Mesenchymal stem cell-based bone tissue engineering for veterinary practice.

Bone tissue engineering has been widely studied and proposed as a promising platform for correcting the bone defects. The applications of mesenchymal stem cell (MSC)-based bone tissue engineering have been investigated in various in vitro and in vivo models. In this regard, the promising animal bone defect models have been employed for illustrating the bone regenerative capacity of MSC-based bone tissue engineering. However, most studies aimed for clinical applications in human. These evidences suggest a knowledge gap to fulfill the accomplishment for veterinary implementation. In this review, the fundamental concept, knowledge, and technology of MSC-based bone tissue engineering focusing on veterinary applications are summarized. In addition, the potential canine MSCs resources for veterinary bone tissue engineering are reviewed, including canine bone marrow-derived MSCs, canine adipose-derived MSCs, and canine dental tissue-derived MSCs. This review will provide a basic and current information for studies aiming for the utilization of MSC-based bone tissue engineering in veterinary practice.

Introduction

Reconstruction of critical-sized bone defects is a challenging procedure for orthopedic surgeon. Various procedures have been introduced for bone defect treatment in clinic for example, autologous bone grafts, allogeneous bone grafts, and xenogeneous bone grafts, but the clinical outcomes are varied and lead to donor site morbidities (Shahgoli and Levine, 2011; Silber et al., 2003). Osteoconductive biomaterials are also available in clinical treatment. However, a therapeutic outcome of bone repair and regeneration in large defects or in compromised host is still unsatisfied. Hence, bone tissue engineering has been proposed as a promising tool for correcting simple and complicated bone defects (Black et al., 2015). Studies supporting the potential application of mesenchymal stem cell (MSC)-based bone tissue engineering employing in vitro and laboratory animal models have been reported (Perez et al., 2018). However, most of the evidences are focused on human application. In this review, fundamental principles and potential applications of MSC-based bone tissue engineering for veterinary practice are summarized. In addition, the canine models exhibited many features which are valuable as model manners for further human application (Pascual-Garrido et al., 2018). Animal models for bone tissue engineering have involved in various bone defect and bone disease models which are able to accelerate the translation of knowledge to clinical practice in both human and veterinary applications (McGovern et al., 2018).

Bone tissue engineering

Tissue engineering is a multidisciplinary study founded on the basis of cell biology, developmental biology, bioengineering, and biomaterial science. Tissue engineering approach aims to develop the biological substitutes that restore, maintain, or improve function of target tissues or organs (Bartold et al., 2006; Caddeo et al., 2017; Langer and Vacanti, 1993). Conventionally, three main components are comprised of matrix (scaffolds), applied cues (biochemical or biophysical cues), and cell resources which are required for an establishment of successful tissue engineering (Bartold et al., 2006; Caddeo et al., 2017).

Scaffolds in bone tissue engineering

Scaffolds are functioned as the supportive extracellular matrix (ECM). Scaffolds should support cell adherence, spreading, proliferation, differentiation, maturation, communication, and ECM production (Caddeo et al., 2017). Scaffolds in bone tissue engineering are usually osteoconductive materials which provide an environment for bone formation in vivo. However, osteoinductive materials are also utilized as bone tissue engineering scaffolds since they could promote osteogenic differentiation leading to bone regeneration (Barradas et al., 2011; Blokhuis and Arts, 2011; Habibovic and de Groot, 2007).

Biomaterials for bone scaffolds

Different materials exhibit specific advantages and disadvantages for bone scaffold utilization. In this regard, natural polymers demonstrate high biocompatibility and contain biological activity. However, low mechanical strength and fabrication limitation are main disadvantages. Synthetic materials like ceramics with calcium phosphate-based materials exhibit osteoinductive abilities (Barradas et al., 2011). Calcium and phosphate ions have been shown to induce osteogenic differentiation in various cell types (Ali Akbari Ghavimi et al., 2018; An et al., 2012). Though, some bioceramic materials have a slow degradation rate, resulting in the remaining of materials in newly regenerated bone tissues (Sheikh et al., 2015). Due to flexible mechanical properties, ease for fabrication, and relatively low cost, biodegradable synthetic polymers have been used as a material of choices for bone tissue engineering scaffolds. However, these materials are lack of bioactivities and some materials are toxic upon degradation (Pastino et al., 2018).

To overcome some limitations of specific materials, advanced hybrid biomaterials have been developed. For example, poly(lactic-co-glycolic acid) (PLGA) and PLGA-polycaprolactone (PCL) co-polymer are developed to achieve a designed degradation time (Ulery et al., 2011). PLGA-polyphosphazenes polymer blends are generated to prevent biological incompatibility of the degradation products (Deng et al., 2008; Krogman et al., 2009). Further, polymer-bioceramic composite materials are initiated to improve osteoinductive properties of polymer materials and to control degradation ability of bioceramics. The exemplification of these polymer-bioceramic composite materials is HA-poly (lactic acid) (PLA), HA-PLGA, HA-gelatin, HA-Chitosan, HA-collagen (Chesnutt et al., 2009; Kim et al., 2005; Kim et al., 2006; Rodrigues et al., 2003; Wei and Ma, 2004; Zhang et al., 2003).

Scaffold characters for bone regeneration

Physical scaffolds’ characters have been shown to participate in the osteoinductive and osteoconductive properties. The examples of those physical parameters are porosity, pore size, surface topography, surface roughness, and surface chemical composition. It has been proposed that these physical parameters could directly interact with cells and also indirectly control cells by the absorbed protein on the surface (Polo-Corrales et al., 2014). Scaffolds with suitable microporosity (i.e. pore size, volume, and interconnectivity) support cell proliferation and migration, nutrient and oxygen flow, vascularization, and ECM production (Logeart et al., 2005; Zeltinger et al., 2001). It has been shown that scaffolds with mean pore size >300 μm could efficiently enhance osteoblast proliferation and differentiation throughout the scaffold due to the thorough delivery of oxygen and nutrient within the scaffold (Holtorf et al., 2005; Kühne et al., 1994; Tsuruga et al., 1997; Volkmer et al., 2008). However, scaffolds with small pore size (<200 μm) could only illustrate peripheral osteoblast survival and bone formation (Karageorgiou and Kaplan, 2005). Nano-scale topography of scaffold plays a crucial role in osteoinduction, osteoconduction, and osseointegration (Porter et al., 2009; Ward and Webster, 2006). In this regard, it has been shown that rough surface promotes human osteoblast-like cell spreading and proliferation (Osathanon et al., 2011). Surface chemistry has been shown to participate to cell response. Phosphoric acid treated titanium surface demonstrates similar surface roughness parameters, but a high phosphate and oxygen content is observed on the surface (Osathanon et al., 2016). Results demonstrate that phosphoric acid treated surface promotes cell attachment and cytoskeletal orientation in murine pre-osteoblast cells (Osathanon et al., 2016). Taken all evidences together, physical characteristics of scaffolds exhibit a crucial role in biological responses.

Fabrication and scaffolding techniques

As scaffold architecture is an important factor to regulate cell responses. Thus, various technologies are introduced in order to tightly control the fabrication of scaffolds’ architecture favoring bone formation. Variety of fabrication techniques have been employed for the production of nano-featured scaffolds i.e. electrospinning (Lao et al., 2011; Tuzlakoglu et al., 2011), molecular self-assembly (Mata et al., 2010; Matson and Stupp, 2012), fiber bonding (Tuzlakoglu et al., 2011), and phase separation (Liu and Ma, 2009). With computer-based technology, computer assisted design (Caddeo et al., 2017) computer assisted technology (Black et al., 2015) systems have been utilized for the production of personalized and anatomical-shaped scaffolds to achieve the complex architecture bone tissues (Grayson et al., 2010).

The additional aspect for designing the scaffold is the orthopedic tissue interface (hard tissue-soft tissue interface) which is very important for graft stability and bone regeneration process (Lu et al., 2010). Bone-soft tissue (ligament, tendon, and cartilage) interface is such a complex and multifactorial organization of distinct tissue types required for musculoskeletal homeostasis and physiological functions (Lu et al., 2010; Qu et al., 2015). Designing of layered scaffolds mimicking natural structure and molecular components of bone-soft tissue interface is introduced. It has been reported that three-phase biomaterials composing of ligament-fibrocartilage-bone-like matrixes can support the maintenance of three cell types (fibroblasts, chondrocytes, and osteoblasts) in vitro (Spalazzi et al., 2006a, b). In term of bone-cartilage interface, osteochondral interface is a key component organizing load bearing and force distribution (Lu et al., 2010). Various material-forming techniques have been reported for the production of osteochondral interface-mimicking biomaterials including biphasic scaffolds fabricated with polyglycolic acid (PGA) mesh and PLGA/polyethylene glycol foam (Schaefer et al., 2000), hyaluronan sponge and porous calcium phosphate scaffold (Gao et al., 2001), and photo-polymerized polyethylene glycol-diacrylate hydrogel (Alhadlaq and Mao, 2005). One side of these layered materials is seeded with osteoblasts and another side is seeded with chondrocytes. This co-culture of two cell types on layered materials could simulate the bone-cartilage interface in vitro.

To further enhance the bone regenerative capacity of scaffolds, the ability to induce cell homing has been proposed as another key factor. In this regard, cell homing induction could be achieved by two approaches, scaffold-based and/or cell-based strategies (Shin et al., 2003). Scaffolds can be engineered to be more attractive for the cells, leading to the effective angiogenesis and subsequently bone regeneration. The integration of these strategies has been illustrated in the customized cytokine microdelivery system that was incorporated with PCL-based scaffold (Schantz et al., 2007). The system could sequentially deliver vascular endothelial growth factor (VEGF), stromal cell-derived factor (SDF)-1, and bone morphogenetic protein (BMP)-6 resulting in the enhanced angiogenesis, MSCs migration, and matrix mineralization, respectively (Schantz et al., 2007). Moreover, constructing a bio-adhesive surface on scaffold by immobilizing short protein motifs (e.g. arginine-glycine-aspartic acid (RGD motif) or glycine-phenylalanine-hydroxyproline-glycine-glutamate-arginine (GFOGER) shows the significant effects on cell adhesion, proliferation, and osteogenic differentiation (Garcia and Reyes, 2005).

MSC-based bone tissue engineering for veterinary practice

Applying of appropriate cell resources is necessary for enhancing bone regeneration and neovascularization (Amini et al., 2012). Many studies suggest the use of pluripotent and multipotent stem cells as cell sources to enhance bone regeneration. In this regard, pluripotent stem cells (such as embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) possess a high differentiation potency and are able to differentiate toward osteogenic lineage upon an appropriate stimulation (Ardeshirylajimi and Khojasteh, 2017; Buttery et al., 2001; Kawaguchi et al., 2005; C. Zhang et al., 2017). However, there are some concerns according to their tumorigenicity and immunological incompatibility after transplanted (Amini et al., 2012). Hence, the application of mesenchymal stem cells (MSCs) has been investigated for bone tissue engineering utilization (Amini et al., 2012). Isolations of MSCs from various tissue resources have been reported for example bone marrow (Bianco et al., 2001; Macrin et al., 2017), adipose tissues (Zuk et al., 2002), dental tissues (Miura et al., 2003; Shi and Gronthos, 2003), umbilical cord blood (Rosada et al., 2003), peripheral blood (Kuznetsov et al., 2001), amniotic fluid (In 't Anker et al., 2004), brain (Crisan et al., 2008; Lojewski et al., 2015), skin (Joannides et al., 2004), heart (de Paula et al., 2017), kidney (Leuning et al., 2017a, b), and liver (El-Kehdy et al., 2016). In this review, potential application of three main MSCs in veterinary bone tissue engineering namely a) bone marrow-derived MSCs (BM-MSCs), b) adipose-derived MSCs (AD-MSCs), and c) dental tissue-derived MSCs, is discussed.

Stemness and surface marker expression along with osteogenic gene marker expression upon an in vitro induction by these three canine mesenchymal stem cells were summarized in Table 1 and Table 2, respectively.

Table 1

Stemness and surface marker expression of canine mesenchymal stem cells.

	cBM-MSCs	cDPSCs	cAD-SCs
STRO-1	− (Bearden et al., 2017)	+ (Dissanayaka et al., 2011)	− (Bearden et al., 2017)
Nanog	+ (Bearden et al., 2017)	+ (Dissanayaka et al., 2011)	+ (Bearden et al., 2017)
OCT4	+ (Bearden et al., 2017)	N/D	+ (Bearden et al., 2017)
Sox2	+ (Bearden et al., 2017)	N/D	+ (Bearden et al., 2017)
CD8	+ (Russell et al., 2016)	N/D	+ (Russell et al., 2016)
CD4	+ (Russell et al., 2016)	N/D	+ (Russell et al., 2016)
CD9	+ (Bearden et al., 2017)	N/D	+ (Bearden et al., 2017)
CD11b	N/D	N/D	− (Bahamondes et al., 2017)
CD14	+ (Russell et al., 2016)	N/D	+ (Russell et al., 2016)
CD29	+ (Russell et al., 2016)	N/D	+ (Russell et al., 2016)
CD34	− (Bearden et al., 2017)	N/D	− (Bearden et al., 2017)
CD44	+ (Bearden et al., 2017)	+ (Khorsand et al., 2013)	+ (Bahamondes et al., 2017)
CD45	− (Russell et al., 2016)	− (Dissanayaka et al., 2011)	− (Bahamondes et al., 2017)
CD73	+ (Russell et al., 2016)	+ (Dissanayaka et al., 2011)	+ (Russell et al., 2016)
CD90	+ (Bearden et al., 2017)	+ (Dissanayaka et al., 2011)	+ (Bahamondes et al., 2017)
CD105	+ (Bearden et al., 2017)	− (Dissanayaka et al., 2011)	+ (Bearden et al., 2017)
CD146	N/D	+ (Dissanayaka et al., 2011)	N/D

cBM-MSCs, canine bone marrow-derived mesenchymal stem cells; cDPSCs, canine dental pulp stem cells; cAD-SCs, canine adipocyte stem cells; N/D, no data.

Table 2

Osteogenic gene marker expression by canine mesenchymal stem cells upon an in vitro induction.

	cBM-MSCs	cDPSCs	cAD-SCs
Runx2	Upregulation (Russell et al., 2016)	N/D	Upregulation (Russell et al., 2016)
Osteocalcin	Upregulation (Chung et al., 2012)	N/D	Upregulation (Requicha et al., 2012)
Collagen 1A1	Upregulation (Chung et al., 2012)	Upregulation (Khorsand et al., 2013)	Upregulation (Chung et al., 2012)
Osterix	Upregulation (Kisiel et al., 2012)	N/D	Upregulation (Kisiel et al., 2012)
Osteopontin	Upregulation (Russell et al., 2016)	Upregulation (Khorsand et al., 2013)	Upregulation (Russell et al., 2016)
ALP	Upregulation (Kisiel et al., 2012)	N/D	Upregulation (Russell et al., 2016)

cBM-MSCs, canine bone marrow-derived mesenchymal stem cells; cDPSCs, canine dental pulp stem cells; cAD-SCs, canine adipocyte stem cells; N/D, no data.

Bone marrow-derived mesenchymal stem cells (BM-MSCs)

Human and mouse BM-MSCs have been isolated and well characterized. Human BM-MSCs have been identified as CD10+, CD13+, CD29+, CD44+, CD59+, CD71+, CD73+, CD90 (Thy1)+, CD105+, CD106 (VCAM)+, CD146+, CD166 (ALCAM)+, STRO-1+, CD11a-, CD14-, CD19-, CD31 (PECAM)-, CD34 (C-18)-, CD45-, CD48-, CD135-, and HLA-DR- (Arvidson et al., 2011; Macrin et al., 2017; Peister et al., 2004). Canine BM-MSCs (cBM-MSCs) have been recently isolated and described as the clonogenic cells containing self-renewing and multilineage differentiation capacities. cBM-MSCs are able to differentiate toward osteogenic, chondrogenic, and adipogenic lineages (Chung et al., 2012; Sawangmake et al., 2016; Screven et al., 2014; Spencer et al., 2012). These cells express pluripotent stem cell markers, zinc finger protein 42 (ZFN42 or Rex1) and POU class 5 homeobox 1 (Pou5f1)/octamer-binding transcription factor 4 (Oct4) (Sawangmake et al., 2016). Immunophenotyping characterization by flow cytometry demonstrates that cBM-MSCs express CD44+ and CD90 + but lack of CD14-, CD29-, CD34-, and MHC II- (Screven et al., 2014). Additional analysis of cell surface markers by real-time reverse transcription-polymerase chain reaction (qRT-PCR) revealed that cBM-MSCs were negative for CD45 and positive for CD14, CD29, CD44, CD73, CD90, CD105, and MHC II (Screven et al., 2014). PCR array has been used to analyze the pattern of gene expression in cBM-MSCs comparing with peripheral blood mononuclear cells (PBMCs), and the results suggested the set of upregulated genes (matrix metallopeptidase 2; Mmp2, and platelet-derived growth factor receptor β; Pdgfrβ) and down regulated genes (protein tyrosine phosphatase receptor type C; Ptprc or Cd45, tumor necrosis factor; Tnf, integrin subunit alpha X; Itgax, beta-2-microglobulin; β2m, interleukin 10; Il10, peroxisome proliferator activated receptor gamma; Pparɣ, transforming growth factor beta 1; Tgfβ1), indicating the mesenchymal specific lineage characterization of cBM-MSCs compared with PBMCs (Screven et al., 2014).

In vitro and in vivo osteogenic differentiation potential of cBM-MSCs has been illustrated (Arinzeh et al., 2003; Chung et al., 2012; Sawangmake et al., 2016; Screven et al., 2014). An in vitro osteogenic induction is performed by employing the culture medium containing 40 μg/mL to 50 mg/mL ascorbic acid, 20 nM–100 nM dexamethasone, and 10 μM to 10 mM β-glycerophosphate. The induction period is also varied from 7 to 21 days (Bearden et al., 2017; Chung et al., 2012; Sawangmake et al., 2016; Screven et al., 2014; Spencer et al., 2012). At the end of the induction period, cells were positively stained with Alizarin Red S or Von Kossa silver staining. The upregulation of osteogenic gene markers (core-binding factor alpha 1 (Cbfa1)/Runt-related transcription factor 2 (Runx2), bone gamma-carboxyglutamic acid-containing protein (Bglap)/osteocalcin (Ocn), collagen type 1 alpha 1 (Col1A1), osterix (Osx), and bone sialoprotein I (Bsp-1)/osteopontin (Opn)) along with the increase of alkaline phosphatase activity are illustrated (Bearden et al., 2017; Sawangmake et al., 2016; Screven et al., 2014; Spencer et al., 2012). These evidences imply the osteogenic differentiation ability of cBM-MSCs.

Various factors are shown to influence the osteogenic differentiation ability of cBM-MSCs in vitro. In this respect, β-glycerophosphate is a source of inorganic phosphate for mineralization in vitro. It is cleaved by alkaline phosphatase enzyme resulting in the release of phosphate ions. Increasing of β-glycerophosphate concentration to 20 and 40 mM enhances an upregulation of Osx and Col1A1 mRNA expression without significant effects on alkaline phosphatase activity and mineral deposition ability compared with 10 mM β-glycerophosphate supplementation (Sawangmake et al., 2016). Supplementation of recombinant human bone morphogenic protein-2 (rhBMP-2) in osteogenic inductive medium increases an in vitro alkaline phosphatase activity and mineralization in dose-dependent manner (Bearden et al., 2017).

Potential application of cBM-MSCs in bone tissue engineering has been studied in various canine bone defect models. Segmental long bone defects treated with cBM-MSCs loaded HA-tricalcium phosphate has been reported (Arinzeh et al., 2003). Results show that callus formation is observed throughout the length of the defect at eight weeks after transplantation (Arinzeh et al., 2003). New bone formation entire the implant was found at sixteen weeks post-transplant (Arinzeh et al., 2003). Further, transplantation of cBM-MSCs in canine mandibular segmental bone defect has been investigated (Yuan et al., 2007). In this regard, osteogenically induced autologous cBM-MSCs seeded on biodegradable β-tricalcium phosphate (β-TCP) scaffold could regenerate the mandibular defects compared with the control (Yuan et al., 2007). At 32 weeks post-operation, the completed bone union of cBM-MSCs seeded β-TCP scaffold transplant is achieved as confirmed by morphological, radiographical, and biomechanical analyses (Yuan et al., 2007). Another success application has been shown in canine orbital wall defect model. In this study, a combination of autologous cBM-MSCs and canine bone marrow (cBM) aspirate is seeded onto β-TCP scaffolds. After 24 weeks, a successful bone repair of orbital wall bone defects is observed compared with β-TCP scaffold and blank control (Wang et al., 2014).

Despite the evidences supporting bone regenerative potential of cBM-MSCs, some reports have suggested the burdens in cBM-MSCs application especially the invasive harvesting protocol and the difficulty of cell expansion in vitro (Bearden et al., 2017; Chung et al., 2012; Oryan et al., 2017). cBM-MSCs showed lesser in vitro capabilities comparing with cAD-MSCs according to initial cell isolation yield, colony forming unit (CFU) potency, and proliferation ability (Bearden et al., 2017). It has been reported that hypoxic conditions (5% and 10% oxygen tension) hampered an in vitro proliferation capacity of cBM-MSCs comparing with 21% oxygen tension (Chung et al., 2012). In this regard, strategy to enhance cell isolation yield and proliferation potential of cBM-MSCs is indeed necessitated for improving the potential application of the cells for bone tissue engineering. Our report has recently suggested the beneficial effects of simvastatin in low-dose range, 0.1 and 1 nM, on cBM-MSCs proliferation and pluripotent marker expression in vitro (Nantavisai et al., 2019).

Adipose-derived mesenchymal stem cells (AD-MSCs)

Canine AD-MSCs (cAD-MSCs) have been isolated and characterized (Screven et al., 2014; Spencer et al., 2012). According to the previous publications, morphology of the early passage cAD-MSCs has been described as a homogeneous spindle-shaped phenotype. The late passage morphology (after 5 passages) has changed to a wide and flat appearance (Spencer et al., 2012). The tri-lineage differentiation capacity of cAD-MSCs (adipogenicity, chondrogenicity, and osteogenicity) has been illustrated in vitro (Bearden et al., 2017; Screven et al., 2014; Spencer et al., 2012). Surface marker characteristics of cAD-MSCs analyzed by flow cytometry have been defined as CD9+, CD44+,CD90+, CD105+, MHC I+, CD14-, CD29-, CD34-, CD45-, STRO-1-, and MHC II− (Bearden et al., 2017; Screven et al., 2014). The expressions of pluripotent markers (Nanog homeobox; Nanog, Pou5f1/Oct4, and SRY-box 2; Sox2) by cAD-MSCs have been illustrated using conventional PCR (Bearden et al., 2017). Canine-specific PCR array is used to identify the unique gene expression profiles of cAD-MSCs compared with canine peripheral blood mononuclear cells (cPBMCs). The results suggested that cAD-MSCs showed the similar gene expression pattern as found in cBM-MSCs. Set of upregulated genes comprised Mmp2 and Pdgfrβ while set of downregulated genes included Ptprc or Cd45, Tnf, Itgax, β2m, Il10, Pparɣ, Tgfβ1 (Screven et al., 2014). Additionally, adipose tissues yielded lesser number of isolated cells per gram of tissue when compared with bone marrow source. However, cAD-MSCs showed greater performances comparing with cBM-MSCs in terms of CFU percentage, short- and long-term proliferation, number of population doublings, as well as cell number per culture plate (Bearden et al., 2017).

An in vitro osteogenic differentiation potential of cAD-MSCs has been investigated. Upon the maintenance with osteogenic induction medium containing the supplement combination as described previously in cBM-MSCs section, an in vitro osteogenic differentiation of cAD-MSCs is illustrated by an upregulation of osteogenic mRNA markers (Cbfa1/Runx2, Bglap/Ocn, Col1A1, Osx, and BSP-1/OPN) and the increase of alkaline phosphatase activity. The positive Alizarin Red S or Von Kossa staining was found in osteogenic induced cells compared to undifferentiated control suggesting the mineralization of extracellular matrix (ECM). The osteogenic differentiation potency between cAD-MSCs and cBM-MSCs is comparable especially the low responses of alkaline phosphatase activity of the both cell types upon an in vitro induction (Bearden et al., 2017; Chung et al., 2012; Screven et al., 2014; Spencer et al., 2012). rhBMP-2 supplement during osteogenic induction could enhance an in vitro alkaline phosphatase activity and mineralization in cAD-MSCs similar to those of cBM-MSCs (Bearden et al., 2017). Hypoxic conditions (5% and 10% oxygen tension) decreased an in vitro proliferation capacity and osteogenic differentiation potential of cAD-MSCs comparing with 21% oxygen tension (Chung et al., 2012).

Potential application of cAD-MSCs in veterinary bone tissue engineering has been investigated in canine cranial bone defect model (Cui et al., 2007). In this regard, cAD-MSCs isolated from inguinal subcutaneous fat pad are osteogenically induced and seeded on cuboid coral scaffolds (width 20 mm x height 20 mm x thickness 3 mm). After maintained for 7 days, the cells-seeded scaffolds and control scaffolds are transplanted into the bilateral critical-size parietal bone defects (20 mm × 20 mm) in each animal. Significant bone formation was observed at week 12 post-transplantation in the defects treated with cells-seeded scaffold. At 24 weeks post-transplant, the average bone formation area is 84.19 ± 6.45% and 25.04 ± 18.82% in the defects treated with cells-seeded scaffold and control scaffold, respectively. Histological analyses illustrate that the defects in cells-seeded scaffold transplant are repaired by a typical bone tissue while fibrous tissues are found in control group (Cui et al., 2007). Recent study demonstrates the advantages of technology that could preserve the cell-to-cell interactions and the endogenous ECM connections (Kim et al., 2016). These processes are necessary for osteogenicity and bone regenerative capability of cAD-MSCs (Kim et al., 2016). In this study, PCL/β-TCP composite scaffolds are wrapped with osteogenically induced cAD-MSCs sheets and transplanted into the 15 mm segmental defects of canine radial diaphysis. At 12 weeks post-transplant, the new cone-shaped bone formation is observed, and the newly formed bone mass in the defects treated with cAD-MSCs sheet-wrapped scaffolds is significantly greater than that in cAD-MSCs-seeded and control scaffolds. Histological examinations revealed the newly formed bone had a woven, trabecular appearance surrounding by fibrous connective tissue with no obvious inflammation (Kim et al., 2016). However, it should be noted that the new bone formation in this study model at 12 weeks post-transplant did not completely cover the defect.

Although cAD-MSCs possesses osteogenic differentiation and bone regeneration potential, the study regarding clinical application of cAD-MSCs for bone tissue regeneration is still limited especially in terms of bone defect model variety and effectiveness comparison between autologous and allogenic transplants.

Dental tissue-derived mesenchymal stem cells (dental-MSCs)

Isolation of various human dental-MSCs types have been previously reported e.g. human dental pulp stem cells (hDPSCs), human periodontal ligament stem cells (hPDLSCs), and stem cells from human exfoliated deciduous teeth (SHED) (Egusa et al., 2012a). These dental-MSCs contained plasticity and differentiation potency toward various lineages including osteogenic, chondrogenic, adipogenic, pancreatic, and neurogenic lineages (Egusa et al., 2012a; Osathanon et al., 2014, Sawangmake et al., 2014a, b). In veterinary practice, canine dental pulp stem cells (cDPSCs) from canine premolars and canine periodontal ligament stem cells (cPDLSCs) have been isolated and reported (Dissanayaka et al., 2011; Tsumanuma et al., 2016). cDPSCs illustrate clonogenicity and contained higher proliferation rate compared with hBM-MSCs. Expression of Nanog mRNA, a pluripotent marker, is also observed. The expression of cell surface markers of cDPSCs were not quite correlated with those expressed in hDPSCs. Surface molecule immunophenotyping of cDPSCs using flow cytometry has been illustrated as CD90+ (relative low), STRO-1+ (relative low), CD45-, CD73-, and CD105-. CD146 mRNA expression was found as analyzed by RT-qPCR (Dissanayaka et al., 2011). In addition, previous research has shown that cPDLSCs expressed CD29+, CD90+, CD34- and CD45- (Tsumanuma et al., 2016). The inconsistent surface marker expression might due to the specificity of antibodies used in flow cytometry analysis.

Osteo/odontogenic differentiation potential of cDPSCs has been investigated in vitro. Upon the maintenance with osteo/odontogenic induction medium containing 10 nM 1,25 dihydroxyvitamin D3, osteo/odontogenic differentiation of cDPSCs is illustrated by the increase of alkaline phosphatase enzymatic activity, mineral deposition, and dentin sialoprotein (DSP) expression. Neurogenic and adipogenic differentiation potential of cDPSCs is also illustrated in vitro (Dissanayaka et al., 2011). There is no further analysis regarding osteo/odontogenic mRNA marker expression of cDPSCs undergone an in vitro induction in this study.

This evidence has preliminarily suggested the osteogenic differentiation potential of cDPSCs in vitro. However, according to the regenerative craniofacial/bone therapies in human, two main issues (i.e. material/scaffold technology and dental stem cell technology) have to be addressed to accomplish the veterinary clinical application using dental-MSCs-based bone tissue engineering (Egusa et al., 2012b).

Conclusion

Advantages and disadvantages of three canine mesenchymal stem cells are summarized in Table 3. Most of the studies regarding bone tissue engineering in various animal bone defect models were human application-aimed studies. However, there were numerous of promising studies that supported the application of MSCs for veterinary bone tissue engineering, especially cBM-MSCs. For cAD-MSCs, despite the evidences supporting osteogenic differentiation potential in vitro, the studies regarding an in vivo bone regenerative capacity and material/scaffold technology are indeed necessitated. Additionally, the promising evidences supporting the in vitro and in vivo osteogenic differentiation and bone regenerative potential of dental-MSCs are still lacking. Studies regarding these issues will fulfill the development of technologies required for the cMSC-based bone tissue engineering for veterinary practice.

Table 3

Advantages and disadvantages of canine mesenchymal stem cells.

	Advantages	Disadvantages
cBM-MSCs	- Self-renewal - Multilineage differentiation potential - High yield of isolated cells - Many supporting evidences both in vitro and in vivo	- Invasive harvesting procedure - In vitro cell expansion difficulty
cAD-SCs	- Self-renewal - Multilineage differentiation potential - High proliferation rate - Many supporting evidences in vitro - Simple harvesting procedure - Multiple site collections	- Morphological change (after 5 passages) - Low number of isolated cells - Less supporting evidences in vivo
cDPSCs	- Self-renewal - Multilineage differentiation potential - High proliferation rate - High osteogenic differentiation performances	- Low number of isolated cells - Less evidences both in vitro and in vivo - Sample collection difficulty

cBM-MSCs, canine bone marrow-derived mesenchymal stem cells; cDPSCs, canine dental pulp stem cells; cAD-SCs, canine adipocyte stem cells.

Declarations

Author contribution statement

All authors listed have significantly contributed to the development and the writing of this article.

Funding statement

CS was supported by the Veterinary Clinical Stem Cell and Bioengineering Research Unit, Ratchadaphiseksomphot Endowment Fund, Chulalongkorn University, Thailand; the Chulalongkorn University Office of International Affairs Scholarship for Short-term Research, Thailand; the Government Research Fund, Thailand; and the Asahi Foundation research grant, Japan. TO was supported by the Chulalongkorn Academic Advancement into Its 2nd Century Project, Thailand.

Competing interest statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.