The Role of Mechanical Loading in Tendon Development, Maintenance, Injury, and Repair

Abstract

➤ Tendon injuries often result from excessive or insufficient mechanical loading, impairing the ability of the local tendon cell population to maintain normal tendon function.

➤ The resident cell population composing tendon tissue is mechanosensitive, given that the cells are able to alter the extracellular matrix in response to modifications of the local loading environment.

➤ Natural tendon healing is insufficient, characterized by improper collagen fibril diameter formation, collagen fibril distribution, and overall fibril misalignment.

➤ Current tendon repair rehabilitation protocols focus on implementing early, well-controlled eccentric loading exercises to improve repair outcome.

➤ Tissue engineers look toward incorporating mechanical loading regimens to precondition cell populations for the creation of improved biological augmentations for tendon repair.

Tendon Structure and Function

Tendons connect muscle to bone for the transmission of forces producing joint movement. Composed of primarily type-I collagen fibers in a parallel alignment¹, tendons are viscoelastic, possessing both solid and fluid-like characteristics and exhibiting changes to the stress-strain relationship with respect to the rate at which they are loaded². In addition to type-I collagen, tendons are composed of minor collagens³, including type III, an immature fibrillar collagen that matures into type-I collagen, and type-X collagen, a short-chained collagen found localized in the tendon-to-bone insertion site³. Given the highly organized, hierarchical collagen structure (Fig. 1), tendons exhibit high tensile strength^4-6, allowing for the efficient transmission of large loads, a result of the local cell population to adapt to changes in loading conditions⁷. Further contributing to the structure and biomechanical properties are proteoglycans and glycoproteins, which function to regulate the process of collagen fibrillogenesis and control fibril diameter throughout tendon development and homeostasis^8-13. Studies using genetically manipulated mouse models, in which decorin has been knocked out, have investigated the role of decorin, a small leucine-rich proteoglycan important to tendon structure, and have shown that the absence of decorin results in improper collagen fibril formation and decreases mechanical properties¹³. Undoubtedly, proper tendon structure relies on the interaction of a number of factors to establish normal tendon function.

Fig. 1.

The tendon’s hierarchical structure begins at the molecular level with tropocollagen¹. Approximately five tropocollagen molecules form a microfibril, which then aggregate to create a subfibril¹. Several subfibrils form a single fibril. Multiple fibrils form a tendon fascicle, and fascicles, separated by the endotenon, join to form the macroscopic tendon¹. Tendon fibroblasts, or tenocytes, are found on collagen fibers allowing for the regulation of the extracellular environment in response to chemical and mechanical cues. (Reproduced, with permission of Elsevier, from: Silver FH, Freeman JW, Seehra GP. Collagen self-assembly and the development of tendon mechanical properties. J. Biomech. 2003 Oct;36(10):1529-33, Copyright 2003; and Wang JH. Mechanobiology of tendon. J Biomech. 2006;39(9):1563-82, Copyright 2006.)

Tendon fibroblasts, also referred to as tenoyctes, are the primary cell type regulating tendon homeostasis. These spindle-shaped cells, located along collagen fibers, interact with one another and adjacent collagen fibers, allowing for the formation of collagen cross-links and recognition of chemical and mechanical changes in the extracellular environment¹⁴. Tenocytes are mechanosensitive since they can respond to mechanical loading events by modulating the extracellular environment through the formation and degradation of matrix proteins via a process termed mechanotransduction^14,15. This process involves interactions among extracellular matrix proteins, cell surface receptors, the internal actin cytoskeleton, and signaling molecules, which ultimately regulate protein expression in response to loading alterations¹⁵. While normal physiologic loads are necessary for appropriate tendon development and maintenance, abnormal loading inhibits the capacity of the cell population to maintain homeostasis, contributing to injury¹⁶. Reestablishing these mechanotransductive processes may be key to improving repair outcome following tendon injury¹⁶.

The Role of Loading in Tendon Development and Homeostasis

Tendon Development

Mechanical forces during development are vital to successful limb and musculoskeletal tissue formation during embryogenesis^17-24. Given limitations in technologies and model systems to isolate single mechanical events, investigating the role of tendon loading during embryogenesis is difficult¹⁷. Nevertheless, investigators have shown through in vivo embryonic immobilization studies in chicks that synovial joint development is impaired in the absence of physiologic loads¹⁸. For example, the menisci of the tibiofemoral joint and the plantar tarsal sesamoid of the tibiotarsal joint fail to form, suggesting the inability of tendinous structures to form properly in the absence of mechanical loading and the importance of mechanical stress for proper musculoskeletal development¹⁸.

It is postulated that embryonic and early postnatal growth of tendon relies on the generation of two types of stresses: rapid muscular activity and slow, growth-related elongation of bone²¹. Early in tendon development, the collagen fibril diameter is characterized as a homogeneous distribution of small fibrils (ranging from approximately 40 to 75 nm). Through tendon maturation and force generation during early postnatal growth, the tendon develops a distribution of both large (approximately 100 to 150 nm) and small fibrils (approximately 40 to 75 nm)^23,24. The formation of large fibrils provides the majority of resistance to tensile strength, while the small fibrils negate creep and support improved interfibrillar binding²⁴. Using an in vitro model of embryonic tendon formation, Kalson et al. found that applying a slow, steady strain rate to embryonic chick metatarsal tendon cells produced an increase in collagen fibril diameter and fibril volume fraction, improved cell elongation, and led to increases in both Young’s modulus and ultimate tensile stress compared with unstretched controls²¹.

Tendon Homeostasis

Normal, physiologic loads are required to maintain tendon homeostasis and prevent excessive degradation of the extracellular matrix^25-27. Nabeshima et al. found that culturing nontensioned rabbit patellar tendon explants in the presence of collagenase over a period of twenty hours significantly decreased linear stiffness (p < 0.0001), elongation to failure (p < 0.002), and maximum failure force (p < 0.002) by 80% compared with explants tensioned with constant 4% strain²⁶. Further, Flynn et al. found that reconstituted type-I collagen micronetworks, strained between micropipettes, degraded significantly slower (p < 0.05) than unloaded controls when exposed to mammalian collagenase matrix metalloproteinase 8²⁷. These experiments support the beneficial effect of mechanical load and the need for its incorporation in both clinical postoperative rehabilitation protocols and in tissue-engineering applications²⁷.

It is hypothesized that the cell population of a tendon responds to the application of mechanical stress and modulation of the extracellular matrix through the activation and/or effects of a number of growth factors, including transforming growth factor-β1 (TGF-β1), connective tissue growth factor, and interleuken-6 (IL-6)^28-30. Type-I collagen synthesis and deposition have been linked to increasing expression levels of TGF-β1 both in human Achilles tendon studies and studies investigating the effects of cyclic loading on TGF-β1 expression in human tendon fibroblasts^29,30.

Scleraxis has been identified as a DNA-binding transcription factor critical to embryonic limb tendon formation^31,32. It is the primary regulator of tenocyte differentiation and overall tendon phenotype, while also modulating type-I collagen synthesis in response to the application of stress^33-37. Scott et al. found that applying cyclic load to a bioartificial tendon over a three-week period produced higher scleraxis and type-I collagen expression compared with the nonstimulated control³⁵. Mendias et al. subjected reporter mice that expressed green fluorescent protein under the control of a scleraxis promoter to a six-week treadmill-training program and showed an upregulation of scleraxis and type-I collagen expression for the exercised mice compared with the controls³⁶. Mechanical stress is key in promoting and maintaining a tendon-specific phenotype, and scleraxis may play a role in the adaptation of tendon to physiologic loading³⁷.

Mechanisms of Tendon Injury, Natural Healing, and Repair

Mechanisms of Tendon Injury

The type and prevalence of tendon injuries are dependent on a number of factors including sex, age, normal daily activity levels, overall health, and the circumstances contributing to the injury^38-41. Tendon impairments typically result from an internal tensile overloading event, with acute injuries occurring after one isolated, overloading event and chronic injuries occurring over time through repetitive, excessive loading events^39,41. In contrast, others have suggested that acute injuries are indicative of an underlying chronic impairment that contributed to the injury⁴¹.

Overuse tendon injuries, commonly involving the patellar tendon, Achilles tendon, and the origin of the extensor carpi radialis brevis (tennis elbow), account for 7% (66,575 of 889,980 office visits in 2002) of the musculoskeletal disorders in the United States⁴². Tendinopathy is characterized by a loss of normal tendon architecture, changes to normal tenocyte morphology and apoptosis, alterations in the collagen fibril distribution profile, and neovascularization^43-47. Chronic tendon impairments are frequently attributed to repetitive motion and/or overuse. Excessive loading events lead to microtear formation in the tendon⁴⁸, which, if not repaired properly, may lead to the initiation of inflammatory and degenerative responses. This results in an overall weakened structure and increased propensity for tendon rupture^48-50. In an in vivo tendinopathy model, Nakama et al. found that when the New Zealand White rabbit flexor digitorum profundus muscle was stimulated repetitively for eighty hours, microtears were found in all tendon regions and were significantly greater (p < 0.0001) in the loaded limb compared with the unloaded limb⁵⁰. Other investigators have claimed that underuse of a damaged segment of a tendon may be the source of the chronic impairment⁴⁹. Egerbacher et al. found that a loss of homeostatic tension following stress deprivation correlated with increased cell apoptosis in a rat tail tendon model⁴⁹. Ultimately, tendinopathy may result from a combination of overuse and underuse mechanisms, with overloading creating microtears leading to decreased loading of the cell population resident on the damaged collagen fibers.

Natural Healing

On the basis of studies on horses, rabbits, and rats^39,40,51-56, given that obtaining human biopsy samples can be difficult, it has been shown that following tendon injury, the natural healing process forms scar tissue via a three-stage process: inflammation, matrix production, and remodeling and maturation^39,40. The inflammatory stage initiates the response to injury, typically throughout the first week³⁹. The process is characterized by the development of a fibrin clot to stabilize the site; hemostasis; migration of neutrophils, macrophages, and erythrocytes; and subsequent neovascularization³⁹. The matrix production stage initiates as matrix-producing fibroblasts localized to the injury site begin synthesizing collagen and other extracellular matrix proteins throughout approximately one to four weeks following injury^39,40. Substantial cellular proliferation and matrix production occur; however, the collagen produced is highly disorganized⁴⁰. The final stage, remodeling and maturation, begins approximately four weeks following injury and continues until the tissue is repaired through scar formation³⁹. During this phase, the extracellular matrix is remodeled to create a more organized structure through collagen turnover, realignment, and formation of collagen cross-links³⁹. Cell density and vascularity decrease as the tissue further repairs⁴⁰.

Studies have shown that natural healing leads to tendon biomechanical properties that fail to match normal levels at up to eight weeks, twenty-six weeks, and twelve months following injury in a murine and rabbit central patellar tendon model, and sheep Achilles tendon model, respectively^51-53. Investigators have suggested the poor mechanical properties result from predominantly small fibrils in the resulting repair tissue, compared with the normal distribution of both large and small fibrils observed in normal adult human tendon^54,55, but conflicting results exist. Matthew and Moore showed that, following an extensor digitorum longus tendon transection in a rat model, the collagen fibril diameter distribution remained at approximately 40 nm up to 240 days following injury⁵⁵. In contrast, Lavagnino et al. found that there was no difference between control and stress-deprived rat tail tendons in the number of fibrils per tendon counted, mean fibril diameter, density, or size distribution when cultured in vitro⁵⁶. This may imply that the collagen fibril diameter distribution is not solely responsible, although more work investigating the mechanisms contributing to the ultimate decrease in mechanical properties is needed.

Repair

After tendon injury, operative intervention is often necessary to restore function. Previous work has shown inferior healing when operative repair is delayed following injury^57-59. While current therapies produce functional outcomes in the short term, long-term repair outcome varies with respect to type of injury, injury location, and severity^38,40. A large number of tendon injuries result from ruptures at the tendon-to-bone insertion site. This complex, zonal interface is characterized by the integration of a tendon’s collagen fibers transitioning through a fibrocartilaginous region into the mineralized bone. The differences in material properties of the soft and hard tissue lead to high stress concentrations at this site, contributing to injury^60-64. In an effort to improve tendon-to-bone healing, the application of static or cyclic loading at the insertion site may be necessary to restore the zonal phenotype^60,65-67. Additionally, the repair tension, the amount of tension placed on a tendon to reattach it to bone, is important in recreating the insertion site⁵⁹. Stasiak et al. developed a knee joint fixation system to study tendon and ligament-to-bone healing in a rat model of anterior cruciate ligament reconstruction⁶⁵. The system applies a cyclic stimulus to the knee joint while monitoring the forces generated across the joint. The system has been validated in a preclinical experimental setting but has yet to be implemented clinically⁶⁵. Brophy et al. investigated the effect of cyclic, axial displacement of the femur and tibia on an anterior cruciate ligament reconstruction and found no healing impairment with the application of cyclic load, although the inflammatory response increased in comparison with nonstimulated controls⁶⁶. Further work is needed to elucidate the effects of loading on insertion site repair prior to implementation in a clinical setting.

Clinical Applications

Normal tendon development and homeostasis is closely linked to the degree and pattern of mechanical loading to which the tissue is exposed^47,67. Likewise, tendon injury and degeneration may be related to alterations in the physiologic loading profile^14,47,68-70. Manipulation of the mechanical environment of healing tendon may exert a biologic effect through the mechanotransduction mechanism and holds promise for promoting a repair process that restores normal tendon structure and function. Clinical applications of mechanobiological principles following tendon injury form the basis of rehabilitation protocols^71-73. Programs emphasizing tendon loading may be applied for the reversal of age and disuse-related tendon dysfunction, rehabilitation following tendon overuse injury, and in the development of postoperative physical therapy regimens that optimize healing and function following surgical repair.

Disuse muscle atrophy and weakness are causes of impaired function in older individuals^74-78. While age-related sarcopenia has been repeatedly documented, the effect of aging on tendon biomechanical properties is inconclusive^74,79-83. Methodological differences make study comparisons difficult, although the majority of investigators have suggested that collagen loss occurs in older individuals^84-86. A substantial component of age-related strength loss may be the result of inactivity and can be modified by an appropriate exercise program^87-92. Studies have shown that collagen and elastin production, along with collagen fibril diameter and collagen cross-linking, decrease as a result of age⁸². Consequently, tendon tensile strength and stiffness decrease, contributing to injury. Others have shown that while aging does result in decreased tendon mechanical properties, collagen fibril morphology, packing fraction, and collagen cross-linking remain relatively constant over time^93,94. An in vivo study comparing the patellar tendons of twenty-seven and sixty-five- year-old men found decreased collagen content with increased age, yet there was no difference in the degree of collagen cross-linking between age groups, indicating that alterations in mechanical properties may be a result of other factors, such as reduction in glycoprotein and proteoglycan levels or the inability of the tissue to interact with water appropriately^93,94.

Disuse following immobilization has been associated with decreased levels of extracellular matrix protein expression, alterations in tenocyte morphology, and loss of normal extracellular matrix architecture, resulting in impaired function and healing capacity^14,95-98. Exercise improves the mechanical properties following age and disuse changes^99-101. The ideal exercise program seeks to avoid injury while providing a biologic stimulus to maintain tendon homeostasis and function. Resistance training utilizing 80% of the five-repetition maximum three times per week for fourteen weeks in elderly individuals was shown to result in a 65% to 70% increase in tendon stiffness¹⁰². Similarly, resistance training following a ninety-day period of simulated weightlessness resulted in improved tendon mechanical properties, although this regimen did not fully restore function to the levels before the period of weightlessness⁹⁵. These data suggest that to ameliorate the loss of tendon strength due to aging and/or disuse, incorporating resistance training into exercise regimens may be beneficial⁹⁵.

Eccentric strengthening exercise programs have been advocated as effective treatments for tendon overuse injuries and prevention of reinjury^100-108. Arampatzis et al. demonstrated that exercises involving high tendon strain (mean and standard deviation, 4.72% ± 1.08%) were more effective than those producing low tendon strain (mean, 2.97% ± 0.47%) in triggering an adaptive response in human Achilles tendon¹⁰⁰. Stanish et al. suggested that eccentric exercises prepare patients for return to functional, sports-related activities better than those that emphasize concentric muscle strengthening¹⁰⁸. Multiple studies have demonstrated eccentric exercise regimens to be effective for the treatment of tendinopathy^103,109-115. Ohberg et al. found decreased Achilles tendon thickness and normal tendon structure in response to a twelve-week eccentric training program in patients with chronic Achilles tendinosis¹⁰³.

While the mechanism of action of eccentric exercise is poorly understood, it is theorized that eccentric exercise loads the tendon to a greater magnitude compared with concentric contraction, thereby stimulating a more effective repair response^108,110. Further, eccentric exercise may facilitate remodeling by increasing the number of collagen cross-linkages¹⁰⁹. However, more recently, a study comparing concentric and eccentric exercises for Achilles tendinopathy found no differences in the magnitude of peak tendon load or tendon length change between the regimens¹¹⁶. The authors described the presence of high-frequency oscillations in tendon force during eccentric contractions, which are rare with concentric exercises. These oscillations may modulate the therapeutic effects attributed to eccentric exercise regimen¹¹⁶. While many have advocated eccentric exercise programs, others have suggested that understanding the mechanisms involved in these regimens is necessary to support the efficacy of these programs for treatment^117,118. The optimal load, speed of movement, number of repetitions, and duration of contraction remain to be determined and require further investigation^109,116.

While tendon overuse rehabilitation protocols still require optimization, current therapeutic protocols often employ primarily eccentric exercises. However, for the treatment of lateral elbow tendinopathy (tennis elbow), one proposed muscle-strengthening regimen incorporates a combination of both eccentric and concentric contraction exercises¹¹⁹. In this regimen, the patient undergoes a series of daily, ten-repetition maximum exercises using a light weight (1 to 2-lb [0.45 to 0.91-kg] dumbbell) to strengthen the wrist extensor muscles¹¹⁹. Combining eccentric contractions, in which the muscle group lengthens with wrist palmar flexion, and concentric contractions, in which the muscle group shortens with wrist dorsiflexion, results in a protocol that may decrease muscle tension, leading to a reduction in muscle and tendon soreness¹¹⁹.

Joint mobilization with protection of the newly repaired tendon is a mainstay of rehabilitation following tendon repair^120-122. Depending on injury type and location, controlling joint range of motion restricts the amount of loading possible, thereby preventing reinjury. Joint contractures following flexor tendon repair are a major potential source of functional impairment. A canine model of joint mobilization following tendon repair found that early motion resulted in fewer adhesions with improved tendon gliding and superior tensile strength compared with postoperative immobilization^120-122. Multiple rehabilitation protocols ranging from those emphasizing passive motion, those that incorporate active finger flexion, and those that incorporate combinations of the two, have been utilized postoperatively and are effective in restoring joint motion^123,124. Kitis et al. found improved grip strength, range of motion, and hand function (Disabilities of the Arm, Shoulder and Hand score¹²⁵) when patients were treated with active mobilization with a dynamic splinting protocol compared with those treated with a controlled passive movement regimen¹²⁶. A 2004 Cochrane review failed to detect outcome-based differences when the various controlled motion and/or loading rehabilitation protocols were compared, finding that all produced acceptable and comparable outcomes¹²⁷. While the optimal rehabilitative protocol after repair has yet to be determined, it is clear that controlled tendon loading and movement within a synovial sheath provide a mechanical environment that is beneficial to flexor tendon repair and results in improved functional restoration.

Early mobilization protocols with controlled weight-bearing have been employed following operative and nonoperative treatment of Achilles tendon rupture. This approach is supported by several animal models, in which early motion accelerates the repair process and results in superior tissue quality¹²⁸. Twaddle and Poon found comparable functional results and no significant difference in rerupture rate when controlled early motion was instituted following immobilization for ten days in patients treated with or without surgery¹²⁹. These findings have been investigated in other studies in which accelerated rehabilitation protocols have resulted in improved functional recovery and low rerupture rates^130-133.

Mechanical Loading in Tissue-Engineering Applications

Tissue engineers are incorporating mechanical stimulation to enhance tendon tissue augmentations and replacements^134-146. By mechanically preconditioning the tissue-engineered construct cell population prior to in vivo implantation, the cells may be better equipped to enhance the repair since they have been exposed to the appropriate mechanical environment^134,135. Further elucidating the cellular processes involved in the response to the normal and abnormal mechanical environments will improve tissue engineering therapies.

Given the importance of physiological loading to maintain tendon homeostasis, investigators have shown that applying load promotes a tendon-like phenotype in both two and three-dimensional culture conditions^139-146. Ralphs et al. showed that when tendon cells are cultured on a two-dimensional substrate and subjected to biaxial strain at 1 Hz for eight hours per day for a total of ninety-six hours, cells link together using actin adherens junctions along the principal line of strain to monitor tensile load¹⁴². Garvin et al. found, after seven days of loading, improved biomechanical properties (mean and standard deviation, 327.65 ± 172.03 MPa versus 112.20 ± 6.07 MPa), gene expression results that trended toward normal tendon tissue, and improved cellular alignment and contraction for the loaded bioartificial tendon compared with unstimulated controls¹⁴⁶.

Others have sought to improve tendon repair by creating tissue-engineered constructs using autologous progenitor cells derived from the bone marrow of New Zealand White rabbits seeded on a bovine-derived type-I collagen sponge and incorporating a mechanical stimulation profile at a frequency of 1 Hz, for eight hours per day, at a peak strain of 2.4% for a total of twelve days¹³⁴. The tissue-engineered constructs were implanted into defects in the central third of rabbit patellar tendons, and repair tissues were harvested at twelve weeks following surgery. The results showed that mechanical stimulation improved repair tissue maximum force, linear stiffness, maximum stress, and linear modulus to 70%, 85%, 70%, and 50%, respectively, of the values in the normal, uninjured central third of patellar tendons¹³⁴. Mechanical stimulation increased collagen type-I and type-III gene expression three and four times greater than in nonstimulated controls, respectively¹³⁵. Additionally, the stimulated tissue-engineered constructs were 2.5 times stiffer than nonstimulated controls¹³⁵.

Overview

Tendons are dynamic tissues composed of a cell population capable of responding to mechanical cues by altering the extracellular matrix. While it is known that loading and tension play a large role in overall tendon function, it is still necessary to determine the most suitable methods of incorporating these findings toward improving tendon repair. How does a tendon naturally heal and when is the most effective phase to implement repair procedures and/or intervention? When is the optimal time for incorporating loading regimens in patient rehabilitation protocols? How much and how often should loading regimens be implemented in a clinical setting? Should it be based on type of injury and/or location? How can loading parameters be incorporated when creating tissue-engineered constructs that are best primed for in vivo tendon repair? To answer these questions, further investigation with a focus on improving clinical outcomes is needed.

Supplementary Material

Supporting Data

Disclosure of Potential Conflicts of Interest