Tissue stiffness scaling with fibrillar collagen results from strain-suppressed fibril proteolysis

Introduction: Fibrillar collagens are the most abundant proteins in mammals present in varying amounts in all tissues. Extremely soft embryonic chick hearts or adult mouse brain contain far less fibrillar collagen than highly stiff adult mouse tendons. Interestingly, stiff tissues sustain larger mechanical stresses physiologically than soft tissues. Thus, the pan-tissue scaling observations of stiffness and collagen levels indicate mechanisms of collagen accumulation are determined in part by sustained mechanical stresses ( or strain) dependent degradation of fibrillar collagen. We studied such processes in embryonic chick hearts and tendon as well as other distinct tissues from adult mice including brain, kidney, lung, heart, and muscle.
Materials and Methods: For studying collagen levels in ex-vivo tissues, we employed label-free second harmonic generation (SHG) signal-based imaging. For mechanical perturbations and simultaneous imaging, we designed an environmentally controlled mechano-bioreactor. In case of beating chick hearts (Fig.1), extracellular strain (produced by cellular contractility) was pharmacologically inhibited in the presence or not of matrix metalloproteinase (MMP) inhibitors, and fibrillar collagen levels were measured (on ~hr timescales) using SHG signal and mass-spectrometry (MS) approaches. Tendon fascicles were subjected to heterogeneous strain distribution via either three-point bending with adhesion (Fig.2) or uniaxial deformation mode and exposed to active forms of Matrix Metalloproteinase-1 (MMP-1) or Bacterial collagenase (BC). Strain distribution was measured in fluorescently labelled fascicles by using patterned photo-bleaching and SHG imaging was used to make microstructure assessment. Fibrillar collagen degradation in tissues was tracked based on temporal SHG signal.
Results and Discussion: SHG signal across diverse tissues is observed to scale non-linearly with stiffness, aligning well with results from cellularized collagen gels. Such scaling relationship interestingly captures diverse molecular perturbations to tissues collagen. In case of isolated beating embryonic chick hearts, ~5% strain (change in aspect ratio) was found to retain their collagen mass whereas the inhibition of myosin-II resulted in both the loss of heart contractile strain and increased collagen degradation by endogenous MMPs. Likewise, for physiological deformation of tendon tissue (i.e., ~5-8% strain), strain-dependent decrease in degradation rate was observed, which was independent of type collagenase used (i.e. MMP1 or BC that differ in cleavage site-specificity and cleavage rate of collagen). However, at higher strains, the degradation rate changes were almost independent of strain magnitude changes. Such three-point bending based mechanical deformation of tendon fascicles within the mechano-bioreactor resulted in collagen triple-helix unfolding duly demonstrated by the binding of collagen hybridizing peptide (CHP) and was similar to that of conventional uniaxial tensile deformation. Sequestration of collagen cleavage sites by tendon tissue strain is likely to restrict conformational access to collagenase molecules because the degradation rate was correlated to unfolded collagen content in artificially crosslinked tendon fascicles subjected to mechanical deformation.
Conclusions: Our results have potential to deepen the understanding of collagen matrices in development, adaptation, and disease. Strategies for strain-sculpting of materials for tissue engineering and for biomimetics are thus suggested as are therapeutic approaches to fibrosis and rehabilitation. Mechano-bioreactor design proposed in the present work can provide simultaneous measurements of tissue mechanical strains and structure of micro-scale samples in a high-throughput manner when compared to conventional mechanical testing techniques.