Cellular and Molecular Bioengineering - Poster Session A
Poster U16 - Adaptive Cyclic Strain, that Increases or Decreases as Hierarchical Collagen Fibers Form, Does not Further Improve Maturation in Engineered Ligaments
Associate Professor Virginia Commonwealth University, Virginia, United States
Introduction: Hierarchical collagen fibers are the primary source of strength in tendons and ligaments; however, these fibers do not regenerate after injury or repair, resulting in limited treatment options.1 ,2 While engineered replacements show promise, forming the large hierarchically organized collagen fibers essential to long-term mechanical success remains a challenge. Mechanical cues, including cyclic muscle activity, are critical for tissue development in vivo and have been shown to improve maturation of nanometer-wide fibrils in engineered tissues;2,3,4 however the effect on larger fiber formation is largely unknown. Previously, we developed a novel culture system which guides anterior cruciate ligament (ACL) fibroblasts in high density collagen gels to form native-size hierarchical fibers,1 and we have demonstrated intermittent cyclic load at 5 or 10% strain further improves maturation.5 These constructs are promising ACL replacements, however further maturation is needed to be clinically relevant. Interestingly, when applying intermittent cyclic loading we found a shift in mechanotransduction with changes in organization, with 10% strain driving early improvements in mechanics and composition when cells were in unorganized gels, and 5% strain being more beneficial later in culture once cells were anchored on aligned fibers.5 The objective of this study was to explore whether an adaptive cyclic load that changes in intensity as collagen fibers develop, further improves maturation. Based on our previous results, we hypothesize that progressively decreasing cyclic strain as organization increases will better drive cells to produce more mature hierarchical fibers, resulting in significantly stronger replacements.
Materials and
Methods: To form constructs, type I collagen and neonatal bovine ACL fibroblasts were mixed and cast into 1.5 mm thick sheet gels at 20 mg/mL collagen and 5x106 cells/mL, as previously described.1 Rectangles (8x30 mm) were cut from gels, divided between groups, and cultured for up to 6 weeks. Constructs were clamped into a tensile bioreactor (Fig.1A) and loaded using an established intermittent loading regime4 at 1 Hz for 1 hr, twice daily, every other day (Fig.1B). Controls were stretched at 0, 5, or 10% strain throughout culture, while adaptive load constructs were loaded with a strain that changed as constructs formed aligned fibrils (2 weeks) and fibers (4 weeks, Fig.1C). Specifically, increasing load constructs were loaded with 5% load from weeks 0-2, 7% strain weeks 2-4 and 10% strain weeks 4-6, while decreasing load constructs were loaded with the opposite pattern (Fig.1C). Time points were taken at 0, 2, 4 and 6 weeks, with 0-week constructs collected 24 hours after one loading cycle. Post culture, confocal reflectance was performed to analyze collagen organization and 6-8 images from each construct were analyzed via a custom Fast Fourier transform based MATLAB code1,4,5 to determine alignment (1 = unorganized, 4.5 = aligned) and fiber diameter. DNA, Glycosaminoglycans (GAGs), LOX activity, and collagen content were measured via Picogreen, DMMB, LOX activity, and hydroxyproline (Hypro) assays. Mechanics were analyzed by tensile tests at 0.75% strain/sec to failure. All data are mean ± SD. Significance determined by 1- and 2-way ANOVA with Tukey’s post-hoc (p < 0.05).
Results, Conclusions, and Discussions: All constructs contracted with time in culture, but decreasing load led to less contraction than steady load at 4 weeks, and a trend towards less contraction than steady load constructs at 6 weeks (Fig.1D). All constructs formed aligned collagen fibrils by 2 weeks (data not shown), and larger fibers and fascicles by 6 weeks (Fig.2A). However, adaptive load constructs (increasing and decreasing) appeared to have less collagen crimp and reduced organization, resulting in significantly less alignment and decreased fiber diameters by 6 weeks compared to steady load constructs (Fig.2B). Early in culture, when cells were in unorganized gels, 10% load produced increased GAGs and LOX activity (Fig.3A); however later in culture, once cells were on aligned fibers, steady 5% load significantly increased collagen (hypro). Interestingly, both adaptive loads did not similarly increase collagen (Fig.3A, hypro/WW). Instead, decreasing load resulted in a significant increase in GAG (Fig.3A) and percent weight (Fig.1D) compared to other groups, suggesting that decreasing strain may cause an injurious response. Further, adaptive load did not improve LOX activity by 6 weeks, despite steady 5% and 10% load having increased LOX activity at 6 weeks. Mirroring organization, all constructs had improved elastic mechanics by 6 weeks (Fig.3B). However, steady 5% load had the highest elastic modulus and ultimate tensile strength (UTS), mirroring collagen concentration, and steady 5 and 10% constructs had significantly improved toe moduli compared to other groups, mirroring increased crimp formation.
Contrary to our hypothesis, neither decreasing nor increasing adaptive load, which changed as collagen organization improved, led to further improvements in hierarchical collagen organization or tissue mechanics. Previously, it has been shown that cells respond to changes in load with catabolism, which only shifts to anabolism once cells modify their environment to sense the new load.6 This may indicate that adaptive cyclic load leads to repeated remodeling, and periods greater than 2 weeks is needed between changes in load for improved maturation. This study provides new insight into how adaptive intermittent cyclic loading affects hierarchical fiber formation, which will help to inform rehabilitation protocols and engineered replacements.
