Hydroxylapatite‐collagen hybrid scaffold induces human adipose‐derived mesenchymal stem cells to osteogenic differentiation in vitro and bone regrowth in patients

Abstract Tissue engineering‐based bone graft is an emerging viable treatment modality to repair and regenerate tissues damaged as a result of diseases or injuries. The structure and composition of scaffolds should modulate the classical osteogenic pathways in human stem cells. The osteoinductivity properties of the hydroxylapatite‐collagen hybrid scaffold named Coll/Pro Osteon 200 were investigated in an in vitro model of human adipose mesenchymal stem cells (hASCs), whereas the clinical evaluation was carried out in maxillofacial patients. Differentially expressed genes (DEGs) induced by the scaffold were analyzed using the Osteogenesis RT2 PCR Array. The osteoinductivity potential of the scaffold was also investigated by studying the alkaline phosphatase (ALP) activity, matrix mineralization, osteocalcin (OCN), and CLEC3B expression protein. Fifty patients who underwent zygomatic augmentation and bimaxillary osteotomy were evaluated clinically, radiologically, and histologically during a 3‐year follow‐up. Among DEGs, osteogenesis‐related genes, including BMP1/2, ALP, BGLAP, SP7, RUNX2, SPP1, and EGFR, which play important roles in osteogenesis, were found to be upregulated. The genes to cartilage condensation SOX9, BMPR1B, and osteoclast cells TNFSF11 were detected upregulated at every time point of the investigation. This scaffold has a high osteoinductivity revealed by the matrix mineralization, ALP activity, OCN, and CLEC3B expression proteins. Clinical evaluation evidences that the biomaterial promotes bone regrowth. Histological results of biopsy specimens from patients showed prominent ossification. Experimental data using the Coll/Pro Osteon 200 indicate that clinical evaluation of bone regrowth in patients, after scaffold implantation, was supported by DEGs implicated in skeletal development as shown in “in vitro” experiments with hASCs.


| INTRODUCTION
Bone fractures/injuries can have a highly deleterious impact on patients' quality of life. Therefore, understanding the mechanisms of bone repair is important for both the patients' health and economy. Tissue engineeringbased bone grafts are emerging as a viable treatment to repair and regenerate tissues damaged as a result of diseases. The structure and composition of scaffolds stimulate the classical osteogenic pathways through an active process which occurs in mesenchymal stem cells (MSCs). Investigators are facing a great challenge to design and develop suitable scaffold materials with biological activities for the tissue regeneration in translational and precise medicine. Tissue-engineered approaches to regenerate bone in the craniomaxillofacial region utilize scaffolds to provide structural and biological cues to stem cells to stimulate osteogenic differentiation.
The repair and regrowth of craniofacial tissues are particularly challenging, because they form a complex structure which is composed of hard and soft tissues involved in complex biological functions. Porous hydroxylapatite/collagen (HA/Coll) composite biomaterials are suitable for bone grafting, as well as for bone regeneration. 1,2 The biomaterial known as Pro Osteon 200 is a coral derived porous HA, which is available in granule and block forms. This scaffold has been shown to be highly biocompatible in cellular [3][4][5] and animal models, [6][7][8] and in some clinical applications. 9 Scaffolds used for the bone regrowth can be assayed for specific features, such as osteoinductivity, osteoconductivity, and biocompatibility proprieties. Little is known about biological features and gene induction occurring after biomaterial adding.
In the past decade, significant studies have been carried out using human adipose-derived stromal/stem cells (hASCs) in the bone tissue engineering regeneration/regrowth 10 and tooth periodontal regeneration. 11 hASCs are a type of adult MSCs capable of self-renewal and differentiation toward osteogenic, chondrogenic, and adipogenic. [12][13][14] Indeed, these cells own a multilineage differentiation capacity that is regulated through extracellular signals. The cellular events related to cell adhesion and cytoskeleton organization have been suggested as central regulators of differentiation fate decision. Little is known about the molecular mechanisms which allow hASCs to differentiate. 15 Since the discovery of the hASCs induced-osteogenesis, many studies substantially contributed the knowledge of hASC cell biology and their use as a cell source for bone regeneration. 16 Ideal scaffolds, for hASC bone tissue engineering strategies, should be biocompatible and capable of supporting stem cells growth and differentiation, while providing predictable mechanical and degradation properties during the healing process.
In the present study, new investigations were carried out to evaluate the biocompatibility and osteoinductivity properties of the HA-Coll hybrid scaffold in an "in vitro" model of hASC and in patients undergoing orthognathic surgical procedures associated with malar augmentation. The cellular biology, molecular genetic and epigenetic experiments were carried out by RT-qPCR Array and protein expression, such as osteocalcin (OCN) and alkaline phosphatase (ALP) by immunostaining and ELISA approaches, respectively. The osteogenesis potential of cells was also investigated to verify the matrix mineralization by Alizarin-Red S staining. The biocompatibility of the material was evaluated by cellular viability, morphology, and cytoskeleton architecture, in hASCs grown on the hybrid scaffold.
In a previous study 17 gene expression analyses by RT-PCR technology were carried out at early times post-cell seeding on biomaterial. To this end, mRNAs extracted from cells, hASCs, were comparatively analyzed by RT-PCR at days 3 and 9. hASCs were grown on the scaffold, and plastic vessel (TCPS), used as control. PCR technology demonstrated that biomaterial induces in hASCs upregulation of specific genes, such as SPP1 and CLEC3B, which are involved in bone mineralization and ossification processes, at days 3 and 9 post-cell seeding. In addition, in our experimental conditions, ALP, ON, CHAD, SP7, and Sox9 genes had a higher expression compared with hASCs grown on TCPS, at day 3 postseeding.
In the present study, in order to mimic the long period needed "in vivo" by the bone to regrowth/repair/healing, the experiments were carried out up to day 40.
Patients operated for maxillomandibular malocclusion and/or asymmetry, or for aesthetic reasons, who underwent malar augmentation with porous HA/Coll prostheses (hybrid scaffold), were evaluated for the new bone formation during a 3-year period of follow-up using radiological and histological analyses.
The strength of this article is the combination of in vitro and in vivo evaluations of HA-Coll hybrid scaffold on osteoinductivity and bone regrowth proprieties. In hASCs grown on biomaterial, RT 2 Profiler PCR array "Human Osteogenesis" approach allowed us to analyze the epigenetic profiles of 84 genes belonging to the osteogenetic pathway, at two experimental time points, days 21 and 40.

| Biomaterial
Porous HA-derived scaffold used herein is composed by Granular Pro

| Cell differentiation and matrix mineralization
Osteogenic proteins including Tetranectin/CLEC3B expression, 17 ALP activity, 17 and OCN expression levels were analyzed in hASCs grown on scaffold. ALP activity was determined, at day 40, by a colorimetric Naphthol AS-BI phosphate-based reaction using the Alkaline Phosphatase Detection Kit (Merck Millipore Corporation, Milan, Italy), following the manufacturer's indications, as described. 17 Alizarin Red (Sigma, Milan, Italy, A5533) was used to analyze the matrix mineralization, as described. 17,18 The mineralized substrates were quantified by using a 20% methanol and 10% acetic acid in a water solution (Sigma-Aldrich, Milan, Italy). 24 . To quantify the matrix mineralization, the solution was transferred into cuvettes, whereas the quantity of Alizarin The concentration of total protein was determinate by bicinchoninic acid assay according to manufacturer's instructions. 19 The OCN protein was quantified by the Human Osteocalcin Instant ELISA (Thermo

| Bone regrowth evaluated in maxillofacial patients
The treatment plan, which aimed to achieve functional and esthetic improvement, was to enhance the malar area of patients with inadequate cheekbone projection or facial asymmetry. All orthognathic surgical procedures were carried out as previously described 9

| Histology
Biopsy samples were obtained from implants in three patients who required plates removal at maxillae level 2 years after surgery. The material collected was treated as described. 9 Morphologic and immunohistochemical analyses of bone samples were performed to evaluate remodeling and osteogenetic process. Four-micrometer-thick sections were stained with hematoxylin and eosin, and immunohistochemistry was performed on subsequent sections using two monoclonal antibodies specific for cathepsin K (clone CK4, Novocastra, Newcastle, UK; clone 3F9, Abcam, Cambridge, UK) to stain osteoclasts and for CD56 (clone 123C3.D5, 1:100; Thermo Scientific, Grand Island, New York) to stain osteoblasts, as previously described. 9 2.10 | Statistical analysis

| Scaffold modulates the DEGs implicated in skeletal development
In a previous study, we investigated in a short period of time few specific osteogenic genes, such as ALP, Osteonectin, Transcription factor Osterix, SP7 and CLEC3B, which were reported upregulated in hASCs grown on the scaffold, at day 9 (24). Herein, RT 2 Profiler PCR array was used to analyze the expression of osteogenic genes.  (Table S1, Table S2), respectively.

| Scaffold induces the matrix mineralization and osteogenic expression proteins in hASCs
Recently, we reported a significant increase of matrix mineralization and ALP activity in hASCs grown on the biomaterial, at day 21. 17 In the present investigation, the osteoinductive activity of the biomaterial is highlighted by the matrix mineralization detected in hASCs grown on the scaffold at day 40. Indeed at day 40, the F I G U R E 1 PCR array analyses of osteogenic genes. The gene expression was evaluated in human adipose mesenchymal stem cells grown on Coll/Pro Osteon200 compared with tissue culture polystyrene. A, Induced genes at day 21. In hASC cultures, mRNAs of 22 genes of the osteogenic pathway, that is, ALPL BGLAP BMP1/2, BMPR1B, COL1A1,CSF2/3, EGFR, FGFR1, FGFR2, IGF1, IGF1R, NOG,RUNX2, SOX9, SP7, SPP1, TGFB1, TNFSF1, TWIST1, and VDR were upregulated; one gene, such as VCAM1 was downregulated. B, Induced genes at day 40. In hASC culture, mRNAs of 10 genes of the osteogenic pathway, that is, BMP2, BMPR1B, CSF2, CSF3, EGF, EGFR, ITGA2, SOX9, SPP1, and TNFSF11 were upregulated, whereas nine genes were downregulated (BMP6, CD36, COMP FGFR2, IGF2, ITGAM, NOG TGFB3, and VCAM1) biomaterial favored the matrix mineralization better than the plastic vessel (TCPS), the control ( Figure 3A,C, *P < .05). Moreover, calcium deposits in hASCs grown on OC were higher compared with cells grown on the biomaterial or TCPS (**P <. 0001; Figure 3A CLEC3B protein, which binds Ca 2+ , was investigated because of its potential involvement in the bone mineral metabolism. In hASCs CD105-enriched cell populations, an increased CLEC3B expression was detected in response to osteoinduction. 30 In an earlier investigation, the CLEC3B mRNA expression levels tested upregulated at days 3 and 9. 17

| Scaffold is biocompatible in hASCs
hASCs grown on the biomaterial were investigated for their viability, proliferation, and cytoskeleton organization at days 14, 21, and 40.
hASC-eGFP grown on biomaterial showed a normal cell morphology ( Figure 4A, B, E). 17 The biomaterial demonstrated its biocompatibility up to day 40 in terms of cell adhesion and proliferation. hASC-eGFP cell morphology was indistinguishable from parental hASCs ( Figure 4A, B, E). The cytoskeleton architecture appeared to be well organized, whereas its integrity remains uninfluenced by the scaffold, up to day 40 ( Figure 4C, E). Actin fibers seem to connect the cellular membranes and the cytoskeleton to the scaffold surface with no visible loss or structural displacement. Similar physiologic cytoskeleton architecture was observed by confocal microscopy at day 40 ( Figure 4D).  (Figure 5F). with findings of earlier studies, the size of prosthesis appeared to be stable over time after an initial moderate decrease in volume during the first 18 months. 9 The prosthesis structure was radiotransparent compared with the compact aspect of the zygomatic bone. At 24 months after surgery (T2), the prosthesis seemed to adhere staunchly to the underlying zygomatic bone in all patients ( Figure 6B).

| SEM analysis
The granular structure was still distinguishable, although less evident if compared with the previous healing period. The partial  Figure 6C).
Histological analysis on bone specimens, harvested from three patients requiring plate device removal 2 years after surgery, was carried out. The persistence of porous HA scaffold and macrophages, although without inflammatory infiltrate, was found in samples. In analyzed fields, fibrous stroma was revealed in 50% of the biopsies, whereas new osteogenesis and mature bone was found in 70% of these specimens ( Figure 6D). Immunohistochemical investigations uncovered some cathepsin K protease contained in the cytoplasm of the macrophages, thus indicating the presence of osteoclast activity localized around HA granules ( Figure 6D).
The anti-CD56 antibodies indicated a higher amount of new bone formation at the side of the biopsy sample adjacent to the native bone (deep), confirming the results of histomorphometric analyses (data not shown). Specifically, in agreement with the biopsy data, the presence of mature bone was found prominently at the periosteal side, whereas the presence of new immature bone was detected entirely on the deep layer of the native bone with a bone F I G U R E 6 Scaffold characterization in patients: radiologic and histologic analyses. A-C, Cone-beam tomogram, coronal slice, at T1 (1 month), T2 (24 months), and T3 (36 months) after surgery. A, The prosthesis maintained its granular structure, whereas the granules did not migrate to the surrounding soft tissues. The structure of the prosthesis is radiotransparent compared with the compact portion of the zygomatic bone (T1). B, Cone-beam tomogram, coronal slice, at 24 months after surgery. The prosthesis seems to adhere strongly to the underlying zygomatic bone in patients. The granular structure is still distinguishable, although less evident, whereas the partial radiotransparency evolved to a radiopacity similar to that seen in the compact part of the native bone, making it impossible to distinguish the interface between the prosthesis and bone. C, Cone-beam tomogram, coronal slice, at 36 months after surgery. Progressive loss of definition of the granular architecture, with an almost complete radiopacity and apparent corticalization of the bone in contact with the prosthesis. The interface between the prosthesis and bone at T3 appears indistinguishable. D, Biopsies harvested 24 months after implant placement. Bone maturation gradient can be observed proceeding from the periosteal layer toward the native bone (hematoxylin and eosin stain: magnification ×10). Osteoclasts surrounding hydroxylapatite residual granules (immunohistochemistry with cathepsin K (magnification ×20) maturation gradient proceeding from the periosteal to the deep side. In hASCs grown on biomaterial CLEC3B protein was found to be expressed up to day 40. CLEC3B protein, which binds Ca 2+ , was investigated because of its potential involvement in the bone mineral metabolism. In hASCs CD105-enriched cell populations, an increased CLEC3B expression was detected in response to osteoinduction. 30 In a previous study, we reported that hASCs grown on the biomaterial, in an early phase at days 3 and 9, upregulated specific genes involved in bone mineralization and ossification processes. 17 It is worth recalling that the osteoinductive activity of the biomaterial is highlighted by the matrix mineralization detected in hASCs grown on the scaffold, up to day 40. Previous studies reported that the pivot role, for hASCs osteogenic differentiation, is played by the HA, [44][45][46][47] whereas the collagen mainly supports the survival, proliferation, and chondrogenic differentiation. 48,49 5 | CONCLUSION Important biological processes, underlying the continuous supply of hASCs for bone remodeling, were the osteoblastic, chondrogenic, and osteoclastic inductions stimulated by HA-Coll hybrid scaffold in patients. In conclusion, the hybrid scaffold investigated herein seems to be an excellent biomaterial able to drive bone regrowth and remodeling. This performance is due to Coll/Pro Osteon characteristics which allow to enhance in hASCs the adhesion, morphology, and proliferation, while inducing upregulation of osteogenic genes with improving matrix mineralization and cell viability.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.