Extracellular Matrix Stiffness Regulates Microvascular Stability by Controlling Endothelial Paracrine Signaling to Determine Pericyte Fate

Abstract

BACKGROUND: The differentiation of pericytes into myofibroblasts causes microvascular degeneration, ECM (extracellular matrix) accumulation, and tissue stiffening, characteristics of fibrotic diseases. It is unclear how pericyte-myofibroblast differentiation is regulated in the microvascular environment. Our previous study established a novel 2-dimensional platform for coculturing microvascular endothelial cells (ECs) and pericytes derived from the same tissue. This study investigated how ECM stiffness regulated microvascular ECs, pericytes, and their interactions. METHODS: Primary microvessels were cultured in the tubular microvascular growth medium on 2-dimensional stiff substrates. Stiff ECM was prepared by incubating ECM solution in regular culture dishes for 1 hour followed by PBS wash. Soft ECM with Young modulus of ≈6 kPa was used unless otherwise noted. Bone grafts were prepared from the rat skull. Immunostaining, RNA sequencing, qRT-PCR, Western blotting, and knockdown experiments were performed on the cells. RESULTS: Primary microvascular pericytes differentiated into myofibroblasts (NG2 + αSMA + ) on stiff ECM, even with the TGFβ (transforming growth factor beta) signaling inhibitor A83-01. Soft ECM and A83-01 cooperatively maintained microvascular stability while inhibiting pericyte-myofibroblast differentiation (NG2 + αSMA −/low ). We thus defined 2 pericyte subpopulations: primary (NG2 + αSMA −/low ) and activated (NG2 + αSMA + ) pericytes. Soft ECM promoted microvascular regeneration and inhibited fibrosis in bone graft transplantation in vivo. As integrins are the major mechanosensor, we performed qRT-PCR screening of integrin family members and found Itgb1 (integrin β1) was the major subunit downregulated by soft ECM and A83-01 treatment. Knocking down Itgb1 suppressed myofibroblast differentiation on stiff ECM. Interestingly, ITGB1 phosphorylation (Y783) was mainly located on microvascular ECs on stiff ECM, which promoted EC secretion of paracrine factors, including CTGF (connective tissue growth factor), to induce pericyte-myofibroblast differentiation. CTGF knockdown or monoclonal antibody treatment partially reduced myofibroblast differentiation, implying the participation of multiple pathways in fibrosis formation. CONCLUSIONS: ECM stiffness and TGFβ signaling cooperatively regulate microvascular stability and pericyte-myofibroblast differentiation. Stiff ECM promotes EC ITGB1 phosphorylation (Y783) and CTGF secretion, which induces pericyte-myofibroblast differentiation.

Multiscale computational model predicts how environmental changes and drug treatments affect microvascular remodeling in fibrotic disease

“…These qualitative changes in node activity were then compared to their corresponding experimental findings. For example, since high TGFβ was predicted by the pericyte intracellular signaling model to increase PMT, and the corresponding experiment also demonstrated increased levels of PMT, as evidenced by αSMA and Collagen 1 expression 54 , we concluded that the model prediction matched the experiment. We further validated that under baseline conditions our models were not overly sensitive to any one input, and we explored the effect of high ECM stiffness on global network sensitivity (Supplementary Figure 1).…”

“…There was agreement between model predictions and experimental outputs for 12/14, or 85%, of the tested perturbations. (D) Predictions made by the pericyte network model (“Model” column) about the cell fate outcomes resulting from eight cellular and environmental perturbations relevant to IPF, including high concentrations of PDGF-BB 62, 63 , pericyte contact with a neighboring EC 64, 65 , high ECM stiffness 43, 66 , high FGF & contact with an EC (EC presence) 41, 56 , high concentrations of TGFβ 67, 68 , TGFβ inhibitor 43 , a combination TGFβ inhibitor and high ECM stiffness 43 , and a PDGF inhibitor 69, 70 . Model predictions were compared to published experiments (“Exp” column).…”

“…While stiffness is just one aspect of the dysregulated microenvironment in IPF, according to recent studies, a stiff ECM alone can cause vascular regression. For example, a recent study co-cultured ECs and fibroblasts in a vessel-on-a-chip model of microvasculature and showed that a stiffness of 20 kPa reduced vessel area and promoted PMT, as quantified by the expression of collagen 1 and αSMA 54 . Further, studies analyzing IPF samples report that mature scar is completely void of microvessels [70][71][72] , which is consistent with the extreme reduction in microvessel area in mature fibrosis (20 kPa) that was predicted by our model and validated in our experiments.…”

“…Nintedanib also inhibits fibroblast growth factor receptor (FGFR), which is known to regulate angiogenesis and pericyte recruitment 40, 41 . Additionally, fibrotic lesions, or foci, in IPF undergo progressive and spatially heterogeneous mechanical stiffening of the ECM 42 , which can affect EC and pericyte dynamics 38, 43 . Understanding how dysregulated biochemical and biomechanical signals impact EC-pericyte coupling in IPF necessitates the use of a computational model where the effects of these signals can be explored, both in isolation and in combination.…”

Perivascular cells function as mechano-structural sensors of vascular capillaries

“…Two events are likely to be associated with this phenomenon. Firstly, stiffer matrices have been shown to reduce the differentiation of hMSCs into pericytes in co-cultures with ECs (11,28). Secondly, poorly differentiated pericytes preserved the more migratory phenotype of hMSCs, and were more likely to penetrate the pore spaces in the collagen matrix, thus reducing the confinement of ECs to the engineered channels.…”