The direction and role of phenotypic transition between podocytes and parietal epithelial cells in focal segmental glomerulosclerosis

Abstract

Focal segmental glomerulosclerosis (FSGS) is a podocyte disease. Among the various histologies of FSGS, active epithelial changes, hyperplasia, as typically seen in the collapsing variant, indicates disease progression. Using a podocyte-specific injury model of FSGS carrying a genetic podocyte tag combined with double immunostaining by different sets of podocytes and parietal epithelial cell (PEC) markers [nestin/Pax8, Wilms' tumor-1 (WT1)/claudin1, and podocalyxin/Pax2], we investigated the direction of epithelial phenotypic transition and its role in FSGS. FSGS mice showed progressive proteinuria and renal dysfunction often accompanied by epithelial hyperplasia, wherein 5-bromo-4-chloro-3-indoyl β-d-galactoside (X-gal)-positive podocyte-tagged cells were markedly decreased. The average numbers of double-positive cells in all sets of markers were significantly increased in the FSGS mice compared with the controls. In addition, the average numbers of double-positive cells for X-gal/Pax8, nestin/Pax8 and podocalyxin/Pax2 staining in the FSGS mice were comparable, whereas those of WT1/claudin1 were significantly increased. When we divided glomeruli from FSGS mice into those with FSGS lesions and those without, double-positive cells tended to be more closely associated with glomeruli without FSGS lesions compared with those with FSGS lesions. Moreover, the majority of double-positive cells appeared to be isolated and very rarely associated with FSGS lesions (1/1,997 glomeruli). This study is the first to show the incidence and localization of epithelial cells with phenotypical changes in FSGS using a genetic tag. The results suggest that the major direction of epithelial phenotypic transition in cellular FSGS is from podocytes to PECs and that these cells were less represented in the active lesions of FSGS.

Keywords: cellular FSGS; focal segmental glomerulosclerosis; parietal epithelial cell; phenotypic transition; podocyte.

Fig. 1.

A: proteinuria in NEP25 focal segmental glomerulosclerosis (FSGS) mice and control mice. In the FSGS mice, no overt proteinuria was observed until day 3, and the urinary protein/creatinine ratio showed a significant increase on day 5 and peaked on day 12, compared with controls. B: serum creatinine and blood urea nitrogen (BUN) in FSGS mice and controls. Serum creatinine in FSGS mice was significantly increased on day 12 compared with controls. C: Wilms' tumor (WT1)-positive cells in control mice and FSGS mice. The numbers of WT1-positive cells per glomerulus were 6.94 ± 0.44 in the control mice and 1.32 ± 0.17 in the FSGS mice (*P < 0.001).

Fig. 2.

Representative glomerular profiles in control (A–D) and FSGS (E–H) mice. Periodic acid-methenamine silver (PAM) staining showed marked glomerular collapse with extracapillary hypercellularity resembling collapsing FSGS pathology in FSGS mice (E). Genetic tags visualized by 5-bromo-4-chloro-3-indoyl β-d-galactoside (X-gal) staining (blue) showed limited localization on podocytes in controls and parietal epithelial cells (PECs) were entirely negative for X-gal staining (B). In the FSGS mice, blue tagging was almost absent in sclerotic lesions (F, arrow). Glomerular profiles are shown for double staining for X-gal and Pax8 (C, D, G, and H). In the controls, Pax8 was restricted to the PEC nuclei, and X-gal was restricted in the cell cytoplasm of podocytes. Double-positive cells occasionally existed along Bowman's basement membrane (C and D, arrow). In the FSGS mice, isolated X-gal- and Pax8-coexpressing cells were found (G, arrow), and hypercellular lesions were composed almost entirely of Pax8-positive cells. No X-gal-coexpressing cells were found (H).

Fig. 3.

Double immunostaining in control (A–D) and FSGS (E–H) mice. WT1 staining (brown) showed well-preserved podocytes in the controls (A), whereas WT1-positive cells were significantly reduced in the FSGS mice but occasionally expressed within hyperplastic lesions (E, arrow). WT1/claudin1 immunostaining showed good discrimination between podocytes and PECs in the controls (B), and double-positive cells were found on Bowman's capsule and tuft (F, arrow). Nestin/Pax8 immunostaining in the controls (C) showed distinct localizations of each cell type. In the FSGS mice (G), hyperplastic lesions were positive for only Pax8, and isolated double-positive cells were observed on the tuft (G, arrow). Podocalyxin/Pax2 immunostaining showed well-identified localization in the controls (D). In the FSGS mice, isolated double-positive cells were found on the tuft (H, arrow).

Fig. 4.

Incidence of double-positive cells with different sets of markers per glomerular profile. The numbers of double-positive cells for X-gal/Pax8, WT1/claudin1, nestin/Pax8, and podocalyxin/Pax2 in controls were comparable, and they all were significantly increased in FSGS mice. In FSGS, average numbers of double-positive cells among X-gal/Pax8, nestin/Pax8, and podocalyxin/Pax2 were comparable, whereas those of WT1/claudin1 were significantly increased.