Impaired Podocyte Autophagy Exacerbates Proteinuria in Diabetic Nephropathy

Diabetic nephropathy is a leading cause of end-stage renal disease and is becoming a serious health problem worldwide. The appearance of microalbuminuria, the progression to overt proteinuria, and the resultant renal dysfunction over several years to decades is the typical progressive course of diabetic nephropathy. Recent clinical studies have shown that microalbuminuria and a part of overt proteinuria can be halted and reversed by strict control of glycemia and blood pressure (1–3). However, some patients with diabetes still develop massive proteinuria, resulting in a rapid decline of renal function (4). Thus, a better understanding of the pathogenesis of massive proteinuria in diabetic nephropathy may further improve renal outcomes in patients with diabetes.

During evolution, living organisms developed several systems to overcome starvation in times of scarcity. These systems may be associated with the pathogenesis of diabetes and its vascular complications in times of plenty. Autophagy is an evolutionarily conserved intracellular catabolic process that allows for the degradation of proteins and organelles via the lysosomal pathway (5,6). One major role of autophagy is to degrade proteins and reconstitute the intracellular metabolism to cope with starvation and another is to remove damaged organelles such as mitochondria, peroxisomes, and lysosomes (5–8). Autophagy is thus essential to maintain cell homeostasis under various stress conditions. Furthermore, autophagy has been shown to regulate whole-body glucose and lipid metabolism in mammals, with impaired autophagy involved in the pathogenesis of several metabolic diseases (9–12). However, the role of autophagy in the pathogenesis of diabetic nephropathy remains unclear.

Podocytes are pivotal in maintaining glomerular filtration barrier function; thus, alterations in these cells are associated with massive proteinuria (13,14). Podocytes are well-differentiated cells with no capacity to divide. The intracellular degradation system is thus believed to be important in maintaining podocyte homeostasis. Indeed, clinical and experimental evidence has shown that lysosome dysfunction leads to severe podocyte injury and massive proteinuria (15–17). Interestingly, autophagy activity in podocytes is constitutively high, even under nonstress conditions (18,19), suggesting that the autophagy-lysosome system plays a pivotal role in maintaining podocyte homeostasis and that its alteration is involved in the pathogenesis of diabetic nephropathy.

Thus, this study was designed to determine the role of autophagy in maintaining podocyte homeostasis under diabetic conditions and its involvement in the development of massive proteinuria in diabetic nephropathy.

All procedures in the animal studies were performed in accordance with the guidelines of the Research Center for Animal Life Science of Shiga University of Medical Science. In human studies, all patients provided written informed consent. The protocols for human studies were approved by the scientific-ethical committees of Shiga University of Medical Science and Ikeda City Hospital and adhered to the Declaration of Helsinki guidelines.

Human kidney biopsy specimens were obtained from seven patients with type 2 diabetes with massive proteinuria (>3.5 g/day), four patients with type 2 diabetes with minimal proteinuria (<0.5 g/day), six patients with membranous nephropathy with reversible nephrotic-range proteinuria, three patients with minimal change nephrotic syndrome with reversible nephrotic-range proteinuria, and five patients with IgA nephropathy with minimal proteinuria. All patients with diabetes had diabetes for >10 years as well as retinopathy.

Eight-week-old male C57BL/6 mice were obtained from Clea Japan Inc. (Tokyo, Japan) and 8-week-old Long-Evans Tokushima Otsuka (LETO) and Otsuka Long-Evans Tokushima Fatty (OLETF) rats from Shimizu Laboratory Supplies Co., Ltd. (Kyoto, Japan). The mice were fed either a standard diet (STD; 10% of total calories from fat) or a high-fat diet (HFD; 60% of total calories from fat) for 32 weeks. LETO and OLETF rats were fed ad libitum until 50 weeks old, although some OLETF rats were killed at age 32 weeks. The rodents were fasted overnight and killed (20).

GFP-LC3 transgenic mice provided by Noboru Mizushima (Tokyo University, Tokyo, Japan) were used to analyze autophagy activity (19).

Podocyte-specific Atg5-deficient mice (Podo-Atg5^−/−) were generated by crossbreeding Atg5^f/f mice (21) with Nphs2-Cre transgenic mice (22). Eight-week-old male Atg5^f/f mice were used as a control group. All mice were crossed on a C57BL/6 background. To assess the effects of dietary intervention, Podo-Atg5^f/f mice were fed an STD (n = 5) or HFD (n = 6) and Podo-Atg5^−/− mice were fed an STD (n = 7) or HFD (n = 9) for 32 weeks (23).

Blood glucose concentrations were measured using a Glutest sensor (Sanwa Kagaku, Nagoya, Japan). Plasma insulin was measured by ELISA (Morinaga, Tokyo, Japan). Urinary albumin excretion was measured by immunoblot and ELISA (Exocell, Philadelphia, PA) (20). Glucose and insulin tolerance tests were performed as previously reported (23).

Fixed kidney specimens embedded in paraffin were sectioned at 3-μm thickness. Antibodies to p62 (MBL, Tokyo, Japan), fibronectin (Chemicon, Temecula, CA), F4/80 (Serotec, Oxford, U.K.), WT1 (Santa Cruz Biotechnology, Santa Cruz, CA), podocin (Sigma-Aldrich, St. Louis, MO), synaptopodin (Progen, Heidelberg, Germany), lysosome-associated membrane protein-2 (lamp2; Abcam, Cambridge, U.K.), and ubiquitin (Cell Signaling Technology, Beverly, MA) were used. Transmission and scanning electron microscopic analyses were performed with the Hitachi S-570 and H-7500 (Hitachi, Tokyo, Japan). For semiquantitative analysis of p62 accumulation, the glomerular intensity of staining was rated as grades 1 (none), 2 (minor), 3 (moderate), 4 (severe), and 5 (most severe) (20). More than 10 glomeruli in each mouse or human samples were evaluated. To determine podocyte number, serial kidney sections were stained with WT1 antibody, and WT1-positive cells were counted, with the number calculated using the dissector/fractionator combination method (24–26). Histological analyses were performed by three independent nephrologists in a blinded manner.

Western blot analysis was performed as previously described (20). The membranes were incubated with antibodies against cleaved caspase 3 (Asp175), ubiquitin and Atg7 (Cell Signaling Technology), microtubule-associated protein 1 light chain 3 (LC3; Novus Biologicals, Littleton, CO), β-actin (Sigma-Aldrich), and p62 (MBL).

The mouse podocyte cell line was cultured as previously described (27,28). Differentiated cells were stimulated with high glucose (500 mg/dL), fatty acids (palmitate and oleate, 150 μmol/L each), tumor necrosis factor-α (TNFα, 10 ng/μL), or serum for 24 h. Sera were collected from the indicated rodent models and patients. Complement was depleted by heat inactivation, and cells were incubated in 10% serum for 24 h (29,30), with/without lysosome inhibitors for 1 h (31). Characteristics of the sera used in this study are shown in Supplementary Tables 1 and 2.

An immortalized Atg7-deficient podocyte cell line was generated with the pMESVTS plasmid containing a SV40 large T antigen (32). Glomeruli of podocyte-specific Atg7-deficient (33,34) and wild-type mice were isolated using Dynabeads M-450 Tosylactivated (Invitrogen, Carlsbad, CA) (35). Podocytes were infected with viral supernatant from PLAT-E cells transfected with pMESVTS plasmid (36) and were maintained in RPMI-1640 with 10% FBS at 33°C. The cells were cultured at 39°C to induce differentiation over 7 days.

Human serum samples were taken from patients with type 2 diabetes in the Shiga Prospective Observational Follow-up Study in 2011 and 2012 (2). The preliminary study included 11 patients, 3 with normoalbuminuria (<30 mg/g Cre), 3 with microalbuminuria (30–300 mg/g Cre), 2 with macroalbuminuria (>300 mg/g Cre), and 3 with massive proteinuria (>3.5 g/g Cre). The validation study included 50 subjects, 10 subjects without diabetes, and 40 patients with type 2 diabetes, including 10 with normoalbuminuria, 11 with microalbuminuria, 10 with macroalbuminuria, and 9 with massive proteinuria (Supplementary Table 2). ELISA was used to measure the concentrations of p62 (Enzo Life Science, Farmingdale, NY).

Results are expressed as the mean ± SEM. ANOVA and a subsequent Tukey test were used to determine the significance of differences in multiple comparisons. Student t test was used for comparisons of two groups. A P value <0.05 was considered statistically significant.

To examine the relationships among levels of proteinuria, podocyte damage, and autophagy, kidney biopsy samples taken from patients with diabetic nephropathy and refractory massive proteinuria (>3.5 g/day) (Fig. 1A, patients 1–7) or minimal proteinuria (<0.5 g/day) (Fig. 1C, patients 17–20), membranous nephropathy and minima change with reversible massive proteinuria (Fig. 1B, patients 8–16), and IgA nephropathy with minimal proteinuria (Fig. 1C, patients 21–25) were analyzed. Mesangial expansion in periodic acid Schiff (PAS) staining was observed in all patients with diabetic nephropathy, regardless of proteinuria level (Fig. 1A, patients 1–7, and Fig. 1C, patients 17–20). To examine podocyte injury, we conducted an immunofluorescent study for podocin protein that is a key protein of the slit diaphragm of podocytes (14). The podocin expression pattern was nearly normal in the patients with minimal proteinuria, whether or not they had diabetic nephropathy (Fig. 1C, patients 17–25). In contrast, the podocin expression pattern was granular and irregularly scattered in the patients with massive proteinuria regardless of the underlying disease (Fig. 1A and B, patients 1–16). In addition, decreases in podocin-positive areas were obviously observed in the patients with refractory massive proteinuria due to diabetes (Fig. 1A, patients 1–7).

The protein p62 is a specific target of the autophagy degradation; thus, intracellular accumulation of this protein is indicative of insufficient autophagy (37). Intense accumulation of p62 protein was significantly increased in the glomeruli of the patients with diabetes and massive proteinuria (Fig. 1A–D).

We further confirmed the relationship among disease stage of diabetic nephropathy, podocyte injury, and autophagy insufficiency by utilizing two rodent models of diabetic nephropathy. One model, in mice, involved HFD-induced renal injury, resulting in diabetes-associated minimal proteinuria (23,38). The second model involved OLETF rats, which spontaneously develop hyperglycemia and subsequent massive proteinuria with age (39). Western blot analysis of urine samples and PAS staining showed minimal proteinuria with glomerular hypertrophy in HFD-fed C57BL/6 mice, and age-dependent progression of proteinuria with glomerular sclerosis in OLETF rats (Fig. 2A–C). Podocytes in 50-week-old OLETF rats with massive proteinuria showed a reduction in podocin-positive areas, with severe alterations in foot processes and p62 accumulation (Fig. 2C and E). However, these alterations were not found in the other rodent models (Fig. 2C–E).

These histological results showed that insufficient podocyte autophagy was associated with severe podocyte injury and massive proteinuria in diabetes. To assess the causal association, we used podocyte-specific autophagy-deficient mice.

The protein coded by the Atg5 gene is essential for autophagosome formation (21). Thus, podocyte-specific autophagy-deficient mice were generated by crossbreeding Atg5^f/f mice (21) with Nphs2-Cre transgenic mice (22). LC3, a regulatory protein essential for the induction of autophagy, localizes to autophagosome membranes during activation of autophagy. Thus, autophagy activity in the cells of GFP-LC3 transgenic mice can be monitored as green dots (19). Impaired formation of GFP-LC3 dots was confirmed in the podocytes of podocyte-specific Atg5-knockout mice (podo-Atg5^−/−) crossbred with GFP-LC3 transgenic mice (Fig. 3A).

To assess the effects of podocyte autophagy deficiency on HFD-induced minimal proteinuria, control Atg5^f/f and podo-Atg5^−/− mice were fed an STD or HFD for 32 weeks. During 32 weeks, HFD-fed Atg5^f/f and podo-Atg5^−/− mice showed a similar development of obesity, hyperinsulinemic hyperglycemia, glucose intolerance, and insulin resistance (Fig. 3B–H). HFD-fed Atg5^f/f mice developed minimal albuminuria, and STD-fed podo-Atg5^−/− mice did not show increased urinary albumin excretion (Fig. 3I and J). However, HFD-fed podo-Atg5^−/− mice developed massive albuminuria (Fig. 3I and J).

Histological analysis of the glomeruli of Atg5^f/f mice showed that the HFD increased glomerular size, PAS-positive area, and fibronectin deposition in glomeruli (Fig. 4A and B). These HFD-induced glomerular alterations tended to be greater in podo-Atg5^−/− than in Atg5^f/f mice, but the differences were not statistically significant (Fig. 4A and B). In contrast, renal tubulointerstitial lesions, as determined by hematoxylin-eosin (H-E) staining and immunohistochemical analysis of F4/80-positive macrophages, were exacerbated in HFD-fed podo-Atg5^−/− mice (Fig. 4C and D), suggesting that podocyte-specific autophagy insufficiency under diabetic conditions resulted in massive proteinuria, accompanied by proteinuria-induced tubulointerstitial damage.

There was no evidence of p62 accumulation in the podocytes of HFD-fed Atg5^f/f mice, and GFP-LC3 dot formation was not altered in the podocytes of HFD-fed GFP-LC3 mice (Fig. 4E–H), suggesting that HFD-induced diabetes alone did not affect podocyte autophagy and that autophagy was not related to the onset of HFD-induced minimal proteinuria. In contrast, p62 accumulation was significantly higher in the podocytes of HFD-fed podo-Atg5^−/− mice with massive proteinuria (Fig. 4E and F). These results suggested that insufficient autophagy played a causal role in the progression of proteinuria, from minimal to massive levels, under diabetic conditions.

Transmission electron microscopy showed that both Atg5^f/f and podo-Atg5^−/− mice fed the HFD for 32 weeks resulted in thickening of the glomerular basement membrane (Fig. 5A). STD-fed podo-Atg5^−/− mice showed nearly normal podocyte morphology, whereas HFD-fed podo-Atg5^−/− mice developed severe foot process effacement (Fig. 5A). Furthermore, in scanning electron microscopy, normal foot process structure was drastically disrupted in the podocytes of HFD-fed podo-Atg5^−/− mice (Fig. 5B).

The podocin expression pattern was normally and linearly aligned in STD- and HFD-fed Atg5^f/f mice and in STD-fed podo-Atg5^−/− mice (Fig. 5C). However, the pattern was visible as dots in HFD-fed podo-Atg5^−/− mice (Fig. 5C), similar to that observed in patients with diabetes and massive proteinuria. Podocin internalization is an additional marker of alterations in podocyte foot processes (40). Most podocin (red) and synaptopodin (green) signals were visible as normal capillary pattern in the undamaged podocytes of STD- and HFD-fed Atg5^f/f mice and STD-fed podo-Atg5^−/− mice (Fig. 5D). In contrast, increased podocin internalization, as shown by the increased podocin expression in cytosol and the merged yellow signals, was observed in the damaged podocytes of HFD-fed podo-Atg5^−/− mice (Fig. 5D). In addition, the number of WT1-positive podocytes was significantly reduced in HFD-fed podo-Atg5^−/− mice (Fig. 5E and F). These results suggested that autophagy insufficiency combined with diabetic conditions caused podocyte damage.

High glucose, fatty acids, and TNFα, an inflammatory cytokine, are major pathogenic factors in diabetic nephropathy. However, each of these, as well as their combination, had no effect on autophagy activity, as determined by p62 accumulation and LC3II conversion in cultured podocytes (Fig. 6A–C), suggesting that other factors may alter autophagy in podocytes. Several reports showed that some serum factors altered intracellular signaling pathways and autophagy activity in some conditions (29,30), raising a possibility that some serum factors associated with diabetic massive proteinuria might inhibit autophagy in podocytes.

To determine this possibility, mouse podocytes were cultured with serum from STD- or HFD-fed C57BL/6 mice or 50-week-old LETO or OLETF rats, and LC3 dot formation was assessed by immunofluorescence (Fig. 6D, protocol 1). LC3 dot formation was lower in cultured podocytes stimulated with serum from 50-week-old OLETF rats (Fig. 6E). Insufficient autophagy in cultured podocytes stimulated with serum from 50-week-old OLETF rats (Fig. 6D, protocol 2) was also confirmed by an increase in p62 expression levels and a decrease in LC3II bands (Fig. 6F and G). Furthermore, treatment with OLETF rat serum alone significantly enhanced apoptosis in normal podocytes, as shown by the cleavage of caspase 3 (Fig. 6F and G). These results suggested that some serum factors of 50-week-old OLETF rats inhibited autophagy and caused apoptosis in podocytes.

We excluded a possibility that serum alterations related to massive proteinuria secondarily caused autophagy insufficiency in podocytes. Stimulation with the serum from podo-Atg5^−/− mice fed an HFD had no effect on both p62 accumulation and apoptosis in cultured podocytes (Fig. 6D, protocol 3, and Fig. 6H), even though they showed massive proteinuria, indicating that massive proteinuria–related serum alteration itself was not a cause of autophagy insufficiency.

To further examine a causal relationship between autophagy insufficiency and apoptosis associated with diabetic serum, we used cultured Atg7-deficient podocytes. The protein coded by the Atg7 gene is also essential for autophagosome formation (33). Atg7 protein expression was not observed in cultured Atg7-deficient podocytes, and autophagy deficiency was confirmed by a significant increase in p62 protein and a decrease in LC3II bands (Fig. 6I). These cells were subsequently incubated with serum from STD- or HFD-fed C57BL/6 mice (Fig. 6D, protocol 4). Serum from HFD-fed C57BL/6 mice had no effect on the control Atg7^f/f podocytes, whereas it significantly increased the cleavage of caspase 3 in the Atg7-deficient podocytes (Fig. 6I and J). The study provided further evidence that insufficient autophagy played a causal role in podocyte apoptosis under diabetic conditions.

To determine the intracellular component targeted by autophagy in podocytes under diabetic conditions, intracellular structural changes in the podocytes were analyzed. The podocytes of Atg5^f/f mice had a normal mitochondrial structure with normal autophagosome formation, regardless of diet (Fig. 7A). In the podocytes of STD-fed podo-Atg5^−/− mice, autophagosomes were absent and unidentifiable deposits were observed but mitochondrial structure and nuclear shape were normal (Fig. 7A). Interestingly, the cytosol in podocytes of HFD-fed podo-Atg5^−/− mice was occupied by a number of huge balloon-like structures, suggesting damaged lysosomes (Fig. 7A).

The autophagy system is involved in removing damaged lysosomes and maintaining lysosome homeostasis (8,41), raising a hypothesis that lysosomes damaged by diabetic metabolic loads were targets of autophagy, and that impairment of autophagy by diabetic conditions would result in the accumulation of damaged lysosomes (Fig. 7B). Actually, the levels of expression of lamp2, a membrane marker of lysosomes (42), and ubiquitinated proteins to be degraded by lysosomes were increased in the podocytes of HFD-fed podo-Atg5^−/− mice (Fig. 7C), suggesting that dysfunctional lysosomes accumulated in these podocytes. The accumulation of huge lysosomes with the increase in lamp2-positive areas and the absence of autophagosomes were observed in the podocytes of 50-week-old OLETF rats with massive proteinuria (Fig. 7D).

In addition, ubiquitinated proteins were found to accumulate in Atg7-deficient podocytes stimulated with the HFD mouse serum and normal podocytes stimulated with the 50-week-old OLETF rat serum (Fig. 7E). Furthermore, a double immunofluorescence assay for LC3 and lamp2 revealed a number of large lamp2-positive signals in these cells, with absent or impaired LC3 dot formation (Fig. 7F and G). These results confirmed our hypothesis that systemic diabetic changes injure lysosomes, and that diabetic conditions accompanied by massive proteinuria additionally impair autophagy, leading to the accumulation of damaged lysosomes.

Finally, to determine the disease stage of human diabetic nephropathy associated with the impaired autophagy-lysosome system, the levels of p62 and ubiquitinated proteins were assessed in cultured podocytes stimulated with serum from 11 patients with diabetes with a varying range of proteinuria. Stimulation of cultured podocytes with serum from patients with diabetes and massive proteinuria resulted in increases in p62 and ubiquitinated protein (Fig. 8A).

To validate the relationship between proteinuria progression and autophagy-lysosome dysfunction in 10 subjects without diabetes and 40 patients with diabetes with varying degrees of proteinuria (Supplementary Table 2), the accumulation of p62 protein in cultured podocytes stimulated with serum was assessed by ELISA (Fig. 8B). The levels of p62 protein were significantly higher in cultured podocytes stimulated with the serum from patients with diabetes and massive proteinuria than with serum from other disease stages (Fig. 8C). Diabetes alone did not affect p62 accumulation in the cultured podocytes (Fig. 8C). These results suggested that insufficient autophagy was particularly associated with progression to massive proteinuria in human diabetic nephropathy.

The results presented here have demonstrated that impaired autophagy in podocytes is involved in the pathogenesis of severe podocyte injury, leading to massive proteinuria in diabetic nephropathy. Furthermore, maintenance of lysosome homeostasis by removing damaged lysosomes may be a crucial task of podocyte autophagy under diabetic conditions.

During the past decade, autophagy has been intensively studied in animals with kidney diseases (43). Previous studies using podocyte- and proximal tubular cell–specific autophagy-deficient mice have shown that autophagy insufficiency is related to stress susceptibility (8,44–47), suggesting that autophagy is essential for renoprotection against various pathogenic conditions. Because autophagy is regulated by both nutrients and stress signals, it has been speculated that autophagy is involved in the pathogenesis of diabetic nephropathy (48). The current study utilizing podocyte-specific autophagy-deficient mice provide, for the first time, strong evidence indicating that autophagy is involved in the pathogenesis of diabetic nephropathy.

Autophagy insufficiency was observed specifically in podocytes of patients and rodents with diabetes and massive proteinuria, but not with the other disease stage. Furthermore, HFD-induced diabetes caused minimal proteinuria in mice with normal podocyte autophagy but massive proteinuria with podocyte loss in mice with podocyte-specific autophagy deficiency. These findings indicate that diabetes alone can cause typical glomerular changes leading to minimal proteinuria, whereas complicated impairment of podocyte autophagy during the course of diabetic nephropathy results in podocyte loss, followed by massive proteinuria (Fig. 8D). Thus, insufficient podocyte autophagy may play a pathogenic role particularly in the disease progression to massive proteinuria in diabetic nephropathy. However, this study had several limitations, which may have affected the interpretation of the results. Autophagy deficiency was assessed only by p62 staining and electron microscopic analysis, the number of human kidney biopsy samples was small, and two different rodent species were compared (HFD-fed mice as a model of albuminuria and OLETF rats as a model of massive proteinuria). Additional methods of assessing autophagy deficiency and large clinical cohort and/or animal studies are needed to validate our conclusions.

The mechanism underlying diabetes-related impairment of podocyte autophagy is still unclear. Sera from patients and rodents with massive proteinuria impaired autophagy in cultured podocytes, suggesting that serum factors associated with massive proteinuria in diabetes, but not with diabetes per se, impair podocyte autophagy. The mammalian target of rapamycin complex 1 (mTORC1) is a nutrient-sensing signal that inhibits autophagy (49). Interestingly, mTORC1 activity has been reported to be enhanced in podocytes of humans and animals with advanced diabetic nephropathy (50,51), suggesting that some factors in diabetic serum may activate mTORC1, suppressing podocyte autophagy. Additional studies are required to identify these serum factors and intracellular molecular pathways that inhibit podocyte autophagy, and may contribute to the development of a new therapy for the treatment of refractory diabetic nephropathy. Furthermore, the bioassay involving serum stimulation of cultured cells may be useful in assessing in vivo autophagy activity in human subjects and experimental animals.

Autophagy can degrade many damaged proteins and organelles (7). To date, however, the specific targets of podocyte autophagy have not been identified. Fabry disease, a lysosome disease, causes podocyte damage and proteinuria (15). Moreover, the deletion of a single gene that regulates lysosome function enhanced podocyte damage and proteinuria in mice (16,17). These results indicate that the lysosome is likely important in maintaining podocyte homeostasis. Interestingly, a massive accumulation of lysosomes with abnormal morphology was observed in the podocytes of diabetic rodents with autophagy deficiency. It remains unclear whether the lysosome is also a target of podocyte autophagy under other pathogenic conditions. Our results, however, suggest that damaged lysosomes are an important degradation target of podocyte autophagy at least under diabetic conditions.

Clinically, there are few effective treatments for patients with diabetes and massive proteinuria; thus, these patients often experience a rapid decline in renal function (4). Because proteinuria is the most likely cause of tubulointerstitial lesions that lead to renal dysfunction, new therapeutic agents are needed to halt the stage progression of proteinuria and/or to protect renal tubular cells against proteinuria-related renotoxicity, thus improving renal outcomes in patients with refractory diabetic nephropathy. We recently reported that autophagy insufficiency in proximal tubular cells was associated with the pathogenesis of obesity- and diabetes-mediated exacerbation of proteinuria-induced tubulointerstitial damage (52). Taken together, our previous and present findings suggest that autophagy activation may be effective for patients with diabetes and massive proteinuria and resultant rapid decline of renal function.

In conclusion, autophagy plays a pivotal role in maintaining lysosome homeostasis in podocytes under diabetic conditions. The impairment of autophagy is involved in the pathogenesis of podocyte loss, leading to massive proteinuria in diabetic nephropathy. These findings suggest a new therapeutic strategy for massive proteinuria in patients with diabetic nephropathy.

Acknowledgments. The authors thank Dr. P. Mundel (Harvard Medical School, Massachusetts General Hospital), Dr. N. Kobayashi (University of Ehime), and Dr. Y. Makino (Asahikawa Medical University) for kindly providing the cells and for their very helpful advice on cell cultures. The authors also thank Dr. N. Mizushima (Tokyo University) for donating the Atg5-floxed and GFP-LC3 transgenic mice and Dr. T. Kitamura (Tokyo University) and Dr. M. Komatsu (Niigata University) for providing the PLAT-E cells and the pMESVTS plasmid. The authors are grateful to Y. Omura and N. Yamanaka (Shiga University of Medical Science), H. Mukai, M. Kawai, M. Iwaki, and T. Kanemoto (Ikeda City Hospital), and the Central Research Laboratory of Shiga University of Medical Science for technical assistance, and to all members of the Maegawa Laboratory for their contributions to the Shiga Prospective Observational Follow-up Study.

Funding. This study was supported by Grants-in-Aid for Scientific Research (KAKENHI) from the Japan Society for the Promotion of Science (no. 25713033 to S.K., no. 26293217 to H.M., no. 15K19511 to M.Y., and no. 26670431 to K.A.), from the Takeda Science Foundation (to S.K.), from Banyu Life Science Foundation International (to S.K.), and from the Uehara Memorial Foundation (to S.K.).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. A.T., M.Y., and S.K. performed the experiments and wrote the manuscript. K.Y., K.A., E.-H.K., and S.U. performed the experiments. J.N., M.C.-K., H.A., and S.-i.A. contributed to the study concept, research design, and interpretation of the results. D.K., M.H., H.M., and T.U. gave conceptual advice. N.K., K.H., and H.O. collected human kidney biopsy samples. All authors contributed to discussion and reviewed and edited the manuscript. S.K. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.