A novel renal perivascular mesenchymal cell subset gives rise to fibroblasts distinct from classic myofibroblasts

Perivascular mesenchymal cells (PMCs), which include pericytes, give rise to myofibroblasts that contribute to chronic kidney disease progression. Several PMC markers have been identified; however, PMC heterogeneity and functions are not fully understood. Here, we describe a novel subset of renal PMCs that express Meflin, a glycosylphosphatidylinositol-anchored protein that was recently identified as a marker of fibroblasts essential for cardiac tissue repair. Tracing the lineage of Meflin+ PMCs, which are found in perivascular and periglomerular areas and exhibit renin-producing potential, showed that they detach from the vasculature and proliferate under disease conditions. Although the contribution of Meflin+ PMCs to conventional α-SMA+ myofibroblasts is low, they give rise to fibroblasts with heterogeneous α-SMA expression patterns. Genetic ablation of Meflin+ PMCs in a renal fibrosis mouse model revealed their essential role in collagen production. Consistent with this, human biopsy samples showed that progressive renal diseases exhibit high Meflin expression. Furthermore, Meflin overexpression in kidney fibroblasts promoted bone morphogenetic protein 7 signals and suppressed myofibroblastic differentiation, implicating the roles of Meflin in suppressing tissue fibrosis. These findings demonstrate that Meflin marks a PMC subset that is functionally distinct from classic pericytes and myofibroblasts, highlighting the importance of elucidating PMC heterogeneity.


Figure S3
Meflin (Islr, ISH) expression in normal mouse kidney (P56) a. Cell type-specific expression of Meflin (Islr) in the normal mouse kidney (accession number: GSE52004). Foxd1-lineage interstitial cells showed the highest Meflin expression. Data were analyzed with GEO2R and are presented as the mean ± SEM. n = 3/group. b, c. ISH for Meflin (Islr) in adult mouse (P56) (b) and human (c) kidneys. Human kidney tissue samples were obtained from a patient with minimal change disease (MCD). Boxed regions are magnified in adjacent panels. Arrowheads indicate cells that are positive for Meflin (Islr). Notably, Meflin + PMCs were hardly detected in the glomerular vascular pole in the normal human kidney. The glomerulus shown in the lower panel is from a patient with MCD and mild adjacent sclerotic change. Boxed areas were magnified in adjacent panels. Scale bar, 100 µm. Gl, glomerulus; Tu, urinary tubules; Ar, artery; Vp, vascular pole; V, small vessels. Figure S4. Co-expression of Meflin with classic pericyte markers in the mouse kidney.
Tissue sections of normal mouse kidney were double stained for Meflin and previously known pericyte markers using ISH. Meflin (Islr) mRNA was expressed in stromal cells positive for Pdgfrb, Pdgfra, Cspg4, and Acta2 mRNA, suggesting that Meflin is a marker of pericytes or PMCs (black arrowheads). Notably, erythropoietin (encoded by the Epo gene), a marker of fibroblasts involved in renal fibrosis, was not co-expressed with Meflin (white arrowhead). Boxed areas were magnified in respective lower panels. Scale bar, 50 µm. G, glomerulus; V, small vessels.

Figure S5
Flow cytometry histogram of cells isolated from kidneys  All cells prepared from the kidneys of tamoxifen-administered Meflin-CreER T2 ; LSL-tdTomato mice were sorted for the expression of tdTomato (upper left), followed by staining for the endothelial marker CD31 and the hematopoietic marker CD45. Note that most tdTomato + cells (0.00597 ± 0.0009% of all cells) are negative for CD31 and CD45 (upper right). tdTomato + cells were not detected in cells isolated from the kidneys of WT mice (lower panel), demonstrating specificity of the analysis.

Figure S6
Single-cell RNA seq of adult mouse healthy kidney (   a. Meflin-ZDC mice (n = 4/group) treated with either PBS (upper) or diphtheria toxin (DTx, 3 ng/g; lower) were sacrificed three days after DTx administration, followed by preparation of kidney sections and ISH for Meflin (Islr). Meflin + cells (arrowhead) were significantly decreased in DTx-administered mice compared to PBS-administered control mice. Boxed areas were magnified in adjacent panels. Scale bar, 50 µm. b. The numbers of Meflin + cells in the indicated groups were counted and quantified . Four HPF (400×) per each kidney were evaluated. Data are shown as the mean ± SEM. ***p < 0.001 (unpaired t-test with Welch's correction). Data were analyzed and visualized using R software.

Figure S8
Meflin Meflin-ZDC mice treated with either PBS (left) or diphtheria toxin (DTx, 3 ng/g; middle) and WT mice treated with DTx (right) were sacrificed three days after DTx administration, followed by transmission electron microscopy. Dilated capillaries and degenerated tubular epithelial cells (arrowheads) were observed in DTx-treated Meflin-ZDC mouse kidneys but not in other groups. Scale bar, 20 µm. C, capillary; TL, urinary tubular lumen.  a. Meflin-ZDC mice (n = 4/group) treated with either PBS (upper) or diphtheria toxin (DTx, 3 ng/g; lower) were sacrificed three days after DTx administration, followed by double ISH for Gli1 and Meflin (Islr) on kidney sections. Gli1 + Meflin + cells were observed in PBS-administered control mice (arrowheads), whereas they were rare in DTx-administered mice. In DTx-administered mice, there was an increase in the numbers of Gli1 + Meflin − cells (arrow), being accompanied by an increase in Gli1 expression level. Boxed areas were magnified in adjacent panels. Scale bar, 50 µm. b. The numbers of Gli1 + (left), Gli1 + Meflin + (middle), and Gli1 + Meflin -(right) cells in the indicated groups were counted, followed by quantification. Four HPF (400×) per each kidney were evaluated. Data are shown as the mean ± SEM. *p < 0.05 (unpaired ttest with Welch's correction). Data were analyzed and visualized using R software. a. Meflin-ZDC mice treated with DTx (2 ng/g, i.p.; n = 3/group) were sacrificed three and six days after DTx administration, followed by kidney harvesting for qPCR and immunofluorescence (IF) staining. These mice were fed with Diet gel to reduce gastrointestinal toxicity caused by ablating intestinal Meflin + cells. b. qPCR analysis of diphtheria toxin receptor (DTR, encoded by HsHBEGF) and Meflin (Islr), which were downregulated on day 3 but recovered on day 6. Note that α-SMA (Acta2) expression showed the opposite trend. Data are presented as mean ± SEM. **p < 0.01, ***p < 0.001 (one-way ANOVA followed by Tukey's multiple comparisons test). c. IF analysis for α-SMA showing the proliferation of α-SMA + fibroblasts in the interstitial area (top two panels) on day 3, which was accompanied by vasculature dilatation (bottom two panels). Note that α-SMA + vascular smooth muscle cells (VSMC) were discontinuously found in a mosaic pattern in the tunica media of the vessels on at day 3 (arrowheads). Boxed areas were magnified in lower panels. Scale bar, 50 µm. Multi-omics dataset of diseased mouse kidneys (GSE118341) showed significant, time-dependent increases in Meflin mRNA (blue) and protein (yellow) levels in FAinduced kidney injury (upper panel) and UUO-induced kidney fibrosis (lower panel) models. Data were analyzed with GEO2R and visualized as Z scores using R software.  NRK-49F cells transduced with AcGFP (control) or mouse Meflin (mMeflin) were stimulated with recombinant human TGF-β (2 ng/mL) for 12, 24, and 48 h, followed by qPCR for the indicated genes (n = 3/group) (a) and Western blotting for the indicated proteins (b).