Novel polyglycidol-lipid conjugates create a stabilizing hydrogen-bonded layer around cholesterol-containing dipalmitoyl phosphatidylcholine liposomes

Accepted Manuscript Novel Polyglycidol-Lipid Conjugates Create a Stabilizing Hydrogen-Bonded Layer around Cholesterol-Containing Dipalmitoyl Phosphatidylcholine Liposomes Pavel Bakardzhiev, Denitsa Momekova, Katarina Edwards, Spiro Konstantinov, Stanislav Rangelov PII:

S1773-2247(15)00114-8

DOI:

10.1016/j.jddst.2015.06.019

Reference:

JDDST 58

To appear in:

Journal of Drug Delivery Science and Technology

Received Date: 31 March 2015 Revised Date:

15 June 2015

Accepted Date: 26 June 2015

Please cite this article as: P. Bakardzhiev, D. Momekova, K. Edwards, S. Konstantinov, S. Rangelov, Novel Polyglycidol-Lipid Conjugates Create a Stabilizing Hydrogen-Bonded Layer around CholesterolContaining Dipalmitoyl Phosphatidylcholine Liposomes, Journal of Drug Delivery Science and Technology (2015), doi: 10.1016/j.jddst.2015.06.019. 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.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

3

Novel Polyglycidol-Lipid Conjugates Create a Stabilizing Hydrogen-Bonded Layer around Cholesterol-Containing Dipalmitoyl Phosphatidylcholine Liposomes

4 5

Pavel Bakardzhieva* , Denitsa Momekovab , Katarina Edwardsc , Spiro Konstantinovd , Stanislav Rangelova

1 2

RI PT

6 7 8

a

9

Sofia, Bulgaria, [email protected]

SC

Institute of Polymers, Bulgarian Academy of Sciences, 103-A Acad. G. Bonchev St., 1113

10

b

11

Medical University – Sofia, 2 Dunav St., 1000 Sofia, Bulgaria

12

c

Department of Chemistry – BMC, University of Uppsala, SE 751 23 Uppsala, Sweden

13

d

Department of Pharmacology, Pharmacotherapy and Toxicology, Faculty of Pharmacy,

14

Medical University – Sofia, 2 Dunav St., 1000 Sofia, Bulgaria

M AN U

Department of Pharmaceutical Technology and Biopharmaceutics, Faculty of Pharmacy,

TE D

15

Abstract

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

Hybrid liposomes resulting from co-assembly of dipalmitoylphosphatidylcholine and polyglycidol-derivatized lipids were prepared. The latter were composed of a lipid-mimetic residue to which a linear polyglycidol chain (degree of polymerization, DP, in the 23 – 110 range) was conjugated. Formulations with varying copolymer type and content were prepared by film hydration technique followed by extrusion. The hybrid structures were studied by means of dynamic and electrophoretic light scattering, cryogenic transmission electron microscopy, and fluorescence spectroscopy. Cytotoxicity towards OPM-2 (multiple myeloma) and EJ (human urinary bladder carcinoma) cell lines was assessed as well. Predominantly unilamellar liposomes with mean hydrodynamic diameters in the 113 – 134 nm range and neutral to slightly negative surface potential were prepared. The integrity of liposomes containing copolymers with DP of the polyglycidol chain 23 and 30 was preserved at copolymer contents up to 10 mol %. Bilayer disks were observed at somewhat lower contents of the copolymers of the highest DP of the polyglycidol chain. The hybrid structures were less leaky than the plain liposomes, which was attributed to formation of a strongly hydrogenbonded polyglycidol layer around the bilayer membrane. They exhibited low toxicological potential, favorable physicochemical characteristics, and ability to act as containers for sustained release.

AC C

EP

16

1

ACCEPTED MANUSCRIPT Keywords: Polymer grafted liposomes, Polyglycidol, Hydrogen bonds, Cytotoxicity

35

1. Introduction

36 37 38 39 40 41 42 43 44 45 46 47 48 49

Nanotechnology promises new medical therapies, precise, sensitive and rapid diagnostic tools as well as new materials for tissue engineering. It presents potential opportunities to create medicines that would be less harmful towards normal tissues and more effective towards diseased tissues by using advanced drug delivery systems. The latter can be hybrid nanoparticles and chimeric nanosystems [1-3] resulting from co-assembly of, for instance, an amphiphilic copolymer with other substances such as surfactants, lipids, (co)polymers, proteins, oligo- and polynucleotides. It is remarkable that the introduction of an additional entity even in small amounts in the hybrid nanostructures can be manifested in significant alteration of the aggregate morphology and characteristics and can produce new functionality and properties [4]. The sterically stabilized (Stealth) liposomes are such hybrid structures that are composed of a self-closed lipid bilayer, enclosing part of the surrounding media into their interior, which is surface modified by polymer-derivatized lipids. The lipid residue is anchored in the bilayer membrane, whereas the polymer (typically poly(ethylene glycol), PEG) creates a repulsive barrier around the liposomes.

50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68

The PEG-lipid is introduced in quantities of less than 10 mol % (typically 4 – 6 mol %), which is sufficient to confer colloidal and in vivo stability to the liposomes, to reduce the opsonization (i.e., adsorption of marker macromolecules on the liposomal surface) as well as the uptake of the marked species by the macrophages, which results in prolongation of circulation time [5,6]. The approach to improve the performance and prolong the blood circulation time by creating sterically stabilized liposomes is not free of drawbacks and disadvantages. At certain critical content, the PEG-lipids induce transition from bilayers to micellar phase [7-10], which may occur via intermediate structures such as open bilayer disks. Such structures are useless as far as encapsulation and delivery of water-soluble active substances are concerned. Furthermore, the PEG-lipids, even in small amounts, may give rise to rapid leakage of entrapped water-soluble substances due to spontaneous formation and stabilization of transient holes in the bilayer [11]. In addition, many reports have emphasized that repeated administration of PEG-coated liposomes causes an unexpected immunogenic response, known as “accelerated blood clearance (ABC) phenomenon,” which results in disappearance of the long-circulation characteristics of the liposomes [12-14]. Furthermore, in the case when PEG is conjugated to phosphatidyl ethanolamine a net negative charge is introduced at the bilayer membrane surface, which has been reported to play a key role in complement activation and anaphylotoxin production [15] in addition to substantial alteration of the properties and performance of the liposomes [16,17].

69 70 71 72 73 74

Although the PEG-lipids, which are nowadays commercially available in diversity of lipid anchors and molar masses of the PEG chain, are the most widely studied and extensively used materials to provide Stealth properties to liposomes, other polymer-derivatized lipids may offer important advancement in creation of such hybrid structures. PEG is the polymer of choice to be conjugated to a lipid-mimetic residue, whereas polymers such as poly(Nisopropylacrylamide) and poly(2-alkyl-2-oxazoline) have been occasionally used as PEG

AC C

EP

TE D

M AN U

SC

RI PT

34

2

ACCEPTED MANUSCRIPT substitutes [18-23]. A number of polyglycerol-derivatized phospholipids with degree of polycondensation of the polyglycerol chain below 40 have been synthesized [24,25]. The individual derivatives were incorporated into distearoylphosphatidylcholine/cholesterol liposomes in order to achieve long circulation time in vivo [24] and to prevent induction of the ABC phenomenon against long-circulating liposomes upon repeated administration [25].

80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95

We have recently reported on the synthesis and self-association of a series of novel nonphospholipid polyglycidol conjugates [26]. These completely non-ionic conjugates are composed of a polyglycidol chain of degrees of polymerization (DP) in the 23–110 range attached to a lipid-mimetic residue, consisting of two C12 fully saturated hydrocarbon chains covalently linked to a glycerol skeleton via ether linkages. The structural formula and molecular characteristics of the conjugates are presented in Fig. 1 and Table 1, respectively. Linear polyglycidol is a flexible, hydrophilic, and biocompatible polymer with biological tolerability, comparable to or even better than those of PEG, poly(N-vinylpyrrolidone) and other biocompatible polymers [27]. Polyglycidol is structurally similar to PEG: both consist of a polyether backbone, however, unlike PEG, the polyglycidol chain bears a hydroxymethylene group in each repeating monomer unit (Fig. 1), the presence of which gives platform for further functionalization. In this study we investigate the effects of the novel non-phospholipid polyglycidol conjugates on the physicochemical properties, morphology, and membrane integrity of dipalmitoylphosphatidylcholine/cholesterol liposomes. Cytotoxicity assays against OPM-2 (multiple myeloma) and EJ (human urinary bladder carcinoma) cell lines are performed as well.

M AN U

SC

RI PT

75 76 77 78 79

TE D

96 97

2. Materials and Methods

98

2.1. Materials

1,2-Dipalmitoyl-sn-glycero-phosphocholine (DPPC) (PubChem CID: 452110), cholesterol (CHOL) (PubChem CID: 5997), 5(6)-carboxyfluoresceine (CF), Triton X100 (PubChem CID: 5590), chloroform (PubChem CID: 6212) and methanol (PubChem CID: 887) were purchased from Sigma Aldrich. The synthesis and characterization of the novel non-phospholipid polyglycidol conjugates (hereinafter DDP-(G)n polymers) are presented in details elsewhere [26]. Fig. 1 and Table 1 present the structural formula and molecular characteristics of DDP(G)n polymers, respectively.

106

2.2. Preparation of liposomes

107 108 109 110 111 112

Chloroform solutions of DPPC and CHOL (2:1 molar ratio, 3 mM total lipid concentration) were placed into glass tubes to which a methanol solution of the respective polymer in a defined polymer/lipid molar ratio was added. The solvents were evaporated under a stream of argon and all traces of solvent were removed under vacuum overnight. Phosphate buffer solution (PBS) (pH 7.4) was added to the dry lipid/polymer film and the resulting dispersions were subjected to ten freeze-thaw cycles and then extruded 30 times through polycarbonate

AC C

EP

99 100 101 102 103 104 105

3

ACCEPTED MANUSCRIPT filters of pore size 100 nm using a LiposoFast handle type extruder (Avestin Inc., Canada). For the leakage experiments the hydration step was performed with 100 mM buffered solution of CF (pH 7.4). The unentrapped dye was removed by passing through a Sephadex G50 column (Pharmacia, Uppsala, Sweden), equilibrated with PBS buffered saline (pH 7.4).

117

2.3. Tumor cell lines

118 119 120 121 122 123 124 125

The human tumor cell line OPM-2 (multiple myeloma) was supplied by DSMZ GmbH, Germany, whereas EJ cells (human urinary bladder carcinoma) originated from the American Type Cell Culture, USA. The cells were cultured routinely in cell culture flasks with RPMI1640 liquid medium supplemented with 10% fetal bovine serum (FBS) and 2 mM Lglutamine, housed in an incubator ’BB 16-Function Line’ Heraeus (Kendro, Hanau, Germany) at 37 °C in 5% CO2 humidified atmosphere. The cell cultures were maintained in logarithmic growth phase by supplementation with fresh medium twice weekly. EJ cells were used before seventh passage.

126

2.4. Methods

127

2.4.1. Dynamic light scattering

128 129 130 131 132

The size and size distribution were determined using photon-correlation spectrometer ZetaSizer NanoZS (Malvern Instruments), equipped with a He-Ne laser (633 nm) and a NIBS (Non-Invasive Back Scatter) system, allowing measurements of the scattered light at angle of 173°. The hydrodynamic diameters of the particles were calculated using the Stokes-Einstein equation (1):

134

Dh = kT/(3πηD)

M AN U (1)

EP

135

TE D

133

SC

RI PT

113 114 115 116

where k is the Boltzman constant, η is the solvent viscosity at temperature T in Kelvin and D is the diffusion coefficient. The measurements were performed at 25 °C. Each measurement was performed in triplicate.

139

2.4.2. Electrophoretic light scattering

140 141 142

The ζ-potentials were determined by laser Doppler micro-electrophoresis using a ZetaSizer Nano ZS (Malvern Instruments) apparatus. The ζ-potentials were calculated from the obtained electrophoretic mobility at 25 °C by the Smoluchowski equation (2):

AC C

136 137 138

143 144

ζ = 4πηυ/ε

(2)

145

4

ACCEPTED MANUSCRIPT

158 159 160 161 162 163 164 165 166

RI PT

A small (~ 1 µl) drop of the liposome sample was placed onto a copper grid coated with a perforated polymer film. Excess sample was removed by a blotting procedure performed in a custom-build environmental chamber with controlled temperature and humidity. The thin sample films spanning the holes of the polymer support were vitrified by plunging the grid into liquid ethane held at a temperature just above its freezing point (-183 °C). Afterwards, the vitrified specimen was transferred to a Zeiss EM 902A microscope, operating at 80 kV. The temperature was kept below 108K during both the transfer and the viewing procedures to prevent sample perturbation. More details about sample preparation and cryo-TEM technique can be found elsewhere [28]. 2.4.4. Leakage assay

The leakage from the liposomes was measured using an F 7000 Hitachi fluorescence spectrophotometer. 200 µl of CF-loaded liposomes were added to 2 ml phosphate buffer solution and incubated for 24 h at 37 °C. At concentration 80 mM and above, the fluorescence of CF is self-quenched. As the dye leaks out from the liposomes, it gets diluted and fluorescence intensity increases. The fluorescence intensity was measured at regular time intervals at 520 nm after excitation at 490 nm. The liposomes were destroyed by adding 100 µl of a 10% Triton solution and the total fluorescence intensity (Itotal) was measured. The percentage released CF is calculated according to the equation (3):

167

170 171 172

(3)

EP

169

% leakage = (It/Itotal) x 100

where It is the fluorescence intensity at time t and е Itotal is the fluorescence intensity evaluated after the destruction of liposomes.

AC C

168

SC

149 150 151 152 153 154 155 156 157

2.4.3. Cryo-Transmission Electron Microscopy (Cryo-TEM)

M AN U

148

where η is the solvent viscosity, υ is the electrophoretic mobility, and ε is the dielectric constant of the solvent.

TE D

146 147

173

2.4.5. Cytotoxicity assessment (MTT-dye reduction assay)

174 175 176

The cellular viability and proliferation after exposure to varying concentrations of plain or surface modified DPPC:CHOL liposomes was evaluated by using the standard MTT-dye reduction assay [29] with slight modifications [30]. Exponentially growing cells were seeded in 96-well flatbottomed microplates and after 24 h incubation at 37 °C they were exposed to various concentrations of the tested liposomal formulations for 72 h. For each concentration 8 wells were used. After the incubation with the test compounds 10 ml MTT solution (10 mg/ml in PBS) aliquots were added to each well. The microplates were further incubated for 4 h at 37 °C and the MTT-formazan crystals formed were dissolved by adding 100 ml/well of 5%

177 178 179 180 181

5

ACCEPTED MANUSCRIPT HCHO-acidified 2-propanol. The MTT-formazan absorption was determined using a microprocessor controlled microplate reader (LabeximLMR-1) at 580 nm. Cell survival fractions were calculated as percentage of the untreated control. The cell survival data were normalized as percentage of the untreated control (set as 100% viability). The statistical processing of cytotoxicity data included the Student’s t-test with p ≤ 0.05 set as significance level.

188

3. Results and Discussion

189

3.1. General remarks

190 191 192 193 194 195 196 197 198

The polyglycidol-derivatized lipids, selected for the present study, are polymeric products with DP of the polyglycidol chain in the 23 – 110 range (Figure 1). In aqueous solution they were found to self-associate above certain critical aggregation concentration into small corecorona micelles [26]. In the present work we investigate the hybrid structures resulting from the co-assembly of these conjugates with the lipids 1,2-dipalmitoyl-sn-glycerophosphocholine (DPPC) and cholesterol (CHOL). The samples were prepared at total lipid concentration of 3 mM and DDP-(G)n contents in the 2.5 – 10 mol % range. All samples, independently on their composition and temperature, appeared slightly opalescent and remained stable and optically clear for months.

199

3.2. Light scattering characterization

200 201 202 203 204 205

The dynamic light scattering (DLS) study of plain DPPC:CHOL liposomes and the corresponding hybrid structures incorporating the polyglycidol-lipid conjugates showed monomodal size distribution with polydispersity indeces typically ranging from 0.04 to 0.08. Representative size distributions are shown in Fig. 2. The measurements were performed at 25 °C. The increase of temperature to the physiologically relevant level did not induce any significant alterations of the size distribution patterns or liposomal size.

206 207 208 209 210 211 212 213 214 215 216 217 218 219

The resulting hybrid structures spanned a diameter range from 113 to 134 nm. Representative variations of the hydrodynamic dimensions with copolymer composition and content in the hybrid structures are depicted in Fig. 3. The curve patterns can be broadly divided into two types depending on the length (DP) of the polyglycidol chain. The copolymers DDP-(G)23 and DDP-(G)30 did not noticeably influence the size of the hybrid structures (Fig. 3a and b). In the case of the copolymers with the longest polyglycidol chain – DDP-(G)54 and DDP-(G)110 (Fig. 3c and d), the size of the liposomes was found to gradually increase up to copolymer content of 7.5 and 5.0 mol % for DDP-(G)54 (Fig. 3c) and DDP-(G)110 (Fig. 3d), respectively, due to formation of a thick polymer layer outside the liposome membrane. After that a size reduction was observed, which can be associated with rearrangement to vesicles of higher curvature (smaller size) or formation of smaller structures such as bilayer disks or spherical micelles. Both would give relaxation of the energy stored in the polyglycidol layer around the liposomes, which is expressed in the lateral pressure between the chains. This energy is proportional to the molar mass and grafting density [31]. In that aspect the shift of the events

AC C

EP

TE D

M AN U

SC

RI PT

182 183 184 185 186 187

6

ACCEPTED MANUSCRIPT leading to particle size decrease or rearrangement at lower DDP-(G)110 content compared to DDP-(G)54 is reasonable.

222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240

ζ-potential measurements were performed on the plain and hybrid liposomes and structures in order to extract information on the effective charge of the particles at varying copolymer contents. Data from the electrophoretic light scattering are presented in Figure 4. DPPC is a zwitterionic lipid with zero net molecular charge. When assembled in bilayers, due to the specific spatial orientation of the dipole lipid headgroups, the surface becomes charged and the orientation, hence, the charge magnitude, strongly depends on the ionic strength, temperature, presence of cholesterol or impurities [32,33]. The presence of cholesterol in the liposomal membrane and the hydrogen-bond interactions between the cholesterol OH group and the dipole lipid heads lead to shielding of the positive charge and push the negative phosphatidyl group out from the surface of the bilayer [33]. Therefore, the ζ-potential of the plain DPPC:CHOL liposomes is slightly negative. Upon the incorporation of any of the polyglycidol-lipid conjugates, even at contents as low as 2.5 mol %, an increase in the ζpotential (that is, less negative and slightly positive values) was observed (Fig. 4). These findings can be explained in terms of screening effects of the polymer chains on electric charges or change in the hydration state of the phospholipid bilayer due to the presence of a hydrophilic polymer [34-36]. Although the absolute values of the ζ-potentials are small, no precipitation occurred and, as noted above, the dispersions remained stable for many months. The colloidal stability was attributed to the formation of a polyglycidol layer with repulsive properties around the liposomes.

241

3.3. Aggregate structure and morphology

242 243 244 245 246 247 248 249 250 251 252 253

Membrane destabilization of liposomes has been observed upon incorporation of PEG-lipids above a certain critical content, which depends on the lipid composition and PEG molecular weight. At high contents of polymer-lipid, the interactions between the polymer chains give rise to a considerable lateral pressure. When the latter exceeds the bilayer cohesion, it becomes energetically more favorable to form lipid/polymer-lipid mixed micelles. In case of cholesterol supplemented liposomes or liposomes composed of lipids in the gel-phase state, the transition from the lamellar to the micellar phase proceeds via intermediate states comprising formation of bilayer disks [8,37]. The lipid/PEG-lipid ratio at which these structures appear sets the upper limit of the amount of polymer-lipid that can be incorporated without affecting the liposomes’ integrity and their function as encapsulation carriers. In this section we use cryo-TEM to investigate the structure and morphology of hybrid DPPC:CHOL:DDP-(G)n aggregates.

254 255 256 257 258 259

It is well documented that the plain DPPC:CHOL liposomes are predominantly spherical and unilamellar [38]. However, co-existing fractions of elongated and multilamellar liposomes have been observed as well. In the cryo-TEM images (not shown), they appeared not wellseparated, which indicated the lack of steric stabilization. Representative micrographs of liposomes based on DPPC:CHOL supplemented with increasing amounts of DDP-(G)n are shown in Fig. 5. Based on the results from DLS (see Fig. 3), our expectations were that the

AC C

EP

TE D

M AN U

SC

RI PT

220 221

7

ACCEPTED MANUSCRIPT effects of the conjugates with intermediate polyglycidol chain length on the morphology and morphological transitions of DPPC:CHOL liposomes would be marginal. Perhaps, the most visible effect is that the liposomes appeared well separated implying formation of a protective polyglycidol layer around liposomes. This layer is invisible for the electron beam but when such polymer-covered liposomal surfaces approach each other, they experience a repulsive force due to unfavorable entropy (loss of conformational freedom) associated with compressing the polymer chains between the two surfaces [39]. The liposomes with grafted DDP-(G)23 or DDP-(G)30 at contents of up to 5 mol % are undoubtedly intact (Fig. 5a). Their size and shape are very similar to what is observed in the absence of polyglycidol-derivatized lipid. Upon increasing contents of the latter, liposomes clearly dominate the samples but some disks are occasionally observed. At 5 mol % of DDP-(G)54, similarly to the previous formulations, the liposomes are intact and well separated. The images are clearly dominated by liposomes but a few disks can be seen (less frequent than in 7.5 mol % of DDP-(G)30). At content of 7.5 mol % of DDP-(G)54, however, the liposome integrity was obviously affected: liposomes in co-existence with a very significant population of large disks were observed. The latter were observable mostly face-on on the background of the images (Fig. 5b). The situation is qualitatively similar for the formulations with DDP-(G)110, in which the number of intact liposomes decreases with increasing total content of this conjugate, while the disks adapt smaller size. Obviously, both contribute to the drop in the average dimensions of the particles detected in the light scattering experiments (see Fig. 3c and d ).

280

3.4. Leakage assay

281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301

The basic function of liposomes comprises retaining of the encapsulated drug/active substance en route to its target. The liposomes are of little use if the encapsulated substance is released too rapidly. The generally accepted mechanism of leakage of hydrophilic solutes is via defects and temporary openings (pores) in the membrane. The introduction of other components in the bilayer may influence both the frequency of pore formation and lifetime of the pores, which would result in effects of increasing or decreasing the membrane permeability. A leakage assay using CF was carried out in order to evaluate the effects of the present DDP(G)n polymers on membrane permeability of DPPC:CHOL liposomes. The measurements were performed at 37 °C in a phosphate buffer solution for a period of 6 hours. The results are presented in Fig. 6. Generally, all polymers were found to decrease the membrane permeability: somewhat deviating behavior was observed for DDP-(G)23 and DDP-(G)54 at 5 mol % content (Fig. 6a) but even in these cases, the leakage just slightly increased compared to the plain DPPC:CHOL liposomes. Considering the different hydrocarbon chain lengths of DPPC and the lipid-mimetic anchor of DDP-(G)n, it was expected that the latter would introduce some packing disorder in the membrane resulting in deterioration of permeability. Furthermore, molecules with large hydrophilic volume have been shown to facilitate formation of pores or other membrane imperfections [40]. Instead, the incorporation of the investigated conjugates generally decreased the CF release with the exceptions noted above. On the other hand, reduction of membrane permeability upon incorporation of PEG-lipids and other related copolymers has been observed earlier [16,38,41,42] and interpreted in terms of reducing the probability of pore formation [16]. In addition, pore formation can be reduced in

AC C

EP

TE D

M AN U

SC

RI PT

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

8

ACCEPTED MANUSCRIPT the presence of membrane building elements capable of forming hydrogen bonds in the head group region. Such results have been reported elsewhere [43-45]. It is difficult to quantify the contribution of those oppositely acting factors to the overall membrane permeability. Nevertheless, polyglycidol with its numerous hydroxyl groups is known to promote strong hydrogen bonding: hydrophobic interactions and interactions via hydrogen bonding were equally involved in formation of well-separated spherical nanoparticles composed of discontinuous hydrophobic domains of poly(propylene oxide) randomly distributed in a strongly hydrogen-bonded continuous medium consisting of polyglycidol and water [46-50]. Likewise, we can assume formation of a strongly hydrogen bonded layer around liposomes, which may contribute to the reduction of membrane permeability. To prove this assumption leakage experiments in the presence of urea were carried out (Fig. 7).

313 314 315 316 317 318 319 320 321

Urea is known to disrupt hydrogen bonds in a variety of systems and processes, for example, in surfactant and polymer solutions, polymer hydrogels, denaturation of proteins and DNA, etc. Indeed, the CF release profiles in Fig. 7 look different from those in the absence of urea in Fig. 6. The striking difference is that the leakage curves of the liposomes in which DDP-(G)23 and DDP-(G)54 are incorporated are invariably situated above the curve of the plain DPPC:CHOL liposomes indicating enhancement of the leakage. Obviously, the role of the strongly hydrogen bonded layer that is formed around the liposomes upon incorporation of the DDP-(G)n polymers is substantial since its effects in decreasing the leakage predominate over the effects leading to increasing membrane permeability.

322 323 324 325 326 327 328 329 330

More information can be extracted from the leakage assay by re-plotting the data in Fig. 6 as CF leakage versus polymer content (Fig. 8). The general trend is for a gradual decrease of CF leakage with increasing polymer content with somewhat erratic behavior shown by DDP(G)30. This counterintuitive finding further supports the assumption for formation of a strongly hydrogen bonded layer around the liposomes. Obviously, the increasing grafting density of polyglycidol chains on the liposomal surface and, hence, formation of a dense network of hydrogen bonds, is associated with further hindering of the leakage of the dye despite the destabilization effects that are expected upon the incorporation of such large amounts of polymer in the bilayer.

331 332 333 334 335 336 337 338 339 340 341

To further assess the extent to which the membrane permeability is affected by the incorporation of the tested copolymers DDP-(G)n a leakage assay in 50 % human plasma was performed. The obtained CF release profiles are shown on Fig. 9. Generally, the fluorescence intensity in plasma is considerably lower than in buffer (cf. Figs. 6 and 7) due to quenching of CF fluorescence as a result of Coulombic interactions with plasma proteins [51,52]. As evident from the presented data, the ability of hybrid liposomes to retain the fluorescent marker was superior: the release was less than ca. 27% at the 360th min, while the plain liposomes were found to release more than 35 % of the encapsulated dye for the same time period. Similarly to the release profiles obtained in phosphate buffer, the liposomes stabilized with 10 mol% DDP-(G)54 exhibited lower CF leakage in plasma compared to those modified with 5 mol% of the same copolymer. This finding implied that the presumably denser and

AC C

EP

TE D

M AN U

SC

RI PT

302 303 304 305 306 307 308 309 310 311 312

9

ACCEPTED MANUSCRIPT 342 343 344

more strongly hydrogen bonded polyglycidol layer of the formulation containing 10 mol % of copolymer hindered the detrimental interactions with plasma proteins to a larger extent. 3.5 Cytotoxic study The lack of toxicity is an important requirement for the excipients used in design of novel nanosized drug delivery systems. Considering the proposed utility of the tested copolymers for steric stabilization of liposomes, we performed a cytotoxicity assessment of DPPC:CHOL liposomes either plain or surface-modified with 10 mol % of copolymers on two human cell lines with different cell type and origin, namely OPM-2 (myeloid) and EJ (epithelial).

350 351 352 353 354 355 356

The cell lines were selected as representative models of cellular populations with which the intravenously applied nanosized drug delivery systems will interact. In general, as evident from the concentration-response curves depicted in Fig. 10, all tested liposomal formulations failed to induce any significant decrease in cell viability. The surface modified liposomes were slightly more inhibitory as compared to the plain formulations, but nevertheless, even for the highest tested concentration of 100 µM only slight (less than 15 %) reduction in cell viability was observed as compared to the untreated control.

357

4. Conclusion

358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379

The main conclusion from the present study is that the original polyglycidol-derivatized lipids are able to form hybrid nanostructures with phospholipids such as DPPC. They provide steric stability and possibly Stealth properties of the resulting liposomes by intercalating in the bilayer via their lipid-mimetic residue, whereas the polyglycidol chains create a thin repulsive layer around the membrane. The conjugates with DP of the polyglycidol chain of 23 and 30 can be incorporated in the bilayer membranes at quantities as high as 10 mol % without compromising the liposome dimensions, morphology and integrity. Those of the highest polyglycidol chain length (DP of 54 and 110), however, affect the liposome integrity as coexisting fractions of disks are detected at contents of 7.5 and 5.0 mol %, respectively. A very important structural feature of polyglycidol is the presence of numerous hydroxyl groups, that are able to promote strong hydrogen bonding in the polymer layer around liposomes, which is considered as the main factor that contribute to the reduction of membrane permeability. Apparently, the dense network of strongly hydrogen bonded polyglycidol chains stabilizes the membrane against formation of defects and pores through which the entrapped dye is released. Furthermore, the enhanced stability in blood plasma strongly indicates that the polyglycidol-derivatized lipids confer resistance to interactions with leakage promoting components present in plasma to the hybrid structures. The cytotoxicity study clearly evidences devoid of cytotoxic effects towards cellular populations that are to be exposed to the intravenously applied liposomes. The low toxicological potential together with the favorable physicochemical characteristics of the tested liposomes give the reason to consider the hybrid liposomes surface modified by polyglycidol-derivatized lipids as promising platforms for drug delivery.

AC C

EP

TE D

M AN U

SC

RI PT

345 346 347 348 349

380 10

ACCEPTED MANUSCRIPT Acknowledgement

382 383 384 385 386 387 388 389

This work was supported by the National Science Fund (Bulgaria) project Б 01-25/2012. The paper benefitted from discussions at events of Precision Polymer Materials (P2M) Research Networking Program of the European Science Foundation and from the European Commission project POLINNOVA (Grant Agreement Number 316086). P. Bakardzhiev expresses his gratitude to BG051PO001-3.3.06–0017 project funded within the Operational Program Human Resources Development. We are grateful to Dr. Jonny Eriksson for performing the cryo-TEM analysis. K. Edwards gratefully acknowledges financial support from the Swedish Research Council and the Swedish Cancer Society.

RI PT

381

390

SC

391

References

393 394 395

[1] N. Pippa, E. Kaditi, S. Pispas, C. Demetzos, PEO-b-PCL–DPPC chimeric nanocarriers: selfassembly aspects in aqueous and biological media and drug incorporation, Soft Matter 9 (2013) 40734082

396 397 398

[2] N. Pippa, M. Merkouraki, S. Pispas, C. Demetzos, DPPC:MPOx chimeric advanced Drug Delivery nano Systems (chi-aDDnSs): Physicochemical and structural characterization, stability and drug release studies, Int. J. Pharm. 450 (2013) 1-10

399 400

[3] N. Pippa, F. Psarommati, S. Pispas, C. Demetzos, The shape/morphology balance: a study of stealth liposomes via fractal analysis and drug encapsulation, Pharm. Res. 30 (2013) 2385-2395

401 402

[4] S. Rangelov, S. Pispas, Polymer and Polymer-Hybrid Nanoparticles: From Synthesis to Biomedical Applications. CRC Press. Taylor and Francis Group. Boca Raton, FL, 2014.

403 404

[5] M. C. Woodle, D. Lasic, Sterically stabilized liposomes, Biochem. Biophys. Acta 1113 (1992) 171-199.

405 406

[6] M. C. Woodle, Sterically stabilized liposome therapeutics, Adv. Drug Delivery Rev. 16 (1995) 249-265.

407 408 409

[7] K. Hristova, A. Kenworthy, T. J. McIntosh, Effect of bilayer composition on the phase behavior of liposomal suspensions containing poly(ethylene glycol)-lipids, Macromolecules 28 (1995) 7693–7699.

410 411 412

[8] K. Edwards, M. Johnsson, G. Karlsson, M. Silvander, Effect of polyethyleneglycolphospholipids on aggregate structure in preparations of small unilamellar liposomes, Biophys. J.73 (1997) 258-266.

413 414 415

[9] S. Belsito, R. Bartucci, G. Montesano, D. Marsh, L. Sportelli, Molecular and mesoscopic

AC C

EP

TE D

M AN U

392

properties of hydrophilic polymer-grafted phospholipids mixed with phosphatidylcholine in aqueous dispersion: Interaction of dipalmitoyl N-poly(ethylene 11

ACCEPTED MANUSCRIPT glycol)phosphatidylethanolamine with dipalmitoylphosphatidylcholine studied by spectrophotometry and spin- label electron spin resonance, Biophys. J. 78 (2000) 1420-1430.

418 419

[10] D. Marsh, R. Bartucci, L. Sportelli, Lipid membranes with grafted polymers: physicochemical aspects, Biochim. Biophys. Acta 1615 (2003) 35–59.

420 421

[11] D. Kashchiev, D. Exerowa, Bilayer lipid membrane permeation and rupture due to hole formation, Biochim. Biophys. Acta 732 (1983) 133-145.

422 423

[12] T. Ishida, R. Maeda, M. Ichihara, K. Irimura, H. Kiwada, Accelerated clearance of PEGylated liposomes in rats after repeated injections, J. Control. Release 88 (2003) 35–42.

424 425

[13] T. Ishida, K. Masuda, T. Ichikawa, M. Ichihara, K. Irimura, H. Kiwada, Accelerated clearance of a second injection of PEGylated liposomes in mice, Int. J. Pharm. 255 (2003) 167–174.

426 427 428

[14] R. Saadati, S. Dadashzadeh, Z. Abbasian, H. Soleimanjahi, Accelerated bloodclearance of

429 430 431 432

[15] S. M. Moghimi, I. Hamad, T. L. Andresen, K. Jorgensen, J. Szebeni, Methylation of the

433 434 435

[16] M. Silvander, M., M. Johnsson, K. Edwards, Effects of PEG-lipids on permeability of phosphatidylcholine/cholesterol liposomes in buffer and in human serum, Chem. Phys. Lipids 97 (1998) 15-26.

436 437 438

[17] R. F. Flewelling, W. L. Hubbell, Hydrophobic ion interactions with membranes. Thermodynamic analysis of tetraphenylphosphonium binding to vesicles, Biophys. J. 49 (1986) 531-540.

439 440 441

[18] T. R. Baekmark, T. Wiesenthal, P. Kuhn, T. M. Bayerl, O. Nuyken, R. Merkel, New insights

442 443 444

[19] T. R. Baekmark, T. Wiesenthal, P. Kuhn, A. Albersdörfer, O. Nuyken, R. Merkel, Systematic

445 446 447

[20] T. Wiesenthal, T. R. Baekmark, R. Merkel, Direct evidence for a lipid alkyl chain ordering transition in poly(ethylene oxide) lipopolymer monolayers at the air-water interface obtained from infrared reflection absorption spectroscopy, Langmuir 15 (1999) 6837-6844.

448 449 450

[21] H. Ahrens, T. R.

SC

RI PT

416 417

M AN U

PEGylated PLGA nanoparticles following repeated injections: effects of polymer dose, PEG coating, and encapsulated anticancer drug, Pharm. Res. 30 (2013) 985–995.

phosphate oxygen moiety of phospholipid- methoxy(polyethylene glycol) conjugate prevents PEGylated liposome-mediated complement activation and anaphylatoxin production, FASEB

EP

TE D

J. 20 (2006) 2591–2593.

AC C

into the phase behavior of lipopolymer monolayers at the air/water interface. IRRAS study of a polyoxazoline lipopolyme, Langmuir 13 (1997) 5521-5523.

infrared reflection-absorption spectroscopy and film balance study of the phase behavior of lipopolymer monolayers at the air-water interface, Langmuir 15 (1999) 3616-3626.

Baekmark, R. Merkel, J. Schmitt, K. Graf, R. Raiteri, C. A. Helm, Hydrophilic/hydrophobic nanostripes in lipopolymer monolayers, ChemPhysChem. 2 (2000) 101–106. 12

ACCEPTED MANUSCRIPT [22] D. N. T. Hay, P. G. Rickert, S. Seifert, M. A. Firestone, Thermoresponsive Nanostructures by Self-Assembly of a Poly(N-isopropylacrylamide)-Lipid Conjugate, J. Am. Chem. Soc. 126 (2004) 2290–2291

454 455 456

[23] N. Pippa, A. Meristoudi, S. Pispasb, C. Demetzos, Temperature-dependent drug release from DPPC:C12H25-PNIPAM-COOH liposomes: Control of the drug loading/release by modulation of the nanocarriers’ components, Int. J. Pharm. 485 (2015) 374-382

457 458

[24] K. Maruyama, S. Okuizumi, O. Ishida, H. Yamauchi, H. Kikuchi, M. Iwatsuru, Phosphatidyl polyglycerols prolong liposome circulation in vivo, Int. J. Pharm. 111 (1994) 103-107.

459 460 461 462

[25] A. S. Abu Lila, K. Nawata, T. Shimizu, T. Ishida, H. Kiwada, Use of polyglycerol (PG),

instead of polyethylene glycol (PEG), prevents induction of the accelerated blood clearance phenomenon against long-circulating liposomes upon repeated administration, Int. J. Pharm.

463 464 465

[26] P. Bakardzhiev, S. Rangelov, B. Trzebicka, D. Momekova, G. Lalev, V. M. Garamus, Polyglycidol-Derivatized Lipids. Synthesis and Self-Assembly in Aqueous Solution, RSC Advances 70 (2014) 37208-37219.

466 467

[27] R. K. Kainthan, J. Janzen, E. Levin, D. V. Devine, D. E. Brooks, Biocompatibility testing of branched and linear polyglycidol, Biomacromolecules 7 (2006) 703–709.

468 469

[28] M. Almgren, K. Edwards, G. Karlsson, Cryo transmission electron microscopy of liposomes and related structures, Colloids Surf A: Physicochem. Eng. Aspects 174 (2000) 3-21.

470 471

[29] T. Mosmann, Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays, J. Immunol. Methods 65 (1983) 55–63.

472 473

[30] S. M. Konstantinov, H. Elbl, M. R. Berger, BCR-ABL influences the antileukaemic efficacy of alkylphosphocholines, Br. J. Haematol. 107 (1999) 365–374.

474 475

[31] S. Milner, T. Witten, M. Cates, Theory of the grafted polymer brush, Macromolecules, 21 (1988) 2610-2619.

476 477 478

[32] K. Makino, T. Yamada, M. Kimura, T. Oka, H. Ohshima, T. Kondo, Temperature- and ionic

strength-induced conformational changes in the lipid head group region of liposomes as suggested by zeta potential data, Biophysical Chemistry, 41 (1991) 175-183.

479 480

[33] C. Chen, C. P. Tripp, A Comparison of the Behavior of Cholesterol, 7-Dehydrocholesterol and Ergosterol in Phospholipid Membranes, Biochim. Biophys. Acta, 1818 (2012) 1673-1681.

481 482 483

[34] A. Yoshida, K. Hashizaki, H. Yamauchi, H. Sakai, S. Yokoyama, W. Abe, Effect of lipid with

covalently attached poly(ethylene glycol) on the surface properties of liposomal bilayer membrane, Langmuir, 15 (1999) 2333-2337.

484

[35] G. Cevc, Membrane electrostatics, Biochim. Biophys. Acta, 1031 (1990) 311-382.

RI PT

451 452 453

AC C

EP

TE D

M AN U

SC

456 (2013) 235-242.

13

ACCEPTED MANUSCRIPT [36] D. Momekova, S. Rangelov, S. Yanev, E. Nikolova, S. Konstantinov, B. Romberg, G. Storm, N. Lambov, Long-circulating, pH-sensitive liposomes sterically stabilized by copolymers bearing short blocks of lipid-mimetic units, European J. Pharm. Sci. 32 (2007) 308-317.

488 489 490

[37] M. Johnsson, K. Edwards, Liposomes, disks, and spherical micelles: aggregate structure in mixtures of gel phase phosphatidylcholines and poly(ethylene glycol)-phospholipids, Biophys. J. 85 2003 3839-3847.

491 492 493 494

[38] D. Momekova, G. Momekov, S. Rangelov, G. Storm, N. Lambov, Physicochemical and

495 496

[39] J. Israelachvili, Intermolecular and surface forces, 2nd ed. Academic Press, London; San Diego, 1992.

497 498 499

[40] A. R. Nicholas, M. J. Scott, N. I. Kennedy, M. N. Jones, Effect of grafted polyethylene glycol (PEG) on the size, encapsulation efficiency and permeability of vesicles, Biochim. Biophys. Acta 1463 (2000) 167-178.

500 501 502

[41] S. Rangelov, K. Edwards, M. Almgren, G. Karlsson, Steric stabilization of egg-

RI PT

485 486 487

biopharmaceutical characterization of dipalmitoyl phosphatidylcholine liposomes sterically stabilized by copolymers bearing short blocks of lipid-mimetic units, Soft Matter 6 (2010) 591-

M AN U

SC

601.

phosphatidylcholine liposomes by copolymers bearing short blocks of lipid-mimetic units, Langmuir 19 (2003) 172-181.

[42] D. Momekova, S. Rangelov, N. Lambov, G. Karlsson, M. Almgren, Effects of amphiphilic

copolymers bearing short blocks of lipid-mimetic units on the membrane properties and morphology of DSPC liposomes, J. Dispersion Science Technology 29 (2008) 1106-1113.

506 507

[43] C. M. Gupta, A. Bali, S. Dhawan, Modification of phospholipid structure results in greater stability of liposomes in serum, Biochem. Biophys. Acta 648 (1981) 192-198.

508 509 510

[44] M. Jansen, A. Blume, A comparative study of diffusive and osmotic water permeation

EP

TE D

503 504 505

across bilayers composed of phospholipids with different head groups and fatty acyl chains, Biophys. J. 68 (1995) 997-1008.

[45] M. Silvander, N. Bergstrand, K. Edwards, Linkage Identity is a Major Factor in Determing

the Effect of PEG-ylated Surfactants on Permeability of Phosphatidylcholine Liposomes,

514

[46] S. Rangelov, M. Almgren, S. Halacheva, Ch. Tsvetanov, Polyglycidol-based analogues to

515 516

Pluronic copolymers. Light scattering and cryogenic transmission electron microscopy studies, J. Phys. Chem. C 111 (2007) 13185-13191.

517 518 519

[47] S. Rangelov, S. Halacheva, V. Garamus, M. Almgren, Structural polymorphism exhibited by polyglycidol-based analogues to Pluronic copolymers in aqueous solution, Macromolecules 41 (2008) 8885–8894.

AC C

511 512 513

Chem Phys. Lipids 126 (2003) 77-83.

14

ACCEPTED MANUSCRIPT 520 521 522

[48] S. Halacheva, S. Rangelov, V. Garamus, Structure and Interactions in Large Compound

523 524 525

[49] S. Halacheva, S. Rangelov, Ch. Tsvetanov, Synthesis of Polyglycidol-Based Analogues to

532 533 534

RI PT

[50] S. Halacheva, S. Rangelov, Ch. Tsvetanov, V. M. Garamus, Aqueous Solution Properties of

Polyglycidol-based Analogues of Pluronic Copolymers. Influence of the Poly(propylene oxide) Block Molar Mass, Macromolecules 43 (2010) 772-781. [51] J. H. Rockey, W. Li, J. F. Eccleston, Binding of fluorescein and carboxyfluorescein by

human serum proteins: Significance of kinetic and equilibrium parameters of association in ocular fluorometric studies, Experimental Eye Research 37 (1983) 455-466.

SC

529 530 531

Pluronic L121 – F127 Copolymers. Self-Assembly, Thermodynamics, Turbidimetric and Rheological Studies, Macromolecules 41 (2008) 7699-7705.

[52] I. M. Vlasova, E. M. Bukharova, A. A. Kuleshova, A. M. Saletsky, Spectroscopic

investigations of interaction of fluorescent nanomarkers of fluorescein family with human serum albumin at different values of pH, Current Applied Physics 11 (2011) 1126-1132.

M AN U

526 527 528

Particles Formed by Polyglycidol-Based Analogues to Pluronic Block Copolymers in Aqueous Solution, Macromolecules 40 (2007) 8015-8021.

535

536

538 539

Figure captions

TE D

537

Table 1 Composition and nominal molar masses of the investigated polyglycidol-derivatized lipids.

541 542

Fig. 1 Chemical structure of the investigated polyglycidol-derivatized lipids, DDP-(G)n. DDP: 1,3didodecyl/tetradecyloxy-propane-2-ol; G: glycidol monomeric unit; n = 23 – 110.

543

Fig. 2 Size distribution of DPPC:CHOL liposomes containing 5 mol % of DDP-(G)30.

544 545 546

Fig. 3 Variations of the hydrodynamic dimensions (mean diameter in nm) of hybrid DPPC:CHOL liposomes containing from 0 to 10 mol % of DDP-(G)23 (a), DDP-(G)30 (b), DDP-(G)54 (c) and DDP(G)110 (d).

547 548

Fig. 4 Values of ζ-potential of DPPC:CHOL liposomes grafted with increasing amounts of DDP-(G)n polymers.

549 550 551

Fig. 5 Cryo-TEM images of DPPC:CHOL (2:1 molar ratio) liposomes grafted with 5 mol % of DDP(G)30 (a) and 7.5 mol % of DDP-(G)54 (b). White arrows and black arrows show intact liposomes and disks in face-on positions, respectively. Scale bar is 100 nm.

AC C

EP

540

15

ACCEPTED MANUSCRIPT Fig. 6 Leakage of CF from DPPC:CHOL liposomes stabilized by 5 (a), 7.5 (b), and 10 (c) mol % of DDP-(G)n polymers. Measurements were performed at 37 °C in a phosphate buffer solution at pH 7.4. Data points represent the mean values of at least four independent measurements.

555 556 557

Fig. 7 Leakage of CF from DPPC:CHOL liposomes stabilized by 10 mol % of DDP-(G)23 and DDP(G)54. Measurements were performed at 37 °C in a phosphate buffer solution at pH 7.4 containing 0.5 M urea. Data points represent the mean values of at least four independent measurements.

558

Fig. 8 Fraction of the released CF at t=360 min as a function of polymer type and content.

559 560 561

Fig. 9 Leakage of CF from plain DPPC:CHOL liposomes and DPPC:CHOL liposomes stabilized with 5 and 10 mol % of DDP-(G)54 as a function of time at 37°C in 50 % human plasma. Data points represent the mean values of at least four independent measurements.

562 563 564 565

Fig. 10 Cytotoxicity of DPPC:CHOL liposomes, plain or coated with 10 mol% DDP-(G)30 or DDP(G)110, against EJ (a) and OPM (b) cell lines, as determined by the MTT-dye reduction assay after 72 hours continuous exposure. Each data point represents the arithmetic mean  SD of 8 separate experiments.

SC

RI PT

552 553 554

M AN U

566

567

568

EP

571

AC C

570

TE D

569

16

ACCEPTED MANUSCRIPT Nominal molar mass (g.mol-1)

DDP-(G)23

2130

DDP-(G)30

2650

DDP-(G)54

4420

DDP-(G)110

8570

AC C

EP

TE D

M AN U

SC

RI PT

Composition

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT