Abstract
Skeletal muscle regeneration is a complex process influenced by non-myogenic macrophages and fibroblasts, which acquire different phenotypes in response to changes in the injury milieu or changes in experimental conditions. In vitro, serum stimulates the differentiation of fibroblasts into myofibroblasts, while lipopolysaccharide (LPS) stimulates the polarization of unstimulated (M0) macrophages to acquire an M1 pro-inflammatory phenotype. We characterized these phenotypes using morphology (with circularity as shape descriptor; perfect circularity = 1.0) and phenotype-specific markers. Myofibroblasts (high α-smooth muscle actin [SMA] expression) had high circularity (mean 0.60 ± 0.03). Their de-differentiation to fibroblasts (low α-SMA expression) significantly lessened circularity (0.47 ± 0.01 and 0.35 ± 0.02 in 2% or 0% serum culture media respectively (p < 0.05). Unstimulated (M0) macrophages (no CD86 expression) had high circularity (0.72 ± 0.02) which decreased when stimulated to M1 macrophages (CD86 expression) (LPS; 0.61 ± 0.02; p < 0.05). Utilizing these established conditions, we then co-cultured M1 macrophages with myofibroblasts or myoblasts. M1 macrophages significantly decreased relative myofibroblast numbers (from 223 ± 22% to 64 ± 7%), but not myoblast numbers. This pro-inflammatory co-culture model was used to rapidly screen the following four compounds for ability to prevent M1 macrophage-mediated decrease in myofibroblast numbers: L-NAME (inducible nitric oxide synthase inhibitor), SB203580 (p38 mitogen-activated protein kinase inhibitor), SP600125 (c-Jun N-terminal kinase inhibitor) and LY294002 (phosphoinositide 3-kinase [PI3K] inhibitor). We found that LY294002 rescued myofibroblasts and decreased macrophage numbers. Myofibroblast rescue did not occur with L-NAME, SB203580 or SP600125 incubation. In conclusion, these data suggest a PI3K-associated mechanism whereby myofibroblasts can be rescued, despite simulated pro-inflammatory conditions.