Bone marrow-derived cells contribute to glomerular cell turnover and repair, but the cell types involved are unknown. Whether rat mesenchymal stem cells (MSC) can accelerate recovery from damage in rat mesangioproliferative anti-Thy1.1 glomerulonephritis was studied. After injection into the left renal artery on day 2 after disease induction, fluorescently labeled MSC were detected in 20 to 50% of glomeruli and rare intrarenal vessels but not in the tubulointerstitium, in contralateral kidneys, or in medium controls. In control experiments, injected mesangial cells were detected less frequently in glomeruli in comparison with injected MSC. In nephritic outbred Wistar rats, MSC injection led to an approximately 50% reduction of mesangiolysis on days 4 and 6 after disease induction, accompanied by three-to four-fold higher intraglomerular cell proliferation on day 4 and more rapid mesangial reconstitution as detected by ␣-smooth muscle actin expression. Injection of MSC into tail veins or intra-arterial injection of mesangial cells instead of MSC failed to reproduce any of these findings. In inbred Lewis rats, anti-Thy1.1 nephritis followed an aggravated course with transient acute renal failure. Acute renal failure was ameliorated by MSC injection into the left renal artery on day 2 after disease induction. Again, MSC led to more rapid recovery from mesangiolysis, increased glomerular cell proliferation, and reduction of proteinuria by 28%. Double immunostaining of 5-bromo-2-deoxyuridine-labeled MSC for endothelial, mesangial, or monocyte/macrophage antigens showed that 85 to 95% of MSC that localized in glomeruli on day 6 failed to express these markers. In vitro, MSC secreted high amounts of vascular endothelial growth factor and TGF-␤1 but not PDGF-BB. In conclusion, even low numbers of MSC can markedly accelerate glomerular recovery from mesangiolytic damage possibly related to paracrine growth factor release and not to differentiation into resident glomerular cell types or monocytes/macrophages.

Podocyte loss contributes to the development of glomerulosclerosis. Although podocyte detachment has been recognized as a new mechanism of podocyte loss in glomerular diseases, its time course and relationship to disease activity are not known. Urinary excretion of viable podocytes was quantified in two models of transient glomerular injury, i.e., rats with puromycin aminonucleoside-induced nephrosis (PAN) and mesangioproliferative nephropathy (anti-Thy 1.1 nephritis model), as well as in a model of continuous glomerular injury, i.e., hypertensive nephropathy (5/6-nephrectomy model), and in aging rats. The number of glomerular Wilm's tumor (WT)-1-positive podocytes and the glomerular expression of cell-cycle proteins in vivo were assessed. Urinary podocyte loss occurred in both primary (PAN) and secondary (anti-Thy 1.1 nephritis) in parallel to the onset of proteinuria. However, subsequently proteinuria persisted despite remission of podocyturia. In continuous glomerular injury, i.e., after 5/6-nephrectomy, podocyturia paralleled the course of proteinuria and of systemic hypertension, whereas no podocyturia became detectable during normal aging (up to 12 mo). Despite podocyte detachment of varying degrees, no decrease in glomerular podocyte counts (i.e., WT-1 positive nuclei) was noted in either disease model. Podocyturia in the PAN and anti-Thy 1.1 nephritis model was preceded by entry of glomerular podocytes into the cell cycle, i.e., cyclin D1, cdc2, and/or proliferating cell nuclear antigen (PCNA) expression. Podocyturia is a widespread phenomenon in glomerular disease and not simply a reflection of proteinuria because it is limited to phases of ongoing glomerular injury. The data suggest that podocyturia may become a more sensitive means to assess the activity of glomerular damage than proteinuria.