Layer-by-layer polysaccharide-coated liposomes for sustained delivery of epidermal growth factor

A three-dimensional layer-by-layer (LbL) structure composed by xanthan and galactomannan biopolymers over dioctadecyldimethylammonium bromide (DODAB) liposome template was proposed and characterized for protein drug delivery.

Carbohydrate Polymers 140 (2016) 129–135 Contents lists available at ScienceDirect Carbohydrate Polymers journal homepage: www.elsevier.com/locate/carbpol Layer-by-layer polysaccharide-coated liposomes for sustained delivery of epidermal growth factor Gabriel A.T Kaminski a,b , Maria Rita Sierakowski a , Roberto Pontarolo b , Larissa Antoniacomi dos Santos a , Rilton Alves de Freitas a,∗ a b BioPol, Chemistry Department, Universidade Federal Paraná, R Coronel F H dos Santos, Curitiba 210–81531-980, PR, Brazil CEB, Pharmacy Department, Universidade Federal Paraná, Av Prof Lothário Meissner, Curitiba 3400–80210-170, PR, Brazil a r t i c l e i n f o Article history: Received 14 September 2015 Received in revised form 31 October 2015 Accepted December 2015 Available online 17 December 2015 Keywords: DODAB Xanthan Galactomannan EGF Layer-by-layer Liposomes a b s t r a c t A three-dimensional layer-by-layer (LbL) structure composed by xanthan and galactomannan biopolymers over dioctadecyldimethylammonium bromide (DODAB) liposome template was proposed and characterized for protein drug delivery The polymers and the surfactant interaction were sufficiently strong to create a LbL structure up to layers, evaluated using quartz crystal microbalance (QCM) and zeta potential analysis The polymer–liposome binding enthalpy was determined by isothermal titration calorimetry (ITC) The bilayer of biopolymer-coated liposomes with diameters of 165 (±15) nm, measured by dynamic light scattering (DLS), and ␨-potential of −4 (±13) mV These bilayer-coated nanoparticles increased up to times the sustained release of epidermal growth factor (EGF) at a first order rate of 0.005 min−1 This system could be useful for improving the release profile of low-stability drugs like EGF © 2015 Elsevier Ltd All rights reserved Introduction Therapies based on growth factors (GF) have promising potential in biomedical technology In cases of chronic wounds proactive treatment is needed for healing and GF might provide the necessary stimuli to induce wound closure (Behm, Babilas, Landthaler, & Schreml, 2011) Of the various growth factors, epidermal growth factor (EGF) has been first and most successfully applied for treating wounds It is a polypeptide composed of 53 amino acids that enhance epidermal and mesenchymal regeneration, cell motility, and proliferation (Choi et al., 2012) EGF could be employed to accelerate re-epithelialization, reducing risk of infection and shortening hospitalization However, due to short half-life, rapid dilution in the body, and the fact that EGF receptors are overexpressed in most squamous carcinomas, brain gliomas and breast cancers (Gedda, Olsson, Pontén, & Carlsson, 1996; Chen & Mooney, 2003), supply of exogenous EGF must be applied in a sustained and localized fashion to be effective and safe The encapsulation of EGF in liposomes might be an alternative to maximize their stability and to avoid enzymatic degradation (De˘gim et al., 2011) ∗ Corresponding author Tel.: +55 4133613260; fax: +55 4133613260 E-mail address: rilton@ufpr.br (R.A.d Freitas) http://dx.doi.org/10.1016/j.carbpol.2015.12.014 0144-8617/© 2015 Elsevier Ltd All rights reserved Liposomes have already been studied as delivery systems for GF (De˘gim et al., 2011) Nevertheless, liposomes have the limitations of spilling their contents over time and aggregating (Taylor, Davidson, Bruce, & Weiss, 2005) In order to prevent these events, the liposomes can be coated with polymers, modulating the drug delivery to achieve the desired release kinetics A very efficient technique to form polymeric coatings on twodimensional and three-dimensional systems, such as liposomes, is Layer-by-layer (LbL), introduced by Decher (1997) This technique consists in the alternating depositions of polycations and polyanions, generating a multilayer coating supported mainly, but not exclusively, by the electrostatic interactions or of hydrogen bonds (Wang et al., 1997), covalent bonds (Sun et al., 1998), hydrophobic interactions (Lojou & Bianco, 2004) and van der Waals forces (Sato & Sano, 2005) In order to coat the dioctadecyldimethylammonium bromide (DODAB) cationic liposomes, using the LbL technique, two biopolymers were chosen Xanthan (XAN) an anionic biopolymer produced by Xanthomonas campestris and composed of a (1 → 4)-␤-d-glucan cellulose backbone substituted with an acid trisaccharide in the side chain (Jansson, Kennark, & Lindberg, 1975), and galactomannan (GMC), a neutral biopolymer from Ceratonia siliqua seeds that is composed of a (1 → 4)-␤-d-mannan backbone with (1 → 6)-␣d-galactose substitutions (Dea & Morrison, 1975) These polymers interact positively and synergistically, as previously described and 130 G.A.T Kaminski et al / Carbohydrate Polymers 140 (2016) 129–135 measured by rheology, differential scanning calorimetry and light scattering (Bresolin, Milas, Rinaudo, & Ganter, 1998; Khouryieh, Herald, Aramouni, & Alavi, 2007) In this manuscript, we evaluate a new approach for LbL-threedimensional systems structured with XAN and GMC, for sustained release of EGF We employed DODAB, a cationic lipid with a quaternary ammonium salt as its polar head and two 18-carbon saturated chains, to form positively superficial charged liposomes potentiating the biopolymer LbL coating The liposomes, including those with coatings, were characterized, and the EGF release rates were determined in vitro Material and methods 2.1 Polymer preparation 2.1.1 Polymer purification XAN gum (Sigma-Aldrich) was purified by dialysis through a cellulose membrane (Sigma-Aldrich), first against 0.1 mol L−1 acetic acid for days to ensure that all the molecules would be in the molecular conformation, and then against ultrapure water for days to remove the acetic acid GMC, from locust bean gum from C siliqua seeds (Sigma-Aldrich), was purified by dispersion in ultrapure water at 40 ◦ C overnight and centrifugation at 10,000 × g for 20 at 40 ◦ C in a 4K15 C (Sigma, Osterode am Harz, Germany) centrifuge to precipitate insoluble impurities Ethyl alcohol (99%) was added to the supernatant to achieve a 70% alcohol concentration, and this suspension was centrifuged at 10,000 × g for 20 at ◦ C to precipitate the purified XAN and GMC The polymers were washed with 99% ethyl alcohol, centrifuged as described above and dried at 40 ◦ C 2.1.2 Polymer characterization The polymers were characterized at 0.5 mg mL−1 by high performance size exclusion chromatography (HPSEC) with 0.1 mol L−1 NaNO3 and 200 ppm sodium azide at 0.4 mL min−1 as the mobile phase at 40 ◦ C The system was composed of a UV/vis detector, refractometer, light scattering detector at 7◦ and 90◦ and a differential viscometer detector (Viscotek, Westborough, MA, USA) with an OHpak SB-806 M HQ column (Shodex, New York, NY, USA) The persistence length (Lp ) was determined as previously described for other galactomannans by Salvalaggio, de Freitas, Franquetto, Koop, and Silveira (2015) The zeta potential was determined for polymers dispersions at 0.5 mg mL−1 in ultrapure water using the electrophoretic mobility measured on a Zetasizer Nano-ZS (Malvern, Westborough, MA, USA) at 25 ◦ C with 120 s of stabilization The protein quantification of purified GMC was determined by the Hartree method (Hartree, 1972) Infrared spectroscopy (FTIR), using an attenuated total reflectance (ATR) mode was determined in a VERTEX 70 (BRUKER) with cm−1 of resolution and 4000–600 cm−1 (Supplementary material S1) 2.2 Liposome preparation Liposomes were prepared by a modified method (Alves et al., 2009) DODAB (Sigma-Aldrich, Switzerland) was dispersed in chloroform at mmol L−1 , the solvent was removed by rotary evaporation at 40 ◦ C, and the lipid was resuspended in a ␮g mL−1 EGF (Caregen) solution at 35 ◦ C, similar to Alves et al (2009) and De˘gim et al (2011) The suspension was then submitted to sonication at 25 ◦ C in pulse mode for The EGF-liposome solution was diluted 1:10 in ultrapure water and centrifuged at 10,000 × g for 20 at 25 ◦ C The supernatant was discarded, and the liposomes were resuspended in 0.5 mg mL−1 XAN solution, followed by centrifugation and washing with ultrapure water The same procedure was performed with a 0.5 mg mL−1 GMC solution 2.3 Liposome characterization The hydrodynamic diameters of the coated liposomes were analyzed and compared with those of the plain liposomes by dynamic light scattering (DLS) on a NanoDLS (Brookhaven Instruments, Holtsville, NY, USA) The uncoated liposomes and polymer-coated liposomes were also characterized by their ␨-potential and AFM as described elsewhere The zeta potential of the liposomes was determined for mmol L−1 dispersion of DODAB in ultrapure water using the electrophoretic mobility measured on a Zetasizer Nano-ZS (Malvern, Westborough, MA, USA) at 25 ◦ C with 120 s of stabilization 2.4 Polymer and DODAB/DODAB vesicle interaction 2.4.1 Isothermal titration calorimetry (ITC) Experiments were performed in a VP-ITC (Microcal, Westborough, MA, USA) calorimeter with a normal cell (1.464 mL) at 25 ◦ C The DODAB vesicle dispersion was injected into ultrapure water or into a polymer solution at 0.5 mg mL−1 Each titration consisted of a preliminary ␮L injection followed by 29 subsequent 10 ␮L injections with 600 s intervals between each injection The syringe tip acted as a blade-type stirrer to ensure proper mixing at 300 rpm Data were collected and processed with Origin 7.0 software (OriginLab, Northampton, MA, USA) 2.4.2 Quartz crystal microbalance (QCM) Analyses were performed in triplicate in a SRS QCM200 using the flow cell mode The QCM (Gold/Cr MHz, SRS, Sunnyvale, CA, USA) crystals were cleaned by immersion in 1:3 (v/v) H2 O2 :H2 SO4 for and followed by rinsing with ultrapure water To mimic the liposome coating process, the gold surface was modified with hexanethiol (Sigma-Aldrich) to form a hydrophobic surface and then coated with DODAB by immersion in a mmol L−1 chloroform solution similar to Morita, Nukui, and Kuboi (2006) One milliliter of each polymer solution (0.5 mg mL−1 ) was alternately injected at 0.1 mL min−1 with a syringe pump (KD100, KD Scientific, Holliston, MA, USA), and mL of ultrapure water was injected between each solution 2.4.3 Atomic force microscope (AFM) Images of each layer on the QCM crystal were obtained on a PicoPlus Molecular Imaging microscope (Agilent, Santa Clara, CA, USA) in the intermittent contact mode in air at 25 ◦ C with silicon cantilevers, an oscillating amplitude of 50 to 100 nm and a resonance frequency close to 300 kHz The dynamic tapping mode was used with an oxide-sharpened micro-fabricated silicon ␮-Masch cantilever with a 4.7 N m−1 nominal spring constant and tip curvature radius of less than 10 nm The image processing and root mean square roughness (rms) determination were performed with Gwyddion software (Czech Metrology Institute, Brno-sever, Czech Republic) 2.4.4 Contact angle The angles of the polymer substrates were determined with OCA15+ (DataPhysics, Filderstadt, Germany) device equipped with SCA20 software by the sessile drop method at 25 ◦ C with the delivery of 10 ␮L ultrapure water drops onto the QCM crystal-coated surface G.A.T Kaminski et al / Carbohydrate Polymers 140 (2016) 129–135 2.5 EGF-liposome release kinetics the effect of its adsorption on the standard deviation of the QCM crystal was larger (different amounts of GMC were required) EGF release profiles were determined by HPLC quantification using a Prominence (Shimadzu, Columbia, MD, USA) system with a Symmetry C18 (4.6 × 250 mm, ␮m particles) column A previously described method (Yang, Huang, Wu, & Tsai, 2005) was adapted to use 40% isocratic grade acetonitrile (Fluka) in 0.1% TFA at 40 ◦ C with mL min−1 flow rate and 210 nm UV detection The resulting EGF retention time was 4.8 A volume of 10 mL of Plain EGF-liposomes, XAN-coated EGFliposomes or GMC + XAN-coated liposomes, all in triplicates, were placed in a dialysis bag (12,000 g mol−1 cut off) and immersed in 90 mL of 0.1 mol L phosphate buffer saline (PBS) at pH 5.5 and 35 ◦ C to mimic dermal delivery conditions (Wagner, Kostka, Lehr, & Schaefer, 2003) with magnetic stirring for up to 48 h One milliliter aliquots were collected, at each time point, from the medium and that volume was replaced with fresh PBS Results and discussion 3.1 Polymers All the purified polymers exhibited a unimodal distribution as measured by HPSEC (data not shown) After purification, the XAN and GMC presented the molar mass (Mw ), intrinsic viscosity and radius of gyration (Rg ) as seen in Table The zeta potentials (␨potential) (Table 1) showed a large difference between DODAB and XAN, suggesting that these polymers may interact with each other electrostatically The negative charge on the neutral GMC polymer could be caused by attached proteins, determined as 4% (m/m), that remained even after polymer precipitation with ethanol 3.2 Liposomes-LbL biopolymer coating The size of the liposomes increased (Fig 1a) from 62 (±9) nm to 132 (±22)nm when coated with XAN and to 165 (±15) nm when coated with the XAN and the GMC layers The size of the liposomes, after 48 h was 79 nm, 140 and 195 nm, respectively for plain, XAN and XAN-GMC layers, suggesting an small increase in the size during storage of the liposomes Although there was greater adsorption for the GMC coating, the XAN layer represented a larger increase in the diameter of the liposomes This increase was most likely because of its larger persistence length (Lp ) of 125 nm, similar to the found by Rinaudo (2001), than of GMC (9 nm) and the Coulomb repulsion between negatively charged XAN, resulting in pores on the surface, which were occupied by GMC The plain and coated liposomes can be seen in AFM (Fig 2) It is clear that XAN left pores on the surface, because the first layer only decreased the ␨-potential of the liposomes to approximately + 35 mV (Fig 1b) However, the first GMC layer shielded all the positive charge from DODAB and decreased the ␨-potential of the liposomes to approximately −4 mV, close to the own ␨potential of the polymer Because GMC filled the pores left by XAN, Table Physicochemical characteristics of the polymers and the surfactant: molar mass (Mw ), radius of gyration (Rg ) and zeta potential (␨-potential) Molecules Mw (g mol−1 )* Rg (nm)* ␨-potential (mV)** XAN GMC DODAB 1.03 × 10 0.26 × 106 630 69 52 – −54.1 (±7.7) −5.4 (±6.2) + 66.5 (±9.4) 0.1 mol L-1 NaNO3 and 200 ppm sodium azide as the mobile phase at 0.4 mL min−1 at 40 ◦ C through a Shodex OHpak SB-806 HQ column The zeta potential was determined using the electrophoretic mobility in ultrapure water at 25 ◦ C * 131 3.3 Physical–chemical description of liposomes-LbL biopolymer coating QCM was used to verify that there were interactions between all layers, enabling the formation and maintenance of the LbL structure Because the liposomes were the template for the biopolymers coating, the QCM crystal should be coated with DODAB before polymer deposition To assure the proper DODAB coating, the crystal was first coated with hexanethiol Thus, the thiol group interacted strongly with gold, and its alkyl chain pointed upwards, allowing hydrophobic interactions with the alkyl tails of DODAB Therefore, the cationic amino head on the surface was able to interact with anionic XAN molecules, the first polymeric layer injected The crystal topography was analyzed by AFM (Fig 3), and the surface contact angle was measured with ultrapure water drops after the first layer of XAN and the first layer of GMC, on top of XAN QCM showed the adsorption of both polymers (Supplementary material S2), with greater adsorption for GMC The grater irregularity of the XAN coating, as evidenced by AFM, can explain this finding It appears that the Lp of XAN molecules leaves pores that are filled by GMC, a much more flexible molecule Because GMC has a slightly negative charge, it can also interact with DODAB The contact angle confirmed that all the layers were properly fixed Hexanethiol formed the most apolar surface, and DODAB the most polar, consistent with its greater ␨-potential The contact angle of the XAN layer was slightly greater than that of DODAB, indicating a less polar surface The GMC layer, a neutral polysaccharide, raised the surface contact angle As the LbL process continued, the adsorbed mass proportional to − F, was increased (Fig 1c) It is understandable that the electrostatic interaction is greater than other interactions Thus, the DODAB charge was responsible for the large initial adsorption As shielding decreased the effective positive charge, fewer polymers were adsorbed Although the adsorbed mass decreased, it was sufficient to change the ␨-potential of the nanoparticles at each layer We confirmed that the LbL process was efficient for up to tested layers of each polymer (Fig 1b) Isothermal titration calorimetry (ITC) was used to determine the binding enthalpy between the coating polymers, XAN and GMC, and the DODAB vesicles ITC analysis results in a binding isotherm (Fig 4) that provides precise analytical information, such as the number of free and bound vesicles at different stages of the binding process It is also possible to determine the average stoichiometry of supramolecular surfactant-polymer clusters Fig shows the raw signal curve (top) and the integral of the area at each injection (bottom) The enthalpy measured in the cell is the sum of several different energies, such as the energy of the dilution effect and the energy of the polymers binding onto the surfactant vesicles The energy of the polymers binding to the vesicles is more significant than the sum of the other energies, which represent the small energy changes that occur after all polymer molecules are consumed The enthalpy curves of both polymers exhibit only one cooperative endothermic event at each injection, attributed to the binding of the polymers to the DODAB vesicles The Hcoat ( H attributed to the coating of DODAB vesicles by polymer molecules) was calculated as the difference between the initial H and the average H plateau reached Fig 4a shows that each GMC molecule interacted with approximately 25 DODAB molecules in the vesicle, the molecules inside the vesicles were not considered (neither the flip-flop effect), with 132 G.A.T Kaminski et al / Carbohydrate Polymers 140 (2016) 129–135 Fig (a) Gaussian distribution of nanoparticles showing the increase in hydrodynamic diameter (Dh) caused by the polymer coating, determined by DLS at 90◦ (b) Zeta potential determined by electrophoretic mobility at 25 ◦ C for LbL-coated liposomes after each layer Bars denote the standard deviation (c) Frequency shift tendency determination (n = 3) by QCM for LbL with polymers injected at 0.1 mL min−1 The crystal was washed with ultrapure water between polymer injections a binding enthalpy of 17.6 kJ mol−1 of DODAB At ∼50 ␮mol L−1 of DODAB all GMC molecules were consumed and there is a surplus of free DODAB vesicles In the ITC plot of the DODAB vesicles titration in the XAN solution (Fig 4b), a plateau is observed between surfactant concentrations of 600 and 900 ␮mol L−1 At these concentrations, all XAN molecules were consumed; each XAN molecule coats ∼1500 DODAB molecules in the vesicle with a coating enthalpy of 26.4 kJ mol−1 of DODAB The small remaining enthalpies with further liposomes addition tend to zero and might be attributed to associations in other larger colloidal systems To better comprehend the polymer coatings of the DODAB liposomes, the number of surfactant (Ntot ) molecules that formed each liposome was calculated by the following equation: G.A.T Kaminski et al / Carbohydrate Polymers 140 (2016) 129–135 133 Fig AFM topography × ␮m images of the LbL process on a QCM crystal (a) Gold surface; (b) Hexanethiol; (c) DODAB; (d) XAN; (e) GMC Fig AFM topography image of (a) Plain liposomes; (b) XAN-coated liposomes; (c) GMC + XAN-coated liposomes Ntot = d/2 +4 a d/2 − h (1) where d is the liposomes diameter, 62 nm determined by DLS, h is the bilayer thickness, nm (Correia, Petri, & Carmona-Ribeiro, 2004) and a is the surfactant head group area It was found that approximately 31,000 DODAB molecules associate to form each colloidal structure Therefore, approximately 1250 GMC molecules coated each liposome while, approximately, only 20 molecules of XAN Possible explanations for this great difference include the electrostatic repulsion between adsorbed XAN molecules on the liposome surface, hindering the adsorption of newly arrived XAN molecules The XAN molecules used in this research are larger than the GMC molecules, they have times the molar mass Also, the longer Lp of XAN molecules play a role in the lack of organization of the adsorbed molecules, while the GMC molecules can adjust better on the surface and even interact with each other The large enthalpy observed in our study suggests that the charged sites of XAN molecules are strongly attracted to the DODAB liposomes The larger heat exchange of DODAB-XAN, compared with that of DODAB-GMC, is likely due to the ion-exchange interactions between DODA+ and the negative sites on XAN 3.4 Release kinetics After confirming the formation and structure of the liposomes coated with biopolymers and the physical-chemical parameters of the coating, EGF was encapsulated in the system The entrapment efficiency was 72 ± 3% The efficiency was determined by HPLC quantification through the direct method of lysing liposomes with 3% Triton X-100, described by Liu, Yang, Liu, and Jiang (2008) and an indirect method that measured the EGF in the supernatant of centrifuged liposomes The kinetics data were treated with the Berens and Hopfenberg (1978) model (Eq (2)), suitable for spheres, which showed confidence intervals above 99% for the systems: 134 G.A.T Kaminski et al / Carbohydrate Polymers 140 (2016) 129–135 Fig Thermogram (top) and binding isotherm (bottom) The latter resulted from the integration of the ITC peaks in the former, which were at 25 ◦ C (a) DODAB liposomes titrated in a solution of GMC at 0.5 mg mL-1 (b) DODAB liposomes titrated in a solution of XAN at 0.5 mg mL-1 Mt =1− M∞ F i n=3 n2 exp −4 n2 Dt d2 1.0 − R exp (−kx) 0.8 where D is the diffusion coefficient, k is the first-order relaxation constant, F and R are the fractions of sorption contributed by Fickian diffusion and chain relaxation, respectively, d is the diameter of the nanoparticle and t is time Each layer of coating contributed to the release kinetics, as seen in Fig The plain liposomes presented a constant drug delivery rate (k) of 0.025 min−1 , and the XAN layer reduced the delivery rate to 0.016 min−1 , 1.6 times longer For the liposomes coated with one layer of both polymers, the delivery k decreased to 0.005 min−1 , times smaller than the rate of the plain liposomes Non-linear least square fitting routine was used to fit the release of EGF from the nanoparticles into the above model This model describes the release behavior in terms of Fickian and non-Fickian contributions As the coating took place the R parameter became dominate When the last term of the equation is switched to −X/ , it provides the characteristic relaxation time ( ) The coating thicknesses (L) of the layers were determined as the difference between the nanoparticles hydrodynamic radius (Fig 1a) The Mt /M∞ was plotted against the square root of t/L2 0.6 in cm2 to provide the EGF diffusibility (D), the angular coefficient, in all nanoparticles according to Vogt, Soles, Lee, Lin, and Wu (2004) The and the D values were plotted together against the coating layers and can be seen in Supplementary material (S3) Mt/M (2) 0.4 0.2 0.0 60 120 180 240 300 360 420 480 Time (min) Fig The tendency profile in vitro of EGF from the nanoparticles in PBS at pH 5.5 and 35 ◦ C by dialysis • Plain liposomes, ᭿ XAN coated liposomes and XAN + GMC coated liposomes Bars denote the standard deviation (n = 3) The plain liposomes relaxation time (t) was found to be 40 min, the XAN layer increased to 60 and the XAN and GMC coatings increased to 195 It can be noticed that the XAN layer contributed to a 50% increase in the nanoparticles sustained release, while the GMC over the XAN layer increased it in almost 400% One additional layer of each polysaccharide XAN-GMC reduced completely the drug delivery to zero during 480 of evaluation G.A.T Kaminski et al / Carbohydrate Polymers 140 (2016) 129–135 As the coating became thicker, the relaxation of the polymers in the coating layers became more important for the release of EGF This way, as D decreased, increased exponentially Conclusion The natural interaction between XAN and GMC was demonstrated to be sufficiently strong to provide LbL structure for up to biopolymer layers The association of the anionic XAN and the neutral GMC for liposomes coating provided a synergistic effect that built a cohesive nanoparticle capable of increasing times the sustained delivery of EGF with only one layer of each polymer The EGF release from the nanoparticles was attributed to the polymer chains relaxation by the use of the diffusion and relaxation model The use of these natural polysaccharides in three-dimensional delivery systems provides biocompatibility, ease and abundance at low cost Acknowledgements We acknowledge the Brazilian funding agencies CNPq (Conselho Nacional de Pesquisa, process no 477275/2012-5 and 300343/2010-8, Fundac¸ão Araucária, project 23643, convênio 447/2012, and Rede Nanobiotec/Capes-Brazil, project 34, for financial support We are grateful to Dra Leila Beltramini (Instituto de Física de São Carlos) for ITC analysis, Dr Watson Loh (Instituto de Química UNICAMP) for thermodynamic discussion aid and Dr Lionel Gamarra (Instituto Cérebro, Hospital Albert Einstein) for ␨potential analysis Appendix A Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.carbpol.2015.12.014 References Alves, J B., Ferreira, C L., Martins, A F., Silva, G A B., Alves, G D., Paulino, T P., et al (2009) Local delivery of EGF-liposome mediated bone modeling in orthodontic tooth movement by increasing RANKL expression Life Sciences, 85, 693–699 Behm, B., Babilas, P., Landthaler, M., & Schreml, S (2011) Cytokines, chemokines and growth factors in wound healing Journal of the European Academy of Dermatology and Venereology, 26, 812–820 Berens, A R., & Hopfenberg, H B (1978) Diffusion and relaxation in glassy polymer powders: Separation of diffusion and relaxation parameters Polymer, 19, 489–496 Bresolin, T M B., Milas, M., Rinaudo, M., & Ganter, J L M S (1998) Xanthan–galactomannan interactions as related to xanthan conformations International Journal of Biological Macromolecules, 23, 263–275 Correia, F M., Petri, D F S., & Carmona-Ribeiro, A M (2004) Colloid Stability of Lipid/Polyelectrolyte Decorated Latex Langmuir, 20, 9535–9540 135 Chen, R R., & Mooney, D J (2003) Polymeric growth factor delivery strategies for tissue engineering Pharmaceutical Research, 20, 1103–1112 Choi, J K., Jang, J.-H., Jang, W.-H., Kim, J., Bae, I.-H., Bae, J., et al (2012) The effect of epidermal growth factor (EGF) conjugated with low-molecular weight protamine (LMWP) on wound healing of the skin Biomaterials, 33, 8579–8590 Dea, I C M., & Morrison, A (1975) Chemistry and interactions of seed galactomannans Advances in Carbohydrate Chemistry and Biochemistry, 32, 241–312 Decher, G (1997) Fuzzy nanoassemblies: Toward layered polymeric multicomposites Science, 277, 1232–1237 De˘gim, Z., C¸elebi, N., Alemdaro˘glu, C., Deveci, M., Ưztürk, S., & Özo˘gul, C (2011) Evaluation of chitosan gel containing liposome-loaded epidermal growth factor on burn wound healing International Wound Journal, 8, 343–354 Gedda, L., Olsson, P., Pontén, J., & Carlsson, J (1996) Development and in vitro studies of epidermal growth factor–dextran conjugates for boron neutron capture therapy Bioconjugate Chemistry, 7, 584–591 Hartree, E F (1972) Determination of protein: A modification of the lowry method that gives linear photometric response Analytical Biochemistry, 48, 422–427 Jansson, P E., Kennark, L., & Lindberg, B (1975) Structure of the extracellular polysaccharide from Xanthomonas campestris Carbohydrate Research, 45, 275–282 Khouryieh, H A., Herald, T J., Aramouni, F., & Alavi, S (2007) Intrinsic viscosity and viscoelastic properties of xanthan/guar mixtures in dilute solutions: Effect of salt concentration on the polymer interactions Food Research International, 40, 883–893 Liu, Y., Yang, Y., Liu, X., & Jiang, T (2008) Quantification of pegylated liposomal doxorubicin and doxorubicinol in rat plasma by liquid chromatography/electrospray tandem mass spectroscopy: Application to preclinical pharmacokinetic studies Talanta, 74, 887–895 Lojou, É., & Bianco, P (2004) Buildup of polyelectrolyte-protein multilayer assemblies on gold electrodes Role of the Hydrophobic Effect Langmuir, 20, 748–755 Morita, S., Nukui, M., & Kuboi, R (2006) Immobilization of liposomes onto quartz crystal microbalance to detect interaction between liposomes and proteins Journal of Colloid and Interface Science, 298, 672–678 Rinaudo, M (2001) Relation between the molecular structure of some polysaccharides and original properties in sol and gel states Food Hydrocolloids, 15, 433–440 Salvalaggio, M O., de Freitas, R A., Franquetto, E M., Koop, H S., & Silveira, J L M (2015) Influence of the extraction time on macromolecular parameters of galactomannans Carbohydrate Polymers, 116, 200–206 Sato, M., & Sano, M (2005) Van der Waals layer-by-layer construction of a carbon nanotube 2D network Langmuir, 21, 11490–11494 Sun, J., Wu, T., Sun, Y., Wang, Z., Zhang, X., Shen, J., et al (1998) Fabrication of a covalently attached multilayer via photolysis of layer-by-layer self-assembled films containing diazo-resins Chemical Communications, 17, 1853–1854 Taylor, T M., Davidson, P M., Bruce, D B., & Weiss, J (2005) Liposomal nanocapsules in food science and agriculture Critical Reviews in Food Science and Nutrition, 45, 587–605 Vogt, B D., Soles, C L., Lee, H.-J., Lin, E K., & Wu, W.-L (2004) Moisture absorption and absorption kinetics in polyelectrolyte films: Influence of film thickness Langmuir, 20, 1453–1458 Wagner, H., Kostka, K.-H., Lehr, C.-M., & Schaefer, U F (2003) pH profiles in human skin: Influence of two in vitro test systems for drug delivery testing European Journal of Pharmaceutics and Biopharmaceutics, 55, 57–65 Wang, L Y., Wang, Z Q., Zhang, X., Shen, J C., Chi, L F., & Fuchs, H (1997) A new approach for the fabrication of an alternating multilayer film of poly(4-vinylpyridine) and poly(acrylic acid) based on hydrogen bonding Macromolecular Rapid Communications, 18, 509–514 Yang, C.-H., Huang, Y.-B., Wu, P.-C., & Tsai, Y.-H (2005) The evaluation of stability of recombinant human epidermal growth factor in burn-injured pigs Process Biochemistry, 40, 1661–1665  ... volume of 10 mL of Plain EGF -liposomes, XAN-coated EGFliposomes or GMC + XAN-coated liposomes, all in triplicates, were placed in a dialysis bag (12,000 g mol−1 cut off) and immersed in 90 mL of. .. plain liposomes presented a constant drug delivery rate (k) of 0.025 min−1 , and the XAN layer reduced the delivery rate to 0.016 min−1 , 1.6 times longer For the liposomes coated with one layer of. .. studies of epidermal growth factor? ??dextran conjugates for boron neutron capture therapy Bioconjugate Chemistry, 7, 584–591 Hartree, E F (1972) Determination of protein: A modification of the lowry