Regeneration of temporomandibular joint using in vitro human stem cells: A review

Abstract Temporomandibular joint disorders (TMDs) range from gross anatomic deformities of the disc and hard tissue to functional disturbances. Traditional treatment of TMDs includes physical therapy, use of appliances, pharmacological, surgical and psychological interventions. However, during the late stage of TMDs, conventional management often results in inadequate relief of symptoms. Stem cell‐based tissue regeneration has been studied extensively in joint regeneration, including the Temporomandibular Joint (TMJ). This study aims to review the potential of various human stem cells (HSC) for the regeneration of the TMJ. In vitro studies using human mesenchymal stem cells cultured under different conditions to evaluate regeneration of TMJ related structures were searched on PubMed, EMBASE, Cochrane, and Web of Science up to March 2020. In vitro studies utilized several different types of stem cells under varying conditions. Increased osteogenesis and/or chondrogenesis were noted with stem cell interventions compared to control groups on Alkaline Phosphatase (ALP) activity, Col‐I, Col‐II, Col‐X, RUNX2, LPL, and Aggrecan mRNA expression. This review emphasizes the potential of stem cell therapies in the regeneration of TMJ‐related structures. However, further in vivo studies are required to evaluate the efficacy and safety of these therapies in humans.

The development of the TMJ is different from other joints in that the cartilage of the mandible's condyle is secondary cartilage compared to the articular cartilage found in other joints, which is primary cartilage associated with endochondral ossification that is dominated by hyaline cartilage. Secondary cartilage develops in association with specific bones formed by intra-membranous ossification after the bones are already formed. These developing bones become entirely surrounded by the periosteum, including the areas that eventually form the articular surfaces of the TMJ. The periosteum lining these articular surfaces is gradually transformed into the dense fibrous articular tissues of the TMJ during its early development (Symons, 1965). Mature condylar cartilage consists of four zones: (1) superficial or articular zone of fibrous tissue facing the disc expressing collagen I, (2) proliferative pre-chondroblastic zone expressing collagen I, (3) a chondroblastic zone expressing collagen II, the proteoglycans aggrecan, decorin, chondroitin sulphate PG, and keratin sulphate PG, and (4) a hypertrophic zone adjacent to bone expressing collagen X. Therefore, TMJ has a unique hybrid structure integrating a superficial fibrocartilage layer covering a secondary hyaline cartilage layer (Chen et al., 2020). It is critical for regenerative cell-based therapies specific for the TMJ to reproduce the zonal architecture of the mandibular condylar cartilage (Chen et al., 2020); however, one of the major challenges for developing an effective regeneration therapy is the lack of understanding of the unique formation of the condylar cartilage from the other synovial joints and replicate this zonal architecture (Tanaka & Koolstra, 2008) (Figure 1b). Temporomandibular disorders (TMDs) include arthralgia, localized myalgia, myofascial pain, internal derangements (disc displacement with or without reduction), degenerative joint diseases, subluxation, and headache attributed to TMD (Schiffman et al., 2014). Conservative measures, such as, physical therapy, oral appliance therapy, pharmacotherapy including topical and systemic medications, glucocorticosteroid injections and arthrocentesis could help in the early stages of repair (Dantas & Vivan, 2015;Dimitroulis et al., 1995;Durham et al., 2015). As the degeneration progresses, bony erosion, condylar resorption, and discal rupture/deformation can be seen radiographically (Brooks et al., 1997). When the wear goes beyond repair, in conditions such as severe trauma or systemic conditions (including autoimmune arthritis, connective tissue disorders and idiopathic condylar resorption), patients may need invasive treatment, involving open surgeries with replacement of the whole or parts of the TMJ with autogenous or allogenous materials (Elgazzar et al., 2010). The prognosis of surgeries varies, depending on the severity of the condition, co-morbidities related to patient's health and the practitioners' knowledge and skills (Dimitroulis, 2005).
Damage to TMJ structures is usually irreversible, and commonly used treatment strategies described above cannot restore damaged TMJ tissues. Stem cell-based therapy is sought as an alternative approach to current treatment strategies, according to few recent studies, to repair discal or bony damage from common conditions such as rheumatoid arthritis and osteoarthritis (Cui et al., 2017;Serakinci & Savtekin, 2017). Two main types of stem cells exist: omnipotent embryonic and non-embryonic/adult stem cells (Helgeland et al., 2018), which include hematopoietic and mesenchymal stem cells (MSCs), along with neural, epithelial and skin stem cells. MSCs are multipotent stem cells that can be isolated from many human tissues such as bone marrow, synovium, fat, muscle, and periodontal ligament of teeth (Mao et al., 2006). Cells used for TMJ regeneration include whole bone marrow extracts, synovial-derived MSCs, bone marrow-derived, adipose tissue-derived and fibrocartilage stem cells, chondrocytes, osteoblasts, and fibroblast-like synoviocytes. Bone marrow-derived stem cells are most frequently used, either alone or embedded in natural or synthetic polymer scaffolds (Helgeland et al., 2018;Mao et al., 2006;Puelacher, 2011). In addition, various growth factors such as bone morphogenetic protein 2 (BMP-2), transforming growth factor-beta 2 (TGF-beta 2), connective tissue growth factor (CTGF), transforming growth factor-beta 1 (TGF-beta 1), and platelet-rich plasma (PRP) have been used alone or in combination with cells and/or scaffolds to regenerate discal or osteochondral TMJ tissues (Helgeland et al., 2018;Puelacher, 2011). PRP is used most often due to its various growth factors, such as platelet-derived epidermal growth factor (PD-EGF), platelet-derived growth factor (PDGF), bone morphogenetic protein (BMP), transforming growth factor (TGF), insulin-like growth factor (IGF), vascular endothelial growth factor (VEGF), endothelial cell growth factor (ECGF), basic fibroblast growth factor (bFGF) (Coskun et al., 2019;Foster et al., 2009;Kılıç et al., 2019;Wang et al., 2018).
Studies have shown that human MSCs can promote TMJ tissue damage repair, suppress the inflammatory response, and modulate the immune system (El Qashty et al., 2018). Transplanted MSCs can seed in the target tissue or migrate to the target tissue and differentiate into mature cells to help in tissue repair (Cui et al., 2017;Serakinci & Savtekin, 2017). These cells also possess antiinflammatory and immunomodulatory properties, which can speed up the healing process in the TMJ (van Poll et al., 2008). Growth factors such as IGF, bFGF, VEGF secreted by stem cells can participate in various levels of tissue regeneration, most importantly, stimulate bone regeneration (Tomoyasu et al., 2007). A few in vivo human studies (Carboni et al., 2019;De Riu et al., 2019;Howlader et al., 2017)  -593 current literature regarding human stem cell based TMJ regeneration to assess the efficacy of stem cells for TMJ regeneration/repair in vitro and to explain the necessity for further research for better patient care. Studies involving experimental animal models and/or animal stem cells were excluded due to two recent reviews published on this topic (Helgeland et al., 2018;Liu et al., 2014).

| Research question
The PICOS question for this research was: Exclusions: Any in vivo study, study using animal models, animal stem cells, or not related to TMJ regeneration was excluded.

| Data extraction and management
One author (S.G.) scanned the articles retrieved from the application of the search strategy, and acquired the full manuscript if the study met the inclusion criteria or when a definite decision could not be made regarding inclusion or exclusion based on the title and abstract only.
The reason for exclusion of the studies was also recorded. The full-text articles were analyzed for inclusion/exclusion, and relevant data was extracted by the same review author (S.G.) using a previously prepared data extraction form. The form from the reviewer was then reviewed by the senior author (R.E.). The form included for each study: the study design, characteristics of the sample (sample size, inclusion/exclusion criteria), interventions, control groups, and outcomes.

| Statistical analyses
Results of in vitro studies are presented in tabular form. Estimated risk ratios were calculated as the relative expression in the stem cells group divided by the relative expression in the controls for those studies reporting ALP activity, GAG content, Col-I, relative Col-I, Col-II, Col-X, SOX9, RUNX2, LPL, and Aggrecan mRNA expression in the stem cell-based interventions and the control groups.

| Results of the search
Through the preliminary search strategy by database on March 3, 2020, 303 references were found, and further 22 more records were discovered through other sources like searching references of included studies and reviews. After duplicate elimination, 244 references were analyzed individually by review authors (S.G. and R.E.), and based on the abstracts and titles, 186 articles were excluded and 40 articles were included. Of those 186 studies, reasons for exclusion were as follows: reviews/editorials (n = 64), animal studies (n = 25), duplicates (n = 5), not TMJ related (n = 45), no stem cells intervention (n = 39), not TMJ regeneration (n = 8). The full texts of these 40 manuscripts were analyzed for inclusion individually by the same authors, and nine manuscripts were found relevant for inclusion (in vitro). Reasons for exclusions were as follows: the authors used animal cells (n = 18), or it was an animal model study (n = 1), or not TMJ related (n = 7), or not TMJ regeneration (n = 2), or in vivo study (Carboni et al., 2019;De Riu et al., 2019;Howlader et al., 2017) (n = 3). PRISMA flowchart shows a summary of our results ( Figure 2).

| Study characteristics
In total, nine in vitro studies were identified using HSCs to regenerate TMJ structures with a sample size varying from 3 to 19. The features of included in vitro studies are summarized in Table 2. Brady et al. (2011) and Legemate et al. (2016) used bone marrow derived MSCs in their research. Other stem cells that were studied include human umbilical cord matrix stem cells (Bailey et al., 2007), synovial fluid-derived stem cells (Koyama et al., 2011;Liu et al., 2017;Sun et al., 2014;Yao et al., 2018), dental pulp stem cells (Bousnaki et al., 2018), and periodontal ligament-derived MSCs .

| Summary of results reported on in vitro studies
Results reported on included in vitro studies for ALP activity, Glycosaminoglycan (GAG) content, collagen I (Col-I), relative Col-I, Col-II, relative Col-X, SOX9, RUNX2, LPL, and Aggrecan mRNA expression in the stem cell-based intervention groups and control EMBASE (searched up to March 3, 2020) #1: "temporomandibular joint"/exp OR "temporomandibular joint" OR "temporomandibular joint disorder"/exp OR "temporomandibular joint disorder" #2: "stem cell" OR "stem cell transplantation" OR "mesenchymal stem cell" OR "bone marrow cell" OR "hematopoietic stem cell" #3: #1 and #2 #4: #3 AND "human"/de AND ("article"/it OR "article in press"/it OR "conference paper"/it OR "review"/it) #5: #4 AND ("case report"/de OR "clinical article"/de OR "clinical trial topic"/de OR: "comparative study"/de OR "evidence based practice"/de OR "human"/de OR "in vivo study"/de OR "major clinical study"/de OR "practice guideline"/de OR "systematic review"/de) groups are reported in Table 4. As compared to controls, ALP activity was significantly increased in stem cells cultured in chondrogenic or osteogenic medium (Koyama et al., 2011;Sun et al., 2014). Significant increases in GAG content were observed by two studies (Legemate et al., 2016;Zhang et al., 2014) but not Bailey et al. (2007), who found no significant difference of GAG content of HUCM cultured in chondrogenic medium and control medium. Col-I mRNA levels were increased in dental pulp stem cells in a chondrogenic medium as compared to an expansion medium (Bousnaki et al., 2018). Human MSC cultured with growth factors showed increased Col-I relative mRNA in the outer band of scaffolds, while no difference was noted in inner bands with or without growth factors (Legemate et al., 2016).
Periodontal ligament-derived MSC-conditioned medium increased Col-I relative mRNA level of chondrocytes .
Relative Col-II mRNA expression and Col-X mRNA expression were increased in both studies that tested it (Bousnaki et al., 2018;Koyama et al., 2011;Zhang et al., 2014). Relative SOX9 mRNA expression was increased in the studies by Bousnaki et al. (2018) and Zhang et al. (2014). Relative RUNX2 mRNA expression was increased in studies by Liu et al. (2014) and Sun et al . (2014). Brady et al. (2011) also examined the expression of SOX9 and RUNX, without the control group. Finally, relative LPL expression was increased in both studies that tested it (W. Liu et al., 2017;Sun et al., 2014).

| DISCUSSION
Large joints' regeneration is challenging due to the limitations of tissue engineering techniques and the complex anatomy of large joints. However, TMJ might be the first to benefit from current advances of tissue regeneration due to its small size (Brady et al., 2011).
Although TMJ regeneration is still at its early stage, and a few pioneer studies have used stem cells to treat TMD patients (Carboni et al., 2019;De Riu et al., 2019;Howlader et al., 2017), most TMJ regeneration studies were conducted in vitro. Bailey et al. (2007) compared polyglycolic acid (PGA) scaffolds seeded with human umbilical cord-derived mesenchymal-like stem cells (HUCM) and PGA scaffolds seeded with TMJ condylar chondrocytes. Authors found that after 4 weeks of culture, HUCM scaffolds contain more collagen I and GAG protein content than chondrocytes scaffolds (Bailey et al., 2007). Currently, there is no consensus regarding markers for SFMSCs or tissue of origin of these cells (Sun et al., 2014;Yao et al., 2018).

| Synovial fluid-derived cells
According to few studies, intra-articular bleeding and early stages of osteoarthritis can increase synovial fluid MSC (Koyama et al., 2011;Sun et al., 2014). Jones et al. (2004) proposed that disrupted joint structures may be the source of these cells. Sun et al. (2014) suggested that disrupted articular cartilage or bones may not be the origin of SFMSCs in TMJ because they obtained SFMSCs from patients with TMD without tissue damage. In addition, SFCs with almost identical characteristics as SFMSCs were primarily from intima of TMJ, which emphasize that SFMSCs could be from TMJ intima as well. SFCs could be a source for MSC-based TMJ regeneration (Sun et al., 2014;Yao et al., 2018). Liu et al. (2017) further evaluated how inflammation affects the multi-lineage differ- have no beneficial effect in some arthritis models, but, rather, aggravated arthritis symptoms (Djouad et al., 2005).

| Dental pulp stem cells
In contrast to the above papers that studied multi-lineage differentiation potential of various stem cells, Bousnaki et al. (2018) and Legemate et al. (2016) Legemate et al. (2016) developed a 3-dimensional (3D)-printed scaffold to replicate anisotropic collagen orientation of fibrocartilaginous matric distribution of human TMJ disc. The authors aimed to reproduce the region-dependent anisotropic tensile properties of the human TMJ disc with this specific design to restore function better. in osteogenic and chondrogenic media were seeded in the biphasiccompressed gel. The construct was cultured for 7 days before osteo-and chondro-differentiation were analyzed (Brady et al., 2011). Authors found that a week after culture, distinct bonelike and cartilage-like zones were identified in the compressed biphasic matrix, as demonstrated by von Kossa staining, Alcian blue staining, and expression of ALP, BSP, RUNX2, aggrecan, and SOX-9.

| CONCLUSIONS
In summary, in vitro studies utilized several different types of stem cells under different conditions. Increased osteogenesis and/or chondrogenesis were noted with stem cell interventions compared to 602 -GONG ET AL.
control groups on ALP activity, Col-I, Col-II, Col-X, RUNX2, LPL, and Aggrecan mRNA expression. This review emphasizes the potential of stem cell therapies in the regeneration of TMJ-related structures.
However, due to the heterogeneity of stem cells used, experimental conditions, outcome assessments and the limited number of studies available, we could not conclude that one type of human stem cells or a culturing condition is better than others. The field of TMJ regeneration using stem cell therapy is still at its infancy and further RCTs are required in order to evaluate the efficacy and safety of these therapies in humans.