Investigating Metalloproteinases MMP-2 and MMP-9 Mechanosensitivity to Feedback Loops Involved in the Regulation of In Vitro Angiogenesis by Endogenous Mechanical Stresses

Abstract

Angiogenesis is a complex morphogenetic process regulated by growth factors, but also by the force balance between endothelial cells (EC) traction stresses and extracellular matrix (ECM) viscoelastic resistance. Studies conducted with in vitro angiogenesis assays demonstrated that decreasing ECM stiffness triggers an angiogenic switch that promotes organization of EC into tubular cords or pseudo-capillaries. Thus, mechano-sensitivity of EC with regard to proteases secretion, and notably matrix metalloproteinases (MMPs), should likely play a pivotal role in this switching mechanism. While most studies analysing strain regulation of MMPs used cell cultured on stretched membranes, this work focuses on MMP expression during self-assembly of EC into capillary-like structures within fibrin gels, i.e. on conditions that mimics more closely the in vivo cellular mechanical microenvironment. The activity of MMP-2 and MMP-9, two MMPs that have a pivotal role in capillaries formation, has been monitored in pace with the progressive elongation of EAhy926 cells that takes place during the emergence of cellular cords. We found an increase of the zymogen proMMP-2 that correlates with the initial stages of EC cords formation. However, MMP-2 was not detected. ProMMP-9 secretion decreased, with levels of MMP-9 kept at a rather low value. In order to analyse more precisely the observed differences of EAhy926 response on fibrin and plastic substrates, we proposed a theoretical model of the mechano-regulation of proMMP-2 activation in the presence of type 2 tissue inhibitor of MMPs (TIMP-2). Using association/dissociation rates experimentally reported for this enzymatic network, the model adequately describes the synergism of proMMP-2 and TIMP-2 strain activation during pseudo-capillary morphogenesis. All together, these results provide a first step toward a systems biology approach of angiogenesis mechano-regulation by cell-generated extracellular stresses and strains.

Appendix

1.1 Nonlinear Models of Matrix Metalloproteinases Kinetics on Fibrin Gel and Plastic Substrates

The nonlinear model we propose is based on the biochemical reactions depicted in Fig. 9, with the correspondence between model variables and the proteases and their complexes being as follow; m1(t) and Plam1(t) represent MT1-MMP concentrations, t2(t) and Plat2(t) TIMP-2 concentrations, pm(t) and Plapm(t) proMMP-2 concentrations, mt(t) and Plamt(t) MT1-MMP:TIMP-2 concentrations, pmt(t) and Plapmt(t)) proMMP-2 :TIMP-2 concentrations.

All variables with prefix Pla- refer to experiments conducted on plastic substrate, while other variables correspond to cell cultured on fibrin gel. In this latter case, the strain-dependent production rates Φ _pm (t) of proMMP-2 and Φ _t2(t) of TIMP-2 are modelled by the nonlinear function of the lacunae expansion ration s(t) (Fig. 3):

$$ \Phi_{pm} (t) = \frac{{\alpha t^{2} }}{{\left( {\beta + t^{4} } \right)}}\frac{{\left( {1 + A_{pm} s(t)} \right)}}{{\left( {B_{pm} + s(t)^{2} } \right)}};\quad \Phi_{t2} (t) = \frac{{\left( {\Phi_{t20} \sqrt t + A_{t2} s(t)} \right)}}{{\left( {B_{t2} + s(t)^{2} } \right)}} $$

For simplicity, the nonlinear differential system is given below only for variables in fibrin gel experiments and reads:

$$ \left| \begin{gathered} \frac{dm1}{dt} = \Phi_{m1} - k^{eff} \left( {m1} \right)^{2} - k_{mt2}^{on} \left( {m1} \right)(t2) + k_{mt2}^{on} K_{mt2}^{I} (mt) \hfill \\ \frac{dt2}{dt} = \Phi_{t2} - k_{mt2}^{on} \left( {m1} \right)(t2) + k_{mt2}^{on} K_{mt2}^{I} (mt) - k_{mpt2}^{on} \left( {pm} \right)(t2) + k_{mpt2}^{of} (pmt) \hfill \\ \frac{dpm}{dt} = \Phi_{pm} - k_{mpt2}^{on} \left( {pm} \right)(t2) + k_{mpt2}^{of} (pmt) \hfill \\ \frac{dmt}{dt} = k_{mt2}^{on} \left( {m1} \right)(t2) - k_{mt2}^{on} K_{mt2}^{I} (mt) \hfill \\ \frac{dpmt}{dt} = k_{mpt2}^{on} \left( {pm} \right)(t2) - k_{mpt2}^{of} (pmt) \hfill \\ \end{gathered} \right. $$

A mirror differential system describes the evolution of proteases and inhibitor for cell cultured on plastic substrate (Pla- variables), with each equation being the same except for the production rates Φ* _pm (t) of proMMP-2 and Φ* _t2(t) of TIMP-2 that no-longer depend on lacunae expansion. Considering no influence of s(t) in the previous expressions, one gets:

$$ \Phi_{pm}^{*} (t) = \frac{{\alpha t^{2} }}{{\left( {\beta + t^{4} } \right)}};\quad \Phi_{t2}^{*} (t) = \left( {\Phi_{t20} /B_{t2} } \right)\sqrt t $$

Model simulations have been conducted starting with zero initial conditions for all variables and with imposed association/dissociation constants k ^on and k ^of taken from the literature and compiled in the paper of Karagiannis and Popel (2004). Parameters defining strain-dependent rates have being identified from our experimental data. All these values are indicated in Table 1.