Natural polymeric microspheres for modulated drug delivery

Accepted Manuscript Natural polymeric microspheres for modulated drug delivery

Costantino Del Gaudio, Valentina Crognale, Gianpaolo Serino, Pierluca Galloni, Alberto Audenino, Domenico Ribatti, Umberto Morbiducci PII: DOI: Reference:

S0928-4931(17)30571-4 doi: 10.1016/j.msec.2017.02.051 MSC 7385

To appear in:

Materials Science & Engineering C

Received date: Revised date: Accepted date:

30 July 2016 16 December 2016 14 February 2017

Please cite this article as: Costantino Del Gaudio, Valentina Crognale, Gianpaolo Serino, Pierluca Galloni, Alberto Audenino, Domenico Ribatti, Umberto Morbiducci , Natural polymeric microspheres for modulated drug delivery. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Msc(2017), doi: 10.1016/j.msec.2017.02.051

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

ACCEPTED MANUSCRIPT

Natural polymeric microspheres for modulated drug delivery Costantino Del Gaudio a, Valentina Crognale a, Gianpaolo Serino b, Pierluca Galloni c, Alberto Audenino b, Domenico Ribatti d,e, Umberto Morbiducci b

Department of Enterprise Engineering “Mario Lucertini”, University of Rome “Tor

PT

a

RI

Vergata”,Intrauniversitary Consortium for Material Science and Technology (INSTM), Research Unit Tor Vergata, Rome, Italy.

Department of Mechanical and Aerospace Engineering, Politecnico di Torino, Turin, Italy

c

Department of Chemical Science and Technology, University of Rome ‘‘Tor Vergata’’, Rome,

NU

SC

b

Italy

Department of Basic Biomedical Sciences, Neuroscience and Sensory Organs, Unit of Human

MA

d

Anatomy and Histology, University of Bari, Italy

National Cancer Institute “Giovanni Paolo II”, Bari, Italy

Corresponding author

PT E

D

e

CE

Costantino Del Gaudio, PhD

Department of Enterprise Engineering

AC

University of Rome “Tor Vergata” Via del Politecnico1, 00133, Rome, Italy Tel. +39-6-72594480 Fax +39-6-72594328 Email: [email protected]

ACCEPTED MANUSCRIPT Abstract

Microspheres can be regarded as a suitable platform for the development of ad hoc drug delivery systems, since the targeted release of a therapeutic agent can effectively contribute to support and improve a pharmacological protocol. However, several crucial factors related to the selection of

PT

materials, drugs and fabrication techniques should be critically analyzed in order to enhance the

RI

expected performance. Dealing with highly compatible materials, e.g. naturally-derived polymers

SC

and “green” reagents, can be a valid approach. For this aim, gelatin, chitosan and blend microspheres were produced by emulsion technique simply using distilled water and olive oil.

NU

Necessarily, due to the intrinsic instability of the selected materials in aqueous environment, microspheres were cross-linked with genipin, an extremely low cytotoxic agent, at three different

MA

concentration (i.e., 0.1, 0.5, 1% w/v). Collected microspheres were then loaded with methylene blue (MB), as drug model. Morphological analysis revealed homogeneous microspheres characterized by

D

an average diameter comprised in the range 42-54 μm. In vitro MB temporal delivery was assessed

PT E

until complete release, which occurred in about 3 days for gelatin and 30 days for chitosan microspheres. Nanoindentation analysis was performed to evaluate how polymers and genipin

CE

influenced the mechanical properties of microspheres. Finally, the effect of released MB was investigated by means of chicken embryo chorioallantoic membrane assay, highlighting anti-

AC

angiogenic properties for gelatin differently from chitosan and blend microspheres.

Keywords: Gelatin, chitosan, microspheres, drug delivery, angiogenesis.

ACCEPTED MANUSCRIPT 1. INTRODUCTION

The development of a controlled and targeted drug delivery system represents one of the main challenges for the definition of an effective pharmacological therapy. To obtain the maximum efficacy, the selected agent should be delivered to the target tissue by optimizing both the dose and

PT

the release period, thus causing minimal side effects. Most pharmaceutical formulations are

RI

systemically administered and, therefore, not specifically directed to the target organ or tissue. This

SC

implies that several biological barriers have to be crossed, such as organs, cells and even intracellular compartments, often causing undesirable side-reactions, and concurring to partially

NU

inactivate the drug [1]. It is also well established that the conventional therapeutic approach is limited by, e.g., fluctuating drug levels and poor efficacy [2]. In addition, it is not possible to

MA

control drug concentration and bioavailability, thus limiting the expected therapeutic outcome. In this regard, the development of ad hoc drug delivery systems can support and improve alternative

D

clinical treatments, providing an effective release over time. In fact, systems with reproducible and

PT E

predictable release kinetics do not necessitate multiple administrations and assure a higher drug concentration at the target site. For this aim, microspheres are generally regarded as a valuable

CE

option, especially when made up of bioresorbable polymers. Both synthetic and natural polymers can be considered, though naturally-derived ones can offer an enhanced response owing to an

AC

intrinsic biological affinity. Here, gelatin and chitosan were selected as two representative polymers to be used for a microsphere-based therapeutic strategy. Gelatin, derived from collagen denaturation, is non-immunogenic, bioresorbable, non-cytotoxic, and available at relatively low cost [3]. In addition, it has been found to have cell-affinitive and enzyme-cleavable domains [4], so it can be broken down by cellular action through the secretion of specific matrix metalloproteinases [5]. Chitosan is a natural linear biopolyaminosaccharide and is obtained by alkaline deacetylation of chitin, which is the second most abundant polysaccharide after cellulose [6]. Its degradation leads to the release of amino sugars, which can be incorporated into glycosaminoglycan and glycoprotein

ACCEPTED MANUSCRIPT metabolic pathways or excreted [7]. In vivo tests have proven that chitosan-based biomaterials do not have any remarkable inflammatory or allergic reaction following implantation, injection, topical application, or ingestion in the human body [8]. However, both polymers are not stable in aqueous environment and this limitation can impair their mid-, long-term efficacy, deeply affecting the releasing characteristics. In order to overcome this drawback, cross-linking represents a possible

PT

solution, which implies, however, a critical selection of the chemicals to be usedin order to

RI

minimizetoxic side-effects. In this study, genipin was considered as a cross-linker agent, because of

SC

its properties: (i) it is naturally derived from the fruits of Gardenia Jasminoides Ellis, (ii) it is about 10,000 times less toxic than glutaraldehyde, and (iii) it elicits a moderate in vivo inflammatory

NU

response [9-11].

To assess the suitability of this proposal, methylene blue (MB), a hydrophilic tricyclic

MA

phenothiazine drug, was selected, which is the very first fully synthetic compound ever used for different clinical applications, including photodynamic therapy and antimicrobial treatment [12-16].

D

A specific role for MB can be also highlighted in angiogenesis, a key-process that should be strictly

PT E

evaluated, being implicated, e.g., in both normal development, tumor growth and metastasis, inflammation and wound repair, and intra-abdominal adhesions after surgical procedures [17]. In

CE

this regard, the potential of MB has been already assessed, inducing a significant reduction on histopathologically determined adhesion markers and affecting angiogenesis through platelet-

AC

derived growth factor [18]. This means that MB could represent a suitable pharmaceutical option, especially if topically delivered, when a tailored strategy to control new blood vessels formation is needed. Moreover, and related to the need to finely control the delivery of a specific therapeutic agent, the influence of (i) a selected material on drug release was here investigated, and (ii) a straightforward approach was proposed in order to modulate MB release by simply modifying gelatin/chitosan ratio in the blend formulation.

ACCEPTED MANUSCRIPT Morphological, thermal and chemical characteristics of gelatin, chitosan and blend microspheres were preliminarily evaluated. A mechanical characterization of the microspheres was also provided

AC

CE

PT E

D

MA

NU

SC

RI

PT

by applying nanoindentation. Finally, drug delivery and angiogenic assays were performed.

ACCEPTED MANUSCRIPT 2. MATERIALS AND METHODS

2.1 Materials Gelatin (type A, from porcine skin), chitosan (molecular weight: 50,000-190,000 Da) and MB (C16H18ClN3S·3H2O) were supplied by Sigma-Aldrich. Acetic acid was supplied by Carlo Erba

PT

Reagenti. Phosphate buffer saline (PBS) tablets were supplied by Gibco, Invitrogen Corporation.

RI

Genipin was supplied by Wako. Gelatin sponges (Gelfoam®) were supplied by Upjohn (Kalamazoo

SC

Mich, USA).

NU

All materials and reagents were used as received.

2.2 Microspheres preparation

MA

Microspheres were prepared by water-in-oil emulsion. Gelatin powder was added to distilled water, heated at 80 °C and gently stirred until complete dissolution, to produce a 10% w/v polymeric

D

solution. In order to cross-link gelatin, genipin at three different concentrations (0.1%, 0.5% or 1%

PT E

w/v), was added to the solution and stirred at 750 rpm for 1 min at 80 °C. The obtained solution was then poured into a test tube and mixed with a vortex mixer for 30 seconds at 2500 rpm.

CE

Using a syringe fitted with a 22 G needle, the resulting solution was injected into 200 ml of olive oil, which had been preheated at 80 °C. The resulting emulsion was stirred, at the same temperature,

AC

for 1 h at 800 rpm. Microspheres were collected onto a 22 µm mesh nylon filter in vacuum condition, and washed with acetone to remove oil residuals. Finally, the microspheres were air dried. Chitosan microspheres were prepared using the same protocol as reported above, with the exception that the polymer was dissolved in aqueous acetic acid solution (1% v/v) to produce a 2% w/v polymeric solution.

ACCEPTED MANUSCRIPT Gelatin-chitosan microspheres were produced according to the same method of fabrication, but this time different volumes of the above prepared solutions were mixed to obtain two different gelatin/chitosan blend systems, labelled blend 5/1 and blend 5/3, respectively.

2.3 Morphological analysis

PT

The morphological characterization of microspheres was performed by means of scanning electron

RI

microscopy (SEM; Leo Supra 35). Samples were sprinkled onto metal stubs using a double sided

SC

carbon tape and sputter coated with gold. The average microsphere diameter and size distribution were calculated from SEM micrographs by measuring about 100 microspheres randomly selected

NU

(ImageJ, NIH).

SEM investigation was also carried out to morphologically assess possible microsphere

MA

modifications during prolonged incubation period. For this aim, microspheres were soaked into PBS and after 10 and 30 days were recovered, washed with distilled water, dried and then observed

PT E

D

according to the same procedure above reported.

2.4 Fourier transform infrared spectroscopy analysis

CE

The chemical structure of microspheres was investigated by means of Fourier transform infrared spectroscopy (FTIR),by using a Perkin Elmer Spectrum 100. The samples were ground with KBr to

AC

prepare pellets by means of a hydraulic press. FTIR spectra were collected in the range 4000-400 cm-1 at a resolution of 4 cm-1.

2.5 Differential scanning calorimetry Differential scanning calorimetry (DSC) was carried out to investigate the thermal properties of microspheres. The samples were placed in aluminium pans and heated at 10 °C/min by means of a differential scanning calorimeter equipped with a thermal analysis data system (Nestch DSC 200 PC). DSC measurements were performed, under nitrogen atmosphere, within a specific heating

ACCEPTED MANUSCRIPT range, i.e., 30-250 °C for gelatin microspheres and 30-350 °C for chitosan and chitosan-gelatin microspheres, respectively. An empty aluminium pan was used as reference.

2.6 Mechanical characterization of microspheres through micro-compression testing Nanoindentation was performed to estimate the mechanical properties of gelatin, chitosan, and

PT

blend 5/1microspheres. In detail, the Nanoindenter XP (Agilent/MTS company), characterized by a

RI

theoretical force resolution of 50 nN and a theoretical displacement resolution lower than 0.01 nm,

SC

was adopted, and a specific protocol based on micro-compression tests was developed for the characterization of the mechanical behaviour of microspheres. For this aim, a flat end punch with a

NU

diameter of 500 µm was selected for indentation. Remembering that in nanoindentation, when the contact between specimen and indenter tip is detected, the displacement is measured as the load is

MA

applied, here the applied compression test was characterized by three steps: loading, hold and unloading.

D

During the loading and unloading phases the velocity of the indenter punch was set at a constant

PT E

value of 10 nm s-1. The hold phase consisted of a stabilization period along which the maximum value of the load was maintained for a period of 5s. The maximum value of the load was obtained,

diameter.

CE

for all microspheres, in correspondence of an indentation depth equal to the 5% of their initial

AC

On each measured load-displacement curve as obtained from the compression tests on single microspheres, the Hertz model [19] was applied and the Young’s modulus was estimated according to: 3

F

4 R E H 3 1   2  2 

(1)

where F is the load, H the displacement, E the Young’s modulus, υ the Poisson’s ratio and R the radius of the microsphere. The value of the Poisson’s ratio was assumed to be equal to 0.5 [20,21]. The least squares method was applied to estimate the value of E on the experimental data (relative

ACCEPTED MANUSCRIPT to the sole loading phase) of single microspheres by applying Eq. 1. Error analysis, in terms of mean percentage error, was performed to evaluate the quality of the fitting [22]. For each group of microspheres,40 indentation tests were carried out. Outliers have been excluded by applying the modified Thompson’s Tau method.

PT

2.7 Methylene blue loading

RI

MB powder was dissolved in PBS (0.1%w/v). Aliquots were withdrawn from the stock solution in

SC

order to treat cross-linked microspheres. The resulting suspensions were incubated at 37 °C for 1 day, at the end of this period the excess of MB solution was pipetted and the loaded microspheres

NU

were recovered and dried before to start the subsequent characterization. In order to evaluate the drug loading potential, loaded microspheres were soaked into 10 mL of PBS

MA

and ultrasonicated for 1 h, aliquots were then withdrawn and the MB concentration was measured by means of UV–Vis Spectrophotometer (UV-2450, Shimadzu, Japan) at 665 nm. This procedure

D

was repeated until 100% MB release. Drug loading was evaluated as the ratio between the amount

PT E

of released MB(mg), and the dry weight of microspheres (g).

CE

2.8 Characterization of drug release properties of microspheres Weighed samples of microspheres were immersed in 5 mL of PBS and kept into an incubator at 37

AC

°C. Aliquots of released solution were withdrawn periodically for measurements and replaced with fresh PBS in order to maintain sink conditions. MB concentration in the medium was measured by UV–Vis Spectrophotometer (UV-2450, Shimadzu, Japan) at 665 nm until the 100% MB temporal release was achieved. The percentage of cumulative drug release (% w/w) was investigated as a function of incubation time.

2.9 Assessment of microspheres angiogenic properties

ACCEPTED MANUSCRIPT The chicken embryo chorioallantoic membrane (CAM) assay was used as in ovo model to evaluate the angiogenic properties of MB loaded microspheres. Firstly, 7 mg of microspheres were incubated in 5 ml of PBS at 37 °C for 1 day, then the treated PBS was recovered to be used for CAM assays. Fertilized White Leghorn chicken eggs (n = 3 for each condition) were incubated under constant humidity at 37 °C. On incubation day 3, a square window was opened in the shell to detach the

PT

developing CAM after removal of 2-3 ml of albumin. The window was sealed with a glass, and the

RI

eggs were returned to the incubator. At day 8 of incubation, 1 mm3 sterilized gelatin sponges

SC

(Gelfoam®) were loaded with MB-released PBS and placed on the CAM. 1 mm3 sterilized gelatin sponges containing vehicle alone (PBS) were used as negative controls, while sponges loaded with

NU

MB stock solution were used as positive controls, as previously described [23,24]. All procedures were performed under sterile conditions. CAMs were examined daily until day 12 and

MA

photographed in ovo with a stereomicroscope equipped with a camera and image analyzer system (Olympus Italia, Milan, Italy). At day 12 the angiogenic response was evaluated as the number of

PT E

D

vessels converging toward the implants.

2.10 Statistics

CE

Results are expressed as mean ± standard deviation. Assays were performed in triplicate. Data analysis was performed with nonparametric tests (SPSS 19.0, SPSS Inc., USA). Statistical analysis

AC

to assess differences between groups was performed in two steps. First, data were compared by using the Kruskal-Wallis nonparametric test. If significant differences were found, groups were compared individually by using the Mann-Whitney U test. Statistical significant level was set at p< 0.05.

ACCEPTED MANUSCRIPT 3. RESULTS AND DISCUSSION

3.1 Microspheres dimension characterization Polymeric microspheres resulted to be homogenous and free of defects (Fig. 1). The average dimensions are summarized in Table I, no significant differences were observed comparing groups.

PT

Due to the preparation method here considered, the resulting microspheres were affected by a clear

RI

degree of dispersion. The dynamic process of droplet formation by emulsion approaches a steady

SC

state droplet size distribution within a period of several minutes or longer, typically 10-30 minutes. The collection of homogenous microspheres depends on several parameters, but generally, a

NU

uniform mixing process provides uniform droplet size [25].Droplet size has an effect on many expected characteristics, such as drug release, therefore a narrow size distribution concurs to a

MA

subsequent homogeneous response. However, polymeric microspheres within the here obtained size range can support a number of different applications. It was reported that microspheres in the range

D

20-100 µm can be used for intramuscular administration, as they can be retained in the interstitial

PT E

tissue acting as sustained release depots [4].In addition, Adhirajan et al [26] reported that microspheres within the same range can be also subcutaneously injected. The obtained

CE

microspheres can be also used for intra-articular drug delivery or incorporated into scaffold, an approach often adopted in tissue engineering. In this regard, Larsen et al. [27] reviewed the

AC

potential administration of microspheres (1-70 µm) following the intra-articular route, whereas Kojima et al.[28] succeeded in developing helical engineered cartilage equivalents for a functional tracheal replacement, including TGF-β2 loaded gelatin microspheres (75 µm average diameter) into nonwoven fibrous polyglycolic acid mesh seeded with bone marrow stromal cells. Kim et al. [29] developed a porous chitosan scaffold containing chitosan microspheres, ranging from 0.2 to 1.5 µm, for the successful regeneration of damaged cartilage. In vitro tests demonstrated that this approach concurred to promote cell proliferation and enhance extracellular matrix production compared to the control case (scaffolds without TGF-β1-loaded microspheres).

ACCEPTED MANUSCRIPT SEM analysis revealed that chitosan and blend microspheres, irrespective of polymer ratio, did not present substantial degradation during the observation period (data not shown). Conversely, gelatin microspheres were affected by a morphological modification which was related to genipin content and soaking period in PBS (Fig. 2). These results suggested that the presence of chitosan effectively contribute to deal with a stable formulation, which can be in turn ascribed to a more effective cross-

PT

linking compared to the gelatin case. The more stable formulation due to the chitosan concerns the

RI

presence of an amino group for each component of the polysaccaride chain. The number of NH2 in

SC

chitosan is greater respect to that present in protein as gelatin, because in proteins the NH2 groups available for the reaction with genepin are those of lisine and arginine or from the terminal amino

NU

acid This observation can contribute to define and/or select stable polymer substrates within a

MA

predictable timeframe, strictly related to the therapeutic application.

3.2 Spectroscopic characterization

D

In order to investigate the effect of cross-linking on possible modifications of the structural

PT E

rearrangement of polymeric chains, FTIR analysis was carried out. Acquired spectra of gelatin microspheres are shown in Fig. 3A. Neat gelatin showed the characteristic amide A peak at 3300

CE

cm-1, related to the N─H stretching mode [30]. A peak shift to higher wavenumbers, directly related to genipin concentration, was detected for cross-linked gelatin microspheres. Typical peaks of

AC

amide I (predominantly C=O stretching mode, with contributions from in-phase bending of the N─H bond and stretching of the C─N bond) and III (C─N stretching mode) were observed at 1650 cm-1 and 1240 cm-1, respectively [30,31], with no significant shifts for all the investigated cases. No relevant differences were detected for the amide II peak at about 1540 cm-1(N─H bending mode). FTIR spectra of chitosan microspheres are shown in Fig.3B. An absorption band at about 3400 cm-1, corresponding to the stretching vibration of ─NH2 and ─OH groups, was clearly detected [32]. It can be observed a peak broadening that decreases with genipin concentration, showing a sharper signal. Similarly, the peak at about 2925 cm-1 [32], representative of C─H stretch vibration, was

ACCEPTED MANUSCRIPT more pronounced in cross-linked chitosan microspheres compared to uncross-linked ones. Neat chitosan showed a peak at 1650 cm-1 characteristic of amide I (C═O) [33] which appeared broader for cross-linked microspheres, and could be attributed to NH2 group deformation [34]. Around 1425 cm-1 neat chitosan showed a peak related to the C─N stretching mode[34], which was slightly shifted to 1435 cm-1 and more pronounced for cross-linked chitosan microspheres. A peak at 1150

PT

cm-1, which corresponded to C─O─C asymmetric stretching, characterized both uncross-linked and

RI

cross-linked chitosan[35,36]. Moreover, cross-linked chitosan presented a more pronounced peak at 1070 cm-1 than uncross-linked one, being representative of C─O stretching vibration[32]. Finally,

SC

the absorption band at about 896 cm-1 is characteristic of saccharide structure of chitosan [32].

NU

Blend microspheres were prepared by mixing different content of gelatin and chitosan. Blending two or more polymers is a simple and effective technical approach to develop alternative

MA

biomaterials characterized by a combination of tailored properties of parent polymers. The analysis of FTIR spectra can allow to assess the interactions that possibly take place as a result of this

D

fabrication procedure [37]. FTIR spectra of blend microspheres are shown in Fig. 3C, compared to

PT E

those of gelatin and chitosan, cross-linked with the same genipin concentration (i.e., 1% w/v). The results obtained cannot indicate the formation of new groups in blend microspheres, obviously

CE

owing to their same main functional groups [38].The spectra are mainly similar to those of gelatin, the latter being the dominant component. Only a shift from about 3400 to 3300 cm-1 due to N─H

AC

and O─H stretching vibration was detected, while the other characteristic gelatin peaks were observed in blend spectra without significant differences.Typically, a shift of the broad 3600-3000 cm-1 band towards lower wavenumbers has been described as indicative of water-mediated hydrogen bonding [39]. It was likely that the energy of interactions involving NH and OH groups in hydrogen-bonding with H2O decreased, because the chains of both polymers required conformational changes to be able to form a blend [39].

3.3 Thermal characterization

ACCEPTED MANUSCRIPT DSC analysis was performed to evaluate the incidence of the cross-linking process on the collected polymeric microspheres. The acquired results are summarized in Table II, while spectra are shown in Fig. 4. Collageneous materials exhibit an endothermic peak, which can be referred to the break-up of hydrogen bonds and rearrangement of the triple helix into a random configuration [40]. This peak

PT

has often been termed as denaturation temperature (Td) [41] and its rise indicates an increase in the

RI

protein network strength, with the consequence of an increase in the cross-linking degree [42].

SC

Cross-linked gelatin microspheres were characterized by a thermal shift of the endothermic peak at higher temperatures, being related to the interchain reaction induced by genipin. Similarly, Bigi et

NU

al. (2002) [43] investigated the thermal properties of gelatin films cross-linked with genipin at different concentrations, from 0.07% to 2%, which represented the limit of solubility in the

MA

examined conditions. The results showed a slight but appreciable increase of Td, from 86 to 98 °C, indicating that the cross-linking treatment enhanced the thermal stability of the investigated films.

D

Yao et al. [44] considered the glass transition temperature as an indicator of cross-linking degree,

PT E

which was taken as the temperature of the midpoint of the increment of the specific heat capacity of the transitions. Calorimetric measurements were carried out on cross-linked gelatin at four different

CE

genipin weight percentages, i.e. 0.05, 0.1, 0.5, and 1.0 wt %. The authors found that the glass transition temperature increased with the concentration of genipin, while no significant shift was

AC

observed when the concentration of the cross-linker exceeded 0.5 wt%. In addition, a thermal analysis performed by Baiguera et al. [45] also highlighted the influence of cross-linking on the increase of the main endothermic peak temperature of genipin-treated gelatin scaffolds, produced by electrospinning, from 117.2 °C to 121.0 °C for neat and cross-linked mats, respectively. Chitosan microspheres were characterized by two different thermal events. A prominent endothermic peak, corresponding to the loss of water, and an exothermic peak due to the decomposition of the polymer were clearly observed. Polysaccharides usually have a strong affinity for water and these macromolecules may have disordered structures, which can be easily hydrated.

ACCEPTED MANUSCRIPT It is well known that the hydration properties of polysaccharides depend on the primary and supramolecular structures [46]. Therefore, the endothermic feature related to the evaporation of bound water was expected to reflect the molecular changes due to the cross-linking reaction [47]. As reported in Table II, its position slightly decreased from neat chitosan to 1% w/v genipin crosslinked chitosan microspheres. The subsequent recorded exothermic transition can be related to the

PT

decomposition of polymer networks [47], which occurred at higher temperatures for uncross-linked

RI

chitosan compared to cross-linked ones. In agreement with the here reported results, Yang et al.

SC

[48] observed that the exothermic decomposition peak of neat chitosan membrane was at about 290 °C, whereas that of glutaraldehyde cross-linked chitosan membrane was at about 271 °C, indicating

NU

that the thermal stability was reduced. These results can be confirmed by thermogravimetric analysis, which showed that the cross-linked chitosan started to decompose at lower temperature

MA

compared to uncross-linked one [49,50]. A possible explanation for the decrease of thermal stability was proposed by Neto et al. [49] and can be related to the formation of intracross-linking reactions

D

between polysaccharides chains. These reactions might interfered with previously existing attractive

PT E

hydrogen bonds in those regions where cross-linking occurred. As a consequence, the cross-linked polymer structure weakened, reducing its thermal stability.

CE

Finally, regarding blend 5/1 microspheres, a clear endothermic peak in the range 100-110 °C and an exothermic shoulder at about 255-257 °C were observed, the latter being related to the chitosan

AC

content, which was only one fifth with respect to gelatin and can be mainly related to the degradation of this polymer [51]. Increasing the chitosan content, blend 5/3 microspheres showed again a distinct exothermic event, similarly to chitosan, and characterized by a comparable trend in the thermal position of the related peak. Differently from the previous case, in which genipin concentration seemed not to significantly affect the measured thermal properties, in this case it is possible to observe a direct relationship with the cross-linker, even for the endothermic peak. In particular, the thermal behaviour of blend 5/3 showed peculiar characteristics if related to parent polymers in which a synergistic effect seems to be obtained leading to an increased stability both

ACCEPTED MANUSCRIPT for the endothermic and exothermic events. Reasonably, this result needs to be further investigated in order to fully evaluate the incidence of the materials involved in this formulation, and their possible mutual relationship.

3.4 Mechanical characterization compression

load-displacement

curves

as

obtained

from

PT

Explanatory

nanoindentation,

RI

representative of all the tests performed on single microspheres, are presented in Fig. 5.The load-

SC

displacement curves, irrespective of composition and genipin concentration, are characterized by an hysteresis loop which is typical of viscoelastic materials. More in detail, the hysteresis loops

NU

indicate that part of the energy accumulated by the deformed microsphere during the loading phase is dissipated, independent of the microsphere preparation, as a consequence of the degrees of

MA

freedom of polymer chains. Therefore, roughly speaking, higher energy loss does indicate a more marked cross-linking, definitely resulting in a higher number of constraints to polymer chains to

D

move, as previously reported [52]. According to this, it can be observed that in general

PT E

microspheres exhibit an hysteresis loop with an increasing area (and the highest strength, as a consequence), when genipin concentration increases from 0.1 to 1% (Fig. 5).

CE

It should be noticed that, during nanoindentation compression tests, the punch is in intimate contact with the microsphere. However, in the final stage of the unloading phase the velocity at which the

AC

indenter is withdrawn to the initial position becomes higher than the deformation recovery rate of the microsphere, thus leading to a contact loss. This is the reason why the curves in Fig. 5 seem to be characterized by a residual deformation. This phenomenon is due to the viscous behaviour of polymers used for the fabrication of the here presented microspheres. The result of a more quantitative characterization of the mechanical properties of the microspheres is summarized in Fig. 6, where the average values of the Young’s modulus for gelatin, chitosan and blend cross-linked microspheres, as obtained by fitting the experimental data with Eq 1, are shown.

ACCEPTED MANUSCRIPT In general, the mean value of the Young modulus of the different microsphere preparations ranges between 1.7 GPa (blend 5/1, 0.5%w/v) and 3.43 GPa (gelatin, 1%w/v), which is a moderate range of variation. In parallel, the statistical analysis highlights that polymers and cross-linker at different concentrations can result in significantly different (even if moderate) mechanical behaviour of microspheres. In particular, it can be noticed from Fig. 3 that: (i) in gelatin microspheres, the

PT

progressive increase in the percentage of genipin over 0.5% results in a statistically significant

RI

stiffening, (ii) chitosan microspheres become significantly stiffer for genipin concentration higher

SC

than 0.1%, thus highlighting a more marked role of cross-linking, with respect to gelatin, in determining microspheres mechanical properties, and (iii) increasing the percentage of genipin from

NU

0.1 to 1% in blend 5/1 composition leads to statistically significant differences, thus highlighting that blending could be used to appropriately modulate microspheres mechanical properties, even

MA

within a moderate range of variation. Error analysis highlights that there is an overall satisfactory agreement between Hertz model and experimental data (the obtained mean percentage errors are

D

below the 6.9%), thus confirming that Eq. 1 can provide a robust estimation of microspheres Young

PT E

modulus.

The strategy for the mechanical characterization adopted in this study, based on nano-

CE

indentation, is applied to the final product, i.e., the microspheres. With this approach an estimate of the viscous and elastic properties of the microsphere, as a whole, was provided. In

AC

detail, nano-indentation makes the application of large deformations to the microsphere possible, thus allowing the evaluation of its mechanical properties both in the linear and in the non-linear field. This is because loads made available in nano-indentation analysis are in the order of magnitude of mN, markedly far from the loads applied using, e.g., AFM, which is having large application for the characterization of microspheres [53,54]. However, it should be reminded here that in the presence of a multiphase material, if a precise value of the hardness of the material is needed, nano-indentation is the technique of election to be applied, AFM being more destined to get a mapping of the surface.

ACCEPTED MANUSCRIPT The implications for the applied strategy are evident, i.e., the implemented nano-indentationbased test bench allowed the characterization of the mechanical behavior of the microsphere, as dictated by its structural composition and shape, and not, e.g., the characterization of the

3.5 Microsphere characterization to MB loading and releasing

PT

material(s) used for its fabrication.

RI

Following the MB loading protocol here reported, collected microspheres revealed a peculiar

SC

response depending on the nature of the materials considered. Irrespective of genipin concentration (i.e., differences were not significant), gelatin was characterized by a drug loading of 22.4±3.4

NU

mg/g, while chitosan 14.8±4.6 mg/g, the lowest value. This result can be explained considering the electrostatic interactions between the drug and the polymers, as discussed below regarding the

MA

releasing profile. Not surprisingly, the blend cases revealed an intermediate behaviour correlated to the gelatin/chitosan ratio, being 19.0±4.3 mg/g for blend 5/1, and 15.9±1.9 mg/g for blend 5/3,

D

respectively.

PT E

The cumulative MB release from polymeric microspheres is shown in Fig. 7. Each system here considered was allowed to completely deliver the encapsulated drug, giving the possibility to

CE

experimentally evaluate the effective therapeutic window associated to a specific drug. Depending on the characteristic nature of gelatin and chitosan in relation with the selected drug, it is possible to

AC

modulate the time course of the release of a therapeutic agent that can be further regulated by properly tailoring the ratio of the two polymers for the blend cases. MB is a cationic compound as well as chitosan. This feature can therefore explain the prolonged time release (28 ± 7 d, p<0.05 with respect to gelatin and blend) due to competitive effect of diffusion and electrostatic interaction. In fact, according to the first route MB should diffuse in the aqueous environment due to a gradient concentration, which acts as a driving force. On the other hand, the electrostatic repulsion between loaded MB and chitosan matrix counterbalanced this process, leading to an opposite force that limited the free diffusion outside each microsphere.

ACCEPTED MANUSCRIPT Gelatin is characterized by an isoelectric point of 7.0-9.0 and, referring to the medium used for this experimental investigation (i.e., PBS, pH=7.4), it is reasonable to assume that no net electrical charge was carried. In this case, diffusion was the main mechanism that regulated the MB delivery, being responsible for the fast and complete release measured within the very first days (2.7 ± 0.5 d, p<0.05 with respect to chitosan and blend). As expected, blend 5/1 microspheres showed an

PT

intermediate behaviour that was affected by the unbalanced gelatin/chitosan ratio, thus providing a

RI

complete release in a short period (10 ± 4 d), but slower when compared to neat gelatin

SC

microspheres. Blend 5/3 microspheres were characterized by a little longer delivery period (15.0 ± 5.5 d), even if not significantly different. The acquired result suggests that the initial burst and

NU

release rate can be modulated by increasing chitosan concentration in blend microspheres, succeeding in delivering the therapeutic agent to the target tissue according to the desired drug

MA

release profile.

In order to prevent a possible low encapsulation efficiency and a possible interference with the

D

cross-linking process and microsphere formation, generally related to one-step fabrication

PT E

processes, cross-linked microspheres were soaked into a MB solution for 1 day at 37 °C. The swelling characteristics derived from the hydrogel-like behaviour acquired by polymeric

CE

microspheres concurred to drug loading that, however, could have been limited in the presence of chitosan (as mentioned above with a similar, but opposite MB entrapping mechanism). This loading

AC

methodology also concurs to preserve the biological activity of drugs or proteins encapsulated within a delivery system so that the expected therapeutic effect can be provided. For instance, it has been demonstrated that activity loss is caused by denaturation and deactivation of proteins during microsphere fabrication, being very sensitive to harsh processing conditions (e.g., heat, organic solvents, and pH level) [55]. Furthermore, a direct drug loading could lead to its extraction from the delivery system during the washing steps to remove solvent residuals, thus resulting in an unsatisfactory drug loading. This issue was addressed by Hejazi and Amiji [56], showing a tetracycline content of 8% within chitosan microspheres, if added in the fabrication process, that

ACCEPTED MANUSCRIPT raised to 69% considering preformed microspheres. A similar protocol was followed by Tu et al. [57], loading MB into alginate microparticles and showing that the medium strongly influenced the release profile. In particular, during the observation period, MB release was incomplete in distilled water, while a near complete release was detected in 0.1 N HCl within 5 min. A complete release (within 30 min) was observed considering NaCl media, being dependent on salt concentration. The

PT

MB loading ratio of anionic starch microspheres cross-linked with sodium trimetaphosphate was

RI

influenced by loading time, dissolution medium, loading temperature and drug concentration,

SC

reaching its maximum in NaCl (0.9%) dissolution medium at room temperature [58]. A direct MB loading approach was considered to prepare starch/gelatin microparticles, but a poor drug loading

NU

efficiency was measured (29-32%), probably due to drug diffusion during microparticle solidification. Moreover, a relevant release was detected within 24 h ranging from 78% to 97%,

MA

depending on the ratio of the two polymers [59].

Finally, in order to verify the role of the temporal extent on the drug loading stage, microspheres

D

were also treated with MB solution for 3 and 7 days. The subsequent release measured at 30 and 60

CE

period (data not shown).

PT E

minutes showed that no significant differences were observed with respect to the 1 day loading

3.6 The chick chorioallantoic membrane assay

AC

The assessment of the influence of MB on angiogenesis was performed by chick chorioallantoic membrane (CAM) assay. Macroscopic observations of CAM treated with MB-loaded sponges showed that only chitosan and blend samples were surrounded by allantoic vessels that developed radially towards the implant in a spoke-wheel pattern. This occurrence was not verified for gelatin and the control case (MB solution incubated at 37 °C), clearly indicating an anti-angiogenic response (Fig. 8). The angiogenic effect on direct blood vessel growth was quantified as the total number of blood converging vessels (Fig. 9). MB is known to be an anti-angiogenic inducer, but this property can be modulated by accurately selecting the delivery substrate. According to the MB

ACCEPTED MANUSCRIPT release analysis, the burst release associated to gelatin microspheres can negatively affect the formation of new blood capillaries, while a more progressive drug delivery related to the presence of chitosan can mitigate this result. Angiogenesis is a crucial biological process that needs to be accurately assessed in order to promote a positive clinical outcome. For this aim, the CAM assay is an effective approach to study the

PT

performance of a pro/anti-angiogenic delivery system, being a valid, rapid, and cost-effective

RI

alternative to mammalian models [60,61]. A pathological disruption of this process can be a

SC

hallmark of both vascular insufficiency (e.g., myocardial or critical limb ischemia) and vascular overgrowth (e.g., malignant tumors, retinopathies, hemangiomas) [17]. Therefore, the tailored

NU

combination of MB with specific polymeric platforms as delivery systems can effectively contribute to improve and support potential therapeutic approaches for the treatment of angiogenesis-related

AC

CE

PT E

D

MA

diseases.

ACCEPTED MANUSCRIPT 4. Limitations of the study The here presented investigation shows a straightforward approach to prepare and characterize microspheres for drug delivery application by modulating the polymer ratio content in order to deal with different release periods. Even if the results clearly suggest the potential of the study, the following limitations should be considered as well.

PT

1. The preparation method does not allow to collect monodispersed microspheres and

RI

this could imply inhomogeneous characteristics, in terms, for instance, of mechanical

SC

or delivery properties;

2. The drug loading procedure could be lead to a MB incorporation in the outer layer of

NU

the microspheres, thus concurring to a substantial burst release (that could be

AC

CE

PT E

D

MA

controlled as shown in Fig. 7).

ACCEPTED MANUSCRIPT 5. CONCLUSIONS

This study presents a straightforward method to prepare microspheres made of natural polymers. In order to stabilize the resulting structures, genipin was used as a cross-linked agent at different concentrations while methylene blue was considered as a model drug to investigate the release

PT

properties of the proposed microspheres, showing a clear relationship with the selected biomaterials.

RI

Specifically, it was observed that (i) the concentration of cross-linking agent cannot or can

SC

moderately affect the mechanical properties of microspheres (in terms of Young’s modulus and depending upon polymers adopted for fabrication) and (ii) the delivery rate from blend

NU

microspheres suggested that it is possible to modulate the temporal release profile by simply varying the gelatin/chitosan ratio. The implication of this different behavior was underlined by the

MA

chick chorioallantoic membrane assay. Methylene blue is largely used for several clinical applications and is also known to have a role in neo-vascularization, typically inducing an anti-

D

angiogenic effect. However, this characteristic can be easily tailored depending on the polymer(s)

PT E

selected and considering that a successful topical delivery therapy is strictly related to the

Acknowledgments

CE

interaction between the polymeric matrix of the carrier and the loaded drug.

AC

CDG acknowledges the Intrauniversitary Consortium for Material Science and Technology (INSTM) for the Research Grant “Sviluppo di idrogeli per applicazioni nel campo della medicina rigenerativa”

ACCEPTED MANUSCRIPT References

[1] S. Harsha, R. Chandramouli, S. Rani S, Ofloxacin targeting to lungs by way of microspheres, Int. J. Pharm. 380 (2009) 127-132. [2] U. Edlund, A.C. Albertsson, Degradable polymer microspheres for controlled drug delivery,

PT

Adv. Polymer Sci.157 (2002) 67-112

RI

[3] Y.Z. Zhang, J. Venugopal, Z.M. Huang, C.T. Lim, S. Ramakrishna, Crosslinking of the

SC

electrospun gelatin nanofibers, Polymer 47 (2006) 2911-2917.

[4] E. Esposito, R. Cortesi, C. Nastruzzi, Gelatin microspheres: influence of preparation parameters

NU

and thermal treatment on chemico-physical and biopharmaceutical properties, Biomaterials 17 (1996) 2009-2020.

MA

[5] L. Solorio, C. Zwolinski, A.W. Lund, M.J. Farrel, J.P. Stegemann, Gelatin microspheres crosslinked with genipin for local delivery of growth factors, J. Tissue Eng. Regen. Med. 4

D

(2010) 514-523.

PT E

[6] R.A.A. Muzzarelli, Genipin-crosslinked chitosan hydrogels as biomedical and pharmaceutical aids, Carbohyd. Polym. 77 (2009) 1–9.

CE

[7] J. Mao, L. Zhao, K. De Yao,Q. Shang, G. Yang, Y. Cao, Study of novel chitosan-gelatin artificial skin in vitro J. Biomed. Mater. Res. A 64 (2003) 301-308.

AC

[8] E. Pianigiani, A. Andreassi, P. Taddeucci, C. Alessandrini, M. Fimiani, L. Andreassi, A new model for studying differentiation and growth of epidermal cultures on hyaluronan-based carrier, Biomaterials 20 (1999) 1689-1694. [9] H.W. Sung, R.N. Huang, L.L. Huang, C.C. Tsai, In vitro evaluation of cytotoxicity of a naturally occurring cross-linking reagent for biological tissue fixation, J. Biomater. Sci. Polym. Ed. 10 (1999) 63-78. [10] H.W. Sung, W.H. Chang, C.Y. Ma, M.H. Lee, Crosslinking of biological tissues using genipin and/or carbodiimide J. Biomed. Mater. Res. A 64 (2003) 427-438.

ACCEPTED MANUSCRIPT [11] K. Zhang, Y. Qian, H. Wang, L. Fan, C. Huang, A. Yin, X. Mo, Genipin-crosslinked silk fibroin/hydroxybutyl chitosan nanofibrous scaffolds for tissue-engineering application, J. Biomed. Mater. Res. A 95 (2010) 870-881. [12] S. Takeoka, T. Ohgushi, H. Nishide, E. Tsuchida, Reduction of methemoglobin via electron transfer across the bilayer membrane of Hb vesicles, Bull. Chem. Soc. Jpn.70 (1997) 1171-1178.

PT

[13] M. Wainwright, K.B. Crossley, Methylene Blue-a therapeutic dye for all seasons?,

RI

J.Chemother.14 (2002) 431-443.

SC

[14] J.P. Tardivo, A. Del Giglio, C.S. de Oliveira, D.S. Gabrielli, H.C. Junqueira, D.B. Tada, D. Severino, M.S. de Fátima Turchiello Rand Baptista, Methylene blue in photodynamic therapy:

NU

From basic mechanisms to clinical applications, Photodiagn. Photodyn. Ther. 2 (2005) 175-191. [15] S. Khan, F. Alam, A. Azam, A.U. Khan, Gold nanoparticles enhance methylene blue-induced

Nanomedicine 7 (2012) 3245-3257.

MA

photodynamic therapy: a novel therapeutic approach to inhibit Candida albicans biofilm, Int. J.

D

[16] C.A. Pereira, R.L. Romeiro, A.C. Costa, A.K. Machado, J.C. Junqueira, A.O. Jorge,

PT E

Susceptibility of Candida albicans, Staphylococcus aureus, and Streptococcus mutans biofilms to photodynamic inactivation: an in vitro study, Lasers Med. Sci. 26 (2011) 341-348.

CE

[17] N. Zacharakis, P. Tone, C.S. Flordellis, M.E. Maragoudakis, N.E. Tsopanoglou, Methylene blue inhibits angiogenesis in chick chorioallontoic membrane through a nitric oxide-independent

AC

mechanism, J. Cell. Mol. Med. 10 (2006) 493-498. [18] A. Boztosun, A.Piçnak, M.I. Kosar, S. Gulturk, A. Cetin, Effects of methylene blue, pentoxyphylline and enoxaparin on postoperative adhesion formation and markers of angiogenesis in a rat uterine horn model, Clin. Exp. Obstet. Gynecol. 39 (2012) 89-95. [19] H. Hertz 1882 Uber die Beruhrung tester Elastische Korper und Uber die Harte (On th Contact of Rigid Elastic Solids and on Hardness) Verhandlungen des Vereinszur Beforderung des (For English translation see Miscellaneous Papers by H Hertz, Eds. Jones and Schott, MacMillan, London, 1896)

ACCEPTED MANUSCRIPT [20] C.X. Wang, C. Cowen, Z. Zhang, C.R. Thomas, High speed compression of single alginate microspheres, Chem. Eng. Sci. 60 (2005) 6649–6657. [21] K. Hidaka, L. Moine, G. Collin, D. Labarre, J.L. Grossiord, N. Huang, K. Osuga, S. Wada, A. Laurent, Elasticity and viscoelasticity of embolization microspheres, J. Mech. Behav. Biomed. Mater. 4 (2011) 2161–2167.

PT

[22] A.J. Wheeler, A.R. Ganji, Introduction to engineering experimentation. (2009) Third Edition.

RI

[23] Ribatti D, Nico B, Vacca A and Presta M 2006 The gelatin sponge-chorioallantoic membrane

SC

assay, Nat. Protoc. 1 85-91

[24] S. Baiguera, P. Macchiarini, D. Ribatti, Chorioallantoic membrane for in vivo investigation of

NU

tissue engineered construct biocompatibility, J. Biomed. Mater. Res. B Appl. Biomater.100 (2012) 1425-1434.

MA

[25] R. Arshady, Albumin microspheres and microcapsules: methodology of manufacturing techniques, J. Control. Release 14 (1990) 111-131.

D

[26] N. Adhirajan, N. Shanmugasundaram, M. Babu M, Gelatin microspheres cross-linked with

PT E

EDC as a drug delivery system for doxycyline: development and characterization, J. Microencapsul.24 (2007) 647-659.

CE

[27] C. Larsen, J. Ostergaard, S.W. Larsen, H. Jensen, S. Jacobsen, C. Lindegaard, P.H. Andersen, Intra-articular depot formulation principles: role in the management of postoperative pain and

AC

arthritic disorders, J. Pharm. Sci. 97 (2008) 4622-4654. [28] K. Kojima, R.A. Ignotz, T. Kushibiki, K.W. Tinsley, Y. Tabata, C.A. Vacanti, Tissueengineered trachea from sheep marrow stromal cells with transforming growth factor β2 released from biodegradable microspheres in a nude rat recipient, J. Thorac. Cardiovasc. Surg.128 (2004) 147-153. [29] S.E. Kim, J.H. Park, Y.W. Cho, H. Chung, S.Y. Jeong, E.B. Lee, I.C. Kwon, Porous chitosan scaffold containing microspheres loaded with transforming growth factor-beta1: implications for cartilage tissue engineering, J. Control. Release 91 (2003) 365-374.

ACCEPTED MANUSCRIPT [30] R. Cortesi, C. Nastruzzi, S.S. Davis, Sugar cross-linked gelatin for controlled release: microspheres and disks, Biomaterials 19 (1998) 1641-1649. [31] C. Del Gaudio, S. Baiguera, M. Boieri, B. Mazzanti, D. Ribatti, A. Bianco, P. Macchiarini, Induction of angiogenesis using VEGF releasing genipin-crosslinkedelectrospun gelatin mats, Biomaterials 34 (2013) 7754-7765.

PT

[32] W. Zeng, J. Huang, X. Hu, W. Xiao, M. Rong, Z. Yuan, Z. Luo, Ionically cross-linked

RI

chitosan microspheres for controlled release of bioactive nerve growth factor, Int. J. Pharm. 421

SC

(2011) 283-290.

[33] J.Kawadkar, M.K. Chauhan, Intra-articular delivery of genipin cross-linked chitosan

NU

microspheres of flurbiprofen: preparation, characterization, in vitro and in vivo studies, Eur. J. Pharm.Biopharm. 81 (2012) 563-572.

MA

[34] M.S. Austero, A.E. Donius, U.G. Wegst, C.L. Schauer, New crosslinkers for electrospun chitosan fibre mats. I. Chemical analysis, J. R. Soc. Interface. 9 (2012) 2551-2562.

D

[35] J.Y. Lai, Y.T. Li, T.P. Wang, In vitro response of retinal pigment epithelial cells exposed to

PT E

chitosan materials prepared with different cross-linkers, Int. J. Mol. Sci. 11 (2010) 5256-5272. [36] V. Chiono, E. Pulieri, G. Vozzi, G. Ciardelli, A. Ahluwalia, P. Giusti, Genipin-crosslinked

898.

CE

chitosan/gelatin blends for biomedical applications J. Mater. Sci. Mater. Med. 19 (2008) 889-

AC

[37] B.A. Ibrahim, K.M. Kadum, Influence of polymer blending on mechanical and thermal properties, Mod. Appl. Sci. 4 (2010) 157-161. [38] H. Chen, Z. Zhao, Y. Zhao, Y. Yang, Fabrication and evaluation of chitosan-gelatin based buckling implant for retinal detachment surgery J. Mater. Sci. Mater. Med.21 (2010) 2887-2895. [39] H. Staroszczyk, K. Sztuka, J. Wolska, A. Wojtasz-Pajak, I. Kolodziejska, Interactions of fish gelatin and chitosan in uncross-linked and cross-linked with EDC films: FT-IR study, Spectrochim. Acta A. 117 (2014) 707-712.

ACCEPTED MANUSCRIPT [40] A. Bigi, G. Cojazzi, S. Panzavolta, K. Rubini, N. Roveri, Mechanical and thermal properties of gelatin films at different degrees of glutaraldehyde crosslinking, Biomaterials 22 (2001) 763768. [41] Y. Su, X. Mo, Genipin crosslinked gelatin nanofibers for tissue engineering, J. Control. Release 152 Suppl 1 (2011) e230-232.

PT

[42] C. Tonda-Turo, P. Gentile, S. Saracino, V. Chiono, V.K. Nandagiri, G. Muzio, R.A. Canuto,

SC

coupling agent, Int. J. Biol. Macromol. 49 (2011) 700-706.

RI

G. Ciardelli, Comparative analysis of gelatin scaffolds crosslinked by genipin and silane

[43] A. Bigi, G. Cojazzi, S. Panzavolta, N. Roveri, K. Rubini, Stabilization of gelatin films by

NU

crosslinking with genipin, Biomaterials 23 (2002) 4827-4832.

[44] C.H. Yao, B.S. Liu, C.J. Chang, S.H. Hsu, Y.S. Chen, Preparation of networks of gelatin and

MA

genipin as degradable biomaterials, Mater. Chem. Phys. 83 (2004) 204–208. [45] S. Baiguera, C. Del Gaudio, E. Lucatelli, E. Kuevda, M. Boieri, B. Mazzanti, A. Bianco, P.

D

Macchiarini, Electrospun gelatin scaffolds incorporating rat decellularized brain extracellular

PT E

matrix for neural tissue engineering, Biomaterials 35 (2014) 1205-1214. [46] F.S. Kittur, K.V. Harish Prashanth, K. Udaya Sankar, R.N. Tharanathan, Characterization of

CE

chitin, chitosan and their carboxymethyl derivatives by differential scanning calorimetry, Carbohydr. Polym. 49 (2002) 185-193.

AC

[47] S.J. Chang, G.C. Niu, S.M. Kuo, S.F. Chen, Preparation and preliminary characterization of concentric multi-walled chitosan microspheres, J. Biomed. Mater. Res. A 81 (2007) 554-566. [48] J.M. Yang, W.Y. Su, Preparation and characterization of chitosan hydrogel membrane for the permeation of 5-Fluorouracil, Mater. Sci. Eng. C 31 (2001) 1002-1009. [49] C.G.T. Neto, J.A. Giacometti, A.E. Job, F.C. Ferreira, J.L.C. Fonseca, M.R. Pereira, Thermal analysis of chitosan based networks, Carbohyd. Polym. 62 (2005) 97–103. [50] D. Thacharodi, K.P. Rao, Propranolol hydrochloride release behaviour of crosslinked chitosan membranes, J. Chem. Technol. Biotechnol. 58 (1993) 177-181.

ACCEPTED MANUSCRIPT [51] W.W. Thein-Han, J. Saikhun, C. Pholpramoo, R.D. Misra, Y. Kitiyanant, Chitosan-gelatin scaffolds for tissue engineering: physico-chemical properties and biological response of buffalo embryonic stem cells and transfectant of GFP-buffalo embryonic stem cells, Acta Biomater. 5 (2009) 3453-3466.

linked actin networks, Phys. Rev. Lett. 101 (2008) 108101.

PT

[52] O. Lieleg, M.M. Claessens, Y. Luan, A.R. Busch, Transient binding and dissipation in cross-

RI

[53] E. Glynos, V. Koutsos, W.N. McDicken, C.M. Moran, S.D. Pye, J.A. Ross, V. Sboros,

SC

Nanomechanics of biocompatible hollow thin-shell polymer microspheres, Langmuir 25 (2009) 7514-75122.

NU

[54] S. Tan, R.L. Sherman Jr, W.T. Ford, Nanoscale compression of polymer microspheres by atomic force microscopy, Langmuir 20 (2004) 7015-7020.

MA

[55] Y. Tabata, Y. Ikada, Surfactant-free preparation of biodegradable hydrogel microspheres for protein release, J. Bioact Compat Polym. 14 (1999) 371-384.

D

[56] R. Hejazi, M. Amiji, Stomach-specific anti-H. pylori therapy. I. Preparation and

PT E

characterization of tetracyline-loaded chitosan microspheres, Int. J. Pharm. 235 (2002) 87–94. [57] J. Tu, S. Bolla, J. Barr, J. Miedema, X. Li, B. Jasti, Alginate microparticles prepared by spray-

(2005) 171-181.

CE

coagulation method: preparation, drug loading and release characterization, Int. J. Pharm. 303

AC

[58] Y. Fang, L. Wang, D. Li, B. Li, B. Bhandari, X.D. Chen, Z. Mao, Preparation of crosslinked starch microspheres and their drug loading and releasing properties, Carbohydr. Polym. 74 (2008) 379-384. [59] T. Phromsopha, Y. Baimark, Preparation of starch/gelatin blend microparticles by a water-inoil emulsion method for controlled release drug delivery, Int. J. Biomater. 2014 (2014) 829490. [60] D. Ribatti, A. Vacca, L. Roncali, F. Dammacco, The chick embryo chorioallantoic membrane as a model for in vivo research on antiangiogenesis, Curr. Pharm. Biotechnol. 1 (2000) 73–82.

ACCEPTED MANUSCRIPT [61] A. Vargas, M. Zeisser-Labouèbe, N. Lange, R. Gurny, F. Delie, The chick embryo and its chorioallantoic membrane (CAM) for the in vivo evaluation of drug delivery systems, Adv.

AC

CE

PT E

D

MA

NU

SC

RI

PT

Drug. Deliv. Rev. 59 (2007) 1162-1176.

ACCEPTED MANUSCRIPT Figure legends

Fig. 1 – SEM micrographs of cross-linked gelatin (A), chitosan (B), blend 5/1 (C) and blend 5/3 (D) microspheres. Scale bar 10 µm.

PT

Fig. 2 – SEM micrographs of cross-linked gelatin microspheres soaked in PBS for 10 and 30 days.

RI

Scale bar 100 µm.

SC

Fig. 3 – FTIR spectra of gelatin (A), chitosan (B), and blend microspheres cross-linked at 1%w/v

NU

genipin concentration (C).

MA

Fig. 4 – DSC spectra of gelatin, chitosan, and blend microspheres cross-linked at 1%w/v genipin

D

concentration.

PT E

Fig. 5 – Explanatory load-displacement curves as obtained from nanoindentation tests performed on(A) gelatin, (B) chitosan, and (C) blend 5/1microspheres. For each polymer-based microsphere,

CE

data related to three different concentrations of cross-linking agent are presented.

AC

Fig. 6 – Average values of the Young’s modulus as obtained for gelatin, chitosan and blend 5/1 microspheres at three different concentrations of genipin. Statistical significant differences (p<0.05) between single microsphere preparations with different genipin concentration are also reported. The bars indicate the statistical influence of genipin and the p-values (1 - 0.0002; 2 - 4.87·10-7; 3 1.38·107; 4 - 2.32·10-9; 5 -0.0004; 6 - 6.23·10-11; 7 - 0.0074) obtained from the ANOVA analysis considering a confidence interval of the 95%.

Fig. 7 – MB release from microspheres cross-linked at 1% genipin concentration.

ACCEPTED MANUSCRIPT

Fig. 8 – Angiogenic response mediated by MB released from polymeric microspheres. Representative examples of chicken chorioallantoic membrane (CAM) to control case (A; MB), gelatin (B), chitosan (C), blend 5/1 (D), and blend 5/3 (E) samples. Scale bar 2 mm.

PT

Fig. 9 – Effect of MB released from polymeric microspheres on the number of converging blood

AC

CE

PT E

D

MA

NU

SC

RI

vessels. * p<0.05 with respect to chitosan, blend cases and negative control.

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Figure 1

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Figure 2

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Figure 3

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Figure 4

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Figure 5

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Figure 6

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Figure 7

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Figure 8

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Figure 9

ACCEPTED MANUSCRIPT

Table I. Average size of gelatin, chitosan and blend cross-linked microspheres

Gelatin

Chitosan

Average size [μm]

Average size [μm]

Blend 5/1

Blend 5/3

Average size [μm]

Average size [μm]

0.1

42.7 ± 17.1

54.0 ± 15.9

50.4 ± 19.9

42.0 ± 18.1

concentration

0.5

44.9 ± 22.1

50.2 ± 22.6

52.3 ± 26.5

43.0 ± 20.5

[% w/v]

1

53.63± 24.8

47.1 ± 27.9

RI

PT

Genipin

46.2 ± 27.7

NU

SC

52.8 ± 24.9

MA

Table II. Characteristic thermal events for gelatin, chitosan and blend cross-linked microspheres

Blend 5/3

Exothermic peak

Endothermic

Exothermic

Endothermic

Exothermic

peak [°C]

peak [°C]

[°C]

peak [°C]

peak [°C]

peak [°C]

peak [°C]

0§

114.5±0.7

117.4±2.7

310.9±0.2

0.1

118±6.3

116.3±2.6

257.3±1.8.*

104.7±6.3

255

118.0±1.1

278.8±5.8

123.6±6.0.*

117.2±0.3

262.1±4.9.*

103.6±3.4

255

139.1±5.6

286.7±3.0

127.5±5.6 *

107.0±1.3 *

269.0±3.8.*

108.2±2.5

257

142.5±8.1

290.0±2.6

0.5

AC

1

PT E

Endothermic

concentration [% w/v]

Blend 5/1

Endothermic

CE

Genipin

Chitosan

D

Gelatin

§

powder

* p<0.05 with respect to control (i.e., neat polymeric microspheres)

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Graphical abstract

ACCEPTED MANUSCRIPT Highlights

• Microspheres were produced with naturally-derived polymers (gelatin, chitosan) and cross-linking agent (genipin). • Collected microspheres were loaded with methylene blue (MB), as drug model.

PT

• Morphological analysis revealed homogeneous microspheres (diameters in the range 42-54 µm).

RI

• In vitro MB complete release occurred (about 3 days for gelatin, 30 days for chitosan microspheres).

NU

SC

• Polymers blending and cross-linking strategies allowed the proper modulation of microspheres mechanical properties.

AC

CE

PT E

D

MA

• Anti-angiogenic behavior was observed for gelatin microspheres.