Home
Search
Collections
Journals
About
Contact us
My IOPscience
Needle-like ion-doped hydroxyapatite crystals influence osteogenic properties of PCL composite scaffolds
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2016 Biomed. Mater. 11 015018 (http://iopscience.iop.org/1748-605X/11/1/015018) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 194.27.18.18 This content was downloaded on 05/03/2016 at 08:31
Please note that terms and conditions apply.
Biomed. Mater. 11 (2016) 015018
doi:10.1088/1748-6041/11/1/015018
PAPER
received
10 September 2015
Needle-like ion-doped hydroxyapatite crystals influence osteogenic properties of PCL composite scaffolds
re vised
28 October 2015 accep ted for publication
29 October 2015 published
29 February 2016
V Guarino1, F Veronesi2, M Marrese1, G Giavaresi2,3, A Ronca1, M Sandri4, A Tampieri4, M Fini2,3 and Luigi Ambrosio1 Institute of Polymers, Composites and Biomaterials, Department of Chemical Sciences & Materials Technology National Research Council of Italy, Mostra D’Oltremare, Pad.20, V. le Kennedy 54, 80125, Naples, Italy 2 Laboratory of Biocompatibility, Innovative Technologies and Advanced Therapies, Department Rizzoli RIT, Italy 3 Laboratory of Preclinical and Surgical Studies, Rizzoli Orthopaedic Institute, Via Di Barbiano 1/10, Bologna, Italy 4 CNR—Institute of Science and Technology for Ceramics (ISTEC), Via Granarolo, 64 48018 Faenza (RA), Italy 1
E-mail: [email protected] and [email protected] Keywords: composite scaffolds, needle-like hydroxyapatite, ion doping, mechanical properties, in vivo studies
Abstract Surface topography and chemistry both play a crucial role on influencing cell response in 3D porous scaffolds in terms of osteogenesis. Inorganic materials with peculiar morphology and chemical functionalities may be proficiently used to improve scaffold properties—in the bulk and along pore surface—promoting in vitro and in vivo osseous tissue in-growth. The present study is aimed at investigating how bone regenerative properties of composite scaffolds made of poly(Ɛ-caprolactone) (PCL) can be augmented by the peculiar properties of Mg2+ ion doped hydroxyapatite (dHA) crystals, mainly emphasizing the role of crystal shape on cell activities mediated by microstructural properties. At the first stage, the study of mechanical response by crossing experimental compression tests and theoretical simulation via empirical models, allow recognizing a significant contribution of dHA shape factor on scaffold elastic moduli variation as a function of the relative volume fraction. Secondly, the peculiar needle-like shape of dHA crystals also influences microscopic (i.e. crystallinity, adhesion forces) and macroscopic (i.e. roughness) properties with relevant effects on biological response of the composite scaffold: differential scanning calorimetry (DSC) analyses clearly indicate a reduction of crystallization heat—from 66.75 to 43.05 J g−1—while atomic force microscopy (AFM) ones show a significant increase of roughness—from (78.15 ± 32.71) to (136.13 ± 63.21) nm—and of pull-off forces—from 33.7% to 48.7%. Accordingly, experimental studies with MG-63 osteoblast-like cells show a more efficient in vitro secretion of alkaline phosphatase (ALP) and collagen I and a more copious in vivo formation of new bone trabeculae, thus suggesting a relevant role of dHA to support the main mechanisms involved in bone regeneration.
Introduction A large variety of bioactive materials (i.e. ceramics [1], glasses [2]) with different chemical groups (i.e. silicates [3], phosphates [4]), have been extensively investigated to improve the biological and biomechanical properties of scaffolds for bone tissue engineering [5]. They offer a unique chance to easily manipulate the chemical composition at the atomic level by the local substitution of ions (i.e. calcium, strontium, magnesium) in the 3D crystal [6], thus enabling one to reproduce the peculiar chemical composition of native mineral phases—the main component of bone tissue © 2016 IOP Publishing Ltd
and teeth—so actively contributing to the osteogenic activity of bone cells. Recent studies demonstrated that ion substituted apatites possess an excellent ability to form in vitro apatite [7], concurring to generate bioactive phases with mechanical properties comparable to those of native HA [8], which are suitable to assembly composite scaffolds in bone tissue engineering [9]. According to the natural bone structure—i.e. apatite crystals embedded in a collagen matrix—bioactive phases have been variously combined with synthetic or natural polymers (i.e. polyesters, polyurethanes, polysaccharides) to reproduce the functionalities of the bone
Biomed. Mater. 11 (2016) 015018
V Guarino et al
extracellular matrix (bECM), overcoming the innate brittleness of the ceramic phases [10, 11]. Among them, aliphatic polyesters such as PCL have been frequently preferred due to the slow degradation properties [12] which allow matching the formation rate of new bone, also preserving a mechanical strength in the range of trabecular bone [13]. In this case, PCL matrix also works as a binder for bioactive particles, providing a protective function which prevents any unexpected brittle failure of the final system [14]. However, this effect may often compromise the bioactive response of scaffold due to a not complete exposure of particles at the interface. So, the tendency of the bioactive particles to form clusters in the polymer matrix and the weak adhesion forces generally occurring among hydrophilic ceramic and hydrophobic polymer phases, still represent the main cause of mechanical failure of this type of composite scaffolds [15]. In this context, recent studies are emphasising the main role of osteo-inductive phases on influencing surface and interface properties (i.e. hydrophobicity, surface charge, roughness) in order to more directly address specific cell activities towards osteogenesys [16]. However, the ability of bioactive particles to affect surface properties of the scaffolds have been poorly focussed and only marginally correlated to in vitro and in vivo scaffolds response [17]. In a previous study, it has been investigated how chemical modification of HA crystals—i.e. doping by Mg2+ and CO32 − ion substitution—may influence spatial distribution of particles, thus affecting the hMSC behaviour in osteogenic way under in vitro and in vivo conditions [18]. More recently, it has been also elucidated the capability of ion doped HA crystals to promote the deposition of a biomimetic coating to improve composite scaffold bioactivity [19] at the pore surface. Here, we propose a comparative study of two different composite scaffolds composed of a PCL porous matrix reinforced by undoped HA (uHA) and Mg2+ ion doped HA (dHA) crystals, respectively. In particular, we have investigated how dHA and uHA with different aspect ratios influence macroscopic (i.e. porosity, mechanical response) and microscopic properties (i.e. roughness, adhesive forces) of scaffolds, thus indirectly addressing in vitro/in vivo response of the composite scaffolds, for the design of implantable devices for bone regeneration.
Experimental methods Materials Poly(ε-caprolactone) (PCL, Mn 42500 Da) was purchased from Sigma-Aldrich (Milan, Italy) and tetrahydrofurane (THF) from Baker (Italy). All chemicals were used as received by product companies. dHA particles were synthesized by ISTEC-CNR (Faenza, Italy) by simultaneously dropwise addition of acid aqueous solution (600 ml) containing 88.8 g H3PO4 (Aldrich, 85% pure) and 200 ml 0.8 M NaHCO3 2
(Merck, A.C.S., ISO) aqueous solution at 40 °C for 3 h, into a basic solution containing 100 g Ca(OH)2 (Aldrich, 95% pure) in 1000 ml of water by adding 48.4 g MgCl2 and 6H2O (Merck, A.C.S., ISO) in order to achieve a molar ratio of mineral phase equal to 5% [7, 18]. Precipitates were aged for 24 h at 25 °C, washed and filtered three times, lyophilized and finally sieved at 150 μm. Moreover, uHA particles (HAP205, theoretical density 3.16 g cm−3) with bimodal size distribution (Φ range = 4–11 μm, Φ 0.5 = 4.02 μm), measured via laser diffractometry were supplied by Plasma Biotal (Tideswell, UK). Scaffold preparation Briefly, 0.20 g of PCL granules were dissolved in 10 ml of THF and kept under magnetic stirring at 40 °C for 24 h. uHA or dHA crystals were added to PCL solution before the solvent extraction by different volume ratios, 13%, 20% and 26% v/v. To impart the required porosity, sodium chloride crystals—300–500 micron in size—with 90/10 salt/polymer volume ratio, were also included and, then, removed by leaching techniques. Briefly, the mixture was placed in anti-adhesive mould to confer the desired shape. Ethanol was used for 24 h to preliminary force the removal of used solvent and, then, bi-distilled water for 7 d to leach out salt crystals and any other contaminants. For all the experimental studies, uHA or dHA loaded specimens were labeled as uHAs and dHAs respectively, whereas unloaded specimens, namely PCL alone, as CTR. Crystal aspect ratio measurements The size and shape of uHA and dHA crystals were analysed by transmission electron microscopy (TEM, TECNAI120—FEI) supported by 2D/3D image analyses. On selected images with higher contrast and best focus, ImageJ Analysis Software was used to estimate particle length and elongation by calculating the shape factor (pAr), defined as the ratio between particle width and length. Samples for TEM were prepared by air-drying a drop of a sonicated ethanol suspension of particles onto a carbon-coated copper grid. Compressive tests At this preliminary stage, mechanical properties under compression were evaluated at room temperature by a dynamometric machine (Instron 4204) equipped with a 100 N load cell and a crosshead speed of 1 mm min−1. Cylindrical specimens with a 2:1 height/diameter ratio (diameter: 12 mm, height: 24 mm) were fabricated according to the ASTM 695/2a standard. In this case, uHAs and dHAs with three different filler/polymer volume ratio, 13:87, 20:80 and 26:74, have been tested in agreement with previous works [12], in order to evaluate the effective reinforcement of inorganic phases, minimizing the clustering effects. The study was carried out through analysis of five specimens for each scaffold type. The elastic modulus E was calculated
Biomed. Mater. 11 (2016) 015018
V Guarino et al
by the initial slope of the stress–strain curve before the plateau region, in the strain range of 0–0.1 mm mm−1. These values were also compared with those carried out via theoretical model based on open cubic cells (COC), by assuming composite scaffolds as high porous solid systems with a continuous and percolating network structure. In this case, elastic moduli were calculated, on the hypothesis of nearly elastic materials [20]. Moreover, densities of the single phases of composite scaffolds were corrected by their relative volume fractions as reported elsewhere [12]. All data were reported as mean ± standard deviation (SD) calculated on 5 specimens for each scaffold typology. MicroCT analysis All the samples were scanned using a SkyScan 1072 micro-CT desk scanner supported by microtomographic reconstruction software (Skyscan) [14]. After a preliminary setting of brightness and image contrast and the manipulation of the grayscale thresholds, a 3D model of the pore network (direct reconstruction) was obtained, for the calculation of morphological parameters. Average Pore Size (APS) and pore size distribution were estimated by CTan Skyscan software whereas further morphometric parameters (i.e. porosity, anisotropy degree (DA)) were calculated by ANT Skyscan software. All data were reported as mean ± standard deviation (SD) calculated on 3 specimens for each scaffold typology. Differential scanning calorimetry Thermal properties were evaluated by differential scanning calorimetry (DSC—TA instrument Q1000) to describe the effect of inorganic phases on scaffold crystallinity. Specimens (5–8 mg) were heated from 30 °C to 120 °C at 10 °C min−1 and, then, cooled from 120 °C to −170 °C at 5 °C min−1. Melting temperatures and enthalpy of fusion were calculated from the first heating while crystallization temperature and enthalpy were calculated from the second cooling ramp. Atomic force microscopy The morphological properties were qualitatively investigated via Atomic Force Microscopy (AFM) in contact/tapping mode by Innova System AFM equipped with the same scanner (Large area scanner INSC090, 90 μm lateral range). The imaging was performed via tapping mode (TM-AFM) on thin scaffold slices— about 100 μm thick—in air, at room temperature (25 °C) by using a RTESPA silicon cantilever (Bruker Corporation, Santa Barbara, USA). Adhesion forces were measured via contact mode by a Si 3N4 AFM cantilever (MPP-31123-10, Bruker, Santa Barbara USA). For all the experiments, MPP and RTESPA tips with the same nominal radius (8–20 nm) were used. Moreover, cantilevers were selected with different spring constants, 0.9–1.8 N m−1 for MPP and 40-80 N m−1 for RTESP, respectively, and resonance frequency of 300–400 kHz for RTESPA and of 20-27 kHz for MPP 3
cantilevers, respectively. Specifically, the resonance frequency and the spring constant of the cantilevers were obtained by the Thermal Tune Calibration on a glass substrate. The images of interest were first captured by AFM raster scanning, processed using the NanoScope Analysis data processing software 1.40 (Bruker Corporation, USA) and then graphically reported in 2D and 3D form. A roughness parameter for the surface Rq, (the rootmean-square height of the surface) was also calculated. Roughness data was acquired by manually applying a rectangular region of interest (ROI) box of (5 × 5)μm2 to different areas of the scaffold surface. In order to obtain an average value of the surface roughness, this operation was repeated several times. The adhesion properties ofloaded scaffolds were characterized by quantitative force spectroscopy via contact mode. The adhesion interaction between tip and sample was determined as force versus Z-piezo displacement and a mean value of each points was considered. All results were reported as mean ± standard deviation (SD). The statistical significance (set at P
Search
Collections
Journals
About
Contact us
My IOPscience
Needle-like ion-doped hydroxyapatite crystals influence osteogenic properties of PCL composite scaffolds
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2016 Biomed. Mater. 11 015018 (http://iopscience.iop.org/1748-605X/11/1/015018) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 194.27.18.18 This content was downloaded on 05/03/2016 at 08:31
Please note that terms and conditions apply.
Biomed. Mater. 11 (2016) 015018
doi:10.1088/1748-6041/11/1/015018
PAPER
received
10 September 2015
Needle-like ion-doped hydroxyapatite crystals influence osteogenic properties of PCL composite scaffolds
re vised
28 October 2015 accep ted for publication
29 October 2015 published
29 February 2016
V Guarino1, F Veronesi2, M Marrese1, G Giavaresi2,3, A Ronca1, M Sandri4, A Tampieri4, M Fini2,3 and Luigi Ambrosio1 Institute of Polymers, Composites and Biomaterials, Department of Chemical Sciences & Materials Technology National Research Council of Italy, Mostra D’Oltremare, Pad.20, V. le Kennedy 54, 80125, Naples, Italy 2 Laboratory of Biocompatibility, Innovative Technologies and Advanced Therapies, Department Rizzoli RIT, Italy 3 Laboratory of Preclinical and Surgical Studies, Rizzoli Orthopaedic Institute, Via Di Barbiano 1/10, Bologna, Italy 4 CNR—Institute of Science and Technology for Ceramics (ISTEC), Via Granarolo, 64 48018 Faenza (RA), Italy 1
E-mail: [email protected] and [email protected] Keywords: composite scaffolds, needle-like hydroxyapatite, ion doping, mechanical properties, in vivo studies
Abstract Surface topography and chemistry both play a crucial role on influencing cell response in 3D porous scaffolds in terms of osteogenesis. Inorganic materials with peculiar morphology and chemical functionalities may be proficiently used to improve scaffold properties—in the bulk and along pore surface—promoting in vitro and in vivo osseous tissue in-growth. The present study is aimed at investigating how bone regenerative properties of composite scaffolds made of poly(Ɛ-caprolactone) (PCL) can be augmented by the peculiar properties of Mg2+ ion doped hydroxyapatite (dHA) crystals, mainly emphasizing the role of crystal shape on cell activities mediated by microstructural properties. At the first stage, the study of mechanical response by crossing experimental compression tests and theoretical simulation via empirical models, allow recognizing a significant contribution of dHA shape factor on scaffold elastic moduli variation as a function of the relative volume fraction. Secondly, the peculiar needle-like shape of dHA crystals also influences microscopic (i.e. crystallinity, adhesion forces) and macroscopic (i.e. roughness) properties with relevant effects on biological response of the composite scaffold: differential scanning calorimetry (DSC) analyses clearly indicate a reduction of crystallization heat—from 66.75 to 43.05 J g−1—while atomic force microscopy (AFM) ones show a significant increase of roughness—from (78.15 ± 32.71) to (136.13 ± 63.21) nm—and of pull-off forces—from 33.7% to 48.7%. Accordingly, experimental studies with MG-63 osteoblast-like cells show a more efficient in vitro secretion of alkaline phosphatase (ALP) and collagen I and a more copious in vivo formation of new bone trabeculae, thus suggesting a relevant role of dHA to support the main mechanisms involved in bone regeneration.
Introduction A large variety of bioactive materials (i.e. ceramics [1], glasses [2]) with different chemical groups (i.e. silicates [3], phosphates [4]), have been extensively investigated to improve the biological and biomechanical properties of scaffolds for bone tissue engineering [5]. They offer a unique chance to easily manipulate the chemical composition at the atomic level by the local substitution of ions (i.e. calcium, strontium, magnesium) in the 3D crystal [6], thus enabling one to reproduce the peculiar chemical composition of native mineral phases—the main component of bone tissue © 2016 IOP Publishing Ltd
and teeth—so actively contributing to the osteogenic activity of bone cells. Recent studies demonstrated that ion substituted apatites possess an excellent ability to form in vitro apatite [7], concurring to generate bioactive phases with mechanical properties comparable to those of native HA [8], which are suitable to assembly composite scaffolds in bone tissue engineering [9]. According to the natural bone structure—i.e. apatite crystals embedded in a collagen matrix—bioactive phases have been variously combined with synthetic or natural polymers (i.e. polyesters, polyurethanes, polysaccharides) to reproduce the functionalities of the bone
Biomed. Mater. 11 (2016) 015018
V Guarino et al
extracellular matrix (bECM), overcoming the innate brittleness of the ceramic phases [10, 11]. Among them, aliphatic polyesters such as PCL have been frequently preferred due to the slow degradation properties [12] which allow matching the formation rate of new bone, also preserving a mechanical strength in the range of trabecular bone [13]. In this case, PCL matrix also works as a binder for bioactive particles, providing a protective function which prevents any unexpected brittle failure of the final system [14]. However, this effect may often compromise the bioactive response of scaffold due to a not complete exposure of particles at the interface. So, the tendency of the bioactive particles to form clusters in the polymer matrix and the weak adhesion forces generally occurring among hydrophilic ceramic and hydrophobic polymer phases, still represent the main cause of mechanical failure of this type of composite scaffolds [15]. In this context, recent studies are emphasising the main role of osteo-inductive phases on influencing surface and interface properties (i.e. hydrophobicity, surface charge, roughness) in order to more directly address specific cell activities towards osteogenesys [16]. However, the ability of bioactive particles to affect surface properties of the scaffolds have been poorly focussed and only marginally correlated to in vitro and in vivo scaffolds response [17]. In a previous study, it has been investigated how chemical modification of HA crystals—i.e. doping by Mg2+ and CO32 − ion substitution—may influence spatial distribution of particles, thus affecting the hMSC behaviour in osteogenic way under in vitro and in vivo conditions [18]. More recently, it has been also elucidated the capability of ion doped HA crystals to promote the deposition of a biomimetic coating to improve composite scaffold bioactivity [19] at the pore surface. Here, we propose a comparative study of two different composite scaffolds composed of a PCL porous matrix reinforced by undoped HA (uHA) and Mg2+ ion doped HA (dHA) crystals, respectively. In particular, we have investigated how dHA and uHA with different aspect ratios influence macroscopic (i.e. porosity, mechanical response) and microscopic properties (i.e. roughness, adhesive forces) of scaffolds, thus indirectly addressing in vitro/in vivo response of the composite scaffolds, for the design of implantable devices for bone regeneration.
Experimental methods Materials Poly(ε-caprolactone) (PCL, Mn 42500 Da) was purchased from Sigma-Aldrich (Milan, Italy) and tetrahydrofurane (THF) from Baker (Italy). All chemicals were used as received by product companies. dHA particles were synthesized by ISTEC-CNR (Faenza, Italy) by simultaneously dropwise addition of acid aqueous solution (600 ml) containing 88.8 g H3PO4 (Aldrich, 85% pure) and 200 ml 0.8 M NaHCO3 2
(Merck, A.C.S., ISO) aqueous solution at 40 °C for 3 h, into a basic solution containing 100 g Ca(OH)2 (Aldrich, 95% pure) in 1000 ml of water by adding 48.4 g MgCl2 and 6H2O (Merck, A.C.S., ISO) in order to achieve a molar ratio of mineral phase equal to 5% [7, 18]. Precipitates were aged for 24 h at 25 °C, washed and filtered three times, lyophilized and finally sieved at 150 μm. Moreover, uHA particles (HAP205, theoretical density 3.16 g cm−3) with bimodal size distribution (Φ range = 4–11 μm, Φ 0.5 = 4.02 μm), measured via laser diffractometry were supplied by Plasma Biotal (Tideswell, UK). Scaffold preparation Briefly, 0.20 g of PCL granules were dissolved in 10 ml of THF and kept under magnetic stirring at 40 °C for 24 h. uHA or dHA crystals were added to PCL solution before the solvent extraction by different volume ratios, 13%, 20% and 26% v/v. To impart the required porosity, sodium chloride crystals—300–500 micron in size—with 90/10 salt/polymer volume ratio, were also included and, then, removed by leaching techniques. Briefly, the mixture was placed in anti-adhesive mould to confer the desired shape. Ethanol was used for 24 h to preliminary force the removal of used solvent and, then, bi-distilled water for 7 d to leach out salt crystals and any other contaminants. For all the experimental studies, uHA or dHA loaded specimens were labeled as uHAs and dHAs respectively, whereas unloaded specimens, namely PCL alone, as CTR. Crystal aspect ratio measurements The size and shape of uHA and dHA crystals were analysed by transmission electron microscopy (TEM, TECNAI120—FEI) supported by 2D/3D image analyses. On selected images with higher contrast and best focus, ImageJ Analysis Software was used to estimate particle length and elongation by calculating the shape factor (pAr), defined as the ratio between particle width and length. Samples for TEM were prepared by air-drying a drop of a sonicated ethanol suspension of particles onto a carbon-coated copper grid. Compressive tests At this preliminary stage, mechanical properties under compression were evaluated at room temperature by a dynamometric machine (Instron 4204) equipped with a 100 N load cell and a crosshead speed of 1 mm min−1. Cylindrical specimens with a 2:1 height/diameter ratio (diameter: 12 mm, height: 24 mm) were fabricated according to the ASTM 695/2a standard. In this case, uHAs and dHAs with three different filler/polymer volume ratio, 13:87, 20:80 and 26:74, have been tested in agreement with previous works [12], in order to evaluate the effective reinforcement of inorganic phases, minimizing the clustering effects. The study was carried out through analysis of five specimens for each scaffold type. The elastic modulus E was calculated
Biomed. Mater. 11 (2016) 015018
V Guarino et al
by the initial slope of the stress–strain curve before the plateau region, in the strain range of 0–0.1 mm mm−1. These values were also compared with those carried out via theoretical model based on open cubic cells (COC), by assuming composite scaffolds as high porous solid systems with a continuous and percolating network structure. In this case, elastic moduli were calculated, on the hypothesis of nearly elastic materials [20]. Moreover, densities of the single phases of composite scaffolds were corrected by their relative volume fractions as reported elsewhere [12]. All data were reported as mean ± standard deviation (SD) calculated on 5 specimens for each scaffold typology. MicroCT analysis All the samples were scanned using a SkyScan 1072 micro-CT desk scanner supported by microtomographic reconstruction software (Skyscan) [14]. After a preliminary setting of brightness and image contrast and the manipulation of the grayscale thresholds, a 3D model of the pore network (direct reconstruction) was obtained, for the calculation of morphological parameters. Average Pore Size (APS) and pore size distribution were estimated by CTan Skyscan software whereas further morphometric parameters (i.e. porosity, anisotropy degree (DA)) were calculated by ANT Skyscan software. All data were reported as mean ± standard deviation (SD) calculated on 3 specimens for each scaffold typology. Differential scanning calorimetry Thermal properties were evaluated by differential scanning calorimetry (DSC—TA instrument Q1000) to describe the effect of inorganic phases on scaffold crystallinity. Specimens (5–8 mg) were heated from 30 °C to 120 °C at 10 °C min−1 and, then, cooled from 120 °C to −170 °C at 5 °C min−1. Melting temperatures and enthalpy of fusion were calculated from the first heating while crystallization temperature and enthalpy were calculated from the second cooling ramp. Atomic force microscopy The morphological properties were qualitatively investigated via Atomic Force Microscopy (AFM) in contact/tapping mode by Innova System AFM equipped with the same scanner (Large area scanner INSC090, 90 μm lateral range). The imaging was performed via tapping mode (TM-AFM) on thin scaffold slices— about 100 μm thick—in air, at room temperature (25 °C) by using a RTESPA silicon cantilever (Bruker Corporation, Santa Barbara, USA). Adhesion forces were measured via contact mode by a Si 3N4 AFM cantilever (MPP-31123-10, Bruker, Santa Barbara USA). For all the experiments, MPP and RTESPA tips with the same nominal radius (8–20 nm) were used. Moreover, cantilevers were selected with different spring constants, 0.9–1.8 N m−1 for MPP and 40-80 N m−1 for RTESP, respectively, and resonance frequency of 300–400 kHz for RTESPA and of 20-27 kHz for MPP 3
cantilevers, respectively. Specifically, the resonance frequency and the spring constant of the cantilevers were obtained by the Thermal Tune Calibration on a glass substrate. The images of interest were first captured by AFM raster scanning, processed using the NanoScope Analysis data processing software 1.40 (Bruker Corporation, USA) and then graphically reported in 2D and 3D form. A roughness parameter for the surface Rq, (the rootmean-square height of the surface) was also calculated. Roughness data was acquired by manually applying a rectangular region of interest (ROI) box of (5 × 5)μm2 to different areas of the scaffold surface. In order to obtain an average value of the surface roughness, this operation was repeated several times. The adhesion properties ofloaded scaffolds were characterized by quantitative force spectroscopy via contact mode. The adhesion interaction between tip and sample was determined as force versus Z-piezo displacement and a mean value of each points was considered. All results were reported as mean ± standard deviation (SD). The statistical significance (set at P
