Sphingomyelinase-like phosphodiesterase 3b expression levels determine podocyte injury phenotypes in glomerular disease

Abstract

Diabetic kidney disease (DKD) is the most common cause of ESRD in the United States. Podocyte injury is an important feature of DKD that is likely to be caused by circulating factors other than glucose. Soluble urokinase plasminogen activator receptor (suPAR) is a circulating factor found to be elevated in the serum of patients with FSGS and causes podocyte αVβ3 integrin-dependent migration in vitro. Furthermore, αVβ3 integrin activation occurs in association with decreased podocyte-specific expression of acid sphingomyelinase-like phosphodiesterase 3b (SMPDL3b) in kidney biopsy specimens from patients with FSGS. However, whether suPAR-dependent αVβ3 integrin activation occurs in diseases other than FSGS and whether there is a direct link between circulating suPAR levels and SMPDL3b expression in podocytes remain to be established. Our data indicate that serum suPAR levels are also elevated in patients with DKD. However, unlike in FSGS, SMPDL3b expression was increased in glomeruli from patients with DKD and DKD sera-treated human podocytes, where it prevented αVβ3 integrin activation by its interaction with suPAR and led to increased RhoA activity, rendering podocytes more susceptible to apoptosis. In vivo, inhibition of acid sphingomyelinase reduced proteinuria in experimental DKD but not FSGS, indicating that SMPDL3b expression levels determined the podocyte injury phenotype. These observations suggest that SMPDL3b may be an important modulator of podocyte function by shifting suPAR-mediated podocyte injury from a migratory phenotype to an apoptotic phenotype and that it represents a novel therapeutic glomerular disease target.

Keywords: FSGS; diabetic nephropathy; glomerular disease; podocyte; proteinuria.

Figure 1.

SMPDL3b expression in FSGS and DKD. (A) Transcriptional analysis of glomerular SMPDL3b expression in 70 patients with DKD and 18 patients with FSGS compared with 32 living donors. Glomerular SMPDL3b mRNA expression is significantly increased in diabetic glomeruli compared with normal controls but showing a trend to be decreased in glomeruli from FSGS patients compared with controls. Numbers reflect fold change in disease compared with living donors. *q<0.05. (B) SMPDL3b protein expression is significantly decreased in podocytes treated with high-risk FSGS patient sera (*P<0.05) and increased in podocytes treated with DKD+ sera (*P<0.05) compared with podocytes treated with the sera from healthy controls (NHS). SMPDL3b protein expression is also significantly increased in podocytes treated with DKD+ sera compared with podocytes treated with high-risk FSGS patient sera (^#P<0.05). GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (C) Immunofluorescence staining using phalloidin (red) indicates a loss of stress fibers in podocytes treated with high-risk FSGS sera. SMPDL3b OE podocytes are protected from high-risk FSGS sera-induced actin cytoskeleton rearrangement. (D) Treatment of podocytes with DKD+ sera induces cell blebbing in normal but not in SMPDL3b KD podocytes. (E) Apoptosis in DKD− and DKD+ sera-treated podocytes is significantly increased compared with podocytes treated with NHS. SMPDL3b KD podocytes are protected from DKD+ sera-induced podocyte apoptosis (**P<0.01).

Figure 2.

Serum suPAR in patients with and mouse models of FSGS and DKD. (A) Serum suPAR levels in 14 high-risk FSGS patients are significantly higher compared with low-risk FSGS patients (*P<0.05). (B) Serum suPAR levels are significantly increased in 34 microalbuminuric and 10 macroalbuminuric DKD patients compared with 30 patients with diabetes but normoalbuminuria (**P<0.01). Serum suPAR levels are further significantly increased in 10 macroalbuminuric compared with 34 microalbuminuric DKD patients (**P<0.01). (C) Serum suPAR levels significantly correlate with the renal disease progression in patients enrolled in the FSGS clinical trial cohort (P<0.01). (D) Longitudinal analysis in patients with diabetes showing that suPAR levels at baseline are not significantly different in the nonprogression group (normo-normo) compared with the progression group (normo-micro) but are significantly elevated in the progression group 7 years later (^##P<0.01). (E) Longitudinal analysis in patients with diabetes shows that suPAR levels at baseline are not significantly different in the nonprogression group (micro-micro) compared with the progression group (micro-macro) but are significantly elevated in the progression group 7 years later (^##P<0.01) and significantly increased compared with the nonprogression group (**P<0.01). (F) Serum suPAR levels significantly correlate with the renal disease progression in diabetic patients (P<0.01). (G and H) suPAR levels in a mouse model of FSGS (ADR-induced nephropathy) and a mouse model of DKD (db/db mice) are significantly increased compared with their respective controls (*P<0.05 ADR versus control; *P<0.05 db/db versus db/m).

Figure 3.

In vitro effects of suPAR treatment of human podocytes on SMPDL3b expression levels and β₃ integrin activation. (A) Treatment of human podocytes with different concentrations of suPAR (0, 0.5, 1, and 2 μg/ml) for different time spans (3, 6, and 24 hours) indicates that suPAR concentrations and time of exposure to suPAR do not change SMPDL3b mRNA expression levels. (B) Western blot analysis followed by (C) quantitative analysis of normal human podocytes treated with different concentrations of suPAR (0, 0.5, and 1 μg/ml) for 6 and 24 hours reveal significant β₃ integrin activation at 24 hours of treatment (*P<0.05). AP5 antibody was used to detect activated β₃ integrin and AP3 antibody was used to detect total β₃ integrin. (D) Normal human podocytes treated with the sera from patients with DKD (top panel) show a characteristic cytoskeletal remodeling phenotype in the form of cell blebbing. The same cytoskeletal phenotype was observed when SMPDL3b OE podocytes were treated with suPAR (bottom panel) but not when human podocytes expressing endogenous levels of SMPDL3b (middle panel) were exposed to suPAR. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Figure 4.

Analysis of mouse models for FSGS and DKD. (A) Immunofluorescence staining with AP5 to detect activated β₃ integrin (green) and synaptopodin (red) of glomeruli from db/db and ADR-treated mice. Original Magnification, ×40. (B) Bar graph analysis showing a significant increase of AP5 staining in podocytes and mesangial cells after treatment of the mice with ADR (**P<0.01) compared with vehicle-treated control mice. However, AP5 staining intensity does not change in db/db mice compared with db/m mice. MFI, mean fluorescence intensity. (C and D) Real-time PCR analysis reveals that Smpdl3b mRNA levels in total kidney cortex are not changed in (C) db/db and (D) ADR-treated mice compared with their respective controls. GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (E) Analysis of the SM content in kidney cortexes shows no differences in db/db and ADR-treated mice compared with their respective controls. w.w., wet weight. (F) Analysis of the ceramide content of kidney cortexes shows significantly decreased ceramide levels in db/db and ADR-treated mice compared with their respective controls. *P<0.05.

Figure 5.

Interaction of suPAR and SMPDL3b. (A) Co-IP experiments performed in HEK293 cells showing an interaction between GFP-SMPDL3b and FLAG-uPAR (F-uPAR; upper panel) but not between GFP-SMPDL3b and FLAG-β₃ integrin (F-β₃; lower panel). FLAG-empty vector (F-C) was used as a negative control (upper panel). (B) Endogenous immunoprecipitation showing interaction of SMPDL3b and suPAR in glomeruli isolated from mice injected with PBS or LPS. E1, eluate from PBS-injected mice; E2, eluate from LPS-injected mice; I1, input (glomerular lysate) from PBS-injected mice; I2, input (glomerular lysate) from LPS-injected mice; IP, immunoprecipitation; WB, Western blot. (C) Competitive Co-IP experiments performed in HEK293 show that increasing amounts of GFP-SMPDL3b interfere with the interaction of FLAG-uPAR/suPAR and GFP-β₃ integrin. However, transfection of GFP-empty vector used as a control does not affect the interaction between uPAR/suPAR and β₃ integrin. GFP, green fluorescent protein.

Figure 6.

Decreased SMPDL3b expression is associated with β₃ integrin activation, resulting in a migratory FSGS-like phenotype in podocytes. (A) Human podocytes were incubated for 24 hours with or without suPAR (1 μg/ml). suPAR treatment (S) significantly increased AP5 and phospho-Src protein expression in podocytes, and these changes are more prominent in suPAR-treated human SMPDL3b KD podocytes compared with control podocytes. However, human SMPDL3b OE podocytes are protected from suPAR-induced increases in AP5 (activated β₃ integrin) and Src expression (*P<0.05). GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (B) Immunofluorescence staining using antibodies for AP5 (green) and AP3 (red) reveals that human SMPDL3b KD podocytes are characterized by activation of β₃ integrin at baseline compared with normal podocytes. Furthermore, although β₃ integrin can be activated by suPAR in control podocytes, human SMPDL3b OE podocytes are protected from suPAR-induced β₃ integrin activation (*P<0.05; **P<0.01). MFI, mean fluorescence intensity. (C) Migration assay showing that suPAR-induced cell migration is significantly increased in control human podocytes and even more accentuated in human SMPDL3b KD podocytes (*P<0.05; **P<0.01). However, suPAR does not increase migration in human SMPDL3b OE podocytes. (D) Rac1 activity is significantly increased in suPAR-treated human SMPDL3b KD podocytes compared with control podocytes and SMPDL3b OE podocytes (*P<0.05, control versus SMPDL3b KD podocytes; ^#P<0.05, control versus SMPDL3b OE podocytes).

Figure 7.

Increased SMPDL3b expression is associated with the suppression of β₃ integrin, resulting in an apoptotic DKD-like phenotype in podocytes. (A) RhoA activity is significantly increased in human SMPDL3b OE podocytes and decreased in SMPDL3b KD podocytes compared with control podocytes (*P<0.05, control versus SMPDL3b KD podocytes; ^#P<0.05, control versus SMPDL3b OE podocytes). (B) Treatment of normal human podocytes with sera from patients with DKD leads to increased RhoA expression, and SMPDL3b KD podocytes are protected from DKD sera-induced increases in RhoA expression (*P<0.01). (C) suPAR treatment significantly increase apoptosis in SMPDL3b OE podocytes, whereas control and SMPDL3b KD podocytes are not susceptible to suPAR-induced apoptosis (*P<0.01). (D and E) Inhibition of β₃ integrin signaling renders (D) normal human podocytes susceptible to suPAR-induced apoptosis, whereas (E) SMPDL3b mRNA expression levels remain unchanged (*P<0.05). cRGD, cyclo-(Arg-Gly-Asp) peptide; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Figure 8.

ASMase inhibitors (AIs) worsen the phenotype in a mouse model for FSGS. (A) Urine albumin (Alb)-to-creatinine ratio is significantly higher in ADR mice compared with control mice, and AI treatment significantly increases proteinuria in ADR mice (*P<0.05, ADR versus control; ^#P<0.05, ADR versus ADR+AI). (B) Segmental sclerosis assessed by periodic acid–Schiff staining is significantly worse in ADR mice than control mice (*P<0.05), and the extent of segmental sclerosis is further significantly increased by AI treatment (^##P<0.01). (C) Glomerular AP5 (activated β₃ integrin) and phospho-Src protein expressions are significantly higher in ADR mice, and these changes are markedly increased after AI treatment (*P<0.05; ^#P<0.05). cRGD, cyclo-(Arg-Gly-Asp) peptide; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Figure 9.

ASMase inhibitors (AIs) ameliorate the phenotype in a mouse model for DKD. (A) Urine albumin (Alb)-to-creatinine ratio is significantly higher in db/db mice compared with db/m mice (**P<0.01), and AI treatment significantly abrogates proteinuria in db/db mice compared with untreated db/db mice (^#P<0.05). (B) Glomerular surface area is increased in db/db mice compared with db/m mice (*P<0.05); however, glomerular surface area is significantly smaller in ASMase-treated db/db mice compared with untreated db/db mice (^#P<0.05) and seems to not be significantly changed compared with db/m mice. (C) Glomerular cleaved caspase-3 expression is significantly higher in db/db mice compared with db/m mice (*P<0.05), and expression is significantly reduced by AI treatment (^#P<0.05). GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (D) Glomerular active RhoA expression is increased in db/db mice (**P<0.01) and attenuated by AI treatment (^#P<0.05). MFI, mean fluorescence intensity.

Figure 10.

Mechanistic model of podocyte injury in FSGS and DKD. In FSGS, increased circulating suPAR together with low or absent SMPDL3b expression lead to αVβ₃ integrin activation, resulting in increased Src phosphorylation and Rac1 activity and ultimately causing a migratory podocyte phenotype. In DKD, increased circulating suPAR in the presence of high SMPDL3b expression leads to competitive binding of SMPDL3b to suPAR, thus preventing αVβ₃ integrin activation but allowing for RhoA activation and increased apoptosis.