Mesenchymal stem cells, used as bait, disclose tissue binding sites: a tool in the search for the niche?

Abstract

We developed an ex vivo approach characterizing renal mesenchymal stem cell (MSC) adhesion to kidney sections. Specificity of MSC adhesion was confirmed by demonstrating a) 3T3 cells displayed 10-fold lower adhesion, and b) MSC adhesion was CXCR4/stromal-derived factor-1 (SDF-1)-dependent. MSC adhesion was asymmetrical, with postischemic sections exhibiting more than twofold higher adhesion than controls, and showed preference to perivascular areas. Pretreating kidney sections with cyclic arginine-glycine-aspartic acid peptide resulted in increased MSC adhesion (by displacing resident cells), whereas blockade of CXCR4 with AMD3100 and inhibition of alpha4beta1(VLA4) integrin or vascular cellular adhesion molecule-1, reduced adhesion. The difference between adhered cells under cyclic arginine-glycine-aspartic acid peptide-treated and control conditions reflected prior occupancy of binding sites with endogenous cells. The AMD3100-inhibitable fraction of adhesion reflected CXCR4-dependent adhesion, whereas maximal adhesion was interpreted as kidney MSC-lodging capacity. MSC obtained from mice overexpressing caveolin-1 exhibited more robust adhesion than those obtained from knockout animals, consistent with CXCR4 dimerization in caveolae. These data demonstrate a) CXCR4/SDF-1-dependent adhesion increases in ischemia; b) CXCR4/SDF-1 activation is dependent on MSC surface caveolin-1; and c) occupancy of MSC binding sites is decreased, while d) capacity of MSC binding sites is expanded in postischemic kidneys. In conclusion, we developed a cell-bait strategy to unmask renal stem cell binding sites, which may potentially shed light on the MSC niche(s) and its characteristics.

Figure 1

A gallery of images depicting MSC-4E cell adhesion to renal sections. A–B: Panoramic views of MSC-4E cell adhesion, stained red with CellTracker CM-Dil, to kidney sections; original magnification ×40. Arrows indicate sites of fluorescently tagged MSC-4E cell adhesion. C–E: MSC-4E cells, stained red with CellTracker CM-DiI, adhered to the cortex, cortico-medullary junction, and medulla, respectively; magnification ×10. F–H: MSC-4E cell adhesion to postischemic sections obtained from Tie-2/GFP mice. Note the proximity of adhesion to the vascular tree (F, G) and glomeruli (F, H); original magnification ×40.

Figure 2

Dynamics of stem/progenitor cell adhesion to postischemic versus control kidney sections. A: Global adhesion of primary and cloned (4E cells) MSC, EPC, and 3T3 fibroblasts to control and ischemic kidney sections on days [D] 1, 3, 5, and 7 postischemia. Data are presented as total cell adhesion to cortex, medulla, and capsule. *P < 0.05 vs. control sections; n = 6. B: Percent increase in MSC-4E cell adhesion to postischemic kidney (as compared with contralateral kidney) cortex, medulla, and renal capsule (solid line) and the percent increase in adhesion after cRGD treatment (hashed line). The attenuation of MSC-4E cell adhesion to postischemic sections on treatment with the CXCR4 antagonist AMD3100 (dotted line). *P < 0.05 vs. 3D, 5D, and 7D; **P < 0.05 vs. 1D and 7D; n = 6.

Figure 3

Adhesion of 3T3 fibroblasts to control and postischemic kidney sections. Adhesion data were obtained in the cortex (A) and medulla (B) under basal conditions and after treatment with either AMD3100 or cRGD (for comparison with MSC see Figure 2). No statistical difference between groups; n = 6.

Figure 4

Co-staining of adhered MSC-4E cells with CD29 and CD44. On completion of adhesion assay, sections were fixed and stained with CD29 conjugated to phycoerythrin (CellTracker green labeled cells) or CD44 conjugated to FITC (CellTracker red labeled cells). Note that adhered cells expressed both markers of MSC. Original magnification, ×40.

Figure 5

Co-localization of SDF-1 expression with adhered MSC-4E cells. Expression of SDF-1 is conspicuous in the ischemic sections and MSC-4E cells tend to co-localize with SDF-1. In contrast, control sections faintly express SDF-1 and MSC-4E cell lodging is not confined to it. Original magnification: top panels (×10), lower panels (×40).

Figure 6

Adhesion assay using MSC obtained from caveolin-1 knockout, transgenic, and wild-type mice. MSC obtained and cloned from transgenic mice overexpressing caveolin-1 demonstrated potentiated adhesion to all regions of the kidney after ischemic insult. MSC void of caveolin-1 demonstrated enhanced adhesion to control sections and no increase in adhesion following ischemic insult. A: cortex; B: medulla; C: capsule. *P < 0.05 vs. control; **P < 0.05 vs. ischemia (wild-type & KO MSC); ***P < 0.05 vs. control (wild-type & TG MSC); n = 6.

Figure 7

Cyclic RGD peptide induces dislodgement of Sca-1 GFP-positive cells from kidney sections. CFU assay (A) of cells displaced from kidney sections by cRGD, as compared with kidney sections that were not treated with cRGD. Displaced cells that formed colonies demonstrated potential to differentiate into (B) osteocytes and adipocytes. After being subject to differentiating medium, displaced MSC were positive for osteogenic staining including alkaline phosphatase staining (left panel) and von Kossa staining (middle panel), and for adipocytic staining with Oil-Red-O (right panel). Images are viewed under ×4 magnification. Quantitative fluorescence intensity of individual samples (C) and average change (D) of Sca-1 GFP after application of cRGD. Decrease in arbitrary fluorescence is indicative of the release of Sca-1 GFP positive cells from kidney sections. *P < 0.05 vs. no treatment (A) or baseline intensity (D).

Figure 8

Effect of neutralizing anti-VLA4, anti-VCAM-1, and anti-N-cadherin antibodies on MSC adhesion. Treatment with neutralizing antibodies against VLA4 and VCAM-1 (antibodies applied either separately or together) resulted in attenuation of MSC-4E cell adhesion to all kidney regions and statistically negated the increase in adhesion on ischemic insult in the cortex and medulla. In contrast, application of neutralizing antibodies against N-cadherin had no effect on MSC-4E adhesion. A: cortex; B: medulla; C: capsule, *P < 0.05 vs. control; **P < 0.05 vs. anti-VLA-4, anti-VCAM, and anti-VLA-4 + anti-N-Cad/anti-VCAM; n = 6.

Figure 9

CXCR4/SDF-1-dependent MSC-4E cell adhesion. Chemokine-dependent adhesion to control and postischemia (1, 3, 5, and 7 days[D]) kidney sections, determined as an AMD3100-inhibitable fraction of basal MSC-4E cell adhesion. CXCR4/SDF-1 dependent MSC-4E cell adhesion was enhanced 1 and 3 days after ischemic insult. A: cortex; B: medulla; C: capsule. *P < 0.05 vs. 7D ischemia; **P < 0.05 vs. 5D ischemia; ***P < 0.05 vs. 1D ischemia; ^†P < 0.05 vs. 1D control; n = 6.

Figure 10

Integrin-dependent MSC-4E cell adhesion. Integrin-dependent adhesion to control and postischemia (1, 3, 5, and 7 days [D]) kidney sections, determined by the numerical value of cRGD-displaced cells expressed as the difference between cRGD-pretreated and basal adhesion. Integrin-dependent adhesion progressively decreased with time after ischemic insult in the cortex and medulla. A: cortex; B: medulla; C: renal capsule. *P < 0.05 vs. 7D ischemia; **P < 0.05 vs. 1D control; n = 6.