Genetic Deletion of the Stromal Cell Marker CD248 (Endosialin) Protects against the Development of Renal Fibrosis

Abstract

Background: Tissue fibrosis and microvascular rarefaction are hallmarks of progressive renal disease. CD248 is a transmembrane glycoprotein expressed by key effector cells within the stroma of fibrotic kidneys including pericytes, myofibroblasts and stromal fibroblasts. In human disease, increased expression of CD248 by stromal cells predicts progression to end-stage renal failure. We therefore, hypothesized that the genetic deletion of the CD248 gene would protect against fibrosis following kidney injury.

Methods: Using the unilateral ureteral obstruction (UUO) model of renal fibrosis, we investigated the effect of genetic deletion of CD248 on post obstructive kidney fibrosis.

Results: CD248 null mice were protected from fibrosis and microvascular rarefaction following UUO. Although the precise mechanism is not known, this may to be due to a stabilizing effect of pericytes with less migration and differentiation of pericytes toward a myofibroblast phenotype in CD248-/- mice. CD248-/- fibroblasts also proliferated less and deposited less collagen in vitro.

Conclusion: These studies suggest that CD248 stromal cells have a pathogenic role in renal fibrosis and that targeting CD248 is effective at inhibiting both microvascular rarefaction and renal fibrosis through modulation of pericyte and stromal cell function.

Fig. 1

CD248 is expressed by a subset of pericytes resting kidney. A CD248 co-localized with recognized markers of pericytes in resting kidney. Representative images are shown at ×630 magnification of WT resting kidney (right panel) with high-powered images (2 × zoom of ×630 original magnification) on the left panel stained with CD31 (endothelium, red), CD248 (blue) and pericyte markers: NG2, αSMA, PDGFRβ and desmin (green). Scale bar 50 μm. * Indicates an example of cell staining positive for both CD248 (blue) and the pericyte marker (green). B Expression of individual pericyte markers, expressed as percentage of CD31 positive vessels positive for PDGFRβ, NG2, αSMA or desmin. Filled bars are WT and open bars are CD248^-/-. Data presented are mean ± SEM from 10 high-powered (×630) fields of view per mouse kidney section (10 × kidney sections were analyzed from 6 individual mouse). N.E. = Not expressed.

Fig. 2

CD248 expression increases following unilateral ureteric obstruction. Expression of CD248 was examined by quantitative PCR on whole kidneys (A). Results are expressed as mRNA fold change mean ± SEM relative to the control group (sham operated day 14; 1-way ANOVA with Bonferroni's post test relative to control, * p < 0.01, ** p < 0.001, n = 6 mice per treatment group). B Whole kidney lysates were also used to demonstrate CD248 protein expression at baseline (no operation performed), following a sham operation on day 14 and after UUO for 3, 7 and 14 days. Ca, b Tissue histology of sham-operated kidney tissue showed CD248 was expressed by mesangial cells of the glomerulus (arrow) and by peritubular cells (*). At 14 days following UUO, tubular dilatation and interstitial expansion accompanied increased CD248 expression in the interstitium (*) in a space expected to be occupied by pericytes and stromal fibroblasts (Cc, d). Magnification ×20 (Ca, c) and ×63 (Cb, d). D Bone marrow chimera studies. UUO-injured tissue from day 14 is shown. Yellow fluorescent protein (green) is expressed in the cytoplasm of infiltrating cells many of which are CD45+ leucocytes (red) but not within resident CD248+ stromal cells (blue). Magnification ×40 (representative images from n = 8 chimeric mice generated). Scale bar 50 μm.

Fig. 3

CD248^-/- mice accumulate less interstitial collagen than WT mice following unilateral ureteric obstruction despite no significant difference in leukocyte infiltration. A Picrosirius red staining of sham-operated and UUO day 14 kidney tissue from WT and CD248^-/- mice (×40 magnification). B Digital quantification of Picrosirius red staining expressed as the percentage area of collagen per high power field. Collagen (red) expression is increased in WT and CD248^-/- mice following injury. CD248^-/- mice deposited significantly less collagen after 14 days of UUO (42% reduction, 2-way ANOVA, n = 6, *** p < 0.001, ** p < 0.01). C Col1a1 RNA expression was measured by quantitative PCR on whole kidneys. Results are expressed as mRNA fold change mean ± SEM relative to the control group (sham operated day 14). Col1a1 RNA expression was significantly reduced in CD248^-/- mice at day 14 following injury compared expression in WT mice (2-way ANOVA, n = 6, * p < 0.05, ** p < 0.01). D Representative confocal images of leucocyte infiltration defined by the expression of CD45 (red), nuclear stain (blue) in kidney tissue sections from WT and CD248^-/- mice undergoing UUO. Magnification ×40. Scale bar 50 μm. E Digital quantification of CD45, CD3 and F4/80 expression in kidney tissue sections from WT and CD248^-/- mice following UUO. Data are expressed as mean ± SEM percentage of positive cells in 10 random fields of view per kidney section at a magnification of ×40 in 3 independent experiments (total n = 30 images analyzed). No significant difference in leucocyte infiltration was observed between WT and CD248^-/- mice.

Fig. 3

CD248^-/- mice accumulate less interstitial collagen than WT mice following unilateral ureteric obstruction despite no significant difference in leukocyte infiltration. A Picrosirius red staining of sham-operated and UUO day 14 kidney tissue from WT and CD248^-/- mice (×40 magnification). B Digital quantification of Picrosirius red staining expressed as the percentage area of collagen per high power field. Collagen (red) expression is increased in WT and CD248^-/- mice following injury. CD248^-/- mice deposited significantly less collagen after 14 days of UUO (42% reduction, 2-way ANOVA, n = 6, *** p < 0.001, ** p < 0.01). C Col1a1 RNA expression was measured by quantitative PCR on whole kidneys. Results are expressed as mRNA fold change mean ± SEM relative to the control group (sham operated day 14). Col1a1 RNA expression was significantly reduced in CD248^-/- mice at day 14 following injury compared expression in WT mice (2-way ANOVA, n = 6, * p < 0.05, ** p < 0.01). D Representative confocal images of leucocyte infiltration defined by the expression of CD45 (red), nuclear stain (blue) in kidney tissue sections from WT and CD248^-/- mice undergoing UUO. Magnification ×40. Scale bar 50 μm. E Digital quantification of CD45, CD3 and F4/80 expression in kidney tissue sections from WT and CD248^-/- mice following UUO. Data are expressed as mean ± SEM percentage of positive cells in 10 random fields of view per kidney section at a magnification of ×40 in 3 independent experiments (total n = 30 images analyzed). No significant difference in leucocyte infiltration was observed between WT and CD248^-/- mice.

Fig. 4

CD248^-/- mice have fewer αSMA+ myofibroblasts. Images showing αSMA expression (red) in injured kidney (14 days post UUO) in both WT and CD248^-/- cells. Nuclei blue. Magnification ×630. Scale bar 50 μm. In WT mice, there was a significant increase in the number of αSMA+ cells at 7 and 14 days following injury compared to sham-operated mice (1-way ANOVA and post hoc, n = 6 mice per time point, ** p < 0.001). In contrast, the increase in αSMA+ cells was not observed in CD248^-/- mice.

Fig. 5

Less pericytes are seen in association with endothelium and in the kidney interstitium in CD248^-/- mice following UUO. Representative immunostaining for CD31 (red), CD248 (white), NG2 (green) and nuclei blue (A) from sham-operated mice (Aa), WT mice 7 days (Ab) and 14 days (Ac) post UUO ×630 magnification. White arrows indicate pericytes not closely associated with endothelium. WT (Ad) and CD248^-/- (Ae) kidney tissue 14 days post UUO (×400 magnification). B Quantification of NG2 positive fibroblasts in WT vs. CD248^-/- mice and NG2+ CD248+ pericytes in WT kidney tissue following injury. Data are presented as mean ± SEM for percentage of positive cells per field of view (10 random fields of view at ×400 magnification selected per treatment group in 3 independent experiments, n = 6 mice per treatment group). The percentage of cells expressing NG2 in kidney tissue of both WT and CD248^-/- mice was significantly increased 3 days (* p = 0.05), 7 and 14 days following UUO (Ψ p ≤ 0.01). The percentage of NG2+ cells in WT mice was significantly (** p ≤ 0.01) increased in kidney tissue from WT vs. CD248^-/- mice at 7 and 14 days post UUO. The percentage of NG2 cells attached to endothelium (defined as co-localization with CD31) was also quantified. In WT mice the percentage of NG2+ CD248+ were significantly increased (** p ≤ 0.01) 7 and 14 days post injury compared to sham-operated mice. There was a significant increase in the number of pericytes (NG2-expressing cells) attached to the endothelium at 7 and 14 days post UUO in CD248^-/- mice compared to WT mice (* p ≤ 0.01, ** p ≤ 0.05).

Fig. 6

Microvascular rarefaction is reduced in CD248^-/- mice. A and B CD31+ staining in UUO kidneys from WT and CD248^-/- mice. Vessel density was assessed at various time points following UUO in sham-operated, WT and CD248^-/- mice (1-way ANOVA, n = 6 mice per group at each time point repeated in 3 independent experiments, * p < 0.05, *** p < 0.001 compared to control). C and D The proliferative capacity and ability to deposit collagen of stromal cells (fibroblasts and mesangial cells) and epithelial cells from WT and CD248^-/- animals were assessed. Stromal cells but not epithelial cells from CD248^-/- mice demonstrated a reduced capacity to proliferate and deposit collagen in vitro (t test, n = 3, * p < 0.05, *** p < 0.01).