Soluble allogenic telocollagen as a direct protein therapeutic: Results of serial injections in a rodent rotator cuff tear model.

HYPOTHESIS: Delivery of soluble allogeneic type I telocollagen (allo-telocollagen) will accelerate and improve the healing of damaged tendons. Our hypothesis draws from known mechanochemical properties of type I collagen that direct its incorporation into damaged connective tissue. We further suggest that allo-telocollagen will raise a minimal immunogenic reaction due to homology within species. METHODS: Seventy-eight shoulders (39 Sprague-Dawley rats) had their supraspinatus tendon surgically detached from its footprint on the humerus and repaired (72 shoulders) or left uninjured (6 shoulders). The repaired tissue was treated with an injection of 100 µl of saline, 10 mg/ml allogeneic atelocollagen (allo-atelocollagen), or 10 mg/ml allo-telocollagen at 0-, 1-, and 2-weeks post-surgery. At 30- and 60-days post-surgery, the tendons were assessed by mechanical testing (failure load, failure stress, stiffness, and relaxation) and by semiquantitative histological scoring. RESULTS: At 30-days post-surgery, the mechanical and histological outcomes were not statistically different. However, at day 60, allo-telocollagen improved the failure strength of the supraspinatus (29.9 ± 4.7 N) relative to saline (20.0 ± 3.5 N; P value <= 0.001) or allo-atelocollagen (23.2 ± 1.5 N; P value = 0.025) treated tendons, and it approached that of uninjured controls (36.9 ± 5.0 N; P value = 0.021). Allo-telocollagen improved the failure stress of the supraspinatus (34.1 ± 9.3 MPa) relative to the saline treated tendons (21.4 ± 6.0 MPa; P value = 0.031; 160% improvement) and was no different than uninjured controls (33.4 ± 9.9 MPa; P value = 0.999) or allo-atelocollagen (32.3 ± 7.4 MPa; P value = 0.977). The stiffness of uninjured controls was far greater than any of injured/treated tendons (>200% stiffer). Histological scoring showed that the allo-telocollagen treated tendons produced better collagen fiber arrangement (1.55 ± 0.17) than saline (2.50 ± 0.29; P value = 0.001) or allo-atelocollagen (2.23 ± 0.28; P value = 0.042) treated tendons and that it did not increase markers of immunogenesis (1.10 ± 0.42) relative to either saline (1.44 ± 0.20; P value = 0.369) or allo-atelocollagen (0.68 ± 0.41; P value = 0.1058). CONCLUSIONS: While all three treatments produced similar results at 30 days, by 60 days, soluble allo-telocollagen clearly separated from the other interventions, yielding better mechanical and histological outcomes in a torn/repaired rotator cuff rat model. Allo-telocollagen treated tendons also approached the failure strength and matched the failure stresses of uninjured control tendons. The data suggest a new use for allo-telocollagen as a deliverable direct protein mechanotherapeutic that can improve both healing quality and speed.

Development of Fluorescently Labeled, Functional Type I Collagen Molecules.

While de novo collagen fibril formation is well-studied, there are few investigations into the growth and remodeling of extant fibrils, where molecular collagen incorporation into and erosion from the fibril surface must delicately balance during fibril growth and remodeling. Observing molecule/fibril interactions is difficult, requiring the tracking of molecular dynamics while, at the same time, minimizing the effect of the observation on fibril structure and assembly. To address the observation-interference problem, exogenous collagen molecules are tagged with small fluorophores and the fibrillogenesis kinetics of labeled collagen molecules as well as the structure and network morphology of assembled fibrils are examined. While excessive labeling significantly disturbs fibrillogenesis kinetics and network morphology of assembled fibrils, adding less than ≈1.2 labels per collagen molecule preserves these characteristics. Applications of the functional, labeled collagen probe are demonstrated in both cellular and acellular systems. The functional, labeled collagen associates strongly with native fibrils and when added to an in vitro model of corneal stromal development at low concentration, the labeled collagen is incorporated into a fine extracellular matrix (ECM) network associated with the cells within 24 h.

Measuring collagen fibril diameter with differential interference contrast microscopy.

Collagen fibrils, linear arrangements of collagen monomers, 20-500 nm in diameter, comprising hundreds of molecules in their cross-section, are the fundamental structural unit in a variety of load-bearing tissues such as tendons, ligaments, skin, cornea, and bone. These fibrils often assemble into more complex structures, providing mechanical stability, strength, or toughness to the host tissue. Unfortunately, there is little information available on individual fibril dynamics, mechanics, growth, aggregation and remodeling because they are difficult to image using visible light as a probe. The principle quantity of interest is the fibril diameter, which is difficult to extract accurately, dynamically, in situ and non-destructively. An optical method, differential interference contrast (DIC) microscopy has been used to visualize dynamic structures that are as small as microtubules (25 nm diameter) and has been shown to be sensitive to the size of objects smaller than the wavelength of light. In this investigation, we take advantage of DIC microscopy's ability to report dimensions of nanometer scale objects to generate a curve that relates collagen diameter to DIC edge intensity shift (DIC-EIS). We further calibrate the curve using electron microscopy and demonstrate a linear correlation between fibril diameter and the DIC-EIS. Using a non-oil immersion, 40x objective (NA 0.6), collagen fibril diameters between ~100 nm to ~ 300 nm could be obtained with ±11 and ±4 nm accuracy for dehydrated and hydrated fibrils, respectively. This simple, nondestructive, label free method should advance our ability to directly examine fibril dynamics under experimental conditions that are physiologically relevant.