Development of a Tissue-Engineered Artificial Ligament: Reconstruction of Injured Rabbit Medial Collateral Ligament With Elastin-Collagen and Ligament Cell Composite Artificial Ligament.

Ligament reconstruction using a tissue-engineered artificial ligament (TEAL) requires regeneration of the ligament-bone junction such that fixation devices such as screws and end buttons do not have to be used. The objective of this study was to develop a TEAL consisting of elastin-coated polydioxanone (PDS) sutures covered with elastin and collagen fibers preseeded with ligament cells. In a pilot study, a ring-type PDS suture with a 2.5 mm (width) bone insertion was constructed with/without elastin coating (Ela-coat and Non-coat) and implanted into two bone tunnels, diameter 2.4 mm, in the rabbit tibia (6 cases each) to access the effect of elastin on the bond strength. PDS specimens taken together with the tibia at 6 weeks after implantation indicated growth of bone-like hard tissues around bone tunnels accompanied with narrowing of the tunnels in the Ela-coat group and not in the Non-coat group. The drawout load of the Ela-coat group was significantly higher (28.0 ± 15.1 N, n = 4) than that of the Non-coat group (7.6 ± 4.6 N, n = 5). These data can improve the mechanical bulk property of TEAL through extracellular matrix formation. To achieve this TEAL model, 4.5 × 106 ligament cells were seeded on elastin and collagen fibers (2.5 cm × 2.5 cm × 80 µm) prior to coil formation around the elastin-coated PDS core sutures having ball-shape ends with a diameter of 2.5 mm. Cell-seeded and cell-free TEALs were implanted across the femur and the tibia through bone tunnels with a diameter of 2.4 mm (6 cases each). There was no incidence of TEAL being pulled in 6 weeks. Regardless of the remarkable degradation of PDS observed in the cell-seeded group, both the elastic modulus and breaking load of the cell-seeded group (n = 3) were comparable to those of the sham-operation group (n = 8) (elastic modulus: 15.4 ± 1.3 MPa and 18.5 ± 5.7 MPa; breaking load: 73.0 ± 23.4 N and 104.8 ± 21.8 N, respectively) and higher than those of the cell-free group (n = 5) (elastic modulus: 5.7 ± 3.6 MPa; breaking load: 48.1 ± 11.3 N) accompanied with narrowed bone tunnels and cartilage matrix formation. These data suggest that elastin increased the bond strength of TEAL and bone. Furthermore, our newly developed TEAL from elastin, collagen, and ligament cells maintained the strength of the TEAL even if PDS was degraded.

Contribution of glycosaminoglycans to the microstructural integrity of fibrillar and fiber crimps in tendons and ligaments.

The biomechanical roles of both tendons and ligaments are fulfilled by the extracellular matrix of these tissues. In particular, tension is mainly transmitted and resisted by protein (collagen, elastin) fibers, whereas compression is opposed by water-soluble glycosaminoglycans (GAGs). GAGs spanning the interfibrillar spaces and interacting with fibrils through the interfibrillar proteoglycans also seem to play a part in transmitting and resisting tensile stresses. Both tendons and ligaments showing similar composition, but different functional roles and collagen array, exhibit periodic undulations of collagen fibers or crimps. Each crimp is composed of many knots of each single fibril or fibrillar crimps. Fibrillar and fiber crimps play a mechanical role in absorbing the initial loading during elongation of both tendons and ligaments, and in recoiling fibrils and fibers when tissues have to return to their original length. This study investigated whether GAGs covalently attached to proteoglycan core proteins directly affect the 3D microstructural integrity of fibrillar crimp regions and fiber crimps in both tendons and ligaments. Achilles tendons and medial collateral ligaments of the knee from eight female Sprague-Dawley rats (90 days old) incubated in a chondroitinase ABC solution to remove GAGs were observed under a scanning electron microscope (SEM). In addition, isolated fibrils of these tissues obtained by mechanical disruption were analyzed by a transmission electron microscope (TEM). Both Achilles tendons and medial collateral ligaments of the rats after chemical or mechanical removal of GAGs still showed crimps and fibrillar crimps comparable to tissues with a normal GAG content. All fibrils in the fibrillar crimp region always twisted leftwards, thus changing their running plane, and then sharply bent, changing their course on a new plane. These data suggest that GAGs do not affect structural integrity or fibrillar crimp functions that seem mainly related to the local fibril leftward twisting and the alternating handedness of collagen from a molecular to a supramolecular level.

Histological and ultrastructural evaluation of the early healing of the lateral collateral ligament epiligament tissue in a rat knee model.

BACKGROUND: In this study, we evaluated the changes which occurred in the epiligament, an enveloping tissue of the ligament, during the ligament healing. We assessed the association of epiligament elements that could be involved in ligament healing. METHODS: Thirty-two 8-month old male Wistar rats were used in this study. In twenty-four of them the lateral collateral ligament of the knee joint was surgically transected and was allowed to heal spontaneously. The evaluation of the epiligament healing included light microscopy and transmission electron microscopy. RESULTS: At the eight, sixteenth and thirtieth day after injury, the animals were sacrificed and the ligaments were examined. Our results revealed that on the eight and sixteenth day post-injury the epiligament tissue is not completely regenerated. Till the thirtieth day after injury the epiligament is similar to normal, but not fully restored. CONCLUSION: Our study offered a more complete description of the epiligament healing process and defined its important role in ligament healing. Thus, we provided a base for new strategies in ligament treatment.

Spatial distribution and orientation of dermatan sulfate in human medial collateral ligament.

The proteoglycan decorin and its associated glycosaminoglycan (GAG), dermatan sulfate (DS), regulate collagen fibril formation, control fibril diameter, and have been suggested to contribute to the mechanical stability and material properties of connective tissues. The spatial distribution and orientation of DS within the tissue are relevant to these mechanical roles, but measurements of length and orientation from 2D transmission electron microscopy (TEM) are prone to errors from projection. The objectives of this study were to construct a 3D geometric model of DS GAGs and collagen fibrils, and to use the model to interpret TEM measurements of the spatial orientation and length of DS GAGs in the medial collateral ligament of the human knee. DS was distinguished from other sulfated GAGs by treating tissue with chondroitinase B, an enzyme that selectively degrades DS. An image processing pipeline was developed to analyze the TEM micrographs. The 3D model of collagen and GAGs quantified the projection error in the 2D TEM measurements. Model predictions of 3D GAG orientation were highly sensitive to the assumed GAG length distribution, with the baseline input distribution of 69+/-23 nm providing the best predictions of the angle measurements from TEM micrographs. The corresponding orientation distribution for DS GAGs was maximal at orientations orthogonal to the collagen fibrils, tapering to near zero with axial alignment. Sulfated GAGs that remained after chondroitinase B treatment were preferentially aligned along the collagen fibril. DS therefore appears more likely to bridge the interfibrillar gap than non-DS GAGs. In addition to providing quantitative data for DS GAG length and orientation in the human MCL, this study demonstrates how a 3D geometric model can be used to provide a priori information for interpretation of geometric measurements from 2D micrographs.

Treatment of ligament laxity by electrothermal shrinkage or surgical plication: a morphologic and mechanical comparison.

Capsular plication or thermal shrinkage can be used to enhance surgical joint stabilization. We compared mechanical or morphologic properties of the medial collateral ligament of the rabbit knee treated by either bipolar radiofrequency electrothermal shrinkage or surgical plication. After 12 weeks, the medial collateral ligaments were procured from treated and contralateral knees to undergo viscoelastic (creep) testing, quantitative transmission electron microscopy, and immunohistochemistry. Creep strain in thermal (1.85% +/- 0.32%) and plicated (1.92% +/- 0.36%) ligaments was almost twice that of the control group (1.04% +/- 0.15%), although there was no difference between treatment modalities. The morphologic parameters of all 3 groups were significantly different (P < .001). The thermal ligaments demonstrated predominantly small fibrils, whereas the plicated group displayed an intermediate distribution of heterogeneous fibrils, suggesting a different pattern of remodeling. Viscoelastic properties are similar after thermal shrinkage or plication, though inferior to those of intact ligaments.

Morphological, biochemical and mechanical features of the cranial cruciate and lateral collateral ligaments in dogs.

The cruciate ligament and the collateral ligament play key roles in stabilization of the knee joint. Cases of serious knee joint problems presented at the, the Veterinary Teaching Hospital of Rakuno Gakuen University, Japan mostly involved rupture of the cranial cruciate ligament (CCL). Disorders in structural and biochemical components of the CCL were thought to be the causes of the knee problems. Morphological, biochemical and biomechanical features of the CCL and the lateral collateral ligament (LCL) were therefore analyzed. In the CCL, fibroblasts with ovoid and enlarged nuclei were observed mainly at the periphery of collagen bundles. The array of collagen fibrils in the LCL was slightly disoriented, but that of the CCL was tight and regular. In the LCL, the major groups of collagen fibrils were those with diameters of 70-80 and 120-130 nm. Most collagen fibrils in the CCL had diameters of 70-80 nm. The mean collagen diameters were 90 nm in the CCL and 105 nm in the LCL. The ratios of the noncollagen area to the area occupied by collagen fibrils were 43% in the CCL and 55% in the LCL. There was no difference between the amounts of HA or between the amounts of DS in two ligaments. However, the amount of CS in the CCL was about 17-times greater than that in the LCL. The expansion of and the resistance to tension exerted onto the CCL were less than those of the LCL. A high concentration of CS and low tensile strength due to small-sized collagen fibrils cause the CCL to rupture easily, especially when overextension of the knee joint occurs.

Ultrastructural comparison of medial collateral ligament repair after single or multiple applications of GaAlAs laser in rats.

BACKGROUND AND OBJECTIVES: To examine single versus multiple applications of a gallium aluminum arsenide (GaAlAs) laser on the ultrastructural morphology of surgically injured medial collateral ligaments (MCLs) in rats. STUDY DESIGN/MATERIALS AND METHODS: Sixteen rats were studied with 12 receiving right MCL transection and 4 receiving sham injury. Group 1 (n = 4) received one session of laser (31.6 J/cm(2)) immediately after injury. Group 2 (n = 4) received 9 doses of transcutaneous laser (3.5 J/cm(2)). The controls (Group 3, n = 4) received one session of placebo laser, while the sham Group 4 (n = 4) received no treatment. Ultrastructural analyses were done with electron microscopy at 3 weeks. RESULTS: The mass-averaged diameters of collagen fibril in the core and periphery of MCLs treated with multiple laser were larger than the control and those with single laser treatment (P < 0.05). However, the sham injured group had larger fibrils than all other groups (P < 0.05). CONCLUSIONS: The repairing MCLs had smaller collagen fibrils than the sham injured ligaments. Multiple laser treatments enhanced the collagen growth in the repairing MCLs at 3 weeks after injury, which are superior to a single treatment with similar dosage.

Investigation of the collagen fibril distribution in the medial collateral ligament in a rat knee model.

This study compared the collagen fibril diameter distribution among six anatomical sites of the rat medial collateral ligament (MCL). Ultrathin MCL sections from 4 male Sprague-Dawley rats were examined electron microscopically. With an automated quantitation method, 41,638 fibrils were measured and compared among the periphery and core regions of the femoral, middle, and tibial portions of the MCL. Results demonstrated significant difference (p < .0033) in mean fibril diameter distribution among the six sites. The mass-averaged diameters of the core and peripheral fibrils were between 175.53 to 190.82 nm and 88.47 to 109.18 nm, respectively, with the peripheral fibrils more homogeneous in size. The fibrils occupied 36.7% to 57.1% of the cross-sectional area of the ligament. About 50% of the fibrils had an oblique factor of 0.8-1.0, implying that most fibrils were aligned longitudinally. This study has provided a detailed profile of the collagen fibril distributions in rat MCL.

Scanning electron microscopic characterization of healing and normal rat ligament microstructure under slack and loaded conditions.

The objective of this study was to observe and compare behavior of the collagen fiber microstructure in normal and healing ligaments, both in situ and ex vivo, in order to add insight into the structure-function relationship in normal and healing ligaments. Fifty-two ligaments from 26 male rats were investigated. Eleven animals underwent surgical transection of both medial collateral ligaments (MCLs) (22 ligaments), which were allowed to heal for a period of 2 weeks. An additional 15 animals (30 ligaments) were used as normals. Ligaments were placed into six groups: Slack (n = 6 control, n = 6 healing), Reference (n = 4 control, n = 4 healing), Loaded (n = 4 control, n = 4 healing), 15 degrees Flexion (n = 4 control, n = 4 healing), 120 degrees Flexion (n = 4 control, n = 4 healing), and Tissue Strain vs. Flexion Angle (n = 8 normals). All ligaments, except those in the Tissue Strain vs. Flexion Angle group, were prepared for scanning electron microscopy. Tissues were harvested, mounted in a load frame, and chemically fixed in one of five states: (1). slack, (2). reference (onset of loading), (3). loaded, (4). 15 degrees knee flexion, or (5). 120 degrees knee flexion. After fixation the tissues were prepared for electron microscopy (SEM). The micrographs from the slack, reference, and loaded groups show fiber straightening with loading in normal ligaments as well as in both scar and "retracted" regions of healing ligaments. Collagen fibers' diameter and crimp patterns were dramatically changed in the scar region of healing ligaments: Width decreased from 19.4 +/- 1.7 microm to 6.5 +/- 2.1 microm (p <.000001), period from 51.4 +/- 15.1 microm to 11.0 +/- 2.4 microm (p <.000001), and amplitude from 9.8 +/- 0.8 microm to 3.9 +/- 0.8 microm (p <.000001). Normal ligaments fixed in situ show wavy regions at 120 degrees but less so at 15 degrees flexion. Healing ligaments fixed in situ show regions of fiber waviness in the scar region at 120 degrees and also at 15 degrees flexion, indicating ligament laxity persists toward both extremes of the range of motion. The data suggest that straightening of crimped fibers is a functionally relevant phenomenon, not only in normal but also in healing ligaments.

Healing of subfailure ligament injury: comparison between immature and mature ligaments in a rat model.

This study evaluated biomechanical properties of healing ligament following subfailure (grade II) injury by comparing young and mature animals in a rat lateral collateral ligament (LCL) model. One randomly selected LCL was stretched in situ using a custom designed device in eighteen young (21 days) and eighteen skeletally mature (8 months) male rats. Animals were euthanized at 0, 7, and 14 days post-surgery, and ligament ultimate stress, strain at failure and laxity were determined (n = 6 pairs per group). At time 0 after introduction of stretch injury, ligament laxity was present in both groups. The mature rats had 54 +/- 9% strength of the control while the immature rats had 58 +/- 11% of the strength of the control, representing a consistent and significant injury. The immature and mature ligaments showed similar patterns of cellular damage post-injury and had similar modes of mechanical failure. Ligament laxity decreased in each group as healing time increased, however ligament laxity did not completely recover in either group after 2 weeks of healing. After 7 and 14 days of healing, the mature rats, respectively, had only 63 +/- 14%% and 80 +/- 8% strengths of the controls while the immature rats had 94 +/- 6% and 94 +/- 10%. Hence, mechanical data showed that immature animals recovered their strength after a grade II sprain at a faster rate than mature animals. However, ligament laxity was still present in both groups two weeks after the injury and was not completely removed by growth in the immature group. These findings are clinically relevant since joint laxity after injury is common, and these results may explain the presence of continued instability in a joint injured at a young age. Hence, this study, with a new injury model, showed differences in ligament healing associated with maturity and quantified the clinically observed persistance of ligament laxity.

Collagen fibril D-period may change as a function of strain and location in ligament.

The purpose of this study was to determine if the characteristic banding pattern (D-period) of collagen fibrils from rabbit medial collateral ligaments changes as a function of gross ligament strain and, if so, whether the changes are location dependent (insertion versus midsubstance). Femur-medial collateral ligament-tibia complexes were strained to 0, 8, or 12% and immediately chemically fixed in situ. Samples were taken from the medial collateral ligament midsubstance and bony insertions, and prepared for and observed under a transmission electron microscope. D-period length was measured and found to increase (albeit not significantly so, p=0. 1) as a function of gross strain for samples obtained from the insertion sites but not for samples obtained from the ligament midsubstance. Results suggested that ligament strains are inhomogeneous at the ultrastructural level.

Collagenase degradation decreases collagen fibril diameters--an in vitro study of the rabbit medial collateral ligament.

Based on the similarity of fibril diameters in healing and grafted ligaments, it has been speculated that all small fibrils represent newly synthesized collagen. Alternatively, small fibrils in grafts could be due to enzymatic degradation of endogenous large fibrils. This study examined the effect of collagenase on collagen fibril diameters in normal NZW rabbit MCLs. Midsubstance MCL slivers were incubated in buffer for 72 or 144 h for comparison with slivers incubated in buffer containing 4 units/ml bacterial collagenase. The samples were examined under TEM for fibril diameter analysis. Mean fibril diameters of 3-day and 6-day collagenase-treated MCLs were significantly reduced, resembling 40-week scar values. These results suggest that collagenase treatment can alter collagen fibril diameter and shape in normal rabbit MCL, thus it is possible that despite their similarity to ligament scars, that at least some small fibrils in ligament grafts may be enzymatically reduced endogenous fibrils.

Fibrous structure and connection surrounding the metacarpophalangeal joint.

The fibrous components of the metacarpophalangeal (MP) joint including the palmar plate, the collateral ligament and the dorsal plate were studied with particular attention paid to the fibrous structure of the fibrous tendon sheath and the deep transverse metacarpal ligament. The tough fibrillar structure around the MP joint, especially the force nucleus, consisted of three types of mixed fibers: the fibrous tendon sheath of the A1 pulley, the deep transverse metacarpal ligament, and the palmar plate. The tendon sheath was located on the ulnar side in the index and middle fingers, on the central position in the ring finger, and on the radial side in the little finger. These fibrous connections among the fingers formed a transverse arch in the hand. The palmar plate of the MP joint was relatively rigid and appears to function as a cushion when flexed. A fold-like protrusion of the synovial layer of the palmar plate of the MP joint had a meniscoid function, which was larger than that of the proximal interphalangeal joint. The capsule of the MP joint was thicker at the dorsal area, forming a dorsal plate, which is a sliding floor of the extensor mechanism and has a meniscoid function for joint congruity. The main lateral stabilizer consisted of collateral ligaments and accessory collateral ligaments anchored to the palmar plate. These structures act together as a "phalangeal cuff", connecting the proximal phalanx to the metacarpal head and stabilizing the MP joint.

Ultrastructural immunolocalization of type-VI collagen and chondroitin sulphate in ligament.

Immunological methods were used to determine the identity of the major components comprising a network of electron-dense seams (described by the authors in a previous work) within the extracellular matrix of medial collateral ligament (MCL) from humans and rabbits. Tissue obtained from MCL midsubstance was subjected to pre-embedding labelling with colloidal gold at the electron microscopic level with monoclonal antibodies (MAbs) against type-VI collagen and chondroitin sulphate (CS), before and after digestion with chondroitinase ABC and testicular hyaluronidase. Tissue labelled with anti-type-VI MAbs showed gold conjugates attached to the microfilamentous component of the seams both before and after enzyme digestion, which confirmed the identity of the beaded microfilaments as type-VI collagen. Treatment of the tissue with anti-CS MAbs resulted in labelling of undigested tissue only. In these treatments, gold particles were found attached to granules that were interspersed throughout the network of type-VI microfilaments. Both the granules and gold labels were absent from the network following enzyme digestion. Thin nonbeaded microfilaments that did not label with anti-type-VI MAbs also were present within the seams. The loss of these nonbeaded microfilaments following enzyme digestion suggested that they might represent strands of hyaluronan. The codistribution and sequestering of type-VI collagen and CS within discrete seams or channels suggests that these regions of the MCL midsubstance may contain higher concentrations of water than the surrounding dense fibrillar matrix.