Dual release of growth factor from nanocomposite fibrous scaffold promotes vascularisation and bone regeneration in rat critical sized calvarial defect

Accepted Manuscript Full length article Dual release of growth factor from a nanocomposite fibrous scaffold promotes vascularisation and bone regeneration in a rat critical sized calvarial defect Shruthy Kuttappan, Dennis Mathew, Jun-ichiro Jo, Ryusuke Tanaka, Deepthy Menon, Takuya Ishimoto, Takayoshi Nakano, Shantikumar V. Nair, Manitha B. Nair, Yasuhiko Tabata PII: DOI: Reference:

S1742-7061(18)30453-7 https://doi.org/10.1016/j.actbio.2018.07.050 ACTBIO 5599

To appear in:

Acta Biomaterialia

Received Date: Revised Date: Accepted Date:

26 April 2018 2 July 2018 27 July 2018

Please cite this article as: Kuttappan, S., Mathew, D., Jo, J-i., Tanaka, R., Menon, D., Ishimoto, T., Nakano, T., Nair, S.V., Nair, M.B., Tabata, Y., Dual release of growth factor from a nanocomposite fibrous scaffold promotes vascularisation and bone regeneration in a rat critical sized calvarial defect, Acta Biomaterialia (2018), doi: https:// doi.org/10.1016/j.actbio.2018.07.050

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.

Dual release of growth factor from a nanocomposite fibrous scaffold promotes vascularisation and bone regeneration in a rat critical sized calvarial defect Shruthy Kuttappan1, Dennis Mathew1, Jun-ichiro Jo2, Ryusuke Tanaka2, Deepthy Menon1, Takuya Ishimoto3, Takayoshi Nakano3, Shantikumar V. Nair1, Manitha B. Nair1*, Yasuhiko Tabata2* 1

Amrita Center for Nanosciences and Molecular Medicine, Amrita Institute of Medical Sciences and Research Center, Amrita Vishwa Vidyapeetham, India 2 Laboratory of Biomaterials, Institute for Frontier Life and Medical Sciences, Kyoto University, Japan 3 Division of Materials and Manufacturing Science, Graduate School of Engineering, Osaka University, Osaka, Japan

*Dr. Manitha B. Nair (Corresponding Author) Amrita Center for Nanosciences and Molecular Medicine, Amrita Institute of Medical Sciences and Research Center, Amrita Vishwa Vidyapeetham, India E-mail: [email protected] *Prof. Yasuhiko Tabata (Co-Corresponding Author) Laboratory of Biomaterials, Department of Regeneration Science and Engineering, Institute for Frontier Life and Medical Sciences, Kyoto University, Japan E-mail: [email protected]

1

Abstract A promising strategy for augmenting bone formation involves the delivery of multiple osteoinductive and vasculogenic growth factors locally. However, success depends on sustained growth factors release and its appropriate combination to induce stem cells and osteogenic cells at the bony site. Herein, we have developed a nanocomposite fibrous scaffold loaded with fibroblast growth factor 2 (FGF2) and bone morphogenetic protein 2 (BMP2) and its ability to promote vascularisation and bone regeneration in critical sized calvarial defect was compared to the scaffold with vascular endothelial growth factor (VEGF) + BMP2. Simple loading of growth factors on the scaffold could provide a differential release pattern, both in vitro and in vivo (VEGF release for 1 week where as BMP2 and FGF2 release for 3 weeks). Among all the groups, dual growth factor loaded scaffold (VEGF+BMP2 & FGF2+BMP2) enhanced vascularisation and new bone formation, but there was no difference between FGF2 and VEGF loaded scaffolds although its release pattern was different. FGF2 mainly promoted cell migration, whereas VEGF augmented new blood vessel formation at the defect site. This study suggests that biomimetic nanocomposite scaffold is a promising growth factor delivery vehicle to improve bone regeneration in critical sized bone defects.

Keywords:

Growth

factor

delivery;

Electrospun

Neovascularisation; Bone regeneration

2

scaffold;

Critical

sized

defect;

1. Introduction Angiogenesis is an essential physiological process that occurs during the repair and regeneration of damaged bone. The newly formed blood vessels not only provide nutrients and growth factors, but also act as delivery route of stem cells and progenitors to the defect site [1]. Most of the synthetic bone grafts fail in bridging critical sized defects because of its inefficiency to facilitate vascularisation [2]. An established strategy to promote vascularisation and new bone formation is the administration of growth factors such as bone morphogenetic protein (BMPs), vascular endothelial growth factor (VEGF) and fibroblast growth factor 2 (FGF2) at the orthotopic region.

VEGF is a potent and specific angiogenic cytokine produced at the fracture site by numerous cell types including endothelial cells and osteoblasts [3]. It enhances neovascularisation and new bone formation during intramembraneous and endochondral ossification [4]. Similarly, FGF2 is a potent angiogenic inducer that modulates cellular functions through RAS/MAP kinase pathway. Studies in mouse models showed that the disruption of FGF2 gene resulted in reduced bone formation [5–7]. It is also demonstrated that FGF2 induces VEGF expression in vascular endothelial cells through autocrine and paracrine mechanism. Both these molecules also serve as a chemo attractants for mesenchymal stem cells to the bone regeneration site. Another important growth factor required for bone regeneration and remodeling is BMP2 that belongs to transforming growth factor β super family [8]. BMP2 plays predominant roles in different stages of endochondral ossification process such as osteogenesis, remodeling of bone and postnatal bone homeostasis [9]. The expression of VEGF and FGF2 is favorable during early phase of healing, while BMP2 to later stage processes [10], but exact sequence of release of these factors is unknown. Many studies have established that combined delivery of growth factors (VEGF / BMP2, FGF2 / BMP2) promote angiogenesis and bone formation [11,12]. Recent studies are focused on local and sustained release of growth factors through 3D scaffolds. However, release profile of growth factors varies based on the nature of the scaffolds chosen [13,14]. Furthermore, an ideal scaffold for bone tissue engineering should mimic the extracellular matrix of native bone with properties such as osteoconductivity, porosity, biodegradability, and mechanical strength, which in turn favor the cellular infiltration and tissue regeneration [15].

3

Native bone is a composite biomaterial comprised of nanohydroxyapatite matrix (65%), which is reinforced by collagen fibrils (35%). This interaction is known to significantly dictate the strength and toughness of bone. Moreover, the presence of trace elements like silicon plays an important role in angiogenesis and early development of bone [16]. In order to mimic the chemical composition and fibrous architecture of native bone, we have fabricated a composite scaffold of silica coated nanohydroxyapatite (nanoHA)-gelatin reinforced with electrospun poly(L-lactic acid) (PLLA) yarns [17]. The material was found to be porous, mechanically stable and biodegradable. The mechanical strength was dictated by the length and weight percentage of fibrous yarns within the matrix. The coating of nanoHA with an amorphous layer of silica could enhance material degradation and release Si ions, which have shown to augment osteogenic and endothelial cellular response in segmental bone defect. When the same scaffold was cultured with BMP2 expressing genetically engineered MSCs, it could accelerate osteoid deposition in bone defects [18]. In this study, our aim was to investigate the suitability of nanocomposite fibrous scaffold as a growth factor delivery system (dual delivery: VEGF+BMP2 & FGF2+BMP2 and single delivery: VEGF / FGF2 / BMP2) for augmenting vascularisation and new bone formation

To our knowledge, there are no reports that compare the effect of VEGF and FGF2 alone or in combination with BMP2 on vascularisation and bone tissue regeneration. We hypothesize that the scaffold mediated local delivery of FGF2 or VEGF might improve endothelial and mesenchymal stem cell migration and proliferation, resulting in new blood vessel and bone formation in critical sized calvarial defect, which would be enhanced in presence of BMP2. The release of these molecules was quantified in vitro and in vivo using radiolabelled growth factors. The composite scaffold without growth factors was used as the control.

2. Materials and Methods 2.1 Preparation of growth factor loaded composite scaffolds Nanocomposite fibrous scaffold (CS) was prepared as reported earlier [17] (Fig. 1). Briefly, PLLA (14% w/v) (Mw=100 kDa, Good Fellow, UK) fibrous yarns were developed by electrospinning, using a modified collector [19]. NanoHA was synthesized through aqueous 4

precipitation method utilizing 0.3 M diammonium hydrogen phosphate (Merck, USA) and 0.5 M calcium chloride (Merck, USA). An amorphous layer of silica was coated over nanoHA using 0.1 mM tetraethylorthosilicate precursor (Sigma, India) at alkaline pH. The synthesized nanorods were blended with gelatin (Type A) (HiMedia, India) to attain a final weight percentage of 35 gelatin: 65 nanoHA. The slurry was extruded into suitable moulds over which PLLA yarns were aligned; freezed at -20°C; lyophilized; cross linked using 1% (w/v) gluteraldehyde solution (GA) (Sigma Aldrich, USA) and sterilized by ethylene oxide for 12 h at 37°C. The final scaffolds thus obtained (8 mm diameter x 1.5 mm thickness) were incorporated with recombinant human growth factors BMP2 (kindly provided by Astellas Pharma Inc., Japan), FGF2 (kindly provided by Kaken Pharmaceutical, Japan) and VEGF (purchased from BioLegend, USA) separately and in combination. For this, the growth factors were dissolved in phosphate buffered saline (PBS, pH 7.4) (Nissui Ltd, Japan), loaded onto composite scaffolds and kept at 4°C overnight. In single loaded groups, the concentration of each growth factor used for in vitro studies was 0.5 µg, whereas for release and in vivo studies, it was 5 µg. In combination groups, ratio between VEGF and BMP2 / FGF2 and BMP2 was calculated based on previous reports [20,21] and the details are given in Table 1. Six groups of scaffolds were included in the study: Scaffold alone - CS (2) BMP2 loaded scaffolds - CSB (3) VEGF loaded scaffolds - CSV (4) FGF2 loaded scaffolds - CSF (5) BMP2 and VEGF loaded scaffolds in a ratio of 3:1 - CSBV and (6) BMP2 and FGF2 loaded scaffolds in a ratio of 2:1 - CSBF. Radiolabelled growth factors were used for release studies, while nonradiolabelled factors were used for biocompatibility studies.

2.2 In vitro growth factor release The release of growth factors from the scaffolds (CSB, CSV and CSF) was analyzed using radiolabelled growth factors. For this, human recombinant BMP2 / VEGF / FGF2 was radioiodinated separately by the conventional chloramine-T method as described previously [22]. Briefly, 0.5 mg / mL growth factor stock was prepared in equal volume of potassium PBS (pH 7.5) and sodium chloride (Nacalai Tesque Inc., Japan). From this, 200 μL of growth factor was mixed with 5 μL of Na125I (NEZ-033H, >12.95 GBq / mL) (Perkin Elmer Life Sciences Inc., USA). Chloramine-T (2 mg / mL) (Nacalai Tesque Inc., Japan) and sodium metabisulphite 5

solution (4 mg / mL) (Nacalai Tesque Inc., Japan) were prepared and added to the above mixture one after the other. A PD-10 desalting column (GE Healthcare Life Sciences, UK) was used to separate the uncoupled free 125I molecules from the 125I-labelled growth factor. Aqueous solution containing 5 μg 125I-labeled BMP2 / VEGF / FGF2 was dropped onto the composite scaffolds as described above and incubated in 100 mM PBS (pH 7.4) with Ca2+/Mg2+ ions at 37ºC. After 24 h, 1 µg / mL collagenase D (Sigma Aldrich, USA) was added to PBS; the supernatant was collected at different time points (Days 1, 3, 7, 14, 21 and 28) and the radioactivity was measured with the gamma counter (ARC-301B, Aloka Co., Ltd., Japan). Natural decay of

125

I

was considered to compensate the entire radioactivity obtained. The percent release was calculated by dividing the radioactivity obtained at a time by the original radioactivity of the scaffold incorporating

125

I-labeled growth factor. The release of growth factors in PBS without

collagenase enzyme was also measured in the same manner. 2.3 In vivo growth factor release The release of growth factors from the scaffolds (CSB, CSV and CSF) was analyzed in vivo using radiolabelled growth factors. For this, the scaffolds loaded with 125I-labeled BMP2 / VEGF / FGF2 (5 μg) were subcutaneously implanted into the back of 6 week old female ddY mice (Shimizu Laboratory Supply Inc., Japan). 100 µL of aqueous solution of

125

I-labeled BMP-2 /

VEGF / FGF2 subcutaneously injected into the same position of the mice was taken considered as the control. At different time intervals (Days 1, 3, 7, 14, 21 and 28), the implanted or injected area (a rectangular strip of 3*5 cm2) of skin was excised. A filter paper was used to wipe around the area and absorb any radio-iodinated growth factors present at the site. The radioactivity of remaining scaffolds, retrieved skin, and filter paper were measured using gamma counter to study the in vivo BMP2 / VEGF / FGF2 retention. In order to calculate the remaining percent radioactivity, the radioactivity obtained was divided by the original radioactivity of the scaffolds incorporating 125I-labeled growth factors. Natural decay of 125I was considered to compensate the entire radioactivity obtained. The experiment was independently performed in 3 samples per experimental group.

2.4 In vitro cytocompatibility studies 6

Mesenchymal stem cells (MSCs) isolated from rat adipose tissue were utilized for the cytocompatibility studies [18]. Briefly, the rat abdominal tissue was harvested and digested with 0.5% collagenase type I in PBS and incubated in tissue culture flask added with α-minimal essential medium (GIBCO, Invitrogen, USA) containing 10% fetal calf serum (FCS) (Hyclone laboratories, Inc., UT), 100 U / mL penicillin and streptomycin (Nacalai Tesque Inc., Japan) at 37°C and 5% CO2. The cells were characterized with positive markers like CD90 and 105 by flow cytometry (data not shown). The scaffolds were then seeded with MSCs (passage 3) (5 x10 4 cells per scaffold) and cultured for 21 days. Similarly, the cytocompatibility of endothelial cells on growth factor loaded scaffolds were investigated using human umbilical vein endothelial cells (HUVEC) purchased from Cambrex Bio Science Walkersville, Inc. (Walkersville, MD) for 3 days. The cells were cultured on six groups of scaffolds (5 x104 cells) supplemented with Iscove’s Modified Dulbecco’s Medium (GIBCO, Invitrogen, USA) containing FCS, penicillin, streptomycin (as mentioned above) and large vessel endothelial supplement (GIBCO, Invitrogen, USA) at 37°C for 3 days. 2.4.1 Cell adhesion The adhesion of MSCs and HUVEC after 24 h was analyzed using scanning electron microscope (SEM). For this, the cells on the scaffold were fixed using 1%gluteraldehyde, dehydrated by ethanol (increasing concentrations), sputtered with gold and observed under SEM (operating voltage 8 kV). 2.4.2 Cell migration The ability of MSCs and HUVEC to migrate towards growth factor loaded scaffolds were analyzed using Transwell inserts (Corning, NY). The cells were serum starved for 6 h; seeded (1 x 105 cells MSCs) on the upper chamber of a transwell insert, which was then placed in 24 well plate containing scaffolds and incubated for 6 h in serum free α-MEM at 37ºC. As a control, the scaffold without growth factors was used. After 6 h, the migrated cells in the lower chamber of transwell insert was fixed with 4% paraformaldehyde; stained with DAPI (Sigma-Aldrich, USA) and viewed under fluorescent microscope (BZX710; Keyence Co., Ltd., Japan) at 10× magnification. The migrated cells were quantified using BZX710 analyser equipped with the microscope. Average of 5 fields per sample was imaged for the quantification. 2.4.3 Cell proliferation 7

The proliferation of MSCs and HUVEC was studied using DNA assay on day 7, 14 and 21. For this, the cells on the scaffold were lysed with 30 mM sodium citrate-buffered saline solution (pH 7.4) containing 0.2 mg / mL sodium dodecyl sulfate for 1 h at 37°C. Equal amount of cell lysate and Hoechst 33258 stain (1µl / mL) (Nacalai Tesque Inc., Japan) were added on a black bottom 96-well plate and the fluorescence was measured by spectrometer (F-2000, Hitachi Ltd, Japan) at an excitation / emission wavelength of 355 and 460 nm respectively. Finally, the cell proliferation was calculated with a calibration line constructed with known cell concentration after 8 h of cell seeding (time standardized for cell attachment). 2.4.4 Osteogenic differentiation 2.4.4.1 ALP activity The activity of alkaline phosphatase (ALP), an early marker of osteogenic differentiation, was measured on the basis of the hydrolysis of p-nitrophenyl phosphate to p-nitrophenol on day 7, 14, and 21 using a Lab Assay ALP kit (Wako pure chemical Ltd, Japan). For this, 20 µL of cell lysate prepared (as described above) was added to 100 µL of ALP reaction buffer (pH 9.8) and incubated at 37°C for 15 min. The absorbance was measured at 405 nm, after addition of stop solution of 0.2M NaOH. Calibration line was constructed with known concentration of ALP enzyme. 2.4.4.2 Calcium assay The level of calcium in the cell lysate was analysed on day 7, 14 and 21 using calcium E-HA kit (Wako pure chemical Ltd, Japan). Briefly, 1M HCl and cell lysate in equal volume (4 µL each) was incubated at 4°C for 4 h; added with ethanolamine buffer (400 µL) and incubated at 37°C for 5 min in dark. Then 200µL of R2 buffer (colouring reagent) was added; incubated at 37°C for 5 min in dark and the absorbance was measured at 570 nm. A calibration line was constructed from different known concentration of calcium standard. 2.4.4.3 BMP2 release The BMP2 (pg / mL) release in the culture medium, in which cells were cultured on the scaffolds, was measured using enzyme linked immunosorbant assay (ELISA) (Research And Diagnostic Systems, Inc., USA). For this, the medium was collected on day 7, 14 and 21 and incubated for 2 h in the wells coated with anti-rat BMP2 antibody. The wells were washed; added with biotinylated anti rat-BMP2 antibody at 37ºC for 60 min and added with avidin8

biotin-peroxidase complex working solution for 45 min at 37ºC. Finally, colouring reagent was added into wells in dark for 30 min followed by stop solution. The absorbance was read at 450 nm immediately. The concentration was calculated with a calibration line plotted using known concentration of rat BMP2. 2.4.5 Endothelial Functionality 2.4.5.1 Nitric Oxide Release Nitric oxide (NO) secreted on day 1 and 3 was quantified by Griess-Romijn reagent (Wako pure chemical Ltd, Japan). For this, equal amount of cell culture supernatant and griess reagent (100 mg / mL) was incubated at room temperature for 10 min and the absorbance was read at 540 nm. A calibration line plotted with sodium nitrite was used as standard. 2.4.5.2 VEGF Release The release of VEGF in the culture medium was analyzed with ELISA (Research And Diagnostic Systems, Inc., USA). For this, the medium was collected on day 1 and 3 and ELISA was conducted as per manufacturer’s instructions (similar to protocol mentioned for BMP2 ELISA). The concentration was calculated with standard rat VEGF. 2.5 In vivo regeneration studies 2.5.1 Animal surgical procedure Wistar rats (4-5 months old male, weighing approximately 250 g) were used for the study. The animals were anesthetized with an intraperitoneal injection of pentobarbital sodium solution (3540 mg / kg body weight). Then a sagittal incision was created on the scalp till the middle sagittal crest after which the periosteum was dissected bluntly. A critical sized defect (8 mm diameter x 1.5 mm thickness) was made at the center of the dorsal calvarium utilizing a dental surgical drilling unit and trephine. The calvarial disk was removed and composite scaffold with and without growth factors were implanted within the defect (scaffolds loaded with 5 μg BMP2 / FGF2 / VEGF and its combination) were implanted (Fig. S1). Sham was also included in the study as a negative control. The defect sites were closed with interrupted vicryl resorbable sutures (4-0) and injected with sterile saline (10 mL / kg / h of surgery) subcutaneously. The animals were housed in soft-bedded cages after the surgery with free access to food and water. 2.5.2 Micro-CT analysis for vascularisation 9

The tissues after 4 weeks of implantation were analyzed for vascularisation by micro-CT (InspeXio SMX–100CT micro focus X–ray CT system, Shimadzu, Japan) after injecting with a contrast agent as mentioned earlier [11]. Briefly, the descending aorta of the rat was clamped after making an incision on the chest. Subsequently, the left ventricle was penetrated with an angiocatheter. The inferior vena cava was cut and perfused with 20 mL of heparanized saline (100 U / mL at 2 mL / min) (Otsuka Pharmaceutical Co., Ltd., Japan) utilizing a syringe pump. Microfil solution (20 mL of Microfil: diluent in a volume ratio of 4:5 with 5% curing agent) (Flowtech, Carver, MA) was perfused at 2 mL / min and permitted to set overnight at 4°C. The implant tissue was then harvested and fixed in 10% neutral buffered formalin for 7 days. Afterwards, the samples were decalcified with 5% formic acid (Nacalai Tesque Inc., Japan) for 4-5 days and analyzed for blood vessel formation by micro-CT. For this, a cylindrical volume of interest (VOI) (8 mm diameter and 1.5 mm height) was contoured in the defect and the samples were scanned at 20 μM resolution with 0.5 mm aluminum filter. The voltage and current used were 50 kV and 200 μA respectively. Volumetric reconstruction and imaging were conducted using Necron and CT-analyser software. The blood vessel formation (fold increase in the vessel area with respect to sham) was quantified using Image J software (Version 1.47, USA). 2.5.3 Histology and histomorphometric analysis For the histological examination, the post-implanted tissues at 4 and 12 weeks were fixed in 10% formalin for 7 days; decalcified with 5% formic acid for 4-5 days; dehydrated with increasing concentration of ethanol and embedded in paraffin using standard histolopathological protocols. The sections were rehydrated and stained using hematoxylin and eosin (H&E) to analyse the cellular infiltration and newly formed bone. Low magnification images (1x) images were captured under a microscope (BZX710, Keyence Co., Ltd., Japan) and stitched into a combined image with the advanced stitching module armed with the microscope. Further, the percentage area of matured bone at the centre of the defect was measured by Image J software. For this, high magnification images (20x) from the middle region of the defect was taken. The percentage area of the matured bone formed at the defect site was quantified by marking the total area of image as 100%.

2.6 Ethical statement 10

All the animal experiments were approved by animal experiment committee of Institute for Frontier Life and Medical Science, Kyoto University and the experiments were conducted according to the Institutional Guidance of Kyoto University on Animal Experimentation. 2.7 Statistical analysis The results obtained in every analysis were expressed as mean of all values ± standard deviation. One-way or two-way analyses of variance (ANOVA) were conducted depending on the experiments and were considered significant if p-values were less than 0.05.

3. Results 3.1 In vitro growth factor release NanoHA was developed and coated with an amorphous layer of silica. These core-shell nanorods were blended with gelatinuous matrix and reinforced with PLLA yarns in random manner as reported earlier. The scaffold was porous with pore size in the range of 50 – 350 µm and porosity of 58.8% ± 7.3%. The electrospun nanoyarns were well integrated within the matrix that contributed to an overall compressive strength of 28±3.5 MPa (Fig. 1). Figure 2 shows the cumulative in vitro time profile of radiolabelled BMP2, VEGF and FGF2 from composite scaffold scaffolds immersed in PBS (with or without collagenase). There was a burst release at 24 h followed by a sustained release for 30 days. BMP-2 and FGF2 release reached a maximum of 4± 0.29% and 8 ± 0.39% of the total loaded dose respectively within 24 h, while VEGF release was quicker and reached 33± 1.82% release during this time. Afterward, the concentration of growth factors released from the scaffolds was significantly very less (unmeasurable) when it is immersed in collagenase free PBS. On the other hand, when collagenase was supplemented to PBS, the release profiling was different. The overall release of BMP2 was slow during the first one week followed by a sustained release, where as FGF2 and VEGF showed a sustained release pattern after 24 h. Nevertheless the concentration of three growth factors released from the scaffolds was different. A total amount of 37±4% BMP2 and 26±0.67% FGF2 was released within 30 days, while VEGF release was considerably high (61± 5%).

3.2 In vivo growth factor release 11

To study in vivo growth factor release, the time course of radioactivity of

125

I-labeled growth

factors remaining in the scaffold after subcutaneous implantation was determined (Fig. 2). This was compared to free

125

I-labeled growth factor solution injected at the site. The growth factors

incorporated in the scaffold was released in sustained manner for longer time period than free BMP-2 injected. The period of release was dependent on the type of the growth factor incorporated into the scaffold. During 24 h, 23±3% of BMP-2, 42± 6% of FGF2 and 85±2% of VEGF was released. Later also, VEGF release was faster (100% within 7 days, while BMP2 and FGF2 release was there until 20 days. Nevertheless, free growth factors were retained only for 24 – 48 h at the site of injection. 3.3 In vitro cytocompatibility studies Rat MSCs and HUVEC were used to evaluate the cytocompatibility of the growth factor loaded scaffolds (single as well as dual growth factors) in comparison to those devoid of it. Scanning electron micrographs revealed that both type of cells could adhere and spread on the surface of all the group of scaffolds after 24 h, without blocking its pores (Fig. 3A&B). Further, the migration of cells towards the scaffold analyzed by transwell migration assay was imaged with DAPI staining (Fig. S2) and quantified using image J. The scaffolds could induce the migration of stem cells (Fig. 3C). Nonetheless, CSF significantly influenced higher recruitment of MSCs when compared to CS, CSB and CSV. Among all groups, the highest chemotactic response was shown by CSBV and CSBF and there was no difference between those two groups. In case of HUVEC, the migration was better towards growth factor loaded scaffolds, especially CSV and CSBV groups (Fig. 3D). The proliferation of MSCs on single and dual growth factor loaded scaffolds was evaluated by DNA assay on day 7, 14 and 21 (Fig. 4A). The cells proliferated on all the group of scaffolds and the proliferation rate increased as the culture period prolonged. The proliferation of cells on growth factor loaded groups (both single and combination) was considerably higher than CS group on day 14 and 21 (CSBF=CSBV>CSB=CSV=CSF=CS). There was no disparity amid single growth factor loaded groups (CSB, CSV and CSF) or among dual growth factor loaded groups (CSBV and CSBF) except that on CSBF on day 14. In contrast, there was no extensive increase in the endothelial cell populations with culture period (Fig. 4D). Moreover, the presence

12

of growth factor did not influence cell proliferation, except the combination of VEGF/BMP2 and FGF2/BMP2 on day 3. 3.3.4 Osteogenic Differentiation The osteogenic differentiation was verified by ALP activity, calcium deposition and BMP2 release by culturing MSCs in the absence of osteogenic supplements. The maximum ALP activity was shown by the cells on CS and CSV on day 14 followed by a decline in its performance (Fig. 4B). Nonetheless, the activity was earlier (day 7) on other growth factor loaded scaffolds. Among all groups, the highest enzyme activity was shown by CSB, CSF, CSBV and CSBF, with preference towards latter two groups. The timely release of calcium verified the osteogenic maturation and its mineralization (Fig. 4C). Similar to ALP activity, calcium was deposited by the cells on all groups of scaffolds and there was no substantial difference in its concentration from day 7 to 21. However, the highest release was shown by CSB, CSBF and CSBV groups when compared to the scaffold alone group. Figure S3 shows the amount of BMP2 released by osteogenic cells cultured on scaffold with and without growth factors. Here, the maximum BMP2 release was shown by CSB, CSBV and CSBF groups on day 7 and 14. In short, all the studies demonstrated that the osteogenic differentiation was significantly higher on dual growth factor loaded scaffolds followed by CSB groups when compared to scaffold alone group. 3.3.5 Endothelial Functionality The functionality of HUVEC was analyzed by VEGF and NO release (Fig. 4E&F). The angiogenic factors were secreted by the cells and its level was increased from day 1 to 3. While comparing, the growth factor loaded groups exhibited higher endothelial funcionality with respect to CS, but there were no difference between CSB, CSV or CSF. The maximum amount of VEGF and NO was secreted by cells on dual growth factor loaded (CSBV and CSBF) at both time points. 3.4 In vivo regeneration studies 3.4.1 Micro CT analysis for vascularization Figure 5 depicts the micro-CT data and fold increase of newly formed blood vessel at 4 weeks with respect to sham within the volume of interest. New vasculature through the scaffolds was increased when compared to sham group. Neo-vascular networks of varying thickness were seen 13

throughout the defect area irrespective of the groups. Nonetheless, in growth factor loaded groups, the blood vessel formation was considerably higher, especially on CSV, CSBV and CSBF (p<0.001). Among all, dual growth factor loaded groups demonstrated significantly higher score for vascular blood vessel area and there was no difference between CSBV and CSBF. 3.4.2 Histology and histomorphometric analysis The histological images of bone regenerated in the cranial defects at 4 and 12 weeks of implantation was shown in figures 6 and 7. There was no sign of material dislocation or inflammation in any groups throughout the study period. The bone union occurred in all the groups with host bone integration, except in sham (Fig. 6). The sham group showed a thin film of fibrous tissue formation throughout the length of the defect (Fig. S4). The new bone formation occurred from the peripheral region towards the mid region of the defect in par with the degradation of the scaffold. The organization of newly formed bone and collagen deposition improved as the implantation period increased. At 4 weeks, the cellular infiltration, osteoid deposition and immature bone formation was seen in CS, CSV and CSF, but the level of cellular infiltration was enhanced in CSF groups (Fig. 7). At 12 weeks, organized mature lamellar bone was more prominent in CSV and CSF groups. Interestingly, in CSB, CSBV and CSBF, the percentage of lamellar bone was considerably higher and faster at both the time points, with preference towards dual growth factor loaded groups (Fig. 8). There was no difference between CSBV and

CSBF groups.

In brief,

the performance among

the groups

was

CSBF~CSBV>CSB>CSV~CSF>CS>Sham.

4. Discussion Bone regeneration is a cascade of events mediated by various angiogenic and osteoinductive growth factors like BMP2, VEGF or FGF2. Many studies have shown the importance of local delivery of these growth factors at the defect site to enhance vascularisation and bone regeneration [23]. Herein, the selection of growth factor delivery vehicle is important, which should maintain a local growth factor concentration within the therapeutic window for a sufficient period of time to allow stem cells, endothelial cells and osteoprogenitor cells to migrate to the target site. In this study, the applicability of a biomimetic nanocomposite fibrous

14

scaffold to facilitate the release of angiogenic and osteogenic growth factors and to enhance vascularisation and bone regeneration in critical sized bone defect was investigated. Retention of growth factors in the delivery vehicle is imperative to enable its prolonged release at the defect site. In this study, we have utilized a composite matrix containing silica coated nanoHA-gelatin (Type A gelatin with isoelectric point 9) reinforced with PLLA fibers. Generally, increasing electrostatic attraction between growth factors and the scaffold would increase the binding of proteins onto biomaterials [24]. Previous experiments in our lab demonstrated that the surface charge of the composite scaffold is negative (mainly imparted by amorphous silica in the matrix) [25] and hence it can bind with positively charged growth factors. But the level of interaction may vary depending on the nature of growth factor used. Herein, the percentage of initial release of VEGF was very high (33±1% in vitro release and 85±2% in vivo release) followed by FGF2 (8±0.39% in vitro release and 42± 6% in vivo release) and BMP2 (4±0.29% in vitro release and 23±3% in vivo release). Additionally, the release period of each growth factor from the scaffold was dissimilar. For example, 100% VEGF was released within 7 days in vivo, while BMP2 and FGF2 release was there until 20 days of implantation. The disparity in the release rate of growth factors from the same scaffold (VEGF>FGF2>BMP2) is due to the variation in the isoelectric point of each growth factor. A previous study demonstrated significantly faster release of bovine serum albumin (BSA) (pI = 5.8) than VEGF (pI = 8.5) followed by BMP2 (pI = 8.5-9.2) from negatively charged hyaluronic acid hydrogels [26]. In addition, the interaction between growth factors and polymeric molecules play a critical role. Kinetic and thermodynamic analysis of the interaction between recombinant BMP2 and collagen matrix suggest that the binding between them is precise with affinities in the order of 103 to 104 M-1 and not as a result of adsorption or aggregation [27]. Similar to collagen, BMP2 can bind with gelatin through hydrophobic interaction and its retention in basic gelatin hydrogels was observed when implanted subcutaneously in mice [28]. Moreover, a higher electrostatic interaction between FGF2 and Type A gelatin in comparison to type B gelatin was also reported [29]. Hence, in our study, the electrostatic and hydrophobic interaction of BMP2 and FGF2 with gelatin and silica may have facilitated its retention in the scaffold and thereby its release for 20 days. In contrast, faster release of VEGF implies its non-specific interaction with composite matrix. 15

There was a change in release pattern when collagenase enzyme was added after 24 h. The concentration of BMP2 and FGF2 released from the scaffold was high in presence of collagenase enzyme. Separate in vitro studies have shown that composite matrix degrade partially within 2-3 weeks in presence of collagenase enzyme (data not shown). Since gelatin does not degrade via hydrolysis, initial release of BMP-2 / FGF2 in PBS (within 24 h) was due to diffusion. However in presence of collagenase, gelatin degrades, which in turn lead to an increased release of BMP2 / FGF2 that bound to gelatin, signifying degradation mediated release. On contrary, there was no considerable difference in VEGF release in presence or absence of collagenase enzyme, implying that its release mechanism was independent of matrix degradation (100% released within a week). While comparing, release of growth factors was faster under in vivo than in vitro scenario. This disparity may be due to complex in vivo environment containing multiple matrix degrading enzymes (MMP2, MMP9), cell types etc that are involved in the healing process [30]. Moreover, variation in factors like pH or temperature would change secondary or tertiary structure of growth factors [31]. Although there was difference in the percentage, the pattern of growth factor release was same under in vitro and in vivo conditions (VEGF>FGF2>BMP2). It is reported that VEGF and FGF2 acts as initiators of angiogenesis and fracture healing, while BMP-2 is required at later stages for the functional maturation of the blood vessels as well as bone remodeling [10].

Studies have illustrated that BMP2 enhances local bone formation in combination with FGF2 / VEGF, but it is dependent on the dose and ratio of each growth factor [20,32]. Charles et al reported an enhanced effect from the combination of a low dose of FGF-2 with high dose of BMP2 on calvarial bone defect healing. In another study, when the effects of different ratios of BMP2 and FGF2 (1:1, 2:1, 4:1, or 8:1) were compared, a ratio of 2:1 showed augmented osteogenic differentiation and ectopic bone formation in rats when compared to other groups. Similar results were seen with the mixture of BMP2 and VEGF. BMP2 to VEGF ratios of 3 to 1 increased microvascular density within the newly formed bone in an ectopic site in rabbit model, whereas, increased proportions of VEGF impaired bone formation [21]. Based on these data, we

16

have loaded 1:3 ratio of VEGF: BMP2 and 1: 2 ratio of FGF2:BMP2 on the scaffold for the combination therapy, both in vitro and in vivo.

Primarily, the effect of single and dual growth factor loaded scaffold on migration, proliferation and osteogenic induction of MSCs was studied in vitro. The mesenchymal stem cells adhered, proliferated and differentiated on nanocomposite fibrous scaffold, but its effect was augmented on growth factor loaded groups. When bone fracture occurs, MSCs are recruited and mobilized into damaged bone, which is mainly mediated by chemotactic growth factors. Among single growth factor loaded scaffolds, CSF was better than CSB and CSV in recruiting MSCs. Nevertheless, the scaffolds with FGF2 did not cause any additional impact on the proliferation or osteogenic maturation in vitro, which was verified through the activity of early marker like ALP as well as late mineralization marker (calcium release). This was similar to previous report, which demonstrated that bFGF can increase the migration of MSCs through activation of the Akt/protein kinase B pathway [33], but it has no influence on the proliferation or differentiation of precursor stem cells directly [34]. On the other hand, CSB group induced the differentiation of stem cells into osteoblasts without providing any osteogenic supplements. It is proven that BMPs are involved in the commitment of adipose derived stem cells through Smad canonical pathway as well as JNK/p38, MAPK/ERK, PI3K/AKT pathways [35]. Interestingly, the migration, proliferation and osteogenic differentiation were considerably enhanced on CSBV and CSBF and there was no difference between them. The co-ordinated interaction between FGF and BMP signaling in differentiating osteoblast was shown by earlier work wherein, the expression of BMP antagonist noggin gene expression was inhibited by exogenous FGF2 [36]. Similarly, the cellular response was higher on CSBV groups. Our results were consistent with a previous study, which demonstrated enhanced adhesion, proliferation and osteogenic differentiation of rat bone marrow derived mesenchymal stem cells on PLGA/gelatin composite scaffold consisting of BMP2 and VEGF [37]. In case of HUVEC, all types of growth factors were effective in enhancing the migration and functionality of cells. There was no difference between VEGF, BMP2 or FGF2, but coadministration of BMP2 with VEGF / FGF2 had better impact on endothelial cells. It is reported that BMP2 stimulates phosphorylation of eNOS (enzymes catalyzing the production of nitric 17

oxide) through protein kinase A pathway and thereby activates endothelial cells [38]. In the same way, FGF2 has shown to directly bind to the integrin ανβ3 promoting EC spreading and functionality [39]. In binary combination (FGF2 + BMP2 or VEGF+ BMP2), an increase in HUVEC response was seen. This was similar to previous works wherein the combined dose could increase the distribution of HUVEC in G2/M and S phases than G0/G1 phase. In these cases, several signaling pathways like AKT/P13K and PKC pathway or Ras-MEK-MAPK or Smad 1/5 gets activated simultaneously [10]. Further, critical sized calvarial defect model was established for studying vascularisation and bone regeneration potential of the scaffold in comparison to growth factor loaded scaffolds (single and dual). The composite fibrous scaffold has found to be osteoconductive and osteointegrative that lead to new bone formation, mainly from the peripheral region of the defects. The silicon ions released from the scaffold could induce angiogenesis, scaffold degradation and new bone formation in critical sized bone defects, which was in corroboration with our previous reports. Nonetheless, the new blood vessel and bone formation was considerably better in growth factor loaded groups (CSBF~CSBV > CSV > CSF ~ CSB > CS). Among single growth factor loaded groups, CSV was better in inducing new blood vessel formation, although the release of VEGF from the scaffold was there for 7 days. Recent studies have shown that VEGF stimulate not only VEGF receptors in endothelial cells, but also PDGF receptors in MSCs, which provoke MSC differentiation into an endothelial phenotype [40]. Thus multiple pro-angiogenic pathways triggered by VEGF in different cells may have facilitated significant blood vessel formation in CSV groups. In CSF and CSB group, vascularisation was enhanced when compared to scaffold alone, while its level did not reach up to CSV groups. This happened as these factors act as an indirect regulator of angiogenesis, not as direct endothelial cell specific mitogen [41,42]. Interestingly, angiogenesis was highest in combination group (CSBV and CSBF), signifying that the need of multiple factors to promote the formation of stable vessels. In presence of FGF, the adaptor protein like Shc is activated, which in turn stimulate receptor tyrosine kinase-dependent VEGF gene expression [43]. Similarly, BMP-2 stimulates VEGF secretion through p-P38, p-ERK and AKT/m-TOR signaling pathway and it can amplify the effects of VEGF on angiogenic activity [44].

18

When bone regeneration potential of single growth factor loaded scaffolds were compared, better performance was shown by CSB groups in comparison to CSV or CSF groups. Superiority in bone regeneration was obtained as BMP2 released in sustained manner from the composite fibrous scaffold. In clinics, a high dose of approximately 12 mg per treatment of recombinant human BMP-2 is delivered with the aid of an absorbable collagen sponge to obtain bone healing [45]. However, lack of specific binding affinity of collagen scaffold for BMPs result in an initial burst release soon after the implantation leading to many side effects [46]. With our delivery system, bone regeneration was achieved at lower BMP-2 doses (1 μg initially followed by 300 – 400 ng towards later stages), which were several order of magnitude lower than the clinical doses. The capability to regenerate bone was enhanced when BMP2 was delivered in combination with VEGF or FGF2. In CSBV system, in addition to the activity of BMP2, VEGF stimulates the invasion of new blood vessels, leading to bone regeneration as there is a cross talk between angiogenesis and osteognesis [47]. Similarly, in CSBF system, more stem cells would be migrated at the defect site due to the chemotactic effects of FGF-2, which in turn enhance bone regeneration. Nevertheless, the effectiveness of these morphogens critically depends on their concentration in the microenvironment of target cells and the stability of the protein.

The main limitation of the current study is the usage of only one dose of each growth factor. Thus the amount of growth factor needed to saturate binding sites is not known at this time. In future, more studies should be done with different dose of growth factors loaded onto the scaffold.

Conclusion In this study, we have developed a nanocomposite fibrous scaffold that could release growth factors (Dual system: VEGF+BMP2 / FGF2+BMP2; Single system: BMP2 / VEGF / FGF2) for promoting neovascularisation and new bone formation in critical size defect. Simple loading of growth factors on the scaffold could provide a differential release pattern, both in vitro and in vivo (VEGF release for 1 week where as BMP2 and FGF2 release for 3 weeks). The scaffold with VEGF or FGF2 alone could not regenerate bone completely, but administration of BMP2 exhibited positive effects on bone formation. In combination of BMP2 with FGF2 or VEGF, an 19

enhanced effect on angiogenesis and bone formation was noticed, but there was no difference between VEGF+BMP2 and FGF2+BMP2 loaded scaffolds. Thus the appropriate environment provided by the nanocomposite scaffold and its ability to release growth factors for a particular period was conducive to regulate the cellular activities and thereby tissue regeneration and remodeling. The growth factor could be loaded in a simple manner, which would ensure ease of use for the end-user, especially for the surgeon treating a patient in an operating room.

Acknowledgments Financial support for this work came from the India-Japan Cooperative Science Programme (IJCSP) and the Thematic Projects on Frontiers of Nano Science and Technology (TPF-Nano) Grant, Department of Science and Technology, Government of India. The Senior Research fellowship for the first author (Ms. Shruthy Kuttappan) was provided by Council of Scientific and Industrial Research (CSIR), Government of India. The authors thank Ms. Anitha A for the technical assistance in scaffold preparation. Disclosure Statement No competing financial interests exist.

References [1]

M. Sukul, T.B.L. Nguyen, Y.-K. Min, S.-Y. Lee, B.-T. Lee, Effect of Local Sustainable Release of BMP2-VEGF from Nano-Cellulose Loaded in Sponge Biphasic Calcium Phosphate on Bone Regeneration, Tissue Eng. Part A. 21 (2015) 1822–1836. doi:10.1089/ten.tea.2014.0497.

[2]

M. Lovett, K. Lee, A. Edwards, D.L. Kaplan, Vascularization Strategies for Tissue Engineering, Tissue Eng. Part B Rev. 15 (2009) 353–370. doi:10.1089/ten.teb.2009.0085.

[3]

W. Götz, C. Reichert, L. Canullo, A. Jäger, F. Heinemann, Coupling of osteogenesis and angiogenesis in bone substitute healing - A brief overview, Ann. Anat. 194 (2012) 171– 173. doi:10.1016/j.aanat.2011.10.002.

[4]

Y.-Q. Yang, Y.-Y. Tan, R. Wong, A. Wenden, L.-K. Zhang, A.B.M. Rabie, The role of vascular endothelial growth factor in ossification, Int. J. Oral Sci. 4 (2012) 64–68. doi:10.1038/ijos.2012.33. 20

[5]

B. Thisse, C. Thisse, Functions and regulations of fibroblast growth factor signaling during

embryonic

development,

Dev.

Biol.

287

(2005)

390–402.

doi:10.1016/j.ydbio.2005.09.011. [6]

A. Montero, Y. Okada, M. Tomita, M. Ito, H. Tsurukami, T. Nakamura, T. Doetschman, J.D. Coffin, M.M. Hurley, D. Gospodarowicz, P. Hauschka, A. Mavrakos, M. Iafrati, S. Doleman, M. Klagsbrun, Y. Shing, M. Hurley, Disruption of the fibroblast growth factor2 gene results in decreased bone mass and bone formation., J. Clin. Invest. 105 (2000) 1085–93. doi:10.1172/JCI8641.

[7]

G. Seghezzi, S. Patel, C.J. Ren, A. Gualandris, G. Pintucci, E.S. Robbins, R.L. Shapiro, A.C. Galloway, D.B. Rifkin, P. Mignatti, Fibroblast Growth Factor-2 (FGF-2) Induces Vascular Endothelial Growth Factor (VEGF) Expression in the Endothelial Cells of Forming Capillaries: An Autocrine Mechanism Contributing to Angiogenesis, J. Cell Biol. 141 (1998) 1659–1673.

[8]

G. Chen, C. Deng, Y.P. Li, TGF-β and BMP signaling in osteoblast differentiation and bone formation, Int. J. Biol. Sci. 8 (2012) 272–288. doi:10.7150/ijbs.2929.

[9]

V. Rosen, BMP2 signaling in bone development and repair, Cytokine Growth Factor Rev. 20 (2009) 475–480. doi:10.1016/j.cytogfr.2009.10.018.

[10] Y. Bai, Y. Leng, G. Yin, X. Pu, Z. Huang, X. Liao, X. Chen, Y. Yao, Effects of combinations of BMP-2 with FGF-2 and/or VEGF on HUVECs angiogenesis in vitro and CAM angiogenesis in vivo, Cell Tissue Res. 356 (2014) 109–121. doi:10.1007/s00441013-1781-9. [11] Z.S. Patel, S. Young, Y. Tabata, J.A. Jansen, M.E.K. Wong, A.G. Mikos, Dual delivery of an angiogenic and an osteogenic growth factor for bone regeneration in a critical size defect model, Bone. 43 (2008) 931–940. doi:10.1016/j.bone.2008.06.019. [12] G. Gronowicz, E. Jacobs, T. Peng, L. Zhu, M. Hurley, L.T. Kuhn, Calvarial Bone Regeneration Is Enhanced by Sequential Delivery of FGF-2 and BMP-2 from Layer-byLayer Coatings with a Biomimetic Calcium Phosphate Barrier Layer, Tissue Eng. Part A. 23 (2017) 1490–1501. doi:10.1089/ten.tea.2017.0111. [13] C.-H.W. Hemin Nie,Beng Wee Soh, Yin-Chih Fu, Three-Dimensional Fibrous PLGA/HAp Composite Scaffold for BMP-2 Delivery, Biotechnol. Bioeng. 99 (2008) 21

223–234. doi:10.1002/bit.21517. [14] D.H.R. Kempen, L. Lu, T.E. Hefferan, L.B. Creemers, A. Maran, K.L. Classic, W.J.A.Dhert, M.J. Yaszemski, Retention of in vitro and in vivo BMP-2 bioactivity in sustained delivery vehicles for bone tissue engineering, Biomaterials. 29 (2008) 3245– 3252. doi:10.1016/j.biomaterials.2008.04.031. [15] S. Kuttappan, D. Mathew, M.B. Nair, Biomimetic composite scaffolds containing bioceramics and collagen/gelatin for bone tissue engineering - A mini review, Int. J. Biol. Macromol. 93 (2016) 1390–1401. doi:10.1016/j.ijbiomac.2016.06.043. [16] K. Dashnyam, A. El-Fiqi, J.O. Buitrago, R.A. Perez, J.C. Knowles, H.-W. Kim, A mini review focused on the proangiogenic role of silicate ions released from silicon-containing biomaterials, J. Tissue Eng. 8 (2017) 1–13. doi:10.1177/2041731417707339. [17] A. Anitha, J. Joseph, D. Menon, S. V. Nair, M.B. Nair, Electrospun Yarn Reinforced NanoHA Composite Matrix as a Potential Bone Substitute for Enhanced Regeneration of Segmental

Defects,

Tissue

Eng.

Part

A.

23

(2017)

345–358.

doi:10.1089/ten.tea.2016.0337. [18] S. Kuttappan, A. Anitha, M.G. Minsha, P.M. Menon, T.B. Sivanarayanan, L.S. Vijayachandran, M.B. Nair, BMP2 expressing genetically engineered mesenchymal stem cells on composite fibrous scaffolds for enhanced bone regeneration in segmental defects, Mater. Sci. Eng. C. 85 (2018) 239–248. doi:10.1016/j.msec.2018.01.001. [19] J. Joseph, S. V. Nair, D. Menon, Integrating Substrateless Electrospinning with Textile Technology for Creating Biodegradable Three-Dimensional Structures, Nano Lett. 15 (2015) 5420–5426. doi:10.1021/acs.nanolett.5b01815. [20] L. Wang, Y. Huan, K. Pan, X. Jiang, C. Liu, Osteogenic Responses to Different Concentrations / Ratios of BMP-2 and bFGF in Bone Formation, Ann. Biomed. Eng. 38 (2010) 77–87. doi:10.1007/s10439-009-9841-8. [21] W.X. Cai, L.W. Zheng, C.L. Li, L. Ma, M. Ehrbar, F.E. Weber, R.A. Zwahlen, Effect of different rhBMP-2 and TG-VEGF ratios on the formation of heterotopic bone and neovessels, Biomed Res. Int. 2014 (2014). doi:10.1155/2014/571510. [22] J. Ratanavaraporn, H. Furuya, H. Kohara, Y. Tabata, Synergistic effects of the dual release of stromal cell-derived factor-1 and bone morphogenetic protein-2 from hydrogels 22

on

bone

regeneration,

Biomaterials.

32

(2011)

2797–2811.

doi:10.1016/j.biomaterials.2010.12.052. [23] W. Zhang, X. Wang, S. Wang, J. Zhao, L. Xu, C. Zhu, D. Zeng, J. Chen, Z. Zhang, D.L. Kaplan, X. Jiang, The use of injectable sonication-induced silk hydrogel for VEGF165and BMP-2 delivery for elevation of the maxillary sinus floor, Biomaterials. 32 (2011) 9415– 9424. doi:10.1016/j.biomaterials.2011.08.047. [24] W.J. King, P.H. Krebsbach, Growth factor delivery: How surface interactions modulate release in vitro and in vivo, Adv Drug Deliv Rev. 64 (2012) 1239–1256. doi:10.1016/j.addr.2012.03.004. [25] A. Anitha, D. Menon, T.B. Sivanarayanan, M. Koyakutty, C.C. Mohan, S. V. Nair, M.B. Nair, Bioinspired Composite Matrix Containing Hydroxyapatite-Silica Core-Shell Nanorods for Bone Tissue Engineering, ACS Appl. Mater. Interfaces. 9 (2017) 26707– 26718. doi:10.1021/acsami.7b07131. [26] J. Patterson, R. Siew, S. Herring, A. Lin, R. Guldberg, P. Staytona, Hyaluronic Acid Hydrogels with Controlled Degradation Properties for Oriented Bone Regeneration., Biomaterials. 31 (2010) 6772–6781. doi:10.1016/j.biomaterials.2010.05.047. [27] R. Morin, D. Kaplan, B. Perez-Ramirez, Bone morphogenetic protein-2 binds as multilayers to a collagen delivery matrix: An equilibrium thermodynamic analysis, Biomacromolecules. 7 (2006) 131–138. doi:10.1021/bm050461i. [28] Z.S. Patel, M. Yamamoto, H. Ueda, Y. Tabata, A.G. Mikos, Biodegradable gelatin microparticles as delivery systems for the controlled release of bone morphogenetic protein-2, Acta Biomater. 4 (2008) 1126–1138. doi:10.1016/j.actbio.2008.04.002. [29] H. Layman, M.G. Spiga, T. Brooks, S. Pham, K.A. Webster, F.M. Andreopoulos, The effect of the controlled release of basic fibroblast growth factor from ionic gelatin-based hydrogels on angiogenesis in a murine critical limb ischemic model, Biomaterials. 28 (2007) 2646–2654. doi:10.1016/j.biomaterials.2007.01.044. [30] A.H. Annor, M.E. Tang, C.L. Pui, G.C. Ebersole, M.M. Frisella, B.D. Matthews, C.R. Deeken, Effect of enzymatic degradation on the mechanical properties of biological scaffold materials, Surg. Endosc. 26 (2012) 2767–2778. doi:10.1007/s00464-012-2277-5. [31] S.N.V. Haihong Fan Tracy Chen ,Donald O’Keefe , C. Russell Middaugh, Effects of pH 23

and polyanions on the thermal stability of FGF20, Mol. Pharm. 4 (2007) 232–240. doi:10.1021/mp060097h. [32] L.F. Charles, J.L. Woodman, D. Ueno, G. Gronowicz, M. Marja, L.T. Kuhn, M. Implantology, Effects of Low Dose FGF-2 and BMP-2 on Healing of Calvarial Defects in Old Mice, Exp Gerontol. 64 (2015) 62–69. doi:10.1016/j.exger.2015.02.006. [33] A. Schmidt, D. Ladage, T. Schinköthe, U. Klausmann, C. Ulrichs, F.-J. Klinz, K. Brixius, S. Arnhold, B. Desai, U. Mehlhorn, R.H.G. Schwinger, P. Staib, K. Addicks, W. Bloch, Basic Fibroblast Growth Factor Controls Migration in Human Mesenchymal Stem Cells, Stem Cells. 24 (2006) 1750–1758. doi:10.1634/stemcells.2005-0191. [34] A. Fakhry, C. Ratisoontorn, C. Vedhachalam, I. Salhab, E. Koyama, Effects of FGF-2 / -9 in calvarial bone cell cultures : differentiation stage-dependent mitogenic effect , inverse regulation of BMP-2 and noggin , and enhancement of osteogenic potential, Bone. 36 (2005) 254–266. doi:10.1016/j.bone.2004.10.003. [35] Q. Chen, P. Shou, C. Zheng, M. Jiang, G. Cao, Q. Yang, J. Cao, N. Xie, T. Velletri, X. Zhang, C. Xu, L. Zhang, H. Yang, J. Hou, Y. Wang, Y. Shi, Fate decision of mesenchymal stem cells: Adipocytes or osteoblasts?, Cell Death Differ. 23 (2016) 1128– 1139. doi:10.1038/cdd.2015.168. [36] B.A. Boswell, L.S. Musil, Synergistic interaction between the fibroblast growth factor and bone morphogenetic protein signaling pathways in lens cells., Mol. Biol. Cell. 26 (2015) 2561–72. doi:10.1091/mbc.E15-02-0117. [37] G. An, W.-B. Zhang, D.-K. Ma, B. Lu, G.-J. Wei, Y. Guang, C.-H. Ru, Y.-S. Wang, Influence of VEGF / BMP-2 on the proliferation and osteogenetic differentiation of rat bone mesenchymal stem cells on PLGA / gelatin composite scaffold, Eur. Rev. Med. Pharmacol. Sci. 21 (2017) 2316–2328. [38] A. Gangopahyay, M. Oran, E.M. Bauer, J.W. Wertz, S.A. Comhair, S.C. Erzurum, P.M. Bauer, Bone morphogenetic protein receptor II is a novel mediator of endothelial nitricoxide

synthase

activation,

J.

Biol.

Chem.

286

(2011)

33134–33140.

doi:10.1074/jbc.M111.274100. [39] X. Yang, L. Liaw, I. Prudovsky, P.C. Brooks, C. Vary, L. Oxburgh, R. Friesel, Fibroblast Growth Factor Signaling in the Vasculature, Curr. Atheroscler. Rep. 17 (2015). 24

doi:10.1007/s11883-015-0509-6. [40] S.G. Ball, C.A. Shuttleworth, C.M. Kielty, Vascular endothelial growth factor can signal through platelet-derived growth factor receptors, J. Cell Biol. 177 (2007) 489–500. doi:10.1083/jcb.200608093. [41] M. Presta, P. Dell’Era, S. Mitola, E. Moroni, R. Ronca, M. Rusnati, Fibroblast growth factor/fibroblast growth factor receptor system in angiogenesis, Cytokine Growth Factor Rev. 16 (2005) 159–178. doi:10.1016/j.cytogfr.2005.01.004. [42] Y.J. Jih, W.H. Lien, W.C. Tsai, G.W. Yang, C. Li, L.W. Wu, Distinct regulation of genes by bFGF and VEGF-A in endothelial cells, Angiogenesis. 4 (2001) 313–321. doi:10.1023/A:1016080321956. [43] V.P. Eswarakumar, I. Lax, J. Schlessinger, Cellular signaling by fibroblast growth factor receptors,

Cytokine

Growth

Factor

Rev.

16

(2005)

139–149.

doi:10.1016/j.cytogfr.2005.01.001. [44] W. Zuo, P. Zeng, X. Chen, Y. Lu, A. Li, J. Wu, Promotive effects of bone morphogenetic protein 2 on angiogenesis in hepatocarcinoma via multiple signal pathways, Sci. Rep. 6 (2016) 37499. doi:10.1038/srep37499. [45] A.L. Jones, R.W. Bucholz, M.J. Bosse, S.K. Mirza, T.R. Lyon, L.X. Webb, A.N. Pollak, J.D. Golden, A. Valentin-Opran, Recombinant human BMP-2 and allograft compared with autogenous bone graft for reconstruction of diaphyseal tibial fractures with cortical defects: a randomized, controlled trial, J. Bone Jt. Surg. 88 (2006) 1431–1441. doi:10.2106/JBJS.E.00381. [46] A. Julka, A.S. Shah, B.S. Miller, Inflammatory response to recombinant human bone morphogenetic protein-2 use in the treatment of a proximal humeral fracture: A case report, J. Shoulder Elb. Surg. 21 (2012) e12–e16. doi:10.1016/j.jse.2011.06.006. [47] A. Grosso, M.G. Burger, A. Lunger, D.J. Schaefer, A. Banfi, N. Di Maggio, It Takes Two to Tango: Coupling of Angiogenesis and Osteogenesis for Bone Regeneration, Front. Bioeng. Biotechnol. 5 (2017) 1–7. doi:10.3389/fbioe.2017.00068.

25

Figure Legends Fig. 1: Scanning electron micrographs of nanocomposite fibrous scaffold (A) 100x (B) 250x (C) 1000x. The inset shows the photographs of the scaffold. The square shows the region of fibers in the matrix. Fig. 2: In vitro and in vivo release profiles of radiolabeled BMP2, VEGF and FGF2 from composite scaffolds. For in vitro release study, scaffolds with growth factors were incubated in PBS with or without collagenase at 37°C for 30 days. In vivo release profiles were obtained after implantation of scaffold into the back subcutis of ddY mice for 30 days. Fig. 3: Cellular response on nanocomposite scaffold with growth factors. (A&B) - Cell adhesion by scanning electron microscopy (Scale bar: 10 μm) (C&D) -Graph showing number of migrated (C) MSCs and (D) HUVEC from transwell chamber towards nanocomposite scaffold at 6 h (Scale bar – 100 µm). The number of migrated cells were significantly higher on growth factor loaded scaffolds when compared to the scaffold without growth factors (** - p <0.001). In case of HUVEC, the number of migrated cells towards CSV and CSBV was better than CSB (# p<0.05). Fig. 4: Response of MSCs and HIVEC on nanocomposite scaffold. (A) DNA Assay of MSCs Hoechst staining (B) ALP activity showing osteogenic differentiation (C) Calcium Assay showing mineralization (D) DNA Assay of HUVEC - Hoechst staining (E &F) Functionality of HUVEC by NO and VEGF release. The cell proliferation and functionality (MSCs and HUVEC) was better on growth factor loaded scaffolds when compared to CS group at respective time point (* - p<0.05 and ** - p< 0.001). In case of HUVEC, it was higher on dual growth factor loaded groups than single loaded groups (# - p<0.05). Fig. 5: Micro-CT images demonstrating vascularisation through the nanocomposite scaffold (with and without growth factors) implanted in critical sized calvarial defect - 4 weeks. The contrast was provided by microfil injection (A) Representative micro-CT images (Scale bar – 1 mm) (B) Graphical data showing fold increase in angiogenesis with respect to sham. The level of vascularisation was considerably higher in growth factor loaded groups when compared to CS (* - p< 0.05, ** - p <0.001). Among growth factor groups, CSV, CSBV and CSBF were better than CSF and CSB (# - p< 0.05). Fig. 6: Histological view of tissues implanted with nanocomposite scaffold (with and without growth factors) – Low magnification (1X). The tissues were stained with Hematoxylin & Eosin at 4 weeks and 12 weeks time point (Scale bar – 1 mm). Fig. 7: Histological view of tissues (mid region of the defect) implanted with nanocomposite scaffold (with and without growth factors) – High magnification (20X) (Scale bar – 100 µm). 26

The tissues were stained with Hematoxylin & Eosin at 4 weeks and 12 weeks time point. NB denotes new bone. Arrow towards black / grey spots denotes the blood vessels perfused with microfil. Fig. 8: Histomorphometric analysis of percentage of mature bone (lamellar) formed at the defect. The bone formation was significantly higher in CS and growth factor loaded scaffold groups when compared to sham (+ - p <0.05, ++ - p <0.001). When compared to CS, it was higher in growth factor loaded groups, except that in CSF and CSV at 4 weeks (** - p<0.001). The highest bone formation was seen in CSBV and CSBF than CSB (# - p < 0.05, ## - p <0.001).

Table 1. Concentration of growth factors used for in vitro and in vivo studies Groups In vitro studies In vivo studies CSB

0.5 µg

5 µg

CSV

0.5 µg

5 µg

CSF

0.5 µg

5 µg

CSBV (3:1)

0.5 µg BMP2 + 0.167 µg VEGF

5 µg BMP2 + 1.67 µg VEGF

CSBF (2:1)

0.5 µg BMP2 + 0.25 µg FGF2

5 µg BMP2 + 2.5 µg FGF2

27

Statement of Significance Many studies have shown the effect of growth factors like VEGF-BMP2 or FGF2-BMP2 in enhancing bone formation in critical sized defects, but there are no reports that demonstrate the direct comparison of VEGF-BMP2 and FGF2-BMP2.

In this study, we have developed a

nanocomposite fibrous scaffold that could differentially release growth factors like VEGF, BMP2 and FGF2 (VEGF release for 1 week where as BMP2 and FGF2 release for 3 weeks), which in turn promoted neovascularisation and new bone formation in critical size defect. There was no difference in vascularisation and bone formation induced by VEGF+BMP2 or FGF2+BMP2. The growth factor was loaded in a simple manner, which would ensure ease of use for the end-user, especially for the surgeon treating a patient in an operating room.

28

29

30

31

32

33

34

35

36

37