- Email: [email protected]
Controllable release from high-transition temperature magnetoliposomes by low-level magnetic stimulation
Accepted Manuscript Title: Controllable release from non-thermal sensitive magnetoliposomes by low-level magnetic stimulation Author: Romina Spera Francesca Apollonio Micaela Liberti Alessandra Paffi Caterina Merla Rosanna Pinto Stefania Petralito PII: DOI: Reference:
S0927-7765(15)00244-1 http://dx.doi.org/doi:10.1016/j.colsurfb.2015.04.030 COLSUB 7034
To appear in:
Colloids and Surfaces B: Biointerfaces
Received date: Revised date: Accepted date:
5-12-2014 16-3-2015 10-4-2015
Please cite this article as: R. Spera, F. Apollonio, M. Liberti, A. Paffi, C. Merla, R. Pinto, S. Petralito, Controllable release from non-thermal sensitive magnetoliposomes by low-level magnetic stimulation, Colloids and Surfaces B: Biointerfaces (2015), http://dx.doi.org/10.1016/j.colsurfb.2015.04.030 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.
Short Communication
Controllable release from non-thermal sensitive magnetoliposomes by low-level magnetic stimulation
ip t
Romina Spera1, Francesca Apollonio2, Micaela Liberti2, Alessandra Paffi2, Caterina Merla3, Rosanna Pinto3, Stefania Petralito1*
Department of Drug Chemistry and Technology, “Sapienza” University of Rome - P.le Aldo Moro 5, 00185 Rome Italy 2 Department of Information Engineering, Electronics and Telecommunications, “Sapienza” University of Rome - Via Eudossiana 18, 00184 Rome Italy 3 Technical Unit of Radiation Biology and Human Health, Enea Research Center - Via Anguillarese 301, 00123 Rome, Italy
us
cr
1
an
*Correspondence: Stefania Petralito, PhD, Department Of Drug Chemistry and Technology“Sapienza”
Ac ce p
te
d
M
Università di Roma-P.le Aldo Moro 5, 00185 Roma, Italy; phone: +39 0649693265, Fax: +39 0649913133 ; Email: [email protected]
Page 1 of 13
us
cr
ip t
Graphical abstract
an
1.1 Introduction
Ac ce p
te
d
M
Stimuli-response release of a cargo from drug carriers at a specific time and location is one of the most aimed results of drug delivery research. Typically such a control is obtained by changing the environmental conditions (e.g. ultrasound, UV–visible light, temperature, or pH of the bulk medium) [1-5]. Among these, the temperature controlled drug release received special interest due to its noninvasive character [6]. Traditional thermo-sensitive liposomes have been largely studied for this purpose: this kind of vesicles can undergo a solid to fluid phase transition when the temperature rises above the lipid melting point (Tm) leading to an increase in the permeability of membranes and consequently the release of the cargo [7]. Thermo-sensitive magnetoliposomes (MLs) MLs are usually designed to obtain a single burst leakage within a temperature window of tolerable local-regional hyperthermia (41-46°C), but they don’t seem suitable to provide a sustained drug delivery for a prolonged period of time. Recently the temperature-induced control release has been approached using magnetic fields as triggering agent and MLs as drug carriers. Such carriers are hybrid systems of traditional thermo-sensitive liposomes containing superparamagnetic iron oxide nanoparticles (MNPs) and their use as devices for magnetic-controlled delivery of drugs is extensively investigated [8-9]. In particular, when exposed to appropriate magnetic fields (amplitudes of kA/m and frequencies from tens to hundreds of kHz) MNPs generate heat, either from hysteresis losses or from Néel or Brownian relaxa on processes[10]. This magnetically induced heat is transferred to the entire carrier causing structural changes in the bilayer, acting as smart trigger for drug release. However, the thermal-magnetic stimuli can cause damage and safety secondary effects to the surrounding tissues both due to temperature and to magnetically induced eddy currents, limiting their clinical applicability [11]. In spite of some recent promising studies [12], the controllable and reversible release of cargo from MLs by triggering them with magnetic stimulation has not been fully exploited. In particular some recent studies suggested that low magnetic fields may be suitable for drug delivery applications with minimal heating [13-14]. In this work, we present a drug-delivery system based on high-transition temperature liposomes made of hydrated soybean phosphatidylcholine and cholesterol, which enclose coated MNPs in their aqueous core (highTm MLs). The liposome membrane can exist only in the ordered state within the experimental temperature interval and neither spontaneous leakage nor thermal responsiveness can occur up to 50 °C. To probe the membrane permeation and release behavior of high-Tm MLs, 5(6)Page 2 of 13
ip t
carboxyfluorescein was loaded in the high-Tm MLs as hydrophilic model drug. The controllable release of cargo from high-Tm MLs was triggered by on-off low-level magnetic stimulation. The hypothesized release mechanism consists in a mechanical stress on the liposome membrane due to nanoparticles oscillations in its proximity, as similarly postulated in [15-16].It was investigated by means of a numerical model solved with multiphysics simulations that coupled magnetic, mechanical and thermal physical problems. The demonstrated high reproducibility of cycle-to-cycle release induced by the mechanical magnetic trigger suggests the potential of high-Tm MLs for a successful use as “on demand” drug delivery systems.
2. Experimental
cr
2.1. Materials
M
an
us
Hydrogenated soybean phosphatidylcholine (HSPC) Phospholipon 90H from Lipoid GmbH was kindly gifted by AVG Srl. Cholesterol, 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES), 5(6)-carboxyfluorescein (CF), Triton X-100 (TX-100), Sephadex G-50 and hydrochloric acid (HCl) were purchased from Sigma Aldrich. Chloroform was obtained from Merck. Bidistilled water, thiocyanatoiron, 1,2-dichloroethane and ethanol were supplied by Carlo Erba Reagents. Aqueous dispersion of 50 nm carboxymethyl-dextran coated magnetite (Fe3O4) nanoparticles fluidMAG-CMX (MNPs) was obtained from Chemicell GmbH. 2.2. Preparation and Physicochemical Characterization of Liposomes
Ac ce p
te
d
The high-Tm MLs were prepared by the thin lipid film hydration method followed by sequential extrusion as reported in [17]. The thin film of HSPC and cholesterol formed on the flask walls was hydrated with 10 mM HEPES buffer solution (pH=7.4) containing MNPs and 20 mM CF used as a model hydrophilic drug for the release experiments. Control liposomes, without MNPs, have been also prepared (see supplementary information). Repeated extrusion through membrane filters of 0.4 μm and 0.2 μm pore sizes yielded unilamellar liposomes with a narrow size distribution. Following extrusion, the unentrapped dye CF and MNPs was removed by size exclusion chromatography (SEC) on Sephadex G-50 column eluted with HEPES buffer (10 mM ), pH 7.4. All liposome formulations were stored in dark at 4°C and used within 1 week. Hydrodynamic diameter and polydispersity index (PdI) were evaluated by dynamic light scattering (DLS) experiments, whereas ζ-potential was measured by electrophoretic light scattering (ELS) technique (see supplementary information). All measurements were performed with Zetasizer Nano ZS90 (Malvern Instruments Ltd., UK) thermostatically controlled at 25°C. Phospholipid concentration was determined using the phosphorus colorimetric assay [18]. The samples were measured against a blank without substrate, using a double beam UV–Vis spectrophotometer Lambda 25 (Perkin Elmer, USA). The measurements were repeated both before and after SEC purification of extruded samples. MNPs content in MLs was determined through the 8.5% hydrochloric acid assay described by Belikov et al. [19]. The calibration curve was performed with standards solutions of magnetite. The measurements were repeated both before and after SEC purification of extruded MLs. All the data collected were used to calculate magnetite/phospholipid ratio. CF entrapped in the inner aqueous compartment was determined fluorimetrically at 512 nm with excitation at 492 nm after lysis of pre-incubation purified sample with 10% the non-ionic surfactant Triton X-100 by using the spectrofluorometer LS 50B (Perkin Elmer, USA).
Page 3 of 13
Physicochemical characterization of MLs formulations incubated at 4.0 ± 0.5°C was repeated at specific time intervals for at least 30 days in order to test colloidal stability of the magneto vesicles (see supplementary information). 2.3 AMF exposure set-up
cr
ip t
The exposure system used was the same described in details in [14, 20]; briefly it consists of two coaxial magnetic coils of square shape. A high magnetic field homogeneity of 99% is achieved in a volume around the system center where the samples were placed. The coils are connected to a signal generator through a wide band amplifier. (See supplementary information).
us
2.4 Measurement of CF released from liposomes
M
an
To measure the membrane permeation and release behavior of high-Tm MLs, CF, which is a self-quenching and membrane impermeable dye applied as a model hydrophilic drug, was loaded into the high-Tm MLs. In vitro CF release from high-Tm MLs was determined fluorimetrically by monitoring CF fluorescence de-quenching at excitation and emission wavelengths of 492 and 512 nm, respectively. The release was measured both due to an applied low intensity AMF stimulation (20kHz, 60 A/m) and after 3 hrs of continuous heating at selected temperature (25-70 ◦C) in absence of AMF. A water bath was used to control the environmental temperature. See supplementary information).
d
2.5. Differential scanning calorimetry (DSC)
Ac ce p
te
Thermal analysis was performed on HSPC/Chol/MNPs and HSPC/Chol without MNPs mixtures in the same molar ratio used for vesicles preparation. Calorimetric measurements were performed using a DSC131 (Setaram, France) differential scanning calorimeter. At least three heating/cooling cycles in a temperature range from 20 to 70 °C under nitrogen flow (20 ml/min) were performed on all the samples before thermograms were recorded at a rate of 5°C/min. 2.6 Statistical analysis
All experiments were performed at least in triplicate. All statistical analyses were performed with ANOVA one-way using the Bonferroni post-test (Instat software, version 3.0 GraphPAD Software Inc., San Diego, CA) to determine significant differences in the experimental data. P < 0.05 was considered statistically significant.
3.Results and discussion Physicochemical characterization results are shown in Table 1. No differences were observed in size between control liposomes and high-Tm MLs, both arranged in a monomodal distribution with PdI values <0.200. The neutrality of the ζ-potential values approaching to zero is due to the zwitterionic character of HSPC, and for the high-Tm MLs, it suggested the absence of the negatively charged MNPs absorbed onto the external leaflet of the bilayer. Furthermore, the percentage of the lipid molecules found into the high-
Page 4 of 13
Ac ce p
te
d
M
an
us
cr
ip t
Tm MLs is only partially decreased respect to control liposomes, suggesting that the hydration step of liposome preparation was slightly influenced by the presence of MNPs (Table 1). Finally, the possibility to use such high-Tm MLs as smart drug carrier is confirmed by the satisfactory values of CF-loading efficiency. Mean hydrodynamic diameter and PdI observed for at least 30 days confirmed the colloidal stability of the samples, as reported in Fig.1 (A). DSC profile reported in Fig. 1(B) exhibited one endothermic peak at Tm= 56.8°C suggesting that the incorporation of MNPs within high-Tm MLs core didn’t affect the gel-to-liquid crystalline transition of the membrane (see supplementary material). As shown in the Fig. 1(B), CF encapsulated in the high-Tm MLs almost couldn’t be released at lower temperatures (<5%), but it was released sharply when the temperature approaches Tm. The release of CF under AMF exposure has been monitored at 37 °C for 15 h to probe the effect of MNPs on the release properties. During the AMF exposure, the high-Tm MLs solution did not shown an increase in temperature in the bulk solution that was always ΔT<0.5°C, due to the low level AMF exposure. At the same time, high-Tm MLs were tested in sham (without AFM exposure) modality to evaluate the effect on the release profile of the environmental parameters without magnetic induction. Moreover, in order to exclude a possible direct AMF effect on the bilayer, CF release has been monitored in the same conditions also for control liposomes not encapsulating MNPs. As shown in Fig. 2(A), the AMF continuative exposure induced the significant increase in the CF release rate of high-Tm MLs. It demonstrated that the permeability of the bilayer was affected by the embedded MNPs, upon exposure to the AMF. The improved release in AMF cannot be ascertained to local membrane heating generated by MNPs due to magneto-caloric effect because of the high Tm=56.8°C of our MLs while temperature measurements detected an increase below 0.5°C. A local temperature increase seems unfeasible as further confirmed by the multiphysics model and discussed in the following. Conversely, the entrapped MNPs can effectively improve the bilayer permeability as a consequence of their motions in AMF, without affecting the geometrical properties of the liposomes (see supplementary material). The CF release extent was nearly 20% in high-Tm MLs already after 3 h of exposure and remains almost stable up to 15 h. No significant CF release (<5%) was observed for sham exposure and pure AMF exposed liposomes. In order to further prove the effect of AMF on controlled release of cargos from high-Tm MLs, the release of the encapsulated CF was carried out by commutatively switching on (3 h) and off (21 h) the AMF as a function of time. An important feature of the plot in Fig. 2(B) was that CF can be released repetitively from high-Tm MLs. When the AMF was turned on the magneticimpelled motions caused the destabilization of the bilayer improving its permeability, which permitted CF diffusion out of the high-Tm MLs. Thus the CF release from the high-Tm MLs can be explained by the mechanical motions of MNPs leading to the improved bilayer permeability, rather than the destruction of the liposome structure. The repetitious release from high-Tm MLs proved that the liposome structure was retained during AMF treatment. Indeed, high-Tm MLs maintain their geometrical properties unchanged over time (see supplementary material). On–off cycles were repeated at four times in the experiment until 70% of the cargo was successfully detected. This raised the possibility that AMF can be applied to control the repetitious release of high-Tm MLs at specific times, with similar dose during all the triggering steps. Long off periods (21 h) demonstrate the stability of the high-Tm MLs against undesired leakage when AMF is off.
Page 5 of 13
4. Conclusions
Ac ce p
te
d
M
an
us
cr
ip t
In summary, we demonstrated that high-Tm MLs, HSPC/Chol bilayers including in the aqueous core iron oxide nanoparticle MNPs, respond to a low amplitude magnetic field at temperature well below the Tm of the bilayer (Tm=56.8). The high-Tm MLs were demonstrated stable and able to preserve their physical properties over time for at least 30 days. To probe the membrane permeation and release behavior of high-Tm MLs, CF as a model hydrophilic drug was loaded in the high-Tm MLs, with satisfactory values of CF-loading efficiency. The content release from the high-Tm MLs could be triggered repetitively by switching on and off the AMF until almost complete depleting of the carrier. The AMF triggered release properties demonstrated that the cargo release from the high-Tm MLs in AMF was due to reversible and controllable permeability change of the bilayer, rather than the destructure of the high-Tm MLs, that maintain their size over the whole experiment duration. The hypothesized release mechanism is a mechanical stress on the liposome membrane due to nanoparticles oscillations in its proximity. It was investigated by means of a numerical model solved with multiphysics simulations that coupled magnetic, mechanical and thermal physical problems (see supplementary material). Fig.3 reports the magnetic induction (Fig.3 A), the mechanical pressure (Fig.3 B) and the electromagnetic heating (Fig.3 C) of the High-Tm ML model considering the maximum load of two nanoparticles to fully explore a possible thermal mechanism. The model results in a substantial absence of relevant temperature increases due to the AMF (δ T = 0.3 °C after three hours), while evidencing a pressure on the membrane of around tens of bars that is within the threshold of 10-100 bars for mechanical destabilization of bilayers indicated in literature [21]. In principle, mechanically AMF stimulus represents a novel modality to trigger release from magneto carriers and opens the possibility for specific complex drug release regimens. The exposure time of 3 h is longer than those typically used for drug delivery based on thermal sensitive MLs [12-13] but compatible with those of low level magnetic field treatments used in orthopedics for bone and connective tissue regeneration [22,23]. Even though more investigations are necessary, high-Tm MLs can be good candidates for smart nanodevice for an“on-demand” drug release in the presence of a low amplitude AMF.
Acknowledgements
Project depicted is sponsored in part by Sapienza University Research Grant 2013. This activity is performed in the framework of the Joint IIT-Sapienza LAB on Life-NanoScience Project "Novel strategies for the imaging and treatment of brain tumors through targeting cancer stem cellspecific signaling pathways." The authors want to thank Dr Iacopo Zanardi for his help in statistical analyses.
Reference
[1] S. Ganta, H. Devalapally, A. Shahiwala, M. Amiji, A review of stimuli-responsive nanocarriers for drug and gene delivery. J. Control. Release, 126 (2008)187–204 doi: 10.1016/j.jconrel.2007 [2] S. Mura, J. Nicolas, P, Couvreur, Stimuli-responsive nanocarriers for drug delivery. Nature Mater, 12 (2013) 991–1003 doi:10.1038/nmat3776 [3] S. J. Leung and M. Romanowski, Light-Activated Content Release from Liposomes. Theranostics, 2 (2012) 1020–1036. doi: 10.7150/thno.4847
Page 6 of 13
[4] A. Schroeder 1, J. Kost , Y. Barenholz, Ultrasound, liposomes, and drug delivery: principles for using ultrasound to control the release of drugs from liposomes. Chemistry and Physics of Lipids, 162 (2009) 1–16 doi:10.1016/j.chemphyslip.2009.08.003 [5] H. Karanth , RS. Murthy , pH-sensitive liposomes-principle and application in cancer therapy. J. Pharm. Pharmacol., 59 (2007) 469-83 doi: 10.1211/jpp.59.4.0001
ip t
[6] T. Ta, T. M. Porter Thermosensitive liposomes for localized delivery and triggered release of chemotherapy. J. Control. Release, 169 (2013) 112-125 doi:10.1016/j.jconrel.2013.03.036
cr
[7] M. B. Yatvin, J. N. Weinstein, W. H. Dennis, & R. Blumenthal, Design of liposomes for enhanced local release of drugs by hyperthermia. Science, 202 (1978) 1290–1293 doi:10.1126/science.364652
us
[8] P.Pradhan, J. Giri , F. Rieken , C. Koch , O. Mykhaylyk , M- Döblinger , R. Banerjee , D. Bahadur , C. Plank, Targeted temperature sensitive magnetic liposomes for thermochemotherapy. J. Control. Release, 142 (2010) 108–121 doi: 10.1016/j.jconrel.2009.10.002. [9] M.R. Preiss, G.D. Bothun, Stimuli-responsive liposome-nanoparticle assemblies. Expert. Opin Drug Deliv, 8 (2011) 1025-1040 doi: 10.1517/17425247.2011.584868
an
[10] P. Guardia, R. Di Corato , L. Lartigue, C. Wilhelm, A. Espinosa, M. Garcia-Hernandez, F. Gazeau, L. Manna, T. Pellegrino, Water-soluble iron oxide nanocubes with high values of specific absorption rate for cancer cell hyperthermia treatment. ACS Nano 6 (2012) 3080– 3091
M
[11] A. K. Gupta, R. R. Naregalkar, V. D. Vaidya & M. Gupta, Recent advances on surface engineering of magnetic iron oxide nanoparticles and their biomedical applications. Nanomedicine, 2 (2010) 23-39 doi:10.2217/17435889.2.1.23
te
d
[12] D.Qiu and X.An, Controllable release from magnetoliposomes by magnetic stimulation and thermal stimulation. Colloids and Surfaces B: Biointerfaces, 104 (2013) 326–329. doi: 10.1016/j.colsurfb.2012.11.033
Ac ce p
[13] S. Nappini, F. Baldelli Bombelli, M. Bonini, B.Nordèn, and P. Baglioni, Magnetoliposomes for controlled drug release in the presence of low-frequency magnetic field. Soft Matter, 6 (2010) 154–162 doi: 10.1039/B915651H [14] R. Spera, S. Petralito, M. Liberti , C. Merla , G. D'Inzeo , R. Pinto , F. Apollonio, Controlled release from magnetoliposomes aqueous suspensions exposed to a low intensity magnetic field. Bioelectromagnetics, 35 (2014) 309-12. doi: 10.1002/bem.21841 [15] D. Cheng, X. Li, G. Zhang, H. Shi, Morphological effect of oscillating magnetic nanoparticles in killing tumor cells. Nanoscale Research Letters, 9 (2014) 195. doi: 10.1186/1556-276X-9-195 [16] G. Podaru , S. Ogden , A. Baxter , T. Shrestha , S. Ren , P. Thapa , RK. Dani , H. Wang , M.T. Basel , P. Prakash , S.H. Bossmann and V. Chikan, Pulsed Magnetic Field Induced Fast Drug Release from Magneto Liposomes via Ultrasound Generation. J.Phys.Chem.B, 118 (2014) 11715-22. doi: 10.1021/jp5022278. [17] S. Petralito, R. Spera, A. Memoli, G. D'Inzeo, M. Liberti, F. Apollonio Preparation and characterization of lipid vesicles entrapping iron oxide nanoparticles Asia-Pacific Journal of Chemical Engineering, 7 (2012) 335-341, ISSN: 1932-2143. [18] Y. Yoshida, E. Furuya, K. Tagawa, A direct colorimetric method for the determination of phospholipids with dithiocyanatoiron reagent. J. of Biochemistry, 88 (1980) 463–468. [19] V.G. Belikov, A.G. Kuregyan, G.K. Ismailova, Standardization of magnetite. Pharm. Chem. J., 36 (2002) 333–336 doi: 10.1023/A:1020845110683
Page 7 of 13
[20] R. Lodato, C. Merla, R. Pinto, S. Mancini, V. Lopresto, G.A. Lovisolo, Complex magnetic field exposure system for in vitro experiments at intermediate frequencies. Bioelectromagnetics, 34 (2013) 211-219 doi: 10.1002/bem.21758. [21] M. Louhivuoria, H. J. Risselada, Erik van der Giessen, Siewert J. Marrink, Release of content through mechano-sensitive gates in pressurized liposomes, PNAS, 107 (2010) 19856–19860 doi:10.1073/pnas.1001316107
cr
ip t
[22] F. Vincenzi, M. Targa, C. Corciulo, S. Gessi, S. Merighi, S. Setti, R. Cadossi, M.B. Goldring, P.A. Borea, K. Varani, Pulsed electromagnetic fields increased the anti-inflammatory effect of A2A and A3 adenosine receptors in human T/C-28a2 chondrocytes and hFOB 1.19 osteoblasts. PLoS One, 8 (2013) doi: 10.1371/journal.pone.0065561
Ac ce p
te
d
M
an
us
[23] M. De Mattei, K. Varani, F.F. Masieri, A. Pellati, A. Ongaro, M. Fini, R. Cadossi, F. Vincenzi, P. A. Borea, A. Caruso, Adenosine analogs and electromagnetic fields inhibit prostaglandin E2 release in bovine synovial fibroblasts. Osteoarthr. Cartil., 17 (2009) 252-262 10.1016/j.joca.2008.06.002
Page 8 of 13
Fig. 1. Colloidal stability of MLs suspensions stored at ~4 °C over time (A) as determined by measuring changes in the hydrodynamic diameter (squared) and PdI (circles). Temperaturedependent changes (B) of cumulative release of CF (dashed) from MLs and DSC scanning profile of the melting process of HSPC/Chol/MNPs mixture (solid). Values (±S.D.) are the average of 3 independent experiments.
cr
ip t
Fig.2. Effect of AMF exposure (20 kHz, 60 A/m at 37°C) in continuous (A) and ON-OFF (B) modality on the cumulative release of CF from MLs and control liposome. The amount of CF released from MLs over time under isothermal sham exposure is reported for both strategies. Values (±S.D.) are the average of 7 independent experiments. Asterisks indicate significant differences between controls (sham exposure and control liposomes) and MLs (***p <0.001).
Ac ce p
te
d
M
an
us
Fig.3 Magnetic induction (A), mechanical pressure (B) and electromagnetic heating (C) on the model on the High-Tm ML in the maximum load of two.
Page 9 of 13
Table 1. Hydrodynamic diameter (nm), PdI and ζ-potential (mV) values of liposomes and MLs with HSPC 20 mol% Chol. Phospholipid content (%), amount of MNPs entrapped (g/mol HSPC) and CF concentration (µl/mg HSPC) in the final purified samples are also reported. Values (±S.D.) are the average of 3 independent determinations. Sample
Size (nm)
Liposomes MLs
HSPC (%)a
MNPs (g/mol HSPC)
242.7±9.7
0.069±0.004
-6.9±0.1
90.1
-
235.5±6.4
0.148±0.011
-8.7±0.2
78.1
115.03±4.24
CF volume trapped (µl/mg HSPC)
ip t
PdI
ζ-potential (mV)
2.10±0.22 1.47±0.18
Percentage of phospholipid was calculated by (mean concentration after purification/mean concentration before purification) X 100.
Ac ce p
te
d
M
an
us
cr
a
Page 10 of 13
Ac
ce
pt
ed
M
an
us
cr
i
Figure 1
Page 11 of 13
Ac
ce
pt
ed
M
an
us
cr
i
Figure 2
Page 12 of 13
Ac
ce
pt
ed
M
an
us
cr
i
Figure 3
Page 13 of 13
S0927-7765(15)00244-1 http://dx.doi.org/doi:10.1016/j.colsurfb.2015.04.030 COLSUB 7034
To appear in:
Colloids and Surfaces B: Biointerfaces
Received date: Revised date: Accepted date:
5-12-2014 16-3-2015 10-4-2015
Please cite this article as: R. Spera, F. Apollonio, M. Liberti, A. Paffi, C. Merla, R. Pinto, S. Petralito, Controllable release from non-thermal sensitive magnetoliposomes by low-level magnetic stimulation, Colloids and Surfaces B: Biointerfaces (2015), http://dx.doi.org/10.1016/j.colsurfb.2015.04.030 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.
Short Communication
Controllable release from non-thermal sensitive magnetoliposomes by low-level magnetic stimulation
ip t
Romina Spera1, Francesca Apollonio2, Micaela Liberti2, Alessandra Paffi2, Caterina Merla3, Rosanna Pinto3, Stefania Petralito1*
Department of Drug Chemistry and Technology, “Sapienza” University of Rome - P.le Aldo Moro 5, 00185 Rome Italy 2 Department of Information Engineering, Electronics and Telecommunications, “Sapienza” University of Rome - Via Eudossiana 18, 00184 Rome Italy 3 Technical Unit of Radiation Biology and Human Health, Enea Research Center - Via Anguillarese 301, 00123 Rome, Italy
us
cr
1
an
*Correspondence: Stefania Petralito, PhD, Department Of Drug Chemistry and Technology“Sapienza”
Ac ce p
te
d
M
Università di Roma-P.le Aldo Moro 5, 00185 Roma, Italy; phone: +39 0649693265, Fax: +39 0649913133 ; Email: [email protected]
Page 1 of 13
us
cr
ip t
Graphical abstract
an
1.1 Introduction
Ac ce p
te
d
M
Stimuli-response release of a cargo from drug carriers at a specific time and location is one of the most aimed results of drug delivery research. Typically such a control is obtained by changing the environmental conditions (e.g. ultrasound, UV–visible light, temperature, or pH of the bulk medium) [1-5]. Among these, the temperature controlled drug release received special interest due to its noninvasive character [6]. Traditional thermo-sensitive liposomes have been largely studied for this purpose: this kind of vesicles can undergo a solid to fluid phase transition when the temperature rises above the lipid melting point (Tm) leading to an increase in the permeability of membranes and consequently the release of the cargo [7]. Thermo-sensitive magnetoliposomes (MLs) MLs are usually designed to obtain a single burst leakage within a temperature window of tolerable local-regional hyperthermia (41-46°C), but they don’t seem suitable to provide a sustained drug delivery for a prolonged period of time. Recently the temperature-induced control release has been approached using magnetic fields as triggering agent and MLs as drug carriers. Such carriers are hybrid systems of traditional thermo-sensitive liposomes containing superparamagnetic iron oxide nanoparticles (MNPs) and their use as devices for magnetic-controlled delivery of drugs is extensively investigated [8-9]. In particular, when exposed to appropriate magnetic fields (amplitudes of kA/m and frequencies from tens to hundreds of kHz) MNPs generate heat, either from hysteresis losses or from Néel or Brownian relaxa on processes[10]. This magnetically induced heat is transferred to the entire carrier causing structural changes in the bilayer, acting as smart trigger for drug release. However, the thermal-magnetic stimuli can cause damage and safety secondary effects to the surrounding tissues both due to temperature and to magnetically induced eddy currents, limiting their clinical applicability [11]. In spite of some recent promising studies [12], the controllable and reversible release of cargo from MLs by triggering them with magnetic stimulation has not been fully exploited. In particular some recent studies suggested that low magnetic fields may be suitable for drug delivery applications with minimal heating [13-14]. In this work, we present a drug-delivery system based on high-transition temperature liposomes made of hydrated soybean phosphatidylcholine and cholesterol, which enclose coated MNPs in their aqueous core (highTm MLs). The liposome membrane can exist only in the ordered state within the experimental temperature interval and neither spontaneous leakage nor thermal responsiveness can occur up to 50 °C. To probe the membrane permeation and release behavior of high-Tm MLs, 5(6)Page 2 of 13
ip t
carboxyfluorescein was loaded in the high-Tm MLs as hydrophilic model drug. The controllable release of cargo from high-Tm MLs was triggered by on-off low-level magnetic stimulation. The hypothesized release mechanism consists in a mechanical stress on the liposome membrane due to nanoparticles oscillations in its proximity, as similarly postulated in [15-16].It was investigated by means of a numerical model solved with multiphysics simulations that coupled magnetic, mechanical and thermal physical problems. The demonstrated high reproducibility of cycle-to-cycle release induced by the mechanical magnetic trigger suggests the potential of high-Tm MLs for a successful use as “on demand” drug delivery systems.
2. Experimental
cr
2.1. Materials
M
an
us
Hydrogenated soybean phosphatidylcholine (HSPC) Phospholipon 90H from Lipoid GmbH was kindly gifted by AVG Srl. Cholesterol, 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES), 5(6)-carboxyfluorescein (CF), Triton X-100 (TX-100), Sephadex G-50 and hydrochloric acid (HCl) were purchased from Sigma Aldrich. Chloroform was obtained from Merck. Bidistilled water, thiocyanatoiron, 1,2-dichloroethane and ethanol were supplied by Carlo Erba Reagents. Aqueous dispersion of 50 nm carboxymethyl-dextran coated magnetite (Fe3O4) nanoparticles fluidMAG-CMX (MNPs) was obtained from Chemicell GmbH. 2.2. Preparation and Physicochemical Characterization of Liposomes
Ac ce p
te
d
The high-Tm MLs were prepared by the thin lipid film hydration method followed by sequential extrusion as reported in [17]. The thin film of HSPC and cholesterol formed on the flask walls was hydrated with 10 mM HEPES buffer solution (pH=7.4) containing MNPs and 20 mM CF used as a model hydrophilic drug for the release experiments. Control liposomes, without MNPs, have been also prepared (see supplementary information). Repeated extrusion through membrane filters of 0.4 μm and 0.2 μm pore sizes yielded unilamellar liposomes with a narrow size distribution. Following extrusion, the unentrapped dye CF and MNPs was removed by size exclusion chromatography (SEC) on Sephadex G-50 column eluted with HEPES buffer (10 mM ), pH 7.4. All liposome formulations were stored in dark at 4°C and used within 1 week. Hydrodynamic diameter and polydispersity index (PdI) were evaluated by dynamic light scattering (DLS) experiments, whereas ζ-potential was measured by electrophoretic light scattering (ELS) technique (see supplementary information). All measurements were performed with Zetasizer Nano ZS90 (Malvern Instruments Ltd., UK) thermostatically controlled at 25°C. Phospholipid concentration was determined using the phosphorus colorimetric assay [18]. The samples were measured against a blank without substrate, using a double beam UV–Vis spectrophotometer Lambda 25 (Perkin Elmer, USA). The measurements were repeated both before and after SEC purification of extruded samples. MNPs content in MLs was determined through the 8.5% hydrochloric acid assay described by Belikov et al. [19]. The calibration curve was performed with standards solutions of magnetite. The measurements were repeated both before and after SEC purification of extruded MLs. All the data collected were used to calculate magnetite/phospholipid ratio. CF entrapped in the inner aqueous compartment was determined fluorimetrically at 512 nm with excitation at 492 nm after lysis of pre-incubation purified sample with 10% the non-ionic surfactant Triton X-100 by using the spectrofluorometer LS 50B (Perkin Elmer, USA).
Page 3 of 13
Physicochemical characterization of MLs formulations incubated at 4.0 ± 0.5°C was repeated at specific time intervals for at least 30 days in order to test colloidal stability of the magneto vesicles (see supplementary information). 2.3 AMF exposure set-up
cr
ip t
The exposure system used was the same described in details in [14, 20]; briefly it consists of two coaxial magnetic coils of square shape. A high magnetic field homogeneity of 99% is achieved in a volume around the system center where the samples were placed. The coils are connected to a signal generator through a wide band amplifier. (See supplementary information).
us
2.4 Measurement of CF released from liposomes
M
an
To measure the membrane permeation and release behavior of high-Tm MLs, CF, which is a self-quenching and membrane impermeable dye applied as a model hydrophilic drug, was loaded into the high-Tm MLs. In vitro CF release from high-Tm MLs was determined fluorimetrically by monitoring CF fluorescence de-quenching at excitation and emission wavelengths of 492 and 512 nm, respectively. The release was measured both due to an applied low intensity AMF stimulation (20kHz, 60 A/m) and after 3 hrs of continuous heating at selected temperature (25-70 ◦C) in absence of AMF. A water bath was used to control the environmental temperature. See supplementary information).
d
2.5. Differential scanning calorimetry (DSC)
Ac ce p
te
Thermal analysis was performed on HSPC/Chol/MNPs and HSPC/Chol without MNPs mixtures in the same molar ratio used for vesicles preparation. Calorimetric measurements were performed using a DSC131 (Setaram, France) differential scanning calorimeter. At least three heating/cooling cycles in a temperature range from 20 to 70 °C under nitrogen flow (20 ml/min) were performed on all the samples before thermograms were recorded at a rate of 5°C/min. 2.6 Statistical analysis
All experiments were performed at least in triplicate. All statistical analyses were performed with ANOVA one-way using the Bonferroni post-test (Instat software, version 3.0 GraphPAD Software Inc., San Diego, CA) to determine significant differences in the experimental data. P < 0.05 was considered statistically significant.
3.Results and discussion Physicochemical characterization results are shown in Table 1. No differences were observed in size between control liposomes and high-Tm MLs, both arranged in a monomodal distribution with PdI values <0.200. The neutrality of the ζ-potential values approaching to zero is due to the zwitterionic character of HSPC, and for the high-Tm MLs, it suggested the absence of the negatively charged MNPs absorbed onto the external leaflet of the bilayer. Furthermore, the percentage of the lipid molecules found into the high-
Page 4 of 13
Ac ce p
te
d
M
an
us
cr
ip t
Tm MLs is only partially decreased respect to control liposomes, suggesting that the hydration step of liposome preparation was slightly influenced by the presence of MNPs (Table 1). Finally, the possibility to use such high-Tm MLs as smart drug carrier is confirmed by the satisfactory values of CF-loading efficiency. Mean hydrodynamic diameter and PdI observed for at least 30 days confirmed the colloidal stability of the samples, as reported in Fig.1 (A). DSC profile reported in Fig. 1(B) exhibited one endothermic peak at Tm= 56.8°C suggesting that the incorporation of MNPs within high-Tm MLs core didn’t affect the gel-to-liquid crystalline transition of the membrane (see supplementary material). As shown in the Fig. 1(B), CF encapsulated in the high-Tm MLs almost couldn’t be released at lower temperatures (<5%), but it was released sharply when the temperature approaches Tm. The release of CF under AMF exposure has been monitored at 37 °C for 15 h to probe the effect of MNPs on the release properties. During the AMF exposure, the high-Tm MLs solution did not shown an increase in temperature in the bulk solution that was always ΔT<0.5°C, due to the low level AMF exposure. At the same time, high-Tm MLs were tested in sham (without AFM exposure) modality to evaluate the effect on the release profile of the environmental parameters without magnetic induction. Moreover, in order to exclude a possible direct AMF effect on the bilayer, CF release has been monitored in the same conditions also for control liposomes not encapsulating MNPs. As shown in Fig. 2(A), the AMF continuative exposure induced the significant increase in the CF release rate of high-Tm MLs. It demonstrated that the permeability of the bilayer was affected by the embedded MNPs, upon exposure to the AMF. The improved release in AMF cannot be ascertained to local membrane heating generated by MNPs due to magneto-caloric effect because of the high Tm=56.8°C of our MLs while temperature measurements detected an increase below 0.5°C. A local temperature increase seems unfeasible as further confirmed by the multiphysics model and discussed in the following. Conversely, the entrapped MNPs can effectively improve the bilayer permeability as a consequence of their motions in AMF, without affecting the geometrical properties of the liposomes (see supplementary material). The CF release extent was nearly 20% in high-Tm MLs already after 3 h of exposure and remains almost stable up to 15 h. No significant CF release (<5%) was observed for sham exposure and pure AMF exposed liposomes. In order to further prove the effect of AMF on controlled release of cargos from high-Tm MLs, the release of the encapsulated CF was carried out by commutatively switching on (3 h) and off (21 h) the AMF as a function of time. An important feature of the plot in Fig. 2(B) was that CF can be released repetitively from high-Tm MLs. When the AMF was turned on the magneticimpelled motions caused the destabilization of the bilayer improving its permeability, which permitted CF diffusion out of the high-Tm MLs. Thus the CF release from the high-Tm MLs can be explained by the mechanical motions of MNPs leading to the improved bilayer permeability, rather than the destruction of the liposome structure. The repetitious release from high-Tm MLs proved that the liposome structure was retained during AMF treatment. Indeed, high-Tm MLs maintain their geometrical properties unchanged over time (see supplementary material). On–off cycles were repeated at four times in the experiment until 70% of the cargo was successfully detected. This raised the possibility that AMF can be applied to control the repetitious release of high-Tm MLs at specific times, with similar dose during all the triggering steps. Long off periods (21 h) demonstrate the stability of the high-Tm MLs against undesired leakage when AMF is off.
Page 5 of 13
4. Conclusions
Ac ce p
te
d
M
an
us
cr
ip t
In summary, we demonstrated that high-Tm MLs, HSPC/Chol bilayers including in the aqueous core iron oxide nanoparticle MNPs, respond to a low amplitude magnetic field at temperature well below the Tm of the bilayer (Tm=56.8). The high-Tm MLs were demonstrated stable and able to preserve their physical properties over time for at least 30 days. To probe the membrane permeation and release behavior of high-Tm MLs, CF as a model hydrophilic drug was loaded in the high-Tm MLs, with satisfactory values of CF-loading efficiency. The content release from the high-Tm MLs could be triggered repetitively by switching on and off the AMF until almost complete depleting of the carrier. The AMF triggered release properties demonstrated that the cargo release from the high-Tm MLs in AMF was due to reversible and controllable permeability change of the bilayer, rather than the destructure of the high-Tm MLs, that maintain their size over the whole experiment duration. The hypothesized release mechanism is a mechanical stress on the liposome membrane due to nanoparticles oscillations in its proximity. It was investigated by means of a numerical model solved with multiphysics simulations that coupled magnetic, mechanical and thermal physical problems (see supplementary material). Fig.3 reports the magnetic induction (Fig.3 A), the mechanical pressure (Fig.3 B) and the electromagnetic heating (Fig.3 C) of the High-Tm ML model considering the maximum load of two nanoparticles to fully explore a possible thermal mechanism. The model results in a substantial absence of relevant temperature increases due to the AMF (δ T = 0.3 °C after three hours), while evidencing a pressure on the membrane of around tens of bars that is within the threshold of 10-100 bars for mechanical destabilization of bilayers indicated in literature [21]. In principle, mechanically AMF stimulus represents a novel modality to trigger release from magneto carriers and opens the possibility for specific complex drug release regimens. The exposure time of 3 h is longer than those typically used for drug delivery based on thermal sensitive MLs [12-13] but compatible with those of low level magnetic field treatments used in orthopedics for bone and connective tissue regeneration [22,23]. Even though more investigations are necessary, high-Tm MLs can be good candidates for smart nanodevice for an“on-demand” drug release in the presence of a low amplitude AMF.
Acknowledgements
Project depicted is sponsored in part by Sapienza University Research Grant 2013. This activity is performed in the framework of the Joint IIT-Sapienza LAB on Life-NanoScience Project "Novel strategies for the imaging and treatment of brain tumors through targeting cancer stem cellspecific signaling pathways." The authors want to thank Dr Iacopo Zanardi for his help in statistical analyses.
Reference
[1] S. Ganta, H. Devalapally, A. Shahiwala, M. Amiji, A review of stimuli-responsive nanocarriers for drug and gene delivery. J. Control. Release, 126 (2008)187–204 doi: 10.1016/j.jconrel.2007 [2] S. Mura, J. Nicolas, P, Couvreur, Stimuli-responsive nanocarriers for drug delivery. Nature Mater, 12 (2013) 991–1003 doi:10.1038/nmat3776 [3] S. J. Leung and M. Romanowski, Light-Activated Content Release from Liposomes. Theranostics, 2 (2012) 1020–1036. doi: 10.7150/thno.4847
Page 6 of 13
[4] A. Schroeder 1, J. Kost , Y. Barenholz, Ultrasound, liposomes, and drug delivery: principles for using ultrasound to control the release of drugs from liposomes. Chemistry and Physics of Lipids, 162 (2009) 1–16 doi:10.1016/j.chemphyslip.2009.08.003 [5] H. Karanth , RS. Murthy , pH-sensitive liposomes-principle and application in cancer therapy. J. Pharm. Pharmacol., 59 (2007) 469-83 doi: 10.1211/jpp.59.4.0001
ip t
[6] T. Ta, T. M. Porter Thermosensitive liposomes for localized delivery and triggered release of chemotherapy. J. Control. Release, 169 (2013) 112-125 doi:10.1016/j.jconrel.2013.03.036
cr
[7] M. B. Yatvin, J. N. Weinstein, W. H. Dennis, & R. Blumenthal, Design of liposomes for enhanced local release of drugs by hyperthermia. Science, 202 (1978) 1290–1293 doi:10.1126/science.364652
us
[8] P.Pradhan, J. Giri , F. Rieken , C. Koch , O. Mykhaylyk , M- Döblinger , R. Banerjee , D. Bahadur , C. Plank, Targeted temperature sensitive magnetic liposomes for thermochemotherapy. J. Control. Release, 142 (2010) 108–121 doi: 10.1016/j.jconrel.2009.10.002. [9] M.R. Preiss, G.D. Bothun, Stimuli-responsive liposome-nanoparticle assemblies. Expert. Opin Drug Deliv, 8 (2011) 1025-1040 doi: 10.1517/17425247.2011.584868
an
[10] P. Guardia, R. Di Corato , L. Lartigue, C. Wilhelm, A. Espinosa, M. Garcia-Hernandez, F. Gazeau, L. Manna, T. Pellegrino, Water-soluble iron oxide nanocubes with high values of specific absorption rate for cancer cell hyperthermia treatment. ACS Nano 6 (2012) 3080– 3091
M
[11] A. K. Gupta, R. R. Naregalkar, V. D. Vaidya & M. Gupta, Recent advances on surface engineering of magnetic iron oxide nanoparticles and their biomedical applications. Nanomedicine, 2 (2010) 23-39 doi:10.2217/17435889.2.1.23
te
d
[12] D.Qiu and X.An, Controllable release from magnetoliposomes by magnetic stimulation and thermal stimulation. Colloids and Surfaces B: Biointerfaces, 104 (2013) 326–329. doi: 10.1016/j.colsurfb.2012.11.033
Ac ce p
[13] S. Nappini, F. Baldelli Bombelli, M. Bonini, B.Nordèn, and P. Baglioni, Magnetoliposomes for controlled drug release in the presence of low-frequency magnetic field. Soft Matter, 6 (2010) 154–162 doi: 10.1039/B915651H [14] R. Spera, S. Petralito, M. Liberti , C. Merla , G. D'Inzeo , R. Pinto , F. Apollonio, Controlled release from magnetoliposomes aqueous suspensions exposed to a low intensity magnetic field. Bioelectromagnetics, 35 (2014) 309-12. doi: 10.1002/bem.21841 [15] D. Cheng, X. Li, G. Zhang, H. Shi, Morphological effect of oscillating magnetic nanoparticles in killing tumor cells. Nanoscale Research Letters, 9 (2014) 195. doi: 10.1186/1556-276X-9-195 [16] G. Podaru , S. Ogden , A. Baxter , T. Shrestha , S. Ren , P. Thapa , RK. Dani , H. Wang , M.T. Basel , P. Prakash , S.H. Bossmann and V. Chikan, Pulsed Magnetic Field Induced Fast Drug Release from Magneto Liposomes via Ultrasound Generation. J.Phys.Chem.B, 118 (2014) 11715-22. doi: 10.1021/jp5022278. [17] S. Petralito, R. Spera, A. Memoli, G. D'Inzeo, M. Liberti, F. Apollonio Preparation and characterization of lipid vesicles entrapping iron oxide nanoparticles Asia-Pacific Journal of Chemical Engineering, 7 (2012) 335-341, ISSN: 1932-2143. [18] Y. Yoshida, E. Furuya, K. Tagawa, A direct colorimetric method for the determination of phospholipids with dithiocyanatoiron reagent. J. of Biochemistry, 88 (1980) 463–468. [19] V.G. Belikov, A.G. Kuregyan, G.K. Ismailova, Standardization of magnetite. Pharm. Chem. J., 36 (2002) 333–336 doi: 10.1023/A:1020845110683
Page 7 of 13
[20] R. Lodato, C. Merla, R. Pinto, S. Mancini, V. Lopresto, G.A. Lovisolo, Complex magnetic field exposure system for in vitro experiments at intermediate frequencies. Bioelectromagnetics, 34 (2013) 211-219 doi: 10.1002/bem.21758. [21] M. Louhivuoria, H. J. Risselada, Erik van der Giessen, Siewert J. Marrink, Release of content through mechano-sensitive gates in pressurized liposomes, PNAS, 107 (2010) 19856–19860 doi:10.1073/pnas.1001316107
cr
ip t
[22] F. Vincenzi, M. Targa, C. Corciulo, S. Gessi, S. Merighi, S. Setti, R. Cadossi, M.B. Goldring, P.A. Borea, K. Varani, Pulsed electromagnetic fields increased the anti-inflammatory effect of A2A and A3 adenosine receptors in human T/C-28a2 chondrocytes and hFOB 1.19 osteoblasts. PLoS One, 8 (2013) doi: 10.1371/journal.pone.0065561
Ac ce p
te
d
M
an
us
[23] M. De Mattei, K. Varani, F.F. Masieri, A. Pellati, A. Ongaro, M. Fini, R. Cadossi, F. Vincenzi, P. A. Borea, A. Caruso, Adenosine analogs and electromagnetic fields inhibit prostaglandin E2 release in bovine synovial fibroblasts. Osteoarthr. Cartil., 17 (2009) 252-262 10.1016/j.joca.2008.06.002
Page 8 of 13
Fig. 1. Colloidal stability of MLs suspensions stored at ~4 °C over time (A) as determined by measuring changes in the hydrodynamic diameter (squared) and PdI (circles). Temperaturedependent changes (B) of cumulative release of CF (dashed) from MLs and DSC scanning profile of the melting process of HSPC/Chol/MNPs mixture (solid). Values (±S.D.) are the average of 3 independent experiments.
cr
ip t
Fig.2. Effect of AMF exposure (20 kHz, 60 A/m at 37°C) in continuous (A) and ON-OFF (B) modality on the cumulative release of CF from MLs and control liposome. The amount of CF released from MLs over time under isothermal sham exposure is reported for both strategies. Values (±S.D.) are the average of 7 independent experiments. Asterisks indicate significant differences between controls (sham exposure and control liposomes) and MLs (***p <0.001).
Ac ce p
te
d
M
an
us
Fig.3 Magnetic induction (A), mechanical pressure (B) and electromagnetic heating (C) on the model on the High-Tm ML in the maximum load of two.
Page 9 of 13
Table 1. Hydrodynamic diameter (nm), PdI and ζ-potential (mV) values of liposomes and MLs with HSPC 20 mol% Chol. Phospholipid content (%), amount of MNPs entrapped (g/mol HSPC) and CF concentration (µl/mg HSPC) in the final purified samples are also reported. Values (±S.D.) are the average of 3 independent determinations. Sample
Size (nm)
Liposomes MLs
HSPC (%)a
MNPs (g/mol HSPC)
242.7±9.7
0.069±0.004
-6.9±0.1
90.1
-
235.5±6.4
0.148±0.011
-8.7±0.2
78.1
115.03±4.24
CF volume trapped (µl/mg HSPC)
ip t
PdI
ζ-potential (mV)
2.10±0.22 1.47±0.18
Percentage of phospholipid was calculated by (mean concentration after purification/mean concentration before purification) X 100.
Ac ce p
te
d
M
an
us
cr
a
Page 10 of 13
Ac
ce
pt
ed
M
an
us
cr
i
Figure 1
Page 11 of 13
Ac
ce
pt
ed
M
an
us
cr
i
Figure 2
Page 12 of 13
Ac
ce
pt
ed
M
an
us
cr
i
Figure 3
Page 13 of 13