Evaluation of nanostructure and microstructure of bone regenerated by BMP-2-porous scaffolds.

Nano and micro-structure of regenerated bone

Evaluation of nano and micro-structure of bone regenerated by BMP-2-porous scaffolds Carlos del Rosario1, Maria Rodríguez-Evora1, Ricardo Reyes1,2, Alejandro González-Orive3, Alberto Hernández-Creus3, Kevin M Shakesheff4, Lisa J White4, Araceli Delgado1,2, Carmen Evora1,2 *

Conflict of interest: No benefit of any kind will be received either directly or indirectly by the authors

1

Department of Chemical Engineering and Pharmaceutical Technology, University of La

Laguna, 38200 La Laguna, Spain 2

Institute of Biomedical Technologies (ITB), Center for Biomedical Research of the Canary

Islands, University of La Laguna, 38200 La Laguna, Spain 3

Department of Physico-Chemistry, Institute of Materials and Nanotechnology, University of

La Laguna, 38200 La Laguna, Spain 4

Wolfson Centre for Stem Cells, Tissue Engineering and Modelling (STEM), School of

Pharmacy, University of Nottingham, UK

* Corresponding author: Carmen Évora, Ph.D. Department of Chemical Engineering and Pharmaceutical Technology University of La Laguna 38200 La Laguna, Spain Tel: (34) 922318957 Fax: (34) 922318506 E-mail: [email protected]

This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process which may lead to differences between this version and the Version of Record. Please cite this article as an ‘Accepted Article’, doi: 10.1002/jbm.a.35436 This article is protected by copyright. All rights reserved.

Journal of Biomedical Materials Research: Part A Nano and micro-structure of regenerated bone

Abstract In this study, three systems containing BMP-2 were fabricated, including two electrospun sandwich-like-systems of PLGA 75:25 and PLGA 50:50 and a microsphere system of PLGA 50:50 to be implanted in a critical size defect in rat calvaria. The in vivo BMP-2 release profiles of the three systems were similar. The total dose was released during the first two weeks. To evaluate the nano and microstructure of the regenerated bone a multi-technique analysis was used, including stereo microscope, X-Ray; AFM, micro-CT and histological analyses. The progression of bone regeneration was followed at 4, 8 and 12 weeks after the microsphere system implantation whereas the two electrospun systems were evaluated at fixed 12 weeks. All the techniques applied showed high bone regeneration. The average values of bone volume density, bone mineral density, Young’s modulus and the percent of bone repair were approximately 70% of the values of the native bone. Besides, SEM-EDX analysis indicated that the main chemical elements in the new bone were oxygen, calcium and phosphorus in a ratio similar to that of native bone. In comparison, the micro-CT may provide an alternative to histology for the evaluation of bone formation at the defect size.

Keywords: bone regeneration, multi-technique analysis, BMP-2, electrospinning, microspheres, AFM.

2 John Wiley & Sons, Inc. This article is protected by copyright. All rights reserved.

Page 2 of 37

Page 3 of 37


INTRODUCTION Bone is a complex hierarchical tissue that consists mainly of collagen and hydroxyapatite (HA). Structure and micro-nano-architecture are determinant aspects of bone strength and essential elements for the assessment of bone mechanical properties. Thus, information on the nano-architecture of the bone might help to better evaluate the functional properties of new bone formed by induction procedures. In bone regeneration studies and bone characterization, many tools are available to define bone morphology such as histological analysis and histomorphometry, transmission electron microscopy (TEM) and scanning electron microscopy (SEM) or computed tomography (CT). However, none of these techniques provide information about both bone stiffness and morphology. Atomic force microscopy (AFM) is a high resolution imaging method which enables the determination of nanoscale mechanical properties. AFM is a non-destructive method that can be utilised in both dry and liquid medium. Moreover in the last decade, AFM techniques have emerged as powerful tools capable of registering with high accuracy not only nanostructured surface profiles but mechanical properties1. In particular, determining the mechanical properties of complex biological systems and living cells in the molecular scale has attracted the interest of researchers2. The applications of AFM to determine the nano-morphology and nano-mechanical properties of bone were recently reviewed3 and it was highlighted that most of the AFM research pertaining to bone is composed of imaging studies. Those studies have provided fundamental information regarding bone morphology, such as collagen fibril-mineral level4,5. Additionally, images of freshly fractured bovine bones6,7 and human bone7 have contributed to knowledge of the nanostructure of fresh trabecular. AFM has even been used to assess age-related differences in human bones8-11. Recently, studies have focused upon the mechanical characteristics of bone at the nanoscopic scale12-15. Although the values of the



nanomechanical properties measured by AFM are stronger affected by the hydration conditions of the samples3,15 poor attention has been paid to this aspect. AFM has been used to characterize the regenerative effect of osteo-inductive stimulus in bone defects in a limited number of cases16-18. By contrast, micro x-ray computed tomography (micro-CT) is frequently used to evaluate the bone regeneration induced by external factors such as BMP-2 included in different scaffolds applied to different defects 19-22. In the present study, multi-technique analysis was applied to evaluate bone regeneration in a critical size defect in rat calvaria. Three distinct systems releasing BMP-2 were utilized comprising two electrospun sandwich-like systems of PLGA 50:50 and PLGA 75:25 and a PLGA 50:50 microsphere system. The in vivo BMP-2 release kinetics was defined using 125IBMP-2. The bone formed in the defect 12 weeks post-implantation was analyzed by histological and histomorphometric methods, SEM and X-Ray images, EDX technique for chemical composition and micro-CT and AFM parameters. Furthermore, the microsphere system was chosen to test the utility of AFM to assess the progression of bone remodeling (4, 8 and 12 weeks post-implantation) compared to histological and histomorphometric analysis. MATERIALS AND METHODS Material was processed under aseptic conditions. Polymers were sterilized prior to use by γirradiation at 25 kGy from 60Co source (Gamma Sterelization Unit of Aragogamma, Barcelona, Spain). Except for BMP-2, all liquid components and lab instruments were autoclaved (121ºC, 30 min, Auster Selecta, Spain). Fabrication and assembly of the scaffolds PLGA (Resomer® RG504 (PLGA 50:50) or Resomer® RG755S (PLGA 75:25), Evonik Industries, Germany) microspheres containing BMP-2 (Biomedal Life Sciences, Spain) were prepared by the double emulsion (water/oil/water) technique23. Briefly, 200 µL of rhBMP-2 (180 µg) in 0.07% polyvinyl alcohol (PVA) were vortexed with 2 mL of the PLGA methylene


Page 4 of 37

Page 5 of 37


chloride solution (50mg/mL). Then, the organic solvent was evaporated in a 0.1 % PVA (w/v) solution. Some batches were prepared with 125I-BMP-2 to determine BMP-2 encapsulation efficiency and release assays using a gamma counter (Cobra II, Packard). Microsphere size was determined using a Mastersizer 2000 (Malvern Instruments) and microsphere morphology was analyzed by scanning electron microscopy (SEM, Jeol JSM6300). The microsphere system, denoted Ms(50:50), consisted of 25 mg of PLGA microspheres (blend of blank microspheres with microspheres containing 6 µg of BMP-2) mixed with 20 µL of a 15% (w/v) solution of Pluronic F-127 (Sigma, USA). Sandwich-like scaffolds (membrane-microspheres-membrane), denoted S(50:50) and S(75:25), were assembled in situ and consisted of two PLGA electrospun sheets of 1cm2 with 25 mg of microspheres (prepared with the same PLGA) in between them. Electrospun membranes of PLGA 50:50 and PLGA 75:25 were prepared as previously described24. Briefly, 700 µl of a 16% (w/v) hexafluoroisopropanol (HFIP, Fluka, Switzerland) polymer solution was loaded into a syringe and pumped at 1 mL/h. The solutions were electrospun on a rotating collector positioned 12 cm from the needle at 200 rpm and subjected to a high voltage power supply (12 kV). Membranes were characterized in terms of membrane quality and fiber diameter by SEM. Membrane thickness were measured by stereo microscopy (Leica M205 C, Leica LAS, v3 software) and porosity was calculated by gravimetric method as described previously24.

Radioiodination of rhBMP-2 BMP-2 was labeled with 125I according to the iodogen method25. Briefly, in a Pierce® PreCoated Iodination Tube (Thermo Scientific) 25 µL of BMP-2 (1 mg/mL) and 10 µL 125INa (1 mCi) (Perkin-Elmer, USA) adjusted to 100 µL with 0.5 M phosphate buffer, pH 7.0, were



combined. After 15 min incubation at room temperature and 120 rpm, 100 µL of saturated tyrosine solution in PBS was added to eliminate the unreacted 125I. Purification of the reaction mixture was performed using a ZebaTM Spin Desalting Column (Thermo Scientific) following manufacturer’s instructions. Labelling yield and radiolabeling stability of 125I-BMP-2 was checked by thin layer chromatography (TLC) as previously described26. Animal experiments Male Sprague-Dawley rats (250-300 g) were used in this study. All experiments were carried out to conform with E.C. (2010/63/UE) regulations on care and use of animals in experimental procedures and were approved by the local committee for animal studies of the University of La Laguna. The groups of rat studied are summarized in Table 1. At each time point the rats were divided in two groups, one group of 4 rats for AFM, SEM, EDX and micro-CT analysis and the other group of 3 animals for radiological and histological evaluation (Table 1). In the first group the samples were wrapped in gauze previously soaked in PBS and frozen at -20 ºC until testing, as described by27. At each time point, the 3 samples in the second group were fixed in 10% formalin solution (pH 7.4) and decalcified in Histofix® Decalcifier 3D (Panreac, Spain) for histological analysis as previously described24. In addition, the in vivo BMP-2 release assays were carried out in 5 rats per scaffold type. Surgical procedure Surgery was carried out as previously described24,28,29. Briefly, the scaffolds were assembled into the 8mm cranium defects of anesthetized rats (ketamine, 100mg/Kg and xylazine, 10mg/Kg) and the skin was closed and stapled. Atipamezole (1mg/Kg) was injected subcutaneously to reverse anaesthesia. Buprenorphine (0.05 mg/Kg) was injected subcutaneously before surgery and 6 and 24 h post-operation.


Page 6 of 37

Page 7 of 37


In vivo release assays BMP-2 release kinetics was monitored periodically by measuring the remaining 125I-BMP-2 at the calvaria defect site using an external gamma counter (Captus®, Nuclear Iberica) as previously described and validated30. Briefly, the detector was placed on the defect site of sedated rats (Ketamine 75–100 mg/Kg). At each sampling time (0, 1, 4, 7, 10, 14, 21 and 30 days) the readings at 27 keV were taken as the remaining radioactivity. Measurement at time point 0 was considered the dose (100%). The similarity factor f2 was used for release profile comparison31,32. Macroscopic, microscopic and radiological analyses Before sample processing, pictures and radiography images were taken of the extracted rat crania using a stereo microscope (Leica M205 C) and a Philips Optimus X-Ray (44 kV, 3.6 mA/s and 15.4 mSv), respectively. Samples were also observed by SEM and chemical element composition was analyzed using SEM-EDX (SEM, Jeol JSM-6300). Atomic Force Microscopy Imaging and force measurements to characterize nanomechanical properties of rat calvarial bone were performed by AFM. In these studies, each calvarial bone portion was attached to a steel sample puck. When measuring in liquid, up to 100 µL of PBS (pH 7.4) was added to cover the sample prior to being placed into the AFM liquid cell. For measurements performed in air, samples were carefully rinsed in Milli-Q water and dried in N2 atmosphere before using. Quantitative mapping was performed in air and in PBS (pH 7.4) at room temperature using a Nanoscope V controller (Bruker, USA). Images were acquired in Peak Force Tapping Mode (Peak Force-Quantitative Nanomechanics, PFT-QNM) as described by33. Silicon nitride probes (NP-C, Bruker) with a nominal tip radius of 20-60 nm were used in liquid, while silicon tips (TAP 525, Bruker) were chosen to record mechanical features in air conditions.



The spring constant of cantilevers were measured using the thermal tuning method34 and values ranged 0.14 - 0.26 N/ m for NP-C, and 186 - 222 N/ m for TAP 525. Fused silica and mica surfaces were selected as rigid substrates for deflection sensitivity calibration in air and liquid, respectively. The Young´s modulus was calculated by utilizing a DMT and Sneddon model fit35,36. Experimental results were collected from at least seven different areas of every single sample. Nanomechanical properties of three distinct zones in every area were assessed and averaged. Thus, data proceeding from 105 images from two independent sets were collected for every sample. Data processing was carried out using the commercial Nanoscope Analysis (Bruker, USA), WS x M (Nanotec) and Gwyddeon (GNU) software37. Micro-computer tomography evaluation New bone formation at the defect site was also evaluated using a micro-CT scanner (Skyscan 1174, Skyscan, Belgium) operated at 50 kV and 800 µA with Aluminium 1.0 mm filter to reduce bean hardening. Samples were fitted in a rotary holder with the coronal aspect of the calvarial bone in a vertical position and scanned entirely, being rotated 180º in 450 equiangular steps. The spatial resolution for specimen scanning was set to 14.7 µm. The regions of interest (ROI) were obtained by manual tracing of the defect area in horizontally serial sections. Bone mineral density (BMD), bone volume fraction (BV/TV), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp) and trabecular number (Tb.N) were calculated using the Skyscan CT Analyzer v.2 morphometrics software. The results of microCT, AFM, SEM and EDX analysis were compared with native bone samples. Histological and histomorphometrical analyses New bone formation was identified by hematoxylin-erythrosin staining. The degree of new bone mineralization was assessed with light green, orange G, acid fuchsin (VOF) trichrome stain, in which red staining indicates advanced mineralization, whereas less mineralized, newly-formed bone stains blue38. Sections were analyzed by light microscopy (LEICA DM


Page 8 of 37

Page 9 of 37


4000B). Computer based image analysis software (Leica Q-win V3 Pro-image analysis system, Spain) was used to evaluate histomorphometrically all sections per specimen. A circular region of interest (ROI) of 50 mm2, the center of which coincided with that of the defect site, was defined for quantitative evaluation of new bone formation. New bone formation was expressed as a percentage of repair [100×new bone area]/[original defect area within the ROI]. Statistical analysis Statistical analysis was performed with SPSS.19 software using one-way analysis of variance (ANOVA) and a Tukey multiple comparison post-test. Significance was set at p< 0.05. Results are expressed as mean ± standard deviation (SD). RESULTS Scaffold characteristics The mean volume diameter of PLGA50:50 and PLGA75:25 microspheres was 191.1 and 182.7 µm and the BMP-2 encapsulation efficiency was 71.4±5.8% and 78.1±9.0%, respectively. The membrane of the PLGA 50:50 was slightly thicker (147 ± 18 µm) than the PLGA 75:25 membrane (122 ± 20 µm). By contrast, the fiber diameter of PLGA 50:50 (1.21 ± 0.40 µm) was smaller than the PLGA 75:25 (1.71 ± 0.43 µm). Porosities ranging from 76 ± 3.7% of PLGA 50:50 membrane to 69.3 ± 2.9 % of PLGA 75;25 membrane. The sandwich-like scaffold (membrane-microspheres-membrane) assembled and placed in the defect, weighted 32 mg ± 1.7 mg. In vivo BMP-2 release assay BMP-2 release kinetics from the systems was monitored using 125I-BMP-2. According to f2 values, in the range of 56.7-71.9, the release profiles of BMP-2 from S(50:50)-BMP-2, Ms(50:50)-BMP-2 and S(75:25)-BMP-2 were similar. During the first 24 h approximately



30% of the dose, equivalent to 1.9 µg of BMP-2, was initially released. Until the end of the second week, the percentage of BMP-2 released was about 85% at a rate of 4-5% by day (260-325 ng/day), depending on the scaffold. Afterwards, the release rate declined to approximately 0.4% (26 ng) until the end of the assay. Consequently, a release of more than 90% of the BMP-2 was achieved within 30 days. Radiological analysis Radiological images reflected the evolution of the different groups. These images were in concord with macroscopic, histological, micro-CT and AFM analysis. Enhanced defect regeneration was observed in the BMP-2 groups with images from 12 weeks showing differential repair in blank specimens and those treated with BMP-2 (Fig. 1) Macroscopic Findings In the blank groups, a semitransparent tissue was observed macroscopically, whilst groups with BMP-2 show a dense hyaline-like tissue resembling the surrounding bone tissue (Fig. 2a) Histological and histomorphometrical evaluation Histological analysis of the samples from blank groups 4 and 12 weeks post-surgery, showed little repair response in the three scaffold type (Fig. 3). The presence of abundant connective tissue occupied most of the defect site (Fig. 3). Repair rates were in the range of 1-10% during the experimental period (Fig. 4a). In contrast to this, the Ms(50:50)-BMP-2 group at 4 weeks post-implantation showed a reparative response with low bone mineral areas confined to the margins of the defect site (Fig. 3). The percent of repair at this time point was around 20% (Fig. 4b). Eight and twelve weeks post-implantation, the Ms(50:50)-BMP-2 group showed an increase in repair response compared to 4 weeks. The presence of mineral bone occupied more than half of the defect. Repair rates in Ms(50:50)-BMP-2 group after 8 and 12 weeks were 49.6±16.8 % and 54.5±31.4 %, respectively (Fig. 4b).


Page 10 of 37

Page 11 of 37


The analyses of samples 12 weeks post-implantation, showed similar results in the three groups treated with BMP-2. The analysis of the S(75:25)-BMP-2 samples 12 weeks postimplantation scaffold, showed similar results to that observed in the Ms(50:50)-BMP-2 groups (Fig. 3), about 60% of repair (Fig. 4b). The S(50:50)-BMP-2 scaffold showed 12 weeks post-implantation the best result with large areas of newly formed bone (Fig. 3) and a repair rate of approximately 80% (Fig. 4b). The mean repair rate of Ms(50:50)-BMP-2 group was lower than the S(50:50)-BMP-2 group because one animal did not respond. However, the repair rates observed in Ms(50:50)-BMP-2 and S(50:50)-BMP-2 groups were not statistically different. The analysis of the samples in cross-section 12 weeks post-implantation corroborated that observed with horizontal sections, showing connective tissue in the blank groups (Fig. 5) and mineral bone with normal structure in the groups implanted with BMP-2 (Fig. 5). The thickness of the bone cross section differed depending on the scaffold, with thinner bone observed in animals implanted with the microspheres compared to sandwich scaffolds (Fig. 5). AFM and EDX analyses Young’s modulus measurements were performed in dry conditions for each sample, and in wet conditions for the stiffer samples. Empty defect, blank samples and the Ms(50:50)-BMP2 group at 4 weeks post-implantation were impossible to measure under wet conditions because of their gelatinous texture. Noticeably, samples treated with BMP-2 exhibited Young’s moduli close to that of native bone at 12 weeks post-implantation (Fig. 6). The elastic moduli of the Ms(50:50)-BMP-2 group, measured under dry and wet conditions were not statistically different to intact bone (Fig. 6). Detailed analysis of the defect site of Ms(50:50) group, showed the progression of the mineralization process comparing histological images and AFM data (Fig 7). In the initial



post-implantation stage, 4 weeks, almost only mostly collagen fibers (CF) can be detected in the AFM images, see inset in Fig 7b. However, for 12 weeks an increasing number of events related to mineralized bone (MB) corresponding to the highest values of elastic modulus progressively appeared (Fig. 7d). In fact, as the time passes, the number of bright spots (higher modulus values) became bigger (Fig. 7b), resulting in a Young’s modulus significantly larger and closer to that of native bone. Likewise, the same response occurred in the groups treated with the two sandwich-like scaffolds 12 weeks post-implantation (Fig 8). In addition a good correlation (p

Regenerated cellulose scaffolds: Preparation, characterization and toxicological evaluation.

In vivo evaluation of bone microstructure in humans: Clinically useful?

A Mechanobiology-based Algorithm to Optimize the Microstructure Geometry of Bone Tissue Scaffolds.

Evaluation of bone microstructure in CRPS-affected upper limbs by HR-pQCT.

Biodegradable borosilicate bioactive glass scaffolds with a trabecular microstructure for bone repair.

A comparison of bovine bone and hydroxyapatite scaffolds during initial bone regeneration: an in vitro evaluation.

3D printing of novel osteochondral scaffolds with graded microstructure.

Mimicking the nanostructure of bone matrix to regenerate bone.

Regenerated silk fibroin films with controllable nanostructure size and secondary structure for drug delivery.

Fabrication of porous hydroxyapatite scaffolds as artificial bone preform and its biocompatibility evaluation.

Angiogenesis in newly regenerated bone by secretomes of human mesenchymal stem cells.

Relations between tensile impact properties and microstructure of compact bone.

3D Printing of Personalized Artificial Bone Scaffolds.

Porcine Bone Scaffolds Adsorb Growth Factors Secreted by MSCs and Improve Bone Tissue Repair.

Evaluation of posterolateral lumbar fusion in sheep using mineral scaffolds seeded with cultured bone marrow cells.

The influence of disuse on bone microstructure and mechanics assessed by HR-pQCT.

Functionalized carbon nanotube reinforced scaffolds for bone regenerative engineering: fabrication, in vitro and in vivo evaluation.

Cone beam computed tomography evaluation of regenerated buccal bone 5 years after simultaneous implant placement and guided bone regeneration procedures--a randomized, controlled clinical trial.

Regulation of DMT1 on Bone Microstructure in Type 2 Diabetes.

The effect of docosahexaenoic acid on bone microstructure in young mice and bone fracture in neonates.

The correlation between dielectric properties and microstructure of femoral bone in rats with different bone qualities.

Nano Multilayered Scaffolds of PLGA and Collagen by Alternately Electrospinning for Bone Tissue Engineering.

Characterization of bone microstructure using photoacoustic spectrum analysis.

Bone Tissue Engineering: Scaffolds with Osteoinductivity for Bone Regeneration.