Tendon and Ligament Injuries: The Evolving Role of Stem Cells and Tissue Engineering

Aims: Musculoskeletal injuries are a common injury associated with a reduction in quality of life, increased morbidity and social and financial implications. Although surgical reconstruction is a well established option, outcomes are variable. There is a growing body of interest in the potential of mesenchymal stem cells (MSCs) in the management of tendon and ligament injuries. This review aims to summarise the information in the literature on the evolving role of these. Study design: Review Article.


INTRODUCTION
Ligament and tendon injuries form a significant proportion of musculoskeletal injuries. These types of injuries can occur at any age but are more commonly associated with sporting activities, in which case they are called 'overuse injuries' (Tenforde et al., 2011). The incidence of ankle sprains in the US is estimated at 2.15/1000 people. (Waterman BR et al., 2010). Over 200,000 Anterior Cruciate Ligament (ACL) reconstructions per year are carried out in the US (Shelton 2011), and 40,000 in-patient procedures were performed in the USA for rotator cuff repair in 2002 (Oh et al., 2007). In 2007, Achilles tendon injuries had an incidence of 11.3/100,000 per year in the UK (Clayton and Court-Brown, 2008).
Aside from reducing patient quality of life and increasing morbidity, treatment of these injuries are costly; the estimated cost of an ACL surgical repair and subsequent rehabilitation is between $17,000-$25,000 per injury (de Loes et al., 2000). Even after treatment, operative or conservative, the outcome is variable. This is partly due to the fact that the poorly vascularised ligaments and tendons have a limited capacity for healing and, even if healing has occurred, the biomechanical properties of the healed structures are usually inferior to normal tissue (Mammoto et al., 2010;Takasugi, 1978). Laboute et al. (2010) found a rerupture rate of 12.7% for ACL reconstructions performed with hamstring tendon autografts.
Mesenchymal Stem Cells (MSCs) are naturally occurring cells that have the ability to both self-replicate as well as differentiate into another cell type, thus acting like a raw ingredient that can be moulded into any potential tissue type (Rayanmarakkar et al., 2009). As such, the use of MSCs has been linked to the treatment of injuries to tendons and ligaments (Oragui et al., 2011;Siddiqui et al., in press). We review the current status of stem cell research related to tendon and ligament injuries.

METHODOLOGY
A thorough literature review was conducted and articles relating to the use of MSCs to treat tendon and ligament injuries were identified. The search primarily focused on the use of MSCs to stimulate and augment repair of the tissues, but also on the manipulation of the cells in order to promote increased osteointegration at the tendon-bone insertion sites. The emerging role of induced pluripotent stem cells (iPSCs) is also described.

Normal Structure of Tendons and Ligaments
Tendons and ligaments are two distinct tissue types that are sometimes erroneously used interchangeably to describe each other. Both structures have a similar composition. Specialised fibroblasts cells constitute approximately 20% of the tissue. The remaining 80% is extracellular matrix, which consists of 70% water and 30% collagen, ground substance, and elastin.
There are at least 16 types of collagen, but 80 -90 percent of the collagen in the body consists of types I, II, and III. These collagen molecules pack together to form long thin fibrils of similar structure. The various collagens and the structures they form all serve the same purpose, to help tissues withstand stretching. The collagen present in ligaments and tendons is predominantly type I. Type I collagen fibrils have enormous tensile strength so that they can be stretched without being broken. There is also a small amount (less than 10%) of type III collagen, but this is more often found in healing tissues, before most of it is converted to type I collagen. There is also a very small amount of type IV collagen present. The basic structural unit of collagen is a triple-stranded helical molecule packed together side by side (Lodish et al., 2000).
Tendons attach muscle to bones and so act to both transmit muscular forces but also as a store of elastic energy (Fukashiro et al., 1995;Oragui et al., 2011). They are dynamic structures that need to undergo changes in direction of their mechanical properties when crossing joints. In tendons these collagen bundles are usually arranged into fascicles, enveloped by layers of connective tissue: endotenon and then epitenon. Ground substance stabilises the collagen microfibrils, and consists of a variety of molecules including proteoglycans, glycoproteins, and plasma proteins. It has the ability to bind water, resulting in a gel-like substance with increased strength. A small amount of elastin is also produced by fibroblasts. It is highly cross-linked and organised multi-directionally, and can stretch and recoil back to its original shape. The biomechanical properties of tendons are directly related to the organization of the collagen molecules that aggregate to become a super-twisted cord. (Aperacida de Aro et al., 2011) Ligaments connect one bone to another and help stabilize joints and control their movements. During sporting activities, they are subjected to higher stress and strain rates as the joints are subjected to more forceful and exaggerated movements. Additionally, they help coordinate the complex movements required during sporting activities via a proprioceptive input into the nervous system (Frank, 2004). Therefore, they are subjected to multidirectional forces and so the collagen strands that form the ligament lie in variable directions, which concordantly give them an inherent weakness.

Structure and Reasons for Poor Healing
Other than direct trauma to the structural integrity of tendons or ligaments, injury can occur at the sites of bony insertion (osteotendinous junction, OTJ) and at myotendinous junction (MTJ). At the MTJ, tendinous collagen fibrils are inserted into deep recesses formed by myocyte processes, allowing the tension generated by intracellular contractile proteins of muscle fibres to be transmitted to the collagen fibrils. This complex architecture reduces the tensile stress exerted on the tendon during muscle contraction. However, the MTJ still remains the weakest point of the muscle-tendon unit. (Kvist et al., 1991) The OTJ is composed of four zones: a dense tendon zone, fibrocartilage, mineralized fibrocartilage, and bone (Benjamin and Ralphs, 1998). The specialized structure of the OTJ prevents collagen fibre bending, fraying, shearing and failure. (Evans et al., 1990) Oxygen consumption by tendons and ligaments is 7.5 times lower than skeletal muscles (Vailas et al., 1978). Given their low metabolic rate and well-developed anaerobic energy generation capacity, tendons are able to carry loads and maintain tension for long periods, whilst avoiding the risk of ischaemia and subsequent necrosis. The blood supply to both tendons and ligaments is relatively poor. They have, therefore, adapted to have a very low metabolic rate which also results in slow healing after injury (Williams, 1986).
Tendons receive their blood supply from three main sources: the intrinsic systems at the MTJ and OTJ, and from the extrinsic system via the paratenon or the synovial sheath 23. The ratio of blood supply from the intrinsic to extrinsic systems varies from tendon to tendon. For example, the central third of the rabbit Achilles tendon receives 35% of its blood supply from the extrinsic system. (Naito and Ogata, 1983) At the MTJ, perimyseal vessels from the muscle continue between the fascicles of the tendon. However, blood vessels originating from the muscle are unlikely to extend beyond the proximal third of the tendon. The blood supply from the OTJ is sparse, and limited to the insertion zone of the tendon, although vessels from the extrinsic system communicate with periosteal vessels at the OTJ. (Carr and Norris, 1989) Tendon vascularity is compromised at junctional zones and sites of torsion, friction or compression.
The specialised structure of tendons and ligaments thus make them ideally suited to their function, but these hypocellular, relatively avascular structures also result in a relatively poor ability to regenerate and heal after injury.
Due to the frequency of injuries of these structures, their repair has been the subject of extensive research. Due to the poor healing capacity already described, reconstructive surgery is a well recognised treatment option, using autologous, cadaveric, or artificial grafts. However, associated problems with these techniques include donor site morbidity, immunological and infection risks, foreign body reactions to artificial grafts (Hertel et al., 2005;Shelton et al., 2011). In addition, they are technically challenging operation to perform and have found to be highly surgeon dependant in nature. (Cheung et al., 2011)

Tissue Engineering as an Approach to Tendon and Ligament Augmentation and Replacement
When conservative treatment methods fail, more invasive options may be consulted. Surgical repair or replacement is a well established option (Barring et al., 2011). Recently however, the field of tissue engineering has emerged as an attractive treatment option to aid these processes by either delivering cells locally to induce a better quality repair or delivering chemicals locally to induce a better repair Al-Rashid and Khan, 2011).
The United States National Committee on Biomechanics (USNCB) has suggested six principles of tissue engineering, as reported by Butler et al. (2000): 1. Measurement of normal tissue stress and strain patterns in vivo 2. Measurement of normal tissue failure properties 3. Identify a subset of the properties to focus research on 4. Establish a minimum standard that these properties much achieve 5. Investigate which stimuli cells react to in vivo 6. Investigate the role of physical factors on cell stimulation in bioreactors during the production of such engineered tissues MSCs are a form of adult stem cell that are being implicated in the use of tissue repair (Rayanmarakkar et al., 2010). They are autologous in nature and hence carry no risk of immunological reactions within the host. They may be found in non-haematopoietic bone marrow stroma, as well as periosteum, fat, and skin and thus are not associated with the ethical considerations surrounding foetal or embryological sourcing (Mafi et al., 2011;Mohal et al., in press). Stem cells are a self-renewing, slow-cycling cell population that exhibit high clonogenity, low cellular proliferation and the ability to undergo multilineage differentiation (Thanabalasundaram et al., in press). They can be administered as local injections with scaffolds in or around the areas that require regeneration and have shown promising results in vitro. Recent studies also support the use of growth factors and other stimulants to further improve outcomes. (Hertel et al., 2005;Schwitzer et al., 2010). Omae et al. (2009) incised 20 dog infraspinatus tendons, seeded them with MSC and then bundled them into one composite. After 14 days the MSC viability was assessed by a fluorescent tracking marker. Histology showed that the seeded cells aligned between the collagen fibers of the tendon slices. Analysis by qRT-PCR showed higher tenomodulin and MMP13 expression and higher collagen type I expression in the composite.
In another study, cultured, autologous, MSCs were suspended in a collagen gel delivery vehicle. The composite was subsequently contracted onto a pretensioned suture. The resulting tissue prosthesis was then implanted into a 1-cm-long gap defect in the rabbit Achilles tendon. Identical procedures were performed on the contralateral tendon, but only the suture material was implanted. The tendon-implant constructs were evaluated 4, 8, and 12 weeks later by biomechanical and histological criteria. Significantly greater load-related structural and material properties were seen at all time intervals in the mesenchymal stem cell-treated tendons than in the contralateral, treated control repairs (p < 0.05), which contained suture alone with natural cell recruitment. The values were typically twice those for the control tissues at each time interval. Load-related material properties for the treated tissues also increased significantly over time (p < 0.05). The treated tissues had a significantly larger cross-sectional area (p < 0.05), and their collagen fibers appeared to be better aligned than those in the matched controls. (Young et al., 2005) Soon et al. (Soon et al., 2007) conducted a study to analyze the effect of coating allografts with MSCs on the quality and rate of osteointegration at the allograft tendon and bone interface, and the biomechanical properties of these enhanced anterior cruciate ligament (ACL) grafts compared with controls. Bilateral ACL reconstructions using Achilles tendon allografts were performed in 36 rabbits. On 1 limb, the graft was coated with autogenous MSCs in a fibrin glue carrier, while the contralateral limb served as a control with no MSCs. The reconstructions were assessed histologically and biomechanically at 2, 4, and 8 weeks. At 8 weeks, histologic analysis of the controls revealed the development of mature scar tissue resembling Sharpey fibers spanning the tendon-bone interface. In contrast, the MSCenhanced reconstructions showed a mature zone of fibrocartilage blending from bone to the allograft, strongly resembling a normal ACL insertion. On biomechanical testing, the MSCenhanced grafts had significantly higher load-to-failure rates than controls. However, the stiffness and Young's modulus were lower in the treatment group.
An in vivo study by Cao et al. (Cao et al., 2002) using flexor tendons from hens attempted to investigate the feasibility of autologous tenocytes to bridge tendon defects. The isolated tenocytes were expanded in vitro and mixed with unwoven polyglycolic acid fibers to form a cell-scaffold construct and cultured for 1 week before in vivo transplantation into a defect of 3 to 4 cm within the tendon, The defects were bridged either with a cell-scaffold construct in the experimental group (n= 20) or with scaffold material alone in the control group ( n= 20). Specimens were harvested at 8, 12, and 14 weeks postrepair for gross and histologic examination and for biomechanical analysis. In the experimental group, a cordlike tissue bridging the tendon defect was formed at 8 weeks postrepair. At 14 weeks, the engineered tendons resembled the natural tendons grossly in both color and texture. Histologic examination at 8 weeks showed that the neo-tendon contained abundant tenocytes and collagen bundles randomly arranged. At 12 weeks, tenocytes and collagen fibers became longitudinally aligned, with good interface healing to normal tendon. At 14 weeks, the engineered tendons displayed a typical tendon structure hardly distinguishable from that of normal tendons. Biomechanical analysis demonstrated increased breaking strength of the engineered tendons with time, which reached 83 percent of normal tendon strength at 14 weeks. In the control group, polyglycolic acid constructs were mostly degraded at 8 weeks and disappeared at 14 weeks. However, the breaking strength of the scaffold materials accounted for only 9 percent of normal tendon strength.
Juncosa-Melvin et al. (Juncosa-Melvin et al., 2006) introduced MSCs into a gel-sponge composite to examine the effect the cells have on repair, biomechanics and histology in patella tendons of rabbits. Autogenous tissue-engineered constructs were created by seeding MSCs from 15 adult rabbits at 0.1 x 10(6) cells/mL in 2.6 mg/mL of collagen gel in collagen sponges. Acellular constructs were created using the same concentration of collagen gel in matching collagen sponges. These cellular and acellular constructs were implanted in bilateral full-thickness, full-length defects in the central third of patellar tendons. At 12 weeks after surgery, repair tissues were assigned for biomechanical (n = 12 pairs) and histological (n = 3 pairs) analyses. Maximum force and maximum stress for the cellular repairs were about 60 and 50% of corresponding values for the normal central third of the PT, respectively. Likewise, linear stiffness and linear modulus for these cellular repairs averaged 75 and 30% of normal PT values, respectively. By contrast, the acellular repairs exhibited lower percentages of normal PT values for maximum force (40%), maximum stress (25%), linear stiffness (30%), and linear modulus (20%). The same author showed in a further study in 2007 that mechanical stimulation increases the stiffness of stem cell-collagen sponge constructs at 14 days in culture and subsequent rabbit patellar tendon repairs at 12 weeks postsurgery (Juncosa-Melvin et al., 2007). Awad et al. (2003) cultured MSCs and implanted the resulting tissue into rabbit patellar ligament defects, after labelling the MSCs with a fluorescent dye. A defect was made in the contralateral side as a control at the same time. After sacrifice at 12 and 26 weeks they found that the strength of the construct was greater than that of naturally healing tissue in the contralateral side. Compared to healthy tissue, the cultured MSCs had 23-27% strength, whereas the healed tissue managed only 15-19% strength. Original MSCs were still present, as evidenced by the fluorescent stained cells. However, they found ectopic bone formation on the cultured side, a result of poorly constrained cellular differentiation. Inducing in vitro differentiation of the MSCs into tendon-forming cells prior to introduction is thought to reduce the incidence of ectopic bone formation and tumour formation (Lui et al., 2011). They were able to further improve tissue properties by intermittent stimulation by applying increased strain for one second every five minutes for eight hours a day for twelve days, a findings reinforced by Juncosa-Melvin et al. in 2007.

The Additional Benefit of Growth Factors
The benefit of repair using MSCs can be further enhanced by using growth factors (Kanitkar et al., 2011). Hou et al. (2009) investigated the effects of transforming growth factor (TGFβ-1) on Achilles tendon healing in rabbits. They showed that by transfecting the bone marrowed-derived MSCs, the tendons showed higher concentrations of collagen and rapid matrix remodeling. Wei et al. (2011) showed local administration of TGFβ-1/VEGF165 gene-transduced bone MSCs in Achilles allograft replacement of the ACL in rabbits improved vascularisation and promoted mechanical strength on biomechanical testing.

Tendon-Bone Interfaces
As previously mentioned, another site of injury of tendons and ligaments occurs at the site of bony insertion. In fact, this is commonly the site of failure for ACL reconstructions.
An emerging area of interest is to look at the effect of MSCs and the enhancement of the tendon-bone interface. Lim et al. (2004) conducted a study to investigate the rate and quality of graft osteointegration in anterior cruciate ligament (ACL) reconstruction after coating the tendon grafts with MSCs. Bilateral ACL reconstructions using hamstring tendon autografts were performed on 48 adult rabbits. The grafts were coated with MSCs in a fibrin glue carrier in one limb, and fibrin glue only in the other. Assessment was done at 2, 4, and 8 weeks.
Histologic analysis and biomechanical testing were performed. Control reconstructions showed mature scar tissue with some Sharpey's-like fibers spanning the tendon-bone interface at 8 weeks. The MSC-enhanced reconstructions had large areas of cartilage cells at the tendon-bone junction at 2 weeks. By 8 weeks, a mature zone of cartilage was seen gradually blending from bone into the tendon grafts. This zone stained strongly for type II collagen and showed histologic characteristics similar to normal rabbit ACL insertions. Biomechanically, there was no statistical difference between limbs at 2 and 4 weeks. At 8 weeks, the MSC-enhanced grafts had significantly higher failure load and stiffness. The apparent increase in osteointegration was also noted by Soon et al. (2007), as previously described.
Using viral modification, Shahab-Osterloh et al. (2010) showed how the tenogenic, osteo-/chondrogenic properties of MSCs could be manipulated to form stem cell dependant tendon-bone interfaces, entheses. This was finding was also reinforced by Gulotta et al. (2010) looked at the effect of MSCs and the healing of supraspinatus tendons in rats. Membrane type 1 matrix metalloproteinase (MT1-MMP) has found to be upregualted during embryogenesis in areas that develop into tendon-bone insertion sites. 30 animals received adenoviral MT1-MMP transduced MSCs while another 30 received solely MSCs. At 4 weeks, the Ad-MT1-MMP group had significantly more fibrocartilage, higher ultimate load to failure, higher ultimate stress to failure and higher stiffness values compared to the MSC group. More importantly, there was a significantly higher prevalence of fibrocartilage at the tendonbone insertion.
Similarly, Ju (2008) used MSCs derived from synovium to accelerate healing of an Achilles tendon graft in murine models. When the Achilles tendon was grafted into a bone tunnel filled with MSCs, integration of tendon to bone was accelerated. Mihelic et al. (2004) proved that MSC and recombinant human bone morphogenetic protein-7 applied randomly to the bone-tendon interface for ACL reconstruction in goats created areas of dense trabecular network with significantly greater invasion of the tendon fibrous tissue into the bone marrow space compared to a control group that did not receive any. Mechanical testing also showed greater strain resistance to force (368 N) in the knees treated with bone morphogenetic protein-7 than in control specimens (214 N).

Induced Pluripotent Stem Cells as an Alternative Source?
Embryonic stem cells (ESCs) are in vitro representations of the inner cell mass of developing embryos and therefore present a valuable tool for regenerative medicine and serve as models of embryonic development in vitro. Induced pluripotent stem cells (iPSCs) are adult cells that have been genetically reprogrammed to an embryonic stem cell-like state by being forced to express genes and factors important for maintaining the defining properties of embryonic stem cells. Although these cells meet the defining criteria for pluripotent stem cells, it is not known if iPSCs and embryonic stem cells differ in clinically significant ways. Currently there is no evidence to support the use of iPSCs in the treatment of tendon or ligament injury. However, the potential application of these cells for these types of injuries poses an intriguing area of research that needs further evaluation.

DISCUSSION
Injuries to tendons and ligaments pose a challenge to medical professionals. The frequency and potential complexities of the injuries make a generic treatment regimen impossible. The treatment options, whether conservative or surgical, must be acceptable to patient. The aims of any intervention are to allow the tendon or ligaments to heal and allow the patient to return to pre injury levels of activity. From a biomechanical perspective, the aim is to provide a structure which has adequate strength, able to withstand peak forces applied during a variety of activities, ranging from standing to sporting activities.
Although reconstruction of ligaments is still a viable and popular option, results are variable. Artificial substitutes for the ACL have been tried, including carbon fibre, polypropylene, Dacron and polyester but have been unsuccessful for a variety of reasons, including immunological reactions and debris generation causing synovitis (Legnani et al., 2010).
Cadaveric grafts are associated with a low risk of transmission of diseases. One author found that 24/181 cadaveric allografts had positive tissue cultures at the time of implantation despite initial screening tests being negative (Díaz-de-Rada et al., 2003). Even when successfully reconstructed, the final structures still lack the complete characteristics of the original. Artificial substances obviously lack the proprioceptive abilities, however the incorporation of mechanoreceptors may address this problem in the future (Hoffman and Gross, 2007).
Within the field of tissue engineering and the use of stem cells, the answer to this challenge may be hidden. There is a growing body of evidence suggesting the benefits of using naturally occurring cells to stimulate the regenerative process (Mahapatra and Khan, 2011;Malik and Khan, in press). Furthermore, we have seen how the modification of these cells with other growth like factors may further aid the reparative process. Enhancement of tendon graft osteointegration with MSCs is a novel method offering the potential for more physiologic and biomechanically stronger ligament reconstructions. iPSCs may also present an alternative reservoir of cells that could be used to treat these injuries however more work is needed.
One thing that is certain is that the field of tissue engineering is an emerging and rapidly developing area that offers an exciting and safe option for treating a wide variety of diseases and conditions (Kennard et al., 2011). However, it is still in the infancy; the vast majority of studies in this field have been performed either in vitro or vivo but using animal models. How this will translate to humans is still unclear. As yet there are no trials looking at the application of stem cells and engineered tissues for ligaments and tendons in humans.

CONCLUSION
Although in its relative infancy, tissue engineering and the use of stem cells is very likely to hold the answer to the problem of how to improve the healing of tendons and ligaments, to replace damaged tissues or improve osteointegration.

COMPETING INTERESTS
Authors have declared that no competing interests exist.