Recreating Normal Tendon and Ligament Attachments After Injury More than 100,000 anterior cruciate ligament (ACL) repairs for the unstable knee are performed each year.1 Many are done with tendon transfers that rely on a biologic-bone interface attachment. These repairs, and the repair of the supraspinatus tendon for shoulder weakness, do not result in a normal bone-ligament interface. Although tendon-to-bone healing following ACL reconstruction does not lead to the reestablishment of the anatomic insertion, a fibrovascular tissue is formed with Sharpey fibers connecting bone and transplant within the bone tunnel.2 The ACL interface often does not remodel with bone attachment and can lead to graft failure because of loosening.3 The four zones of tendon insertion —tendon, fibrocartilage, calcified cartilage, and bone interface— were first determined by Dolgo-Saburoff in 1929.4 Although the anatomic structures critical for a biomechanically stable tendon in-sertion remain to be determined, it is reasonable to assume that this hierarchical structure is important.5 Recent cell culture studies suggest that an insertion interface might be reestablished by manipulating interaction between cells derived from tendon (eg, fibroblasts) and bone tissue (eg, osteoblasts). Using conditions that promote cartilage and bone development (ie, ascorbic acid, β-glycerophosphate),6 co-cultures of fibroblasts and osteoblasts were shown to express matrix proteins, types I and II collagen, aggrecan, and cartilage oligomeric matrix protein (COMP), all of which are characteristic of fibrocartilage, at the fibroblast-osteoblast interface. Like the ACL, the supraspinatus tendon is surrounded by fluid and synovium, and repair is characterized by disorganized scar and lack of fibrocartilage. Recent studies suggest that the temporal sequence of transforming growth factor (TGF)-β isoforms expressed during development of the supraspinatus insertion sites in mice plays a role in tissue organization and differs from that seen in healing.7 By postnatal day 7, hypertrophic chondrocytes in the interval between humeral head and tendon were present, expressing type X collagen by day 14. The four-zone transition was present at days 21 to 28. This included zone 1, characterized by type I collagen expression; zone 2, fibrochondrocytes characterized by types I and II collagen expression; zone 3, hypertrophic chondrocytes characterized by expression of types II and X collagen; and zone 4, mature bone. By day 56, hypertrophic chondrocytes were gone, but type X collagen persisted. Expression of TGF-β3 and -β1 was observed early in insertion site development in fetal mice, with transition to TGF-β1 at 15.5 days (mice are born at 21 days). Additional studies suggest that a peptide fragment of type I procollagen plays a role in the regulation of type I collagen production by mediating TGF-β expression. The pentapeptide KTTKS promotes release of type I collagen from fibroblasts. Investigation of the role of this peptide in development suggests that it promotes transcription of mRNA for TGF-β and mRNA-stabilizing binding proteins (mRBPs) E1 and K,8 which results in increased type I collagen production by both promotion and stabilization of type I collagen mRNA expression. This protein fragment could play a pivotal role in stimulating the developmental expression of TGF-β and type I collagen and thus might offer therapeutic efficacy in the stimulation of tendon repair. An inviting approach when attempting tendon repair is to replicate the sequence of events seen in development. Manipulation of cytokines and breakdown products in wound repair could lead to the reproduction of developmental anatomy in adult insertion repairs. Fred R.T. Nelson MD New Challenges in the Design of Prosthetic Devices for Amputees The prevalence of lower extremity amputation because of combat injuries is high. Current military combat operations have led to thousands of amputations, with 70% of these considered major limb amputations.9 Historically, the great majority of lower extremity amputations have resulted from complications of venous insufficiency and diabetes, and this patient population has had relatively low expectations of physical function. However, the growing population of amputees injured in military operations is significantly younger and more physically active than is the traditional amputee population. Consequently, these young, active amputees have driven the development of advanced prosthetic devices. Many of these young, active amputees are capable of a very high level of physical function with prosthetic devices. (For example, consider Oscar Pistorius, the bilateral transtibial amputee sprinter who was recently banned from competing in the Olympics.) However, for many unilateral amputees, previous research has indicated that there may be a long-term risk of osteoarthritis (OA) developing in the unaffected contralateral limb.10–12 Thus, for the young amputee who desires to remain active and maintain a high level of physical function for many years, the challenge is in how to avoid OA developing in the contralateral limb. The assessment of new prosthetic devices has largely been based on subjective evaluation and conventional biomechanical analyses. These conventional analyses have often relied on ground reaction forces, joint kinematics, estimated joint forces and moments, and electromyography to characterize differences between the affected and unaffected limbs of unilateral amputees.13–20 For example, Kaufman et al14 recently used a crossover study design to compare gait patterns of transfemoral amputees using a passive mechanical prosthesis and then a microprocessor-controlled knee. (The latter is a device capable of adjusting knee flexion/extension resistance based on walking speed and phase of gait). The investigators used kinematic and force plate data to estimate joint forces and moments. In addition, the study also assessed each device's effect on balance by quantifying postural stability for each patient. The authors concluded that use of the microprocessor-controlled knee resulted in a more nearly normal walking pattern and a significant increase in postural stability. In a similar study, Schmalz et al15 compared the gait patterns of transtibial and transfemoral amputees to those of nonamputees during the more challenging tasks of ascending and descending stairs. This study reported on joint kinematics, ground reaction forces, estimated joint moments, and muscle activity patterns. For several of the outcome measures, the study failed to detect significant differences between the amputees' involved and uninvolved limbs or, even more impressively, between the amputees and the non-amputee control subjects. These studies suggest that differences in joint function between amputees and nonamputees during activities of daily living may be shrinking, at least when quantified using conventional biomechanical outcome measures. Not known, however, is the extent to which conventional biomechanical outcome measures will be able to detect subtle differences in joint mechanics associated with the development of OA. Future research efforts should focus not only on improving prosthetic design and rehabilitation techniques but also on developing more sensitive measures of joint function that are predictive of OA development. It will be necessary to develop more sophisticated and sensitive laboratory techniques to identify the prosthetic devices and rehabilitation protocols that simultaneously provide a high level of joint function while minimizing the development of OA in the contralateral limb. Michael Bey PhD Cartilage Oligomeric Matrix Protein, Collagen Fibril Formation, and Tissue Mechanics Cartilage oligomeric matrix protein (COMP) is a pentameric (five-armed) glycoprotein that is found in cartilage, tendon, bone, and other connective tissues. Mutations of COMP cause pseudoachondroplasia and multiple epiphyseal dysplasia,21,22 but mice with complete knockout of the protein have normal skeletal development.23 Levels of COMP in synovial fluid, serum, and urine are increased in osteoarthritis and rheumatoid arthritis, making it a potential marker for these diseases.24 Recently fibrillogenesis of collagen I and II was shown to be accelerated with COMP acting as a catalyst.25 COMP interacted with more than one collagen molecule, bringing them into close proximity, which is the apparent mechanism that accelerated fibril formation. After a 67-nm repeat fibril was formed, COMP was released undamaged, making it available to repeat the process of binding and forming a fibril. This catalytic process is in itself an interesting nanoscopic biomechanics problem waiting to be solved because native COMP can perform the “capture and release” process, but mutated COMP cannot. Another interesting nanoscopic biomechanical problem involving COMP comes from the observation that COMP binds both aggrecan26 and collagen. This implies that COMP might act as a binding molecule among the aggrecan molecules or, perhaps, act as a binding molecule between aggrecan and collagen II. The binding affinity among these molecules is not sufficiently strong to be considered permanent; therefore, the formation and breaking of the bonds could affect the flow independent of the viscoelasticity of cartilage. At the tissue level, mechanical loading increases the expression of COMP27 and collagen II,28 suggesting that both the catalyst and substrate are regulated together. Higher COMP concentration is associated with thinner collagen fibers in the tendons of young horses,29 which is consistent with this idea. The interrelationship between these molecules, functional loading, age, and tendon development is a very interesting and active research area. In conclusion, COMP is mechanically regulated; is associated with collagen fibril formation in cartilage, bone, and tendon; and is one of several potential biomarkers for arthritis. Research in this area crosses over from biochemistry and biology to directly affect tissue mechanical function through fibril formation and, potentially, through binding among aggrecan and collagen in articular cartilage. This is one of the molecules of which it is truly possible to predict that “more will be revealed.” David P. Fyhrie PhD Paul E. Di Cesare MD