Podocyte Regeneration Driven by Renal Progenitors Determines Glomerular Disease Remission and Can Be Pharmacologically Enhanced

Abstract

Podocyte loss is a general mechanism of glomerular dysfunction that initiates and drives the progression of chronic kidney disease, which affects 10% of the world population. Here, we evaluate whether the regenerative response to podocyte injury influences chronic kidney disease outcome. In models of focal segmental glomerulosclerosis performed in inducible transgenic mice where podocytes are tagged, remission or progression of disease was determined by the amount of regenerated podocytes. When the same model was established in inducible transgenic mice where renal progenitors are tagged, the disease remitted if renal progenitors successfully differentiated into podocytes, while it persisted if differentiation was ineffective, resulting in glomerulosclerosis. Treatment with BIO, a GSK3s inhibitor, significantly increased disease remission by enhancing renal progenitor sensitivity to the differentiation effect of endogenous retinoic acid. These results establish renal progenitors as critical determinants of glomerular disease outcome and a pharmacological enhancement of their differentiation as a possible therapeutic strategy.

Graphical abstract

Figure 1

Podocyte Regeneration Occurs after Podocyte Detachment and Is Not Driven by Mitosis of Pre-existing Podocytes (A) Experimental scheme. (B) Healthy kidney from a NPHS2.iCreERT2;mT/mG mouse stained with anti-SYN (blue). (C and D) Low (top) and high (bottom) magnification of a renal biopsy specimen from a doxorubicin-treated NPHS2.iCreERT2;mT/mG mouse at day 14 (C) and the kidney of the same mouse at day 28 (D) stained with anti-SYN (blue). (E) Number of pre-existing podocytes (detected as green and blue cells, GFP+/SYN+) (top) and number of newly generated podocytes (detected as red and blue cells, TomRed+/SYN+) (bottom) per glomerular section of all glomeruli in the biopsy specimen (day 14) and kidney (day 28) (n = 8) of each mouse; ^∗p < 0.05 (day 28) by Wilcoxon test. (F) Healthy kidney from a NPHS2.iCreERT2;mT/mG mouse stained with anti-WT1 (blue). (G and H) Low (top) and high (bottom) magnification of renal biopsy from a doxorubicin-treated NPHS2.iCreERT2;mT/mG mouse at day 14 (G) and kidney of the same mouse at day 28 (H) stained with anti-WT1 (blue). (I) Number of pre-existing podocytes (detected as green and blue cells, GFP+/WT1+) (top) and number of newly generated podocytes (detected as red and blue cells, TomRed+/WT1+) (bottom) per glomerular section of all glomeruli in the biopsy specimen (day 14) and kidney (day 28) (n = 8) of each mouse; ^∗p < 0.05 (day 28) by Wilcoxon test. Data represent the mean ± SEM; DAPI counterstains nuclei. Scale bar, 20 μm.

Figure 2

Remission of Glomerular Disease in Mice with Adriamycin Nephropathy Is Associated with the Generation of Novel Podocytes (A) Albumin/creatinine levels in persistent proteinuric NPHS2.iCreERT2;mT/mG mice (n = 10) and mice undergoing remission (n = 5). Day 0 corresponds to the time of second doxorubicin injection; ^∗p < 0.05 (day 28) by Mann-Whitney test. (B) Blood urea nitrogen levels in healthy (n = 4), progressed (n = 10), and remitted (n = 5) mice at day 28; ^∗p < 0.05 versus healthy mice and mice with proteinuria remission by Mann-Whitney test. (C and D) Representative kidney of a NPHS2.iCreERT2;mT/mG mouse with persistent proteinuria (C) and of a NPHS2.iCreERT2;mT/mG mouse with proteinuria remission (D) stained with anti-SYN (blue). (E) Number of pre-existing podocytes (detected as green and blue cells, GFP+/SYN+) per glomerular section of all glomeruli in healthy (n = 4) and persistently proteinuric mice (n = 10) and in mice with proteinuria remission (n = 5) at day 28. ^∗p < 0.05 versus healthy mice by Mann-Whitney test. (F) Number of newly generated podocytes (detected as red and blue cells, TomRed+/SYN+) per glomerular section of all glomeruli in healthy (n = 4) and persistently proteinuric mice (n = 10) and in mice with proteinuria remission (n = 5) at day 28. ^∗p < 0.05 versus healthy and persistently proteinuric mice by Mann-Whitney test. (G) Percentage of regenerated podocytes over lost podocytes (n = 10 for proteinuric mice and n = 5 for mice with proteinuria remission). (H and I) Representative kidney of a NPHS2.iCreERT2;mT/mG mouse with persistent proteinuria (H) and of a NPHS2.iCreERT2;mT/mG mouse with proteinuria remission (I) stained with anti-WT1 (blue). (J) Percentage of regenerated podocytes over lost podocytes (n = 10 for proteinuric mice and n = 5 for mice with proteinuria remission). Data represent the mean ± SEM; DAPI counterstains nuclei. Scale bar, 20 μm.

Figure 3

PAX2+ Cells Located in Parietal Epithelium of the Bowman’s Capsule Are Progenitors for Podocytes during Postnatal Kidney Growth (A) Experimental scheme for results shown in (B) and (C). (B) Kidney of Pax2.rtTA;TetO.Cre;R26.Confetti mice induced at P5 for 10 days and sacrificed at 2 weeks of age (n = 8). Multicolor Confetti tracing is shown in red, yellow, and cyan. PAX2+ progenitors are located in parietal epithelium of the Bowman’s capsule and in tubular structures. (C) Higher magnification of a glomerulus showing PAX2+ cells only in parietal epithelium of the Bowman’s capsule. (D) Experimental scheme for results shown in E and F. (E) Kidney of Pax2.rtTA;TetO.Cre;R26.Confetti mice induced at P5 for 10 days when doxycycline was suspended and mice tracked until 5 weeks of age (n = 8). Fluorescent cells were present in parietal epithelium of the Bowman’s capsule as well as inside the glomeruli. (F) Higher magnification of a glomerulus showing fluorescently labeled PAX2+ cells inside the tuft, positively stained with anti-SYN antibody (light blue), demonstrating their differentiation into podocytes. (G) Experimental scheme for results shown in (H). (H) Juxtaposed confocal images from top (outer cortex) to bottom (inner cortex) of an adult Pax2.rtTA;TetO.Cre;R26.Confetti mouse kidney. (I) Experimental scheme for results shown in (J) and (K). (J) Glomeruli with PAX2+ capsule in mice induced at 5 weeks of age for 10 days, then doxycycline was suspended and mice tracked for the following 9 months (n = 6). (K) Higher magnification of a glomerulus with PAX2+ cells in the Bowman’s capsule. G, glomerulus. DAPI counterstains nuclei (white). Scale bar, 20 μm. See also Figure S1.

Figure 4

PAX2+ Progenitors Generate Lesions in Mice with Persistent Proteinuria and Podocytes in Animals with Proteinuria Remission (A) Experimental scheme. (B) Albumin/creatinine levels in persistent proteinuric Pax2.rtTA;TetO.Cre;mT/mG mice (n = 11) and mice undergoing remission (n = 5). Day 0 corresponds to the time of second doxorubicin injection; ^∗p < 0.05 (day 28) by Mann-Whitney test. (C) Percentage of glomeruli with sclerosis in mice with persistent proteinuria (n = 11) and mice undergoing remission (n = 5). ^∗p < 0.05 versus persistently proteinuric mice (Mann-Whitney test). (D) Representative periodic acid-Schiff staining of renal sections of persistently proteinuric mice (left) and of mice undergoing remission (right). (E and F) Representative glomeruli with PAX2+ capsule in healthy Pax2.rtTA;TetO.Cre;mT/mG mice induced at 5 weeks of age (n = 4). (G and H) Absence of basal Cre recombinase activity in the kidney of Pax2.rtTA;TetO.Cre;mT/mG healthy mice maintained with water without doxycycline and analyzed at 12 weeks of age (n = 3) (G) and in the kidney of Pax2.rtTA;TetO.Cre;mT/mG mice with Adriamycin nephropathy (n = 3) maintained with water without doxycycline and sacrificed 28 days after the second doxorubicin injection (H). (I) Representative kidney from a Pax2.rtTA;TetO.Cre;mT/mG mouse with persistent proteinuria. No green cells are present inside the glomeruli. (J) Representative kidney from a Pax2.rtTA;TetO.Cre;mT/mG mouse with persistently nephrotic proteinuria, showing PAX2+ cells in a hyperplastic lesion. (J′) Higher magnification of (J). (K) Representative kidney from a Pax2.rtTA;TetO.Cre;mT/mG mouse with proteinuria remission, where PAX2+ cells are observed inside the glomerular tuft. (K′) Higher magnification of (K). (L) 3D reconstruction of a glomerulus showing PAX2+ cells in epithelium of the Bowman’s capsule and PAX2-derived cells inside the glomerulus. Data represent the mean ± SEM. DAPI counterstains nuclei (white). Scale bar, 20 μm. See also Figure S2 and Movie S1.

Figure 5

PAX2+ Progenitors Generate Fully Differentiated Podocytes in Mice with Proteinuria Remission (A and B) In Pax2.rtTA;TetO.Cre;mT/mG mouse with proteinuria remission, PAX2-derived cells observed inside the glomeruli expressed the podocyte marker podocin (NPHS2, blue) (A) and SYN (blue) (B). (B′) Split images of the PAX2+/SYN+ cell inside the glomerulus shown in (B). (C) Higher magnification of foot processes in the PAX2+/SYN+ cell shown in (B). (D and E) Higher magnification of PAX2+/SYN+ cells with foot processes. (F) Representative image of a PAX2-derived cell inside the glomerulus expressing the podocyte marker WT1 (blue). (G) Left. Quantification of PAX2+SYN+ cells/glomerular section of all glomeruli in mice with persistent proteinuria (n = 11) and mice undergoing remission (n = 5). ^∗p < 0.05 versus persistently proteinuric mice (Mann-Whitney test). Right: percentage of glomeruli with PAX2-derived podocytes inside the glomerular tuft (top) and quantification of PAX2-derived podocytes/glomerular section of positive glomeruli (bottom). (H) Percentage of PAX2-derived podocyte over lost podocytes in animals with persistent proteinuria (n = 11) and mice with proteinuria remission (n = 5). Data represent the mean ± SEM. DAPI counterstains nuclei. Scale bar represents 20 μm, except in (D) and (E) (10 μm).

Figure 6

GSK3s Inhibition Increases Podocyte Regeneration from PAX2+ Progenitors by Enhancing RA Transcriptional Activity (A) Effect of BIO treatment on nephrin mRNA (left) and protein (right) expression in hRPCs. Results are expressed as mean ± SEM obtained in at least four separate experiments. ^∗p < 0.05 by Mann-Whitney test. (B) Representative cell cycle analysis of untreated- (left) and BIO-treated (right) hRPC. One representative experiment out of four independent experiments is shown. (C) Effect of BIO on RAR/RXR mRNA expression in hRPC. Results are expressed as mean ± SEM obtained in at least four separate experiments. (D) Effect of BIO treatment on the transcriptional activity of RARE-reporter plasmid in hRPCs. Results are expressed as mean ± SEM obtained in at least four separate experiments ^∗p < 0.05 by Mann-Whitney test. (E) RARE activity in the glomeruli of RARE-LacZ mice with Adriamycin nephropathy treated with DMSO (left) (n = 6) or BIO (right) (n = 6) (β-gal, blue; claudin1, red). (F) Number of β-Gal+/claudin1+ cells lining the Bowman’s capsule or within the glomerular tuft per glomerular section of all glomeruli in DMSO-treated (n = 6) and BIO-treated mice (n = 6). ^∗p < 0.05 by Mann-Whitney test. Data represent the mean ± SEM. Scale bar, 20 μm.

Figure 7

GSK3s Inhibition Induces Remission of Glomerular Disease and Is Associated with the Generation of Novel Podocytes (A) Albumin/creatinine levels in DMSO- (n = 8) and BIO-treated (n = 8) NPHS2.iCreERT2;mT/mG mice with Adriamycin nephropathy; day 0 corresponds to the time of second doxorubicin injection. ^∗∗p < 0.001 by Mann-Whitney test (day 35). (B) Kidney of DMSO- (left) and BIO-treated (right) NPHS2.iCreERT2;mT/mG mice stained with anti-SYN (blue). (C) Number of newly generated podocytes per glomerular section of all glomeruli in healthy (n = 4), DMSO-treated (n = 8), and BIO-treated (n = 8) mice. ^∗p < 0.05 versus healthy and DMSO-treated mice by Mann-Whitney test. (D) Percentage of regenerated podocytes over lost podocytes (n = 8 for DMSO-treated and n = 8 for BIO-treated mice). (E) Albumin/creatinine levels in DMSO- (n = 8) and BIO-treated (n = 8) Pax2.rtTA;TetO.Cre;mT/mG mice with Adriamycin nephropathy; day 0 corresponds to the time of second doxorubicin injection. ^∗∗p < 0.001 (day 35) by Mann-Whitney test. (F) Blood urea nitrogen levels in DMSO- (n = 8), and BIO- (n = 8) treated mice at day 28; ^∗p < 0.05 versus BIO-treated mice by Mann-Whitney test. (G) Representative periodic acid-Schiff staining of renal sections of DMSO- (left) and BIO-treated (right) mice. (H) Percentage of glomeruli with sclerosis in DMSO- (n = 8) and BIO-treated (n = 8) mice. ^∗p < 0.05 versus DMSO-treated mice, by Mann-Whitney test. (I) Top: absence of PAX2+ (green) cells in the majority of glomeruli of DMSO-treated Pax2.rtTA;TetO.Cre;mT/mG mice (left) and rare PAX2+ (green) cells expressing SYN (blue) inside the glomeruli in DMSO-treated Pax2.rtTA;TetO.Cre;mT/mG mice (right). Bottom: a higher number of PAX2+ cells inside the glomeruli in BIO-treated Pax2.rtTA;TetO.Cre;mT/mG mouse (left) and PAX2+ cells inside the glomeruli express SYN (blue) (right). (J) Quantification of PAX2+/SYN+ cells per glomerular section of all glomeruli in healthy (n = 4), DMSO-treated (n = 8), and BIO-treated (n = 8) mice. ^∗p < 0.05 versus healthy and DMSO-treated mice. (K) Percentage of PAX2-derived podocytes over lost podocytes (n = 8 for DMSO- and n = 8 for BIO-treated mice). Data represent the mean ± SEM. DAPI counterstains nuclei. Scale bar, 20 μm.