Human Umbilical Cord-Mesenchymal Stem Cells Promote Extracellular Matrix Remodeling in Microglia

Abstract

Human mesenchymal stem cells modulate the immune response and are good candidates for cell therapy in neuroinflammatory brain disorders affecting both adult and premature infants. Recent evidence indicates that through their secretome, mesenchymal stem cells direct microglia, brain-resident immune cells, toward pro-regenerative functions, but the mechanisms underlying microglial phenotypic transition are still under investigation. Using an in vitro coculture approach combined with transcriptomic analysis, we identified the extracellular matrix as the most relevant pathway altered by the human mesenchymal stem cell secretome in the response of microglia to inflammatory cytokines. We confirmed extracellular matrix remodeling in microglia exposed to the mesenchymal stem cell secretome via immunofluorescence analysis of the matrix component fibronectin and the extracellular crosslinking enzyme transglutaminase-2. Furthermore, an analysis of hallmark microglial functions revealed that changes in the extracellular matrix enhance ruffle formation by microglia and cell motility. These findings point to extracellular matrix changes, associated plasma membrane remodeling, and enhanced microglial migration as novel mechanisms by which mesenchymal stem cells contribute to the pro-regenerative microglial transition.

Keywords: extracellular matrix; human umbilical cord mesenchymal stem cells; microglia; migration; neuroinflammation.

Figure 1

h-MSCs affect homeostatic and activation marker expression in primary mouse microglia. (A) qPCR analysis of primary mouse microglia exposed to inflammatory cytokines (IL1β, TNFα, and IFNγ) and cocultured with h-MSCs for 48 h in transwells. Compared with those in activated cells, the expression of the proregenerative marker Arg1, the immunomodulating marker Socs3, and the homeostatic marker Tmem119 was significantly increased after h-MSCs exposure. The data are presented as the means ± SEs normalized to nonstimulated cells (fold change of 1). (Kruskal–Wallis multiple comparisons test: Arg1 p = 0.007; one–way ANOVA with Tukey’s multiple comparisons test: Socs3 p < 0.001, Tmem119 p = 0.04; N = 6). The activation marker Clec7a and the inflammatory marker Tnfα are downregulated (one-way ANOVA, Tukey’s multiple comparisons test: Clec7a, p = 0.006; Tnfα, p < 0.001, N = 6), whereas the other inflammatory markers Il1β and Il6 are significantly upregulated in h-MSCs-treated microglia compared with activated microglia (one-way ANOVA, Tukey’s multiple comparisons test: Il1β, p < 0.001; Il6, p < 0.001, N = 6). (B) qPCR analysis of primary mouse microglia magnetically isolated from neonatal mice exposed to inflammatory cytokines (IL-1β, IFNγ) for 3 h and then cocultured with h-MSCs for an additional 3 h. (C) qPCR analysis of primary mouse microglia magnetically isolated from neonatal mice exposed to inflammatory cytokines (IL-1β, IFNγ) for 3 h and then cocultured with h-MSCs for an additional 21 h. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 2

RNA-seq analysis showing extracellular matrix remodeling in microglia cocultured with h-MSCs. (A) Volcano plot of differentially expressed genes in h-MSCs-treated activated microglia vs. activated microglia. Red circles represent genes with a |log2-fold change| >  0.5 and a -log10 Benjamini–Hochberg (BH)-adjusted p value > 1; blue circles represent genes with a -log10 BH-adjusted p value > 1; green circles represent genes with a |log2-fold change| > 0.5; gray circles represent genes with a |log2-fold change| ≤ 0.5 and a -log10 BH-adjusted p value ≤ 1. (B) Bar plot showing all the significant hallmarks identified via GSEA. Orange bars represent positive enrichment; blue bars represent negative enrichment. (C) Heatmap reporting the expression levels of the top leading genes of the 4 positively enriched gene sets and 2 negatively enriched gene sets. TNFα signaling, blue; INF-γ, green; inflammatory response, red; EMT, violet; G2M checkpoint, yellow; mitotic spindle, dark green. (D,E) GSEA plots reporting the top enriched MATRISOME gene set (NABA_CORE_MATRISOME) in h-MSCs-treated activated microglia vs. activated microglia (D) and h-MSCs-treated activated microglia vs. control microglia (E).

Figure 3

RNA-seq analysis showing cell cycle and immune system regulation by microglia cocultured with h-MSCs. (A) Bar plot showing the top 10 Reactome pathways enriched in h-MSCs-treated activated microglia vs. activated microglia. (B) GSEA plots reporting gene sets related to anti-inflammatory cytokine signaling pathways (IL4/IL13 and IL10) and negative regulation of immune system processes in activated h-MSCs-treated microglia vs. activated microglia. (C) Dot plot of positively or negatively enriched GSEA gene sets related to relevant microglial functions.

Figure 4

Impact of h-MSCs on extracellular matrix composition and microglial motility. (A) Representative confocal z-stack projections by Imaris of activated microglia, cultured with or without h-MSCs, and stained for Fn (red), IB4 (green), and DAPI under nonpermeabilizing conditions. Dotted boxes indicate the zoomed region on the right. Scale bars: 20 mm. Scale bar: zoom-in: 20 mm. (B) Corresponding quantification of surface Fn (top; t-test, p < 0.001, N = 6) and extracellular Fn (bottom; t-test, p < 0.001, N = 6). (C,D) Microglia described in A were subjected to live staining with TG2 for 1 h (red) and IB4 for 5 min (green), fixed and stained with DAPI. (C) Representative confocal z-stack projections generated by Imaris showing cellular and extracellular (arrow) TG2 staining. (D) Corresponding quantification of cellular TG2 (top; t-test, p = 0.0152, N = 6) and extracellular TG2 (bottom; t-test, p = 0.0072, N = 6). (E) Representative bright fields of microglia shown in A; live images at 0 and 20 min are shown. The colored traces indicate the paths traveled by the cells. The arrows point to membrane ruffles. (F) Histograms showing the path length, average speed, and number of ruffles of the imaged cells (Mann–Whitney test, p = 0.0026; path length, p = 0.0023; average speed, p = 0.0008; ruffles, activated N = 32, activated + h-MSCs N = 23). * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 5

Effects of h-MSCs on microglial proliferation, phagocytosis, and MCHII and Clec7a expression. (A) Representative confocal images of control, activated, and h-MSCs-treated activated proliferating microglia that were positive for EdU and double-stained for IB4 (green) and DAPI (blue). On the right, the corresponding quantification is shown. Each dot represents the percentage of proliferating cells in a single field. The data are presented as the means ± SEMs (one-way ANOVA, Tukey’s multiple comparisons test, p < 0.001 ctr vs. activated, p < 0.001 h-MSCs vs. ctrl, p = 0.60 h-MSCs vs. activated, N = 3). (B) Representative images of microglia that engulfed fluorescent beads (arrows) under the conditions described in (A) and were stained with IB4 (green) and DAPI (blue). On the right, the data are quantified (one-way ANOVA, Tukey’s multiple comparisons test, p = 0.59 ctr vs. activated, p = 0.20 h-MSCs vs. activated, N = 4). (C) Representative confocal images of microglia as in A stained for MHC-II (red), IB4 (green), and DAPI (blue). On the right, CTCF quantification of MHC-II+ microglia per field are normalized to those in the control group (one-way ANOVA, Tukey’s multiple comparisons test, p = 0.001 ctr vs. activated, p < 0.001 h-MSCs vs. activated, N = 4). (D) Representative images of microglia stained with Clec7a (red), IB4 (green), and DAPI (blue) under the conditions described in A and the corresponding CTCF quantification of Clec7a+ microglia per field, normalized to the control (one-way ANOVA, Tukey’s multiple comparisons test p = 0.001 ctr vs. activated, p < 0.001 h-MSCs vs. activated, N = 4). * p < 0.05, ** p < 0.01, *** p < 0.001.