Genetic deletion of calcium-independent phospholipase A2γ protects mice from diabetic nephropathy

Abstract

Calcium-independent phospholipase A2γ (iPLA2γ) is localized in glomerular epithelial cells (GECs)/podocytes at the mitochondria and endoplasmic reticulum, and can mediate release of arachidonic acid and prostanoids. Global knockout (KO) of iPLA2γ in mice did not cause albuminuria, but resulted in mitochondrial structural abnormalities and enhanced autophagy in podocytes. In acute glomerulonephritis, deletion of iPLA2γ exacerbated albuminuria and podocyte injury. This study addresses the role of iPLA2γ in diabetic nephropathy. Hyperglycemia was induced in male mice with streptozotocin (STZ). STZ induced progressive albuminuria in control mice (over 21 weeks), while albuminuria did not increase in iPLA2γ KO mice, remaining comparable to untreated groups. Despite similar exposure to STZ, the STZ-treated iPLA2γ KO mice developed a lower level of hyperglycemia compared to STZ-treated control. However, there was no significant correlation between the degree of hyperglycemia and albuminuria, and even iPLA2γ KO mice with greatest hyperglycemia did not develop significant albuminuria. Mortality at 21 weeks was greatest in diabetic control mice. Sclerotic glomeruli and enlarged glomerular capillary loops were increased significantly in diabetic control compared to diabetic iPLA2γ KO mice. Glomerular matrix was expanded in diabetic mice, with control exceeding iPLA2γ KO. Glomerular autophagy (increased LC3-II and decreased p62) was enhanced in diabetic iPLA2γ KO mice compared to control. Treatment of cultured GECs with H2O2 resulted in increased cell death in control GECs compared to iPLA2γ KO, and the increase was slightly greater in medium with high glucose compared to low glucose. H2O2-induced cell death was not affected by inhibition of prostanoid production with indomethacin. In conclusion, mice with global deletion of iPLA2γ are protected from developing chronic glomerular injury in diabetic nephropathy. This is associated with increased glomerular autophagy.

Fig 1. KO of iPLA₂γ attenuates development of albuminuria in STZ-induced diabetic nephropathy.

A) Treatment of mice with STZ induces progressive albuminuria (monitored as urine albumin/creatinine) in control mice. N = 9 mice in control (Ctrl) untreated (Untr), 10 in KO Untr, 10 in Ctrl STZ and 15 in KO STZ groups. ****P<0.0001 Ctrl STZ vs KO STZ, P<0.0001 Ctrl STZ vs Ctrl untreated. KO Untr vs KO STZ is not significant (the week 2–21 time points were considered together in the analysis; two-way ANOVA). W, week. B) Albuminuria does not correlate with blood glucose. The mean week 2–21 albumin/creatinine value of each diabetic mouse are plotted against mean blood glucose value of the same mouse. The correlations in each group are relatively weak (P values are not significant). The dashed lines indicate the mean urine albumin/creatinine values of the respective untreated groups. C) Viability of mice. Greatest mortality is evident in the Ctrl STZ mice.

Fig 2. STZ increases glomerular matrix expansion.

Kidney sections were stained with periodic acid-Schiff and glomerular matrix expansion was evaluated with a pixel-counting algorithm. A) Representative photomicrographs. B) Quantification of extracellular matrix. Both STZ-treated control and iPLA₂γ KO mice show significant increases in glomerular matrix, compared to untreated groups, and the increase is greater in control (P<0.02). C) Glomerular cross-sectional area is reduced slightly in iPLA₂γ KO STZ compared to KO untreated mice and tends to be slightly lower in control STZ compared to control untreated. 20 glomeruli in 5–6 mice per group were analyzed. *P<0.05, ****P<0.0001 (ANOVA). Bar = 25 μm (all panels are presented at the same magnification).

Fig 3. Control (Ctrl) STZ mice show changes in glomerular morphology.

A) Kidney sections were stained with antibodies to synaptopodin and collagen IV-α5 (representative immunofluorescence micrographs). B) Quantification of immunofluorescence intensity. There are no significant differences in synaptopodin or collagen immunofluorescence intensities among groups (ANOVA). However, Ctrl STZ mice show an increase in sclerotic glomeruli (A; arrow) and glomeruli with large capillary loops (*) compared to the other groups (quantification in panel C). 5–7 glomeruli/mouse in 4–6 mice per group were analyzed. *P<0.05, **P<0.01 (Chi-squared test). Bar = 25 μm.

Fig 4. Podocyte numbers in control and iPLA₂γ KO mice (WT1-positive cells).

A) Kidney sections were stained with anti-WT1 antibody (representative photomicrographs). B) WT1 counts. There are no significant differences in WT1 counts or WT1/glomerular area counts among groups (ANOVA). 10 glomeruli/mouse in 5–7 mice per group were analyzed; each point is the mean value of a single mouse. Bar = 25 μm.

Fig 5. iPLA₂γ KO mice show increased glomerular autophagy.

A and B) Glomerular lysates were immunoblotted with antibodies to LC3, p62 and ubiquitin (representative immunoblots). C) Signals were quantified by densitometry. LC3-II is increased significantly in untreated and STZ-treated iPLA₂γ KO mice compared to untreated and STZ-treated control. p62 is decreased significantly in STZ-treated iPLA₂γ KO mice. There are no significant differences in protein ubiquitination among groups. There are 6–8 mice per group. *P<0.05, **P<0.01 (ANOVA).

Fig 6. Control (Ctrl) GECs are more susceptible to injury compared to iPLA₂γ KO GECs.

A) Control and iPLA₂γ KO GECs were untreated, or treated with tunicamycin (TM; 5 μg/ml), adriamycin (ADR; 2 μM) or H₂O₂ (1 mM) in high (36 mM) or low glucose (7.8 mM) media for 24 h. At the end of incubations, LDH was measured in cell supernatants and lysates, and percent LDH release was calculated. Control GECs are more susceptible to adriamycin- or H₂O₂-induced cytolysis compared to iPLA₂γ KO GECs. In control GECs, H₂O₂-induced cytolysis is greater in high glucose. 5 experiments performed in duplicate. *P<0.05, ****P<0.0001 (ANOVA). B) Control GECs (high glucose) were untreated or treated with indomethacin (Indo; 10 μM), ADR (0.5 and 1.0 μM), H₂O₂ (1 mM), ADR + Indo or H₂O₂ + Indo for 24 h. No significant effects of Indo on LDH release are observed. 3–5 experiments performed in duplicate. *P<0.05, ****P<0.0001 (ANOVA). C) Effect of autophagy inhibition on LDH release. Control and iPLA₂γ KO GECs were incubated with H₂O₂ (1 mM) in the presence or absence of 2.5 μM SBI0206965 for 24 h. (SBI0206965 did independently induce LDH release.) SBI0206965 augments H₂O₂-induced LDH release to a similar extent in control and iPLA₂γ KO GECs. Values are fold-increase in LDH release in the presence of SBI0206965 compared to its absence (bars indicate median values, P = not significant, Mann Whitney U test). LDH release after H₂O₂ treatment in the absence of SBI020696 was 9.9±2.2% in control GECs and 4.3±0.08% in iPLA₂γ KO GECs. 4 experiments performed in duplicate.