recorded to obtain more information about the surface topography and the release mechanism(s). These two formulations were selected for analysis as they apparently were characterized by two different release mechanisms; whereas non-complexed siRNA was released from 1, siRNA was released from 2e entirely as lipoplexes (Supplementary data, Tables S3–S4). At t = 0, both formulations were spherical and showed smooth, featureless surfaces (Fig. 3A and B). However, the surface of 2e appeared less well-defined already after 1 day of incubation at 37 °C (Fig. 3D), as compared to 1 (Fig. 3C). At day 4 the morphology of 2e was dramatically impaired (Fig. 3F), and no longer distinguishable at day 7 (Fig. 3H) while 1 retained its welldefined morphology both at days 4 and 7 (Fig. 3E and G). In addition, visible erosion of the PLGA matrix was apparent for 1 at day 15 (Fig. 3I) and day 30 (Fig. 3J). This suggests that the siRNA release mechanism for 1 during the first weeks is mainly based on diffusion of siRNA in the PLGA matrix, while erosion-based release of siRNA plays a substantial role from day 15 and onwards. In contrast, the release of siRNA in the physical form of lipoplexes from 2e appears to be related to re-organization events on the surface of the nanoparticles.

P

particles (0.5 mg) were analyzed for each sample. Therefore, the total amount of siRNA used varied between the samples, and the data was normalized for each formulation to the total amount of siRNA present at t = 0 to ease comparison of data. During the initial 48 h, each formulation appeared to have a unique siRNA release dynamics, which was N/P ratio dependent (Fig. 2). The concentration of non-complexed siRNA in the release medium was minimal for all formulations, except for 1 and 2a (Supplementary data, Tables S3–S4). For 2a, approximately 22% of the siRNA payload was rapidly released within 0.5 h. Subsequently, the siRNA was gradually released, resulting in the release of approximately 25% non-complexed siRNA and 3% of lipoplex-bound siRNA during the first 48 h (Fig. 2 and Supplementary data, Tables S3–S4). The total release of siRNA, both as non-complexed and lipoplexbound siRNA, is shown in Fig. 2. Compared to 2a, the burst release in 2e during the first 0.5 h was significantly reduced (approximately 4%, as compared to 22%). Afterwards, the released amount of siRNA gradually increased, and a total of approximately 22% was released after 48 h. Formulations 2b–2d displayed a minimal burst release followed by a slow sustained release; less than 5% of the loaded siRNA was released after 48 h, and almost all the siRNA was released as lipoplexes. Similarly, the release profile for 3 was gradual and little burst release was observed (Fig. 2).

F

0 1 2 5 10 62 1

T

320 321

0:100 15:85 15:85 15:85 15:85 15:85 15:85

C

318 319

Actual loading (μg siRNA/mg LPN)

E

316 317

Theoretical loading (μg siRNA/mg LPN)

R

314 315

Encapsulation efficiency (%)

R

312 313

Zeta-potential (mV)

O

310 311

PDI

C

308 309

Diameter (nm)

N

307

1 2a 2b 2c 2d 2e 3

N/P

U

t1:4 t1:5 t1:6 t1:7 t1:8 t1:9 t1:10

DOTAP:PLGA (w/w)

D

t1:3

E

t1:1 t1:2


Fig. 1. CLSM images of the primary w1/o-emulsion formed during the DESE process in the presence of different amounts of DOTAP (different N/P ratios). Red (A, D, G, J, M): Cy3-EGFP siRNA; green (B, E, H, K, N): NBD-DOTAP; overlay (C, F, I, L, O). Scale bars = 20 μm.


332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351


371 372 373 374 375 376 377 378 379 380 381 382

C

369 370

E

367 368

R

365 366

R

363 364

N C O

361 362

3.6. Nanoparticle morphology

414

Cryo-TEM was performed to further examine the structures identified by SAXS, and the formulations with the most pronounced differences were selected for these studies. The non-loaded PLGA nanoparticles, non-loaded LPNs and siRNA-loaded LPNs (2b, 2d) appeared spherical (Fig. 5), and the apparent particle sizes corresponded well to the particle sizes measured by using DLS (Table 1). The surface of the non-modified PLGA nanoparticles appeared smooth and featureless (Fig. 5A), while the surface of the non-loaded LPNs displayed small irregularities (Fig. 5B). A clear visual difference was observed between the siRNAloaded LPNs at N/P = 2 (2b, Fig. 5C) and the other types of nanoparticles: Highly electron-dense structures co-localized with the nanoparticles, which were easily melted upon electron beam exposure, and more

415 416

O

R O

U

359 360

P

To further shed light on the nanostructural characteristics of the LPNs, 2b–2d and control samples were examined at 25 °C by using SAXS (Fig. 4). An upturn of the SAXS pattern at low q-values was present in all the spectra for the particles. It is attributed to the presence of the submicron sized particles. In the absence of DOTAP and siRNA, the SAXS pattern for the PLGA nanoparticles did not show any indication of the formation of structure, confirming the amorphous state of the PLGA matrix (results not shown). For the non-loaded LPNs (PLGA– DOTAP) two Bragg peaks were apparent (Fig. 4, red arrows); an intense peak at a q-value of about 0.038 Å−1 and a weak, broad peak at a q-value of about 0.078 Å−1. These are most likely the characteristic first and second order reflections of a lamellar phase with a d-spacing value of 161.1 Å. The assignment of the lamellar phase is well in agreement with previously published studies on the propensity of DOTAP to form a lamellar phase in excess water [24]. The spacing recorded for these lamellar phases is significantly larger as compared to the d-spacing of the lamellar nanostructures of DOTAP-based dispersions. The dramatic increase in the structure parameter could be attributed to the interaction of DOTAP with the copolymer PLGA that can induce the formation of a highly swollen lamellar structure [24]. This lamellar phase might also be influenced by residual PVA, which is used as a stabilizer during the formation of the secondary emulsion and exists as a surface layer between the aqueous bulk phase and the hydrophobic PLGA nanoparticle core. This structure is retained in the presence of siRNA, cf. Fig. 4, 2b–2d. The first reflection of a lamellar phase was clearly visible for all three N/ P ratios (Fig. 4, red arrows) and the second reflection of the lamellar phase was also observed for 2b (Fig. 4, red arrow, q = 0.096 Å−1). For 2b–2d a decreased d-spacing was observed compared to the d-spacing for the non-loaded LPNs. Interestingly, an apparent increase of the dspacing values at increased N/P ratios was observed resulting in a

D

353 354

357 358

383 384

E

3.5. Structural characterization of LPNs by SAXS

T

352

355 356

spacing of 131.0 Å at N/P = 2 (2b), 139.5 Å at N/P = 5 (2c) and 151.2 Å at N/P = 10 (2d). Moreover, this reflection appears to be less intense for 2b, as compared to 2c, 2d and the non-loaded LPNs. This suggests that the lamellar structures on the surface get more compressed when more anionic siRNA is present, probably as a result of reduced repulsion between the positively charged DOTAP headgroups and increased electrostatic interaction between siRNA and DOTAP. An additional peak is apparent at a q-value of about 0.129 Å−1 in the SAXS patterns of the siRNA-loaded LPNs (2b–2d, Fig. 4, blue arrows). The corresponding d-spacing value was constant and thus independent of the N/P ratio. This signal most likely indicates the presence of a newly formed lamellar phase with a corresponding d-spacing value of about 48.7 Å. This phase presumably consists of siRNA layers sandwiched between cationic lipid bilayers altogether covering the outer surface of the dispersed polymeric particles. In support of this observation, the SAXS pattern of siRNA–DOTAP complexes at an N/P ratio of 10 showed a single clear peak at a q-value of 0.11 Å−1, which indicates the formation of a lamellar phase with a d-spacing value of about 55.9 Å (data not shown). This is well in agreement with the reported results in the literature for siRNA–DOTAP complexes (approximately 57 Å) [25]. Here, the interbilayer spacing is d = δlipid + δsiRNA + δwater (thickness of the membrane + the monolayer of siRNA + water). This value is also very similar to the d-spacing of lamellar nanostructures of DOTAPbased dispersions and DOTAP/DNA complexes (about 60 Å) [24,26, 27]. It has been reported that for DOTAP in dispersion, d = 54.4 Å and δlipid = 34.1 Å [25]. The slight decrease in the d-spacing value of this lamellar phase in the presence of PLGA could be attributed to the presence of impurities or residual PVA. Further experiments are needed to fully elucidate the effect of siRNA at different N/P ratios on the nanostructures of all investigated LPNs as they appear to have a highly complex structure.

F

Fig. 2. Cumulative release (%) of total siRNA (non-complexed + DOTAP-complexed) from 0.5 to 48 h. Results denote mean ± SD (n = 3).

5

Fig. 3. AFM images (5 μm × 5 μm) of 1 (A, C, E, G, I, J) and 2e (B, D, F, H) after 0 h (A–B), 1 day (C–D), 4 days (E–F), 7 days (G–H), 15 days (I) and 30 days (J) incubation. Scale bars = 1 μm.


385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413

417 418 419 420 421 422 423 424 425 426

449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466

C

447 448

E

445 446

R

443 444

R

441 442

O

439 440

C

437 438

N

435 436

U

433 434

4. Discussion

508

The N/P ratio is apparently decisive for the localization of the siRNA in the primary emulsion during preparation (Fig. 1). This can be explained by attractive electrostatic forces between the negatively charged siRNA molecules initially added to the w1 phase and the positively charged DOTAP molecules present in the o phase. At high N/P-ratios, the siRNA molecules are fully neutralized via electrostatic interaction with the hydrophilic headgroups of the DOTAP molecules, which are present in a molar excess. These charge-neutral and lipophilic complexes translocate into the organic PLGA phase. At the lower N/P ratios (1 and 10), the siRNA molecules are not fully neutralized in the primary emulsion, and siRNA molecules therefore localize mainly in the w1/o interphase due to the net negative charge of the complexes (Fig. 2). We hypothesize that this spatial siRNA localization during preparation of the nanoparticles determines the final localization of the siRNA in the nanoparticles, eventually influencing the specific siRNA release profile and the resulting delivery dynamics. To prove this hypothesis, the physicochemical characteristics, the release profiles and the siRNA delivery dynamics of a library of five formulations (2a–e) were studied, in which the N/P ratio was varied systematically from 1 to 62. Two additional control samples were prepared; one without DOTAP (1) and one where pre-complexed siRNA–DOTAP complexes were encapsulated using the SESE method (3)

509 510

F

The RNAi effect was evaluated in EGFP-H1299 cells and measured by using flow cytometry. In addition, the intracellular siRNA (int-siRNA) was measured by stem–loop RT qPCR to determine the effective delivery of full length siRNA [20,28]. Initial screening experiments showed that 2a induced low levels of silencing, also at high siRNA concentrations, and 2e caused cellular toxicity, even at low siRNA concentrations (results not shown). Therefore, these formulations were excluded from subsequent analyses. The silencing and int-siRNA mediated by 2b–d and 3 were dose and N/P ratio dependent (Fig. 6A–B), and no significant toxicity was observed (Fig. 6C). As negative controls 2c loaded with firefly luciferase (LUC) siRNA and non-loaded 2c were used, and none of the controls induced significant EGFP depletion, as compared to untreated cells (results not shown). These results confirm siRNA-dependent RNAi activation [29–31]. The int-siRNA amount was normalized against the initial amount of siRNA used for transfection (trans-siRNA) to estimate the fraction of initially transfected siRNA delivered in the intracellular compartment. Significantly higher delivery levels were observed for 2b–d, as compared to 1 and 3 (p b 0.001, Fig. 6D), and increasing the N/P ratio significantly enhanced this proportion (compare 2b, 2c and 2d, Fig. 6D). To investigate the correlation between siRNA release and int-siRNA, the int-siRNA was normalized against the amount of siRNA released in TE buffer during the first 48 h (Fig. 2). There was no statistically significant difference between the ratios of released siRNA to int-siRNA for 2b–d, indicating that the siRNA delivery mediated by these formulations is strongly related to the siRNA release kinetics (Fig. 6E). In addition, this ratio was significantly higher in 2b–d, as compared to 1 and 3, but also significantly higher for 3 as compared to 1 (p = 0.05, Fig. 6E). The relative intsiRNA quantification confirmed a strong correlation between the silencing effect and the int-siRNA concentration (p b 0.05, Fig. 6F). This suggests i) that gene silencing is mediated by transfection-competent DOTAP– siRNA lipoplexes [25], ii) that siRNA complexes released by 2b–d possess enhanced transfection efficiency, as compared to DOTAP–siRNA complexes released by 3 and non-complexed siRNA released by 1, and iii) that the DOTAP–siRNA lipoplexes released by 2b–d have comparable physicochemical properties and delivery efficiencies.

468 469

O

431 432

To further examine the behavior of the LPNs under in vitro conditions that reflect the biological environment more closely (i.e. with the presence of serum), a preliminary evaluation of the LPN stability and the siRNA release dynamics was performed by using fSPT and dcFFS. Fluorescent LPNs were prepared containing Alexa488-labeled siRNA (green) and FPR645-labeled PLGA (red). The LPNs were analyzed with regard to their hydrodynamic diameter, the concentration and the siRNA–PLGA co-localization. Formulations 2c–d and 3 underwent aggregation during the first few hours of incubation as indicated by the increase in size and the reduced particle concentration (Table 2). In contrast, 1 and 2b appeared rather stable during the first 3 h of incubation and subsequently aggregated gradually. This stability of 1 and 2b can be attributed to the negative surface charge [32], or to steric stabilization caused by the presence of PVA at the nanoparticle surface, whereas the aggregation of 2c–d and 3 is a result of the positive surface charge. However, the initial increment in size and aggregation dynamics appeared less severe for 2c and 3 as compared to 2d suggesting a particularly poor stability of 2d in serum [33] (Table 2). The release dynamics in TE buffer and TE buffer supplemented with 50% serum after 24 and 48 h was evaluated by dcFFS [34]. Fluorescent peaks appear simultaneously in the green and red channels of the spectrum upon detection of siRNA associated to the PLGA matrix. In contrast, only a peak in the green channel will be observed when siRNA–DOTAP lipoplexes are released from the PLGA nanoparticles. The release of noncomplexed siRNA will only contribute to the baseline fluorescence and is thus not detected. The co-localization index of the green peaks (gCL) was calculated as the percentage of co-localized peaks/min relative to the total peaks/min in the green channel. Therefore, the gCL represents the amount of siRNA–DOTAP complexes tightly associated with the PLGA matrix relative to the total amount of siRNA–DOTAP complexes in the sample (PLGA-associated + released siRNA–DOTAP complexes). Thus a reduced gCL is suggestive of higher amounts of released lipoplexes. Preliminary results in TE buffer showed that 2d had the lowest gCL while 2b and 3 had higher gCLs (Fig. 7A). This indicates a slower release dynamics of siRNA–DOTAP complexes for 2b as compared to 2d, which appears coherent with the release dynamics in Fig. 2. This trend was also observed in the presence of 50% FBS (Fig. 7B), but the gCL values appeared higher, as compared to TE buffer. The reason might be that the released lipoplexes disassemble and re-organize onto larger complexes in the presence of serum.

R O

3.7. Evaluation of RNAi and siRNA delivery dynamics in cell culture

467

P

430

3.8. The presence of serum affects the LPN stability and release dynamics

T

429

extended and loose structures were present in the dispersed phase. Such structures were also present for 2d (Fig. 5D), although they were less pronounced than for 2b.

D

427 428


E

6

Fig. 4. SAXS diffraction patterns for 2b–2d and non-loaded LPNs (PLGA–DOTAP). Arrows indicate Bragg peaks of two lamellar phases (red and blue, respectively).


470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507

511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529


7

Fig. 5. Cryo-TEM micrographs of different types of PLGA nanoparticles. Non-loaded PLGA nanoparticles (A), non-loaded LPNs (B), 2b (C) and 2d (D). Red arrow heads indicate electrondense structures colocalizing with the nanoparticles. Scale bars = 200 nm.

548 549 550 551

F

O

R O

P

D

546 547

E

544 545

T

542 543

C

540 541

E

538 539

R

536 537

R

534 535

addition of increasing amounts of siRNA; decreasing the N/P ratio results in d-spacings of 151.2 Å, 139.5 Å, and 131.0 Å for 2d, 2c and 2b, respectively (Fig. 4). This observation could be explained by compression of the extended lamellar DOTAP structures present on the surface of the nanoparticles. Therefore, the siRNA might compact the lamellar DOTAP structure, shifting the spacing to lower distances at lower N/P ratios [35]. The cryo-TEM images for 2b and 2d also confirm the presence of electron-dense structures co-localizing with the nanoparticles (Fig. 5C–D). Furthermore, LPNs loaded with siRNA displayed a characteristic peak with a Bragg spacing equivalent to 48.7 Å corresponding most likely to a lamellar phase covering the outer surface of the dispersed polymeric nanoparticles consisting of siRNA layers sandwiched between cationic lipid bilayers (Fig. 8A2) [25]. As a consequence of this structural organization, we suggest that the siRNA–DOTAP displacement dynamics mechanisms in vitro can be described as i) release of siRNA–DOTAP complexes from the lamellar DOTAP structures on the surface of the particles (Fig. 8B), eventually combined with ii) diffusion-mediated sustained release (Fig. 8C), and iii) matrix erosion-mediated release (Fig. 8D). Coherently with this model, 1 was characterized by siRNA molecules loosely associated with the surface and entrapped in the polymeric matrix, resulting in a burst release of siRNA, followed by diffusion- and erosion-mediated

N C O

532 533

[10–12]. Formulation-specific physicochemical properties (Table 1), release profiles (Figs. 2 and 7), structural features (Figs. 3, 4 and 5) and transfection efficiencies (Fig. 6) were demonstrated to be highly dependent on the N/P ratio. Consequently, we can propose an interpretation of the data based on nanostructural information to explain the release and delivery mechanism(s) of these formulations. The DOTAP molecules are suggested to mediate the association of siRNA molecules to the nanoparticles via three main mechanisms: i) Packing of siRNA molecules in a loose surface layer consisting of extended lamellar DOTAP structures (Fig. 8A1), ii) sandwiching of siRNA layers between cationic lipid bilayers via electrostatic interaction between the siRNA molecules and the headgroups of the DOTAP molecules, eventually attached to the hydrophobic PLGA matrix by incorporation of the hydrophobic lipid tails (Fig. 8A2), and iii) entrapment of net neutral and lipophilic siRNA–DOTAP complexes in the hydrophobic PLGA matrix (Fig. 8A3). This nanostructural model is supported by the SAXS data (Fig. 4) demonstrating that one phase present at the interphase between the aqueous phase and the LPN matrix consists of lamellar structures (Fig. 8A1), which provide 2c–2e with a net positive zeta-potential (Table 1). The spacing of this structure is large (161.1 Å) for the nonloaded LPNs (Fig. 4). However, this spacing is gradually reduced upon

U

530 531

Fig. 6. EGFP silencing, intracellular siRNA (int-siRNA) quantification and cell viability after 48 h. EGFP silencing (%) (A), int-siRNA quantification normalized to the expression of snoRNA u109 and let7a [arbitrary units (a.u.)] (B), percentage of viable cells (C), fraction of initial siRNA delivered intracellularly represented by the int-siRNA relative to the initial siRNA transfected (trans-siRNA) at a dose of 0.1 mg nanoparticles (D), fraction of released siRNA (rel-siRNA) delivered intracellularly represented by the int-siRNA amount relative to the relsiRNA during 48 h at a dose of 0.1 mg nanoparticles (E) and correlation between EGFP silencing (%) and int-siRNA (a.u.) (F). The color and numbers in the legend represent different doses of nanoparticles: black: 0.025 mg, dark gray: 0.05 mg, medium gray: 0.1 mg, light gray: 0.2 mg. In (F) the red dot represents 1 (0.1 mg), the closed squares represent 2b–d (0.025–0.1 mg), and open squares represent 3 (0.025–0.1 mg). Data represent mean values ± SD (N = 2, n = 3). In A and B red stars indicate results significantly different between successive doses, green stars indicate significant differences between the highest and lowest doses: *p b 0.05. In D and E results significantly different from 2b are indicated by: *p = 0.05, ***p b 0.001.


552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573

8 t2:1 t2:2 t2:3


Table 2 Average hydrodynamic diameter and particle concentration in TE-buffer containing 50% FBS measured by fSPT. The result of the analysis of at least ten videos for each of the two analyzed batches is reported as batch1/batch2. NSA (not suitable for analysis) indicates that the samples contained too few particles for analysis.

t2:4

t=0h

t2:5 t2:6

Diameter (nm)

Number of particles/mg × 1012

Diameter (nm)


Diameter (nm)


Diameter (nm)


234/247 270/301 247/314 336/408 183/236

2.56/2.24 2.20/2.08 2.80/1.70 0.98/1.52 2.26/2.40

236/265 240/295 221/223 332/345 269/303

2.36/2.23 2.54/2.34 1.97/0.41 1.22/1.21 1.56/1.60

313/348 276/385 306/364 369/545 394/434

1.48/0.81 1.40/0.63 0.29/0.29 0.58/0.48 0.21/0.74

401/450 743/853 NSA NSA 489/513

1.24/0.71 1.33/1.03 NSA NSA 0.09/0.09

593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612

5. Conclusions

620

R O

O

F

613

P

A comprehensive study of systematically varied LPNs loaded with siRNA allowed for a thorough mechanistic modeling of the carriermediated intracellular siRNA delivery. The proposed model highlights the importance of the DOTAP and siRNA organization mainly with regard to surface structures. This organization is influenced by the N/P ratio and the presence of excess DOTAP. For such formulations, the siRNA is primarily released in the physical form of siRNA–DOTAP lipoplexes upon surface re-organization. This results in improved transfection efficiency and siRNA delivery in vitro.

D

591 592

stability in biologically relevant media, and they might be suitable candidates for in vivo translation. Highly cationic DOTAP structures have been shown to be responsible for invoking cellular toxicity and inflammatory responses in vivo [40,41]. For this reason, formulations that are characterized by an abundant release of such structures with excess cationic charge a few hours after administration, e.g. 2d and 2e, may raise safety concerns.

E

589 590

t = 48 h

T

587 588

C

585 586

E

583 584

R

581 582

R

579 580

O

577 578

release (Fig. 2) [19]. For 2a, we hypothesize the co-existence of noncomplexed siRNA, which is released as a burst, and siRNA–DOTAP complexes (Fig. 2, Supplementary data, Table S3). The released, noncomplexed siRNA has a poor transfection efficiency, which explains why 1 and 2a appear to be inefficient delivery vectors in vitro (Fig. 6 and results not shown). In contrast, the molar excess of DOTAP in 2b– e most likely results in the formation of siRNA–DOTAP structures. The release of siRNA–DOTAP complexes from the lamellar DOTAP structures on the surface of the particles, which might be visually represented by the morphological changes observed by AFM (Fig. 3), results in the sustained release of lipoplexes (Fig. 2), which are potent transfection agents [25]. This mechanism is suggested to account for enhancing the transfection efficiency of 2b–e (Fig. 6D–E and results not shown) after 24 h of transfection, assuming a comparable cytoplasmic delivery efficiency for the lipoplexes released by 2b–d in the presence of physiological components, as suggested by the strong correlation between intracellular siRNA and gene silencing for this formulation (Fig. 6F) [20], and that the displacement dynamic is rather conserved. Nevertheless, the presence of cellular components and the intracellular fate of the siRNA might also have a major impact on the transfection efficiency, but in the current in vitro model it appears that the siRNA lipoplex displacement dynamic is preponderant, determining the strong correlation observed in Fig. 6F. Preliminary SAXS results suggest that 3 is characterized by different structures, which we are currently analyzing further (data not shown), as compared to 2b–d, which could explain the lower transfection efficiency of 3 (Fig. 6E) [25,36]. The dcFFS results support this hypothesis by showing that the complexes are released both in TE buffer and serum with a similar N/P dependency (Fig. 7). However, in serum-containing medium the released lipoplexes appeared more susceptible to disassembly or re-organization (Fig. 7) as previously discussed [37]. Coherently with recently published results, formulations extensively exposing charged lipid structures at the surface appear more prone to interact with serum proteins, resulting in corona formation and aggregation (Table 2) [38,39]. These results are comparable with the results obtained in human serum (Supplementary data, Table S5). In a therapeutic perspective, the obtained results suggest that 2b and 2c mediate potent siRNA-dependent RNAi activation in vitro through the sustained release of siRNA–DOTAP complexes and promising

C

575 576

N

574

1 2b 2c 2d 3

t = 24 h

Conflict of interests The authors declare that they have no competing interests.

614 615 616 617 618 619

621 622 623 624 625 626 627 628 629 630 631

Acknowledgments

632

We gratefully acknowledge the Danish Agency for Science, Technology and Innovation, the Danish Research Council for Technology and Production Sciences [Grant No. 26-02-0019], the Alfred Benzon Foundation and the Carlsberg Foundation for funding this project. In addition, the research leading to these results has received support from the Commission of the European Communities, Priority 3 “Nanotechnologies and Nanosciences, Knowledge Based Multifunctional Materials, New Production Processes and Devices” of the Sixth Framework Programme for Research and Technological Development (Targeted Delivery of Nanomedicine: NMP4-CT-2006-026668), and the Innovative Medicines Initiative Joint Undertaking under grant agreement n° 115363 resources of which are composed of financial contribution from the European Union's Seventh Framework Programme (FP7/

633

U

t2:7 t2:8 t2:9 t2:10 t2:11

t=3h

Fig. 7. Percentage of co-localized peaks in the green channel (gCL) as a function of incubation time in TE buffer (A) and 50% FBS (B). Data represents mean values of two batches in triplicate measures, and the error bar indicates the SD (N = 2, n = 3).


634 Q3 635 636 637 638 639 640 641 642 643 644 645

9

F


Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jconrel.2014.12.026.

661

References

662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698

[1] L. Aagaard, J.J. Rossi, RNAi therapeutics: principles, prospects and challenges, Adv. Drug Deliv. Rev. 59 (2007) 75–86. [2] S. Colombo, X. Zeng, H. Ragelle, C. Foged, Complexity in the therapeutic delivery of RNAi medicines: an analytical challenge, Exp. Opin. Drug Deliv. 11 (2014) 1481–1495. [3] K. Bruno, Using drug–excipient interactions for siRNA delivery, Adv. Drug Deliv. Rev. 63 (2011) 1210–1226. [4] B. Semete, L. Booysen, Y. Lemmer, L. Kalombo, L. Katata, J. Verschoor, H.S. Swai, In vivo evaluation of the biodistribution and safety of PLGA nanoparticles as drug delivery systems, Nanomedicine 6 (2010) 662–671. [5] M.S. Cartiera, K.M. Johnson, V. Rajendran, M.J. Caplan, W.M. Saltzman, The uptake and intracellular fate of PLGA nanoparticles in epithelial cells, Biomaterials 30 (2009) 2790–2798. [6] H. Katas, E. Cevher, H.O. Alpar, Preparation of polyethyleneimine incorporated poly(D, L-lactide-co-glycolide) nanoparticles by spontaneous emulsion diffusion method for small interfering RNA delivery, Int. J. Pharm. 369 (2009) 144–154. [7] W. Hasan, K. Chu, A. Gullapalli, S.S. Dunn, E.M. Enlow, J.C. Luft, S. Tian, M.E. Napier, P.D. Pohlhaus, J.P. Rolland, J.M. DeSimone, Delivery of multiple siRNAs using lipidcoated PLGA nanoparticles for treatment of prostate cancer, Nano Lett. 12 (2012) 287–292. [8] S. Diez, G. Navarro, C.T. De Ilarduya, In vivo targeted gene delivery by cationic nanoparticles for treatment of hepatocellular carcinoma, J. Gene Med. 11 (2009) 38–45. [9] K. Tahara, T. Sakai, H. Yamamoto, H. Takeuchi, Y. Kawashima, Establishing chitosan coated PLGA nanosphere platform loaded with wide variety of nucleic acid by complexation with cationic compound for gene delivery, Int. J. Pharm. 354 (2008) 210–216. [10] T. Lobovkina, G.B. Jacobson, E. Gonzalez-Gonzalez, R.P. Hickerson, D. Leake, R.L. Kaspar, C.H. Contag, R.N. Zare, In vivo sustained release of siRNA from solid lipid nanoparticles, ACS Nano 5 (2011) 9977–9983. [11] G.B. Jacobson, E. Gonzalez-Gonzalez, R. Spitler, R. Shinde, D. Leake, R.L. Kaspar, C.H. Contag, R.N. Zare, Biodegradable nanoparticles with sustained release of functional siRNA in skin, J. Pharm. Sci. 99 (2010) 4261–4266. [12] G.B. Jacobson, R. Shinde, C.H. Contag, R.N. Zare, Sustained release of drugs dispersed in polymer nanoparticles, Angew. Chem. Int. Ed. Engl. 47 (2008) 7880–7882. [13] B.C.M. te Boekhorst, L.B. Jensen, S. Colombo, A.K. Varkouhi, R.M. Schiffelers, T. Lammers, G. Storm, H.M. Nielsen, G.J. Strijkers, C. Foged, K. Nicolay, MRI-assessed therapeutic effects of locally administered PLGA nanoparticles loaded with anti-

C

E

R

R

N C O

U

655 656

[16]

R O

659 660

653 654

P

Appendix A. Supplementary data

651 652

[15]

inflammatory siRNA in a murine arthritis model, J. Control. Release 161 (2012) 772–780. D.K. Jensen, L.B. Jensen, S. Koocheki, L. Bengtson, D. Cun, H.M. Nielsen, C. Foged, Design of an inhalable dry powder formulation of DOTAP-modified PLGA nanoparticles loaded with siRNA, J. Control. Release 157 (2012) 141–148. S.D. Rose, D.-H. Kim, M. Amarzguioui, J.D. Heidel, M.a. Collingwood, M.E. Davis, J.J. Rossi, M.A. Behlke, Functional polarity is introduced by Dicer processing of short substrate RNAs, Nucleic Acids Res. 33 (2005) 4140–4156. G.B. Jacobson, R. Shinde, R.L. McCullough, N.J. Cheng, A. Creasman, A. Beyene, R.P. Hickerson, C. Quan, C. Turner, R.L. Kaspar, C.H. Contag, R.N. Zare, Nanoparticle formation of organic compounds with retained biological activity, J. Pharm. Sci. 99 (2010) 2750–2755. T.C. Huang, H. Toraya, T.N. Blanton, Y. Wu, X-ray powder diffraction analysis of silver behenate, a possible low-angle diffraction standard, J. Appl. Crystallogr. 26 (1993) 180–184. S.S. Nielsen, K.N. Toft, D. Snakenborg, M.G. Jeppesen, J.K. Jacobsen, B. Vestergaard, J.P. Kutter, L. Arleth, BioXTAS RAW, a software program for high-throughput automated small-angle X-ray scattering data reduction and preliminary analysis, J. Appl. Crystallogr. 42 (2009) 959–964. D. Cun, D.K. Jensen, M.J. Maltesen, M. Bunker, P. Whiteside, D. Scurr, C. Foged, H.M. Nielsen, High loading efficiency and sustained release of siRNA encapsulated in PLGA nanoparticles: quality by design optimization and characterization, Eur. J. Pharm. Biopharm. 77 (2011) 26–35. S. Colombo, H.M. Nielsen, C. Foged, Evaluation of carrier-mediated siRNA delivery: lessons for the design of a stem–loop qPCR-based approach for quantification of intracellular full-length siRNA, J. Control. Release 166 (2013) 220–226. A.N. Spiess, C. Feig, C. Ritz, Highly accurate sigmoidal fitting of real-time PCR data by introducing a parameter for asymmetry, BMC Bioinformatics 9 (2008) 221. M.W. Pfaffl, A new mathematical model for relative quantification in real-time RTPCR, Nucleic Acids Res. 29 (2001) e45. K. Braeckmans, K. Buyens, W. Bouquet, C. Vervaet, P. Joye, D.V. Filip, L. Plawinski, L. Doeuvre, E. Angles-Cano, N.N. Sanders, J. Demeester, S.C. De Smedt, Sizing nanomatter in biological fluids by fluorescence single particle tracking, Nano Lett. 10 (2010) 4435–4442. J.O. Radler, I. Koltover, A. Jamieson, T. Salditt, C.R. Safinya, Structure and interfacial aspects of self-assembled cationic lipid–DNA gene carrier complexes, Langmuir 14 (1998) 4272–4283. N.F. Bouxsein, C.S. McAllister, K.K. Ewert, C.E. Samuel, C.R. Safinya, Structure and gene silencing activities of monovalent and pentavalent cationic lipid vectors complexed with siRNA, Biochemistry 46 (2007) 4785–4792. C.R. Safinya, Structures of lipid–DNA complexes: supramolecular assembly and gene delivery, Curr. Opin. Struct. Biol. 11 (2001) 440–448. S. Weisman, D. Hirsch-Lerner, Y. Barenholz, Y. Talmon, Nanostructure of cationic lipid–oligonucleotide complexes, Biophys. J. 87 (2004) 609–614. A. Cheng, M. Li, Y. Liang, Y. Wang, L. Wong, C. Chen, A.V. Vlassov, S. Magdaleno, Stem–loop RT-PCR quantification of siRNAs in vitro and in vivo, Oligonucleotides 19 (2009) 203–208. J. Zhou, K.T. Shum, J.C. Burnett, J.J. Rossi, Nanoparticle-based delivery of RNAi therapeutics: progress and challenges, Pharmaceuticals (Basel) 6 (2013) 85–107. J. Tabernero, G.I. Shapiro, P.M. LoRusso, A. Cervantes, G.K. Schwartz, G.J. Weiss, L. Paz-Ares, D.C. Cho, J.R. Infante, M. Alsina, M.M. Gounder, R. Falzone, J. Harrop, A.C. White, I. Toudjarska, D. Bumcrot, R.E. Meyers, G. Hinkle, N. Svrzikapa, R.M. Hutabarat, V.A. Clausen, J. Cehelsky, S.V. Nochur, C. Gamba-Vitalo, A.K. Vaishnaw, D.W. Sah, J.A. Gollob, H.A. Burris III, First-in-humans trial of an RNA interference therapeutic targeting VEGF and KSP in cancer patients with liver involvement, Cancer Discov. 3 (2013) 406–417. G.J. Hannon, J.J. Rossi, Unlocking the potential of the human genome with RNA interference, Nature 431 (2004) 371–378. D.M. Charbonneau, H.A. Tajmir-Riahi, Study on the interaction of cationic lipids with bovine serum albumin, J. Phys. Chem. B 114 (2010) 1148–1155. J.L. Betker, J. Gomez, T.J. Anchordoquy, The effects of lipoplex formulation variables on the protein corona and comparisons with in vitro transfection efficiency, J. Control. Release 171 (2013) 261–268.

D

658

649 650

[14]

[17]

T

657

2007–2013) and EFPIA companies' in kind contributions. MAXLab is acknowledged for providing beamtime and the instrument for the SAXS studies. The research leading to the SAXS results has received funding from the European Community's Seventh Framework Programme (FP7/2007–2013) under grant agreement n° 226716. We thank M.L. Pedersen, F. Rose, E. Zagato, G. Dakwar, S.C. de Smedt and A. Hendrix for technical support and valuable scientific discussions. K. Remaut is a post-doctoral fellow of the Research Foundation-Flanders (FWO). The funding sources had no involvement in the study design, in the collection, analysis and interpretation of data, just as they had no involvement in the writing of the report and the decision to submit the paper for publication.

647 648

E

646

O

Fig. 8. Model for the structural characteristics of siRNA-loaded LPNs and their release dynamics. For LPNs the siRNA is loaded in 1) surface lamellar layers, 2) surface-grafted DOTAP–siRNA complexes, and 3) matrix-entrapped siRNA–DOTAP complexes (A). The release from these structures occurs as release of siRNA–DOTAP complexes by disassembly of surface structures (B), sustained release of siRNA–DOTAP complexes by diffusion (C) and matrix erosion (D).

[18]

[19]

[20]

[21] [22] [23]

[24]

[25]

[26] [27] [28]

[29] [30]

[31] [32] [33]


699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761

10

[38] G. Caracciolo, D. Pozzi, A.L. Capriotti, C. Cavaliere, P. Foglia, H. Amenitsch, A. Lagana, Evolution of the protein corona of lipid gene vectors as a function of plasma concentration, Langmuir 27 (2011) 15048–15053. [39] A.L. Capriotti, G. Caracciolo, G. Caruso, P. Foglia, D. Pozzi, R. Samperi, A. Lagana, Differential analysis of “protein corona” profile adsorbed onto different nonviral gene delivery systems, Anal. Biochem. 419 (2011) 180–189. [40] E. Khazanov, D. Simberg, Y. Barenholz, Lipoplexes prepared from cationic liposomes and mammalian DNA induce CpG-independent, direct cytotoxic effects in cell cultures and in mice, J. Gene Med. 8 (2006) 998–1007. [41] Z. Ma, J. Li, F. He, A. Wilson, B. Pitt, S. Li, Cationic lipids enhance siRNA-mediated interferon response in mice, Biochem. Biophys. Res. Commun. 330 (2005) 755–759.

N

C

O

R

R

E

C

T

E

D

P

R O

O

F

[34] K. Remaut, B. Lucas, K. Raemdonck, K. Braeckmans, J. Demeester, S.C. De Smedt, Can we better understand the intracellular behavior of DNA nanoparticles by fluorescence correlation spectroscopy? J. Control. Release 121 (2007) 49–63. [35] G. Subramanian, R.P. Hjelm, T.J. Deming, G.S. Smith, Y. Li, C.R. Safinya, Structure of complexes of cationic lipids and poly(glutamic acid) polypeptides: a pinched lamellar phase, J. Am. Chem. Soc. 122 (2000) 26–34. [36] J. Smisterova, A. Wagenaar, M.C. Stuart, E. Polushkin, G. ten Brinke, R. Hulst, J.B. Engberts, D. Hoekstra, Molecular shape of the cationic lipid controls the structure of cationic lipid/dioleylphosphatidylethanolamine–DNA complexes and the efficiency of gene delivery, J. Biol. Chem. 276 (2001) 47615–47622. [37] K. Crook, B.J. Stevenson, M. Dubouchet, D.J. Porteous, Inclusion of cholesterol in DOTAP transfection complexes increases the delivery of DNA to cells in vitro in the presence of serum, Gene Ther. 5 (1998) 137–143.

U

762 763 764 765 766 767 768 769 770 771 772 773 774 786



775 776 777 778 779 780 781 782 783 784 785