Basic Research on Tendon Repair: Strategies, Evaluation, and Development

Tendon is a fibro-elastic structure that links muscle and bone. Tendon injury can be divided into two types, chronic and acute. Each type of injury or degeneration can cause substantial pain and the loss of tendon function. The natural healing process of tendon injury is complex. According to the anatomical position of tendon tissue, the clinical results are different. The wound healing process includes three overlapping stages: wound healing, proliferation and tissue remodeling. Besides, the healing tendon also faces a high re-tear rate. Faced with the above difficulties, management of tendon injuries remains a clinical problem and needs to be solved urgently. In recent years, there are many new directions and advances in tendon healing. This review introduces tendon injury and sums up the development of tendon healing in recent years, including gene therapy, stem cell therapy, Platelet-rich plasma (PRP) therapy, growth factor and drug therapy and tissue engineering. Although most of these therapies have not yet developed to mature clinical application stage, with the repeated verification by researchers and continuous optimization of curative effect, that day will not be too far away.


INTRODUCTION
As an anatomical structure connecting muscle and bone, tendons are highly resistant to mechanical loads, and capable of transferring, distributing and regulating the forces exerted by muscles to the connected structures. In this way, tendons maintain posture or generate movement (1-3). Tendons, which connect muscles to bones, have high tear resistance and tensile strength and play a crucial role in the stable movement of bones. But in fact, tendons composed of cells and parallel arrays of collagen fibers are often injured and even ruptured (2). The total incidence rate of tendon or ligament injuries is about 1/1,000 per year. Up to 46% of musculoskeletal injuries are reported as tendon injuries, including tendinopathy (4). Due to the lack of sufficient cells and growth factors, tendon healing is slow and the quality is unsatisfactory (5, 6).
The arrangement of tendons is hierarchical. The fibers composed of triple helix type I collagen molecules form fibers, bundles and tendon units in turn. Type I collagen is the most abundant ingredient of the extracellular matrix (ECM) in all soft tissues, including tendons. The special framework of the tendon is counted on the specific parallel organization of type I collagen fibrils, instead of the expression of type I collagen. So far, little is known about the mechanism by which type I collagen fibers drive specific spatial structures in tendons. Growth factors, such as transforming growth factors (TGF)-β and fibroblast growth factors (FGFs), are beneficial to improve the expression of collagen and promote the synthesis of collagen in tendon tissue. The synthesis of type I collagen in tendons also involves mechanical forces: increased load leads to increased collagen content in the tendons and reduced load results in reduced collagen content (1).
Collagen fibers densely arranged in ECM are considered as the primary force transfer unit of the tendon (1). A variety of ECM molecules are involved in the formation of type I collagen, including collagen, elastin and glycoprotein (7). Human tendon tears at 8% strain and plastic deformation occurs at 4% strain (8).
Based on previous surveys, tendon injuries are considered as the most common musculoskeletal disorder for patients seeking medical treatment (9,10). For example, 1 of every ten people and 1 of every two runners are afflicted with Achilles tendinopathy before the age of 45 (11). Tendon injuries can be caused by sudden rupture orchronic process, which is widely known as tendinopathy (2, 3).
Acute mechanical load tends to result in the outbreak of tendon tears (2, 3). In contrast, the pathogenesis of tendinopathy or chronic tendon injury is still controversial and has different definitions, degenerative disease or failure of the healing process (12)(13)(14). Chronic tendon injury or tendinopathy refers to the clinical symptoms of affected tendons, including pain, focal tendon tenderness, decreased strength and reduced activity. There is no macroscopic tear at the time of tendinopathy in comparison with partial or complete tendon tears. The following histological features can characterize tendinopathy: disordered collagen fibers, increase of proteoglycan and glycosaminoglycan content, increase of non-collagenous ECM, cell proliferation and neovascularization (15)(16)(17)(18).
Both chronic and acute tendon injuries have many extrinsic and intrinsic factors (14). Common inherent risk factors for tendon diseases include sexual distinction, age, type 2 diabetes and obesity (14,19). The main extrinsic factor of tendon injury is abnormal tendon load related to physiological exercise and specific working environment. Tendinopathy is thought to be caused by repeated eccentric mechanical loading, while acute tendon injury usually occurs after a single overloading event (2, 3, 13,[19][20][21][22]. Tendon injuries could bring a high burden to patients leading to a significant loss in individual production capacity. Moreover, the quality of patients' life is impaired, and the whole society's healthcare system suffers huge losses (23).
The natural healing process of injured tendons is complex and varies in clinical outcomes according to the anatomical location of the tendon tissue. The wound healing process comprises three overlapping phases: the wound healing process, proliferative phase, and tissue remodeling phase (24). In the inflammatory stage, red blood cells infiltrate into the wounded area, accompanied by white blood cells (leukocytes), and platelets equipped with important growth factors. Tendon cells, also known as tenocytes, recruited to the wound site are stimulated to multiply, especially in the epitenon (24). Inflammatory cells secrete a variety of cytokines to promote tendon healing, stimulate cell migration and angiogenesis. These cytokines include insulin-like growth factor-1 (IGF-1), transforming growth factor-beta (TGF-β), platelet-derived growth factor (PDGF) and basic fibroblast growth factor (bFGF) (25)(26)(27). The second stage, known as the proliferative or repair stage, is characterized by a large amount of synthetic activity under the guidance of macrophages and tenocytes. Macrophages that alter from phagocytic to reparative release growth factors (16,28). The synthesis of collagen I begins to play a leading role in the third and final stage known as the remodeling phase. This phase starts 1-2 months after injury and spans more than 1 year. The repaired tissue is similar to a scar in appearance. The biomechanical properties of the repaired tissue can never restore entirely because of an increase in the water content and a decrease in the quantity and quality of collagen (16). As a result, rupture may occur later, and sometimes a reduction in load capacity is observed, which is due to the tendency of tendons to form adhesions (29,30). The healing process of injured tendons is often longer and the healing intensity is weak in the early stage owing to the lack of cells and the low activity of growth factors (5, 31). In addition, the mechanical properties of the healed tendon were only 70% of that of the pre-injured tendon (32,33).
The development of new treatments is imminent. In this review, we introduce various therapies' research progress and achievements in recent years from various aspects.

TREATMENT OF TENDON INJURIES
The first-line treatment options are different facing acute and chronic tendon injuries. The main purpose of chronic tendon injury treatment is to reduce pain, mainly through local or systemic anti-inflammatory drugs, while treatment of acute tendon injuries aims to repair broken tendons with surgical techniques (3, 34,35).
In terms of tendon treatment, the management of minor injuries may be relatively straightforward and a combination of moderate rest and/or medical intervention is usually enough. Surgery is considered the last resort for tendon disease. The type and location of the injury determine the outcome of reconstruction surgery (3). There are many suture techniques including four-strand cross-lock repair, U-shaped repair, fourstrand Kessler type repair and etc. The four-strand cross-lock and U-shaped repair methods may be better (36). Even though, more than 40% of surgical patients still have complications and limited functionality after tendon surgery (37)(38)(39)(40).
Faced with such situation, it is urgently needed to find out advanced therapies. Several methods have been investigated to enhance the healing process including drugs, growth factors, gene therapy, stem cell therapy, platelet-rich plasma (PRP) therapy, and tissue engineering.

Drugs
In tendon tears, the ruptured tendons generally do not require medical treatment. But it seems to be about to change. In 2012, Bryan et al. found that within 6 weeks, compared with conventional sutures, the tendon repaired with Butyric Acid (BA)-impregnant suture showed better biomechanical properties (41). Although the U.S. Food and drug administration has not yet approved clinical use, BA is a natural carboxylic acid that has been proved to enhance angiogenesis in appropriate concentrations (42). In the early biological healing of tendon repair, the BA-impregnant suture can improve the biomechanical strength of repaired tendon, so as to play a protective role in early rehabilitation. Curcumin contributes to complete tendon healing not only histologically but also biomechanically. That is the reason why it may be an additional drug for the surgical repair of tendon injury (43).
In chronic tendinopathy, the main purpose of drug therapy is to relieve pain, such as NSAIDs (Non-steroidal Antiinflammatory Drugs), but it does not alter long-term outcomes. However, as one of the drugs, glucocorticoid injection has been proved to have significant long-term damage to tendon tissue and cells. A 2014 study showed that the local application of glucocorticoid had obvious adverse effects on tendon cells in vitro, including decreased cell viability, cell multiplication as well as collagen synthesis (44).
Besides, there are also some potential discoveries in recent years. A study in 2018 showed that tendons could acutely alter glucose metabolism, increasing glycolysis and lactic acid synthesis to deal with injury. This study also demonstrated that lactic acid synthesis inhibitors could promote the recovery of biomechanical properties of injured tendons (45). In 2017, Tack et al. demonstrated the role of direct (synthesis rate modification) and indirect (antioxidant) mechanisms in tendon healing in animal models by studying the effects of amino acids and vitamins, effects of amino acids and vitamins (46).

Growth Factors
As mentioned above, growth factors can be used independently or be involved in other kinds of therapies (47). These growth factors control tendon healing with the help of many regulators (2, 48). So far, there is a lot of knowledge about growth factors, such as TGF-β, bFGF, VEGF (vascular endothelial growth factor), and PDGF, through various studies both in vivo and in vitro (49).
There are three major isoforms of TGF-β, which is involved in a variety of cellular pathways (49). It is well-known that TGF-β is engaged in a series of reactions, including initial inflammatory response, collagen synthesis, neovascularization and fibrosis/excessive scar formation (49). TGF-β1 is one of three isoforms expressed by tenocytes, inflammatory cells and infiltrating fibroblasts (50,51). According to the present understanding, TGF-β1 is closely related to the occurrence and development of excessive scar formation. It is worth mentioning that after flexor tendon injury, when TGF-β1 signaling pathway is blocked by antibody or miRNA, the range of finger movement is improved, but the mechanical strength of tendon is downregulated (50,52,53). Type 2 and type 3 of transforming growth factor are vital for tendon formation and have the potential to induce tendon progenitor cells (54,55).
Basic fibroblast growth factor (bFGF), also known as FGF2, is a member of heparin-binding growth factor family. As a single chain polypeptide, bFGF promotes a variety of mitotic and angiogenic activities (5, 56). FGF2 is related to inflammation, cell proliferation, angiogenesis and collagen synthesis during tendon healing (57)(58)(59)(60)(61)(62). The role of exogenous FGF in tendon injury remains controversial. Ectopic FGF2 is considered to be able to promote cell proliferation and angiogenesis during tendon repair, but the improvement of mechanical strength is still unclear (62,63). However, some studies indicating that exogenous administration of FGF2 through a fibrin-heparin-based delivery system cannot improve the mechanical or functional properties of flexor tendon injury in dogs (60).
The VEGF family comprises several isoforms that linked to three tyrosine kinase receptors and their isoforms determine the bioavailability for each receptor (64). During the process of tendon development, VEGF levels ascend. The presence of VEGF in human fetal tendon is considered to be the cause of the differentiation of vascular and non-vascular regions in the tendon (65). Then VEGF levels descended to a steady state in the adult Achilles tendon (66). The lowest increase level of VEGF in adults means a long-term overuse tendon injury (67). It has been confirmed for a long time that the levels of VEGF increase early during tendon healing. VEGF promotes neovascularization by stimulating matrix metalloproteinases, which may degrade connective tissue and enhance angiogenesis (66,(68)(69)(70). Ectopic VEGF delivery can improve the strength of tension injured Achilles tendon. However, VEGF did not significantly up-regulate the expression of the collagen gene. Therefore, it is not necessarily the most critical factor in the synthesis of collagen during intrasynovial tendon healing, it just plays a vital role in angiogenesis during the repair and regeneration of injured tendons.
As a 30-kDa dimer, PDGF contributes to the migration and proliferation of fibroblasts, tenocytes and mesenchymal stem cells related to tissue homeostasis (71). PDGF signaling pathway may be essential to tendon homeostasis. It has been proved that downregulation of PDGF signal inhibited the normal growth response of tendon tissue to mechanical stimulation in adult mice (72). The role of PDGF in tendon development is not well-understood. In many animal and tendon models, exogenous delivery of PDGF contributes to the morphological and biomechanical properties, indicating that PDGF may help to enhance tendon healing (73)(74)(75)(76)(77).
In addition to the above four, many growth factors also play a role in tendon healing, such as CTGF, IGF and EGF (49). Moreover, multiple factors can be applied simultaneously to achieve better results. Recently, growth factors are more likely to be used in conjunction with gene therapy, cell therapy and tissue engineering than in isolation.

Gene Therapy
Despite recent medical advances, tendon tissue repair and regeneration remained a relatively unattainable challenge (78). In the past few decades, many surgeons and researchers have tried some methods to accelerate tendon healing and prevent tendon adhesion, such as biodegradable synthetic biomaterial barrier, inhibitors and gene therapy (79)(80)(81)(82)(83). Among these therapies, gene therapy is considered as a promising method (84).
Gene therapy is the treatment of disease by introducing a foreign nucleic acid (such as DNA or RNA) into a specific cell/tissue. Considering their negative charge and considerable size, these molecules require gene vectors to mediate their transfer (85).
The basic principle of gene therapy in tendon healing is that when a tendon is damaged, genes critical for collagen production cannot be activated in time. In addition, genes that cause excessive scarring during tendon healing may be overexpressed after injury, leading to an unbalanced tendon healing process. For more than a decade, the approach behind gene therapy has been to correct this balance to achieve early healing strength with the lowest possible adhesion around the tendon after direct end-to-end surgical repair (86).
Founded on the concept of offering therapeutic gene sequences, gene transfer procedures can be used to promote tendon healing, because these gene sequences can persistently increase the healing response and restore the tendon function before injury as completely as possible. Due to a deeper understanding of tendon biology, physiology, pathophysiology and tendon repair mechanism, there has been researches for the advantages of gene therapy (87).
The advantage of gene therapy is that the continuous expression of genes can increase the endogenous gene products in target tissues, such as signal molecules and transcription factors (3). Compared with stem cell therapy, gene therapy has a lower incidence of immune responses (88). Gene therapy can produce continuous local production and secretion of proteins. This distinguishing feature allows the restriction of protein delivery associated with short half-lives to be bypassed and makes it possible to regulate the timing of bioactivity cues in tissue sites. Therefore, gene therapy is a promising approach for tendon regeneration (78).
As is shown in Figure 1, there are three essential factors in gene therapy: target gene, gene carrier and target cell. Genes usually have a large molecular weight and a high negative charge density, so cell permeability is low, and gene therapy is limited (89,90). In addition, the safe and effective delivery of nucleic acids into tendon tissues still needs to be considered. Basing on the consideration of biocompatibility and further improving the efficiency of gene therapy, a variety of vectors have been developed, including nanospheres, viral vectors and other delivery systems (91)(92)(93)(94).
The selection of gene delivery vectors for tendon tissue repair and regeneration is an important parameter. Gene delivery vector dramatically affects the efficacy of gene therapy (95,96). Many vectors have been tested for gene therapy, both viral and nonviral. Generally speaking, viral vectors are efficient in gene transduction, but there are safety hazards that are caused by virus insertion mutation and quality control (85,97,98). Although non-viral vectors are relatively safe, there is still an urgent requirement for them to improve transfection efficiency and expression of the transgene to be applied in clinical trials (99).
Many viral vectors such as adenovirus, adeno-associated virus (AAV), and lentivirus have been applied in animals to treat tendon injuries (100-102). As a typical viral vector, Adenovirus (Ad) can transduce many dividing and non-dividing cells, with solid but transient gene expression. Ad carrier is still the most familiar carrier in clinical trials (103,104). In 2003, Ad became the basis for the first approved commercial gene therapy drug in China. Shenzhen GenTech has developed a modified Ad carrier Geneticine that encodes the p53 tumor suppressor gene and has been approved to treat head and neck cancer. Because of the lack of available information on clinical outcomes, the outcomes associated with this therapy remain controversial (105,106). Although Ad is typical in clinical trials, it has several fundamental limitations. First, proteins in the viral capsid activate a strong immune response, and second, cell transduction is dependent on the presence of the CAR (coxsackie and adenovirus receptor) on the cell membrane, which leads to preferential hepatocyte transduction after intravenous injection. Finally, biological activity in the body is limited over time (107,108).
Due to the side effects of viral vectors, non-viral vectors have been paid more and more attention because of their low immunogenicity, safety, flexibility of chemical design, stability and high gene capacity. In addition, non-viral vectors are easily created and chemically altered extensively. Finally and most importantly, the non-viral vector system is not limited by the size of the introduced gene molecule (85). Plasmid DNA (pDNA) is a circular, double-stranded DNA that replicates independently of chromosomal DNA. Based on the specific application, pDNA can be easily designed to express one or more genes. Compared with viral delivery systems, pDNA offers several advantages, including the ability of plasmids to hold significant genes (109). PLGA poly(lactic-co-glycolic acid) nanospheres have been shown to deliver plasmids efficiently to cultured muscle cells and injured tendons. Nanospheres ensured the high level of transgene expression in tendons for at least 6 weeks, and almost no tissue reaction was observed in tendons. It may be a promising non-viral vector for gene therapy (110). Nanospheres can achieve efficient gene-targeted delivery and protect gene sequences from degradation owing to its subcellular structure, which makes it easy to penetrate into the cell through endocytosis, enter the cytoplasmic chamber, and release the contained substances for long, so as to break through the absorption barrier of cell membrane (111-113). Cationic lipids and cationic polymers can also be used as non-viral vectors to carry DNA to form lipoplex and polyplex (114,115). Many natural and synthetic polymers have been successfully used as delivery vectors. Since naturally occurring polymers are isolated from plant, animal or human tissues, vehicles usually developed from these polymers have the advantage of reproducing the critical structure and/or biochemical characteristics of ECM due to their natural derivation. Therefore, naturally occurring polymers have the advantages of low immunogenicity, relatively high abundance and easy processing. Fibrin and collagen are examples of such polymers (116,117). Hydrogels and scaffolds can be formed in a mild manner to encapsulate more sensitive genetic vectors and cells. Synthetic polymers have the unique advantages of repeatability and controllable production, allowing precise operation (117)(118)(119). In addition, nuclear targeting of non-viral vectors can improve transfection efficiency by the application of SV40DNA-targeting sequences (DTS). DTS stimulates nuclear entry by binding nuclear localization signals to transcription factors, thereby encapsulating plasmid DNA and guiding nuclear entry (85). Each vector has its inherent advantages and disadvantages, but the combination of biomaterial carriers can help to make up for the defects. Some studies have used surface coating and/or hydrogel encapsulation to mask Ad from host immune responses in order to compensate for the deficiency of viral vector Ad (78). Some researchers have tried to combine biomaterials such as fibrin and collagen with physical methods to treat musculoskeletal injury and have shown the potential of this approach (120)(121)(122)(123)(124)(125).
Integration of biomaterial systems that appear in gene delivery strategies is central to most MS tissue engineering strategies and plays a role in the delivery system of both viral and nonviral vectors (78,126). These biomaterials offer a combination of mechanical structure, cell support and biochemical signal control. In order to achieve this goal, biomaterial structures have been designed to be similar to natural extracellular matrix in structures (78).
As is shown in Table 1, gene delivery strategies have concentrated on sequences that are significant to tendon healing, such as growth factors, matrix molecules, transcription factors, anti-inflammatory molecules, and signaling molecules (145). It has been proved that bFGF and vascular endothelial growth factor A(VEGFA) are the most effective stimulants (146,147). The transfer of bFGF and VEGFA genes can correct the deficiency of tendon intrinsic healing ability, and promote tendon healing, respectively (62,94). In addition, it is worth mentioning that fibromodulin may be a good substitute for tendon healing growth factor (130).
In conclusion, gene therapy is an attractive strategy to promote tendon healing. The study of adaptive preclinical animal model has proved the feasibility of using this gene-based therapy in the treatment of tendon injury, which provides a reasonable hope for its transformation to patients as soon as possible.

Stem Cell Therapy
In the management of chronic musculoskeletal diseases, regenerative medicine is a new modality which attracts increasing interest (148,149). Stem cells have been widely used in the treatment of musculoskeletal diseases (150,151). Current evidence suggests that stem cell therapy is highly effective on musculoskeletal disorders (152,153). Stem cells are considered as cells that have the ability to divide and self-renew over a long period of time and capable of differentiating in all cell lines (154,155). Stem cells have been assumed to promote regeneration during tendon healing (156,157). The application of stem cell helps regulate inflammation, organize ECM regeneration, and promote tissue regeneration over scarring (158,159). Identification of cell origin and characterization is necessary to achieve more effective repair or regeneration (160).
Stem cell-based therapies have attracted massive attention in the regeneration of defective tissues or organs. At present, stem cell-based regenerative treatment mainly includes isolation and screening, in vitro culture and amplification, and transplantation with or without directional differentiation, which is shown in Figure 2 (161-164). The success of gene therapy with relevance to applications for tendon healing can be improved with the role of the antiaging gene Sirtuin 1 that is critical for healing and regeneration. The activation of this genes is critical to the success of stem cell therapy (165)(166)(167)(168). However cell transplantation has encountered many obstacles in therapeutic translation, including immuno-rejection, pathogen transmission, potential tumorigenesis, problems related to packaging, storage and transportation, and tissues in clinical application and regulatory approval (162,(169)(170)(171)(172)(173).
One of the fundamental problems limiting cell-based therapy is how to deliver cells to their targets and how they are implanted (174). Stem cells are either injected throughout the body into the circulation or injected locally at the injury site.

Systemic administration of stem cells is limited by inefficient
Frontiers in Medicine | www.frontiersin.org  targeting of the injured area. At the same time, local injection may additionally do harm to the injured tissue. Because of implantation problems, survival rate of stem cells after local injection is low (174). With simple drug administration, tendon ruptures provide an attractive opportunity to target the site needed for treatment. Granulation tissue (i.e., the early loose connective tissue formed between the ends of the broken tendon) is composed of new capillaries (175). These newly formed capillaries have special molecular frameworks on surfaces, targeting ligands binding in an organ-specific manner (176). Therefore, in the proliferation phase, a random peptide library is filtered by phage display in vivo on the ruptured Achilles tendon and patellar tendon to look for the systemic administration peptide that is able to home to the injury site (176). At present, the best vascular peptide can home to damage tendons up to 200 times, and has been successfully used to deliver therapeutic agents that promote tissue regeneration (176)(177)(178)(179). After these vascular targeting peptides were applied, the delivery and implantation of MSC to infarcted myocardium increased by four times (179). Therefore, a sizeable microvascular network resulted from a strong angiogenesis reaction at the rupture sites of tendons provides a large number of molecular targets for systemic administration of ligands, and transports the goods to the injured tendon (180). Neovascularization is a sign of tendinopathy Since these vascular targeting peptides can home to ruptured tendons, the platform for transporting stem cells to tendon injuries by using vascular homing peptide technology is basically already in place (174,179).
There are many kinds of stem cells with different characteristics. Table 2 lists some of the categories. Some differences between stem cell populations must be noted.
MSCs have recently been shown to possess more significant and possibly more crucial therapeutic functions in the injury response, such as immune modulation and nutritional activity. Therefore, they are defined as "drugstores" (181). In fact, they can enter the site of inflammation or tissue damage and begin to secrete immunomodulatory and nutritional substances, such as cytokines and growth factors, to rebuild physiological homeostasis to cope with the environment (181). Therefore, whether as a direct participant in the process or the "Drugstore" of bioactive molecules, BMSCs may enhance tissue repair and regeneration, thus restoring normal intraarticular homeostasis. These characteristics, together with the relatively easy isolation and amplification process, make MSCs have great potential in many clinical applications in recent years.
In 2003, 6.4 × 10 5 of autologous bone marrow mesenchymal stem cells (BMSCs) were transplanted into SDFT of horses injured by strain for the first time (160). Isolated from bone marrow, BMSCs are applied and studied most commonly of all adult stem cells. BMSCs have been proven effective in regenerating different tissues, including tendons, either on their own or in combination with scaffolds (182,183).
Adipose tissue-derived mesenchymal stem cells (ASCs) are the most abundant source of mesenchymal stem cells. Additionally, the higher number of mesenchymal stem cells extracted from the same amount of fat compared with bone marrow is another advantage of using ASCs. Because ASCs can produce and secrete components of ECM and cytokines, ASCs, as the most potential cells for cell therapy, have attracted much attention (184). In addition, compared with other mesenchymal stem cells, ASCs have the highest expression level of tendon ECM component and can be a hopeful cell source for the treatment of tendinopathy (185).
A recent study has shown that Tendon-derived MSCs (TDSCs) also known as tendon stem/progenitor cells (TSPCs) is an ideal cell type. Compared with other mesenchymal stem cells, TDSCs can display tendon-like phenotype and express the most tendon-related markers (186,187). TSPCs are pluripotent adult stem cells involved in tendon healing (188,189). It has been reported that TSPCs have clonality, differentiative potential and express specific surface markers of stemness (190). In addition, these TSPCs also express tenogenic markers making them a unique stem cell population. The number of TSPCs decreases with age, which may be the reason for the high prevalence of tendinopathies in the elderly (190,191). In vitro and animal experiments, TSPCs are capable of differentiating in tenocytes (136,192). TSPCs account for about 4% of the total number of tendon cells, (193). However, their use may be limited due to donor site lesions, an insufficient number of long-term cultured cells, and phenotypic drift during in vitro amplification, expansion is required before injection to in vitro (194).
The clinical potential of embryonic-like stem cell (ESC) to treat tendon injuries has been revealed, and the non-tumorigenic bias of these cells remains to be studied in more extended followup studies. Still, they require destruction of embryos to be isolated (195)(196)(197). Moreover, it has been demonstrated that they can differentiate into tendon cells (4).
Stem cells can be utilized to enhance the repair and regeneration of injured tendons. Different techniques can be applied to induce tendon differentiation and a gradual process can be avoidable to accidental differentiation. As shown in Table 2, there are many different strategies for tendon differentiation, such as hypoxia, biomaterials, growth factors, physical stimuli and co-culture system (198).
Despite many expectations, stem cell therapy has also been questioned. As is shown in a review in 2017, stem cell therapy still needs a lot of practice research before clinical application. Furthermore, there are also concerns about the long-term safety of stem cell therapy (199).

Platelet-Rich Plasma (PRP)
The application of PRP is based on changing the molecular environment by providing a super-physiological concentration of platelets (and optional white blood cells) at the lesion/pathological site to simulate the initial stage of healing. The core of PRP research is composed of the molecular complexity of PRPs and their interactions with different forms of host tissue. PRP can manipulate and boost healing if we integrate the molecular knowledge of PRP and healing mechanism that PRP involved in,. Another layer of complexity, which is very challenging for researchers, describes the state of the host tissue: the stage of the disease and the mechanism of pathological tendon changes with which PRP is obliged to interact (200). There is such a probability that PRP composed of high concentrations of platelets, growth factors and cytokines, mainly by changing the molecular microenvironment, may contribute to healing.
One of the early effects of PRP on rats is intermittent inflammation. The activation of pro-inflammatory TNF alpha and NFκB pathways following PRP exposure, along with the expression of genes related to cell proliferation and tendon collagen remodeling have been observed (201). In chronic tendon diseases, the induction of acute inflammation may be a crucial factor in triggering the subsequent regeneration reaction (145). The inflammatory process seems to be under control of PRP from one of the key elements, hepatocyte growth factor (HGF). It is recognized that PRP has anti-inflammatory effect on human chondrocytes by inhibiting NFκB activation (202).
In the past two decades, PRPs have emerged in different medical fields as biotherapy for repairing or regenerating damaged or non-functional tissues. From the beginning, they have been used by highly expected sports medicine physicians and orthopedic surgeons to accelerate tissue healing (203). Anecdotal results from media coverage in elite athletes and other celebrities have increased the appeal of the public (204).
However, scientific understanding of how PRP works and clinical evidence to specific indications are not available yet and the process is time-consuming (200). A recent systematic review has underlined the controversial results of PRP applied in different pathologies. The authors assert that, in terms of current evidence, patellar and lateral elbow tendinopathy are benefited from PRP treatment, while in the treatment of Achilles tendon and rotator cuff, it seems that PRP do not help the conservative treatment or surgery (205). On the other hand, many studies have shown that PRP is conductive to Achilles tendon and rotator cuff disease (206,207). Recently, a trial in the British Medical Journal (BMJ) disappoints PRP-related researchers reported that platelet-rich plasma did not provide any benefit after acute Achilles tendon rupture. PRP as an autologous whole-blood product composed of high concentrations of platelets, growth factors and cytokines is extensively used in sports medicine. The study included 230 adults from 19 British hospitals who suffered from acute Achilles tendon rupture in the previous 12 days without surgery. Patients were randomly given platelet-rich plasma injection or placebo and received standard rehabilitation care. At 24 weeks, there were no differences in tendon function, patient-reported function, pain, goal achievement, or quality of life between the two groups (208).
In summary, the future of PRP in tendon pathology remains open. At least, its role in the early stage of tendon healing has been proved by basic science studies. Still, a lot of preparation, for example, the preclinical premise and specific standardized clinical protocols is needed before it can be used in the clinical stage.

Tissue Engineering
In 1997, Charles Vacanti Laboratory of the University of Massachusetts Medical Center created a mouse with the human ear on its back. The idea of manufacturing renewable tissue from in vitro has ushered in a new epoch in biomedical science, which is known as tissue engineering. As is first 2020 Scaffold (144) introduced in 1987, the concept of tissue engineering was defined to apply the principles and methods of engineering and life sciences to basically understand the structure-function relationship of normal and pathological tissues in mammals, and develop biological substitutes that can restore, keep or improve function (209). Tissue engineering is a multidisciplinary approach aimed at inducing tissue repair, replacement, or regeneration. Tissue engineering involves using a combination of cells, scaffolds, and bioactive molecules to produce functional tissue (210).
The torn tissue will be replaced with tendons from other parts of the body in the early stage, tendons from another person, tendons from different species, or artificial tendons. As shown in Table 3, these grafts include autograft, allograft, xenograft and synthetic grafts (211). The main drawback of autografts is donor site complications. Allografts and xenografts are limited by the availability of donor tissues and the potential risk of immune rejection and pathogen transmission (3). A clinical trial has shown that the application of xenograft still needs to reduce its high infection rate (212). Early synthetic grafts have some disadvantages, such as early re-tear, decrease of mechanical strength with time, insufficient tissue that grows inward and graft debris deposition (213). Although the new generation of grafts shows better performance, the application of these artificial grafts remains controversial because clinical reports have not yet reached a unified and precise conclusion (214).
Because of the problems above, there has been increasing interest in the preparation of tissue-engineered tendons in the past decade. The concept of tissue engineering is to create a safe and effective substitute for tissue injured in the biological environment (215). Tissue-engineered grafts consist of two or all of the three main components: cells, biomaterials/scaffolds, and biomolecules (216). The fabrication of an engineered scaffold is a method of tissue-engineering grafts, which provides a physical environment to regulate the repair and regeneration of damaged tissue. Figure 3 shows the preparation and use of the scaffold. The main function of scaffolds is to offer a physical environment to regulate the healing and regeneration of injured tissues (217).
Stem cells are usually implanted into engineered scaffolds to promote tissue repair and regeneration. Mesenchymal stem cells are natural cells of tendons, so they are widely used. Mesenchymal stem cells are easy to obtain from various tissue sources. Besides, they have the characteristics of antiinflammatory, reducing tissue inflammation, secreting a large number of trophic factors and promote tissue regeneration. That may explain why they are widely used in tendon and ligament tissue engineering (218)(219)(220). Cell secretory products such as exosomes, secretory bodies and platelet-rich plasma (PRP) can also be implanted into engineering tissues to improve the therapeutic effect (189,221,222).

CONCLUSION
Tendon is a complex tissue with unique structure, function and mechanical properties. Damage to this vital connective tissue can lead to severe pain and disability. For this reason, improving tendon healing is an urgent need. There are many research topics in the field of tendon healing and repair. At present, gene therapy, stem cell therapy and others have made some achievements, although they are not up to the standard of clinical application. Although many other studies have focused on tendon healing, there are still many problems to be clarified about this complex process.
A recent review found a strong link between cancer and wound healing (223). Maybe we can further understand of tendon healing and thus find ways to accelerate tendon healing from some cancer studies. For example, some signs show that the temporary closure of the apoptosis regulation pathway can accelerate regular repair (223).

AUTHOR CONTRIBUTIONS
ZL wrote the initial manuscript and revised the manuscript. QY revised the manuscript. YZ provided initial idea of this paper and revised the manuscript. All authors contributed to the article and approved the submitted version.