REDD1 Ablation Attenuates the Development of Renal Complications in Diabetic Mice

Diabetic nephropathy (DN) is a chronic complication of diabetes that progresses to end-stage renal failure. DN represents a growing socioeconomic and global health burden, with the annual incidence of end-stage renal disease among patients with diabetes reaching 1,016 per million in 2015 (1). The Diabetes Control and Complications Trial demonstrated a key role for hyperglycemia in the development and progression of renal complications (2). However, current therapeutic interventions used in combination with glycemic control have not proven sufficient to halt or reverse DN progression (3,4). This is at least in part due to a deficit in understanding of the specific molecular events that contribute to renal pathology.

Diabetes-associated kidney damage includes functional and structural changes in glomerular cells, including podocytes, mesangial cells, and endothelial cells (5). Podocytes are terminally differentiated epithelial cells that preserve glomerular structure and maintain barrier function by allowing the selective filtration of water and solutes, while preventing large macromolecules like proteins from leaking into the urine (6). Among the changes in renal morphology that are caused by diabetes, a reduced number of podocytes per glomerulus, known as podocytopenia, is the strongest predictor of DN progression (7–9). Podocytopenia is observed in patients with early and late DN (7,8), as well as in animal models of type 1 and type 2 diabetes (9,10). A variety of stressors associated with diabetes, including oxidative stress, inflammation, and endoplasmic reticular stress, contribute to podocytopenia (11). Hyperglycemia-induced reactive oxygen species (ROS) have specifically been shown to decrease podocyte viability by triggering caspase-3–mediated apoptosis (12).

Expression of the stress response protein regulated in development and DNA damage response 1 (REDD1) (also known as DDIT4/RTP801) has been linked to the development of oxidative stress in multiple pathological conditions, including diabetic retinopathy (13–15). Our laboratory demonstrated a critical role for REDD1 in the development of diabetes-induced retinal complications and functional deficits in vision (15–18). Furthermore, clinical intervention with an siRNA targeting REDD1 has shown promise for the improvement of vision in patients suffering from diabetic macular edema (19). REDD1 acts at least in part to promote association of PP2A with Akt, leading to site-specific dephosphorylation of the kinase and reduced Akt-dependent suppression of glycogen synthase kinase 3β (GSK3β) (20). Interplay between REDD1 and GSK3β signaling contributes to the development of hyperglycemia-induced oxidative stress via suppression of the Nrf2 antioxidant response (18). GSK3β hyperactivity is associated with DN progression, and GSK3 inhibitors are currently being pursued as potential therapeutic options (21,22). However, a role for REDD1 in DN has never been investigated. Herein, we examined a role for REDD1 in development of renal complication caused by diabetes.

Male wild-type (WT) and REDD1 knockout (KO) mice on a B6;129 background (23) were administered 50 mg/kg streptozotocin (STZ) intraperitoneally for five consecutive days to induce diabetes. Control mice received equivalent volumes of sodium citrate vehicle. Diabetic phenotype was confirmed by fasting blood glucose concentrations >250 mg/dL. For examination of a role for GSK3β, C57BL/6J mice (The Jackson Laboratory, Bar Harbor, ME) were administered STZ as described above and then received daily intraperitoneal injections of either VP3.15 (10 mg/kg; MedChemExpress, Monmouth Junction, NJ) or vehicle (10% DMSO, 0.9% NaCl) during the last 3 weeks of diabetes. At 16 weeks of diabetes, mice were euthanized, and terminal urine collection was carried out from the bladder. Both kidneys were removed, weighed, and processed for examination. All procedures adhered to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Penn State College of Medicine Institutional Animal Care and Use Committee.

Urine protein was separated by SDS-PAGE, transferred to polyvinylidene fluoride membrane, and visualized with reversible protein stain (Thermo Fisher Scientific). Urine creatinine and albumin concentrations were determined with Creatinine (urinary) Colorimetric Assay Kit (Cayman Chemical) and Mouse Albumin ELISA Kit (Eagle Biosciences), respectively.

Renal sections (6 µm) were cut from 10% formalin-fixed, paraffin-embedded (FFPE) kidneys. Periodic acid-Schiff (PAS) and hematoxylin-eosin (H-E) staining were performed as previously described (9). Immunohistochemical analyses were carried out with ImmPRESS HRP Horse Anti-Rabbit IgG Polymer Detection Kit (Vector Laboratories). Heat-induced antigen retrieval was carried out with citrate buffer (0.01 mol/L, pH 6), and endogenous peroxidase activity was quenched through incubation of sections in BLOXALL Blocking Solution. Sections were then blocked with 2.5% horse serum, incubated overnight with appropriate antibodies (Supplementary Table 1), and detected with 3,3′-diaminobenzidine. Tissue sections were counterstained with H-E, and micrographs were captured with an Olympus BX51 brightfield microscope. Fibronectin or nitrotyrosine immunoreactivity was quantified with ImageJ software. WT-1–positive cells in glomeruli (20 per section) were counted.

All quantifications were performed in a masked manner and carried out as previously described (24). We evaluated glomerular mesangial expansion and glomerular volume (GV) in sagittal kidney sections by examining 30–50 cortical glomeruli per mouse (five mice per group). The index of the mesangial expansion was defined as ratio of mesangial area to glomerular tuft area. Glomerular area (GA) was quantified with ImageJ and converted to GV with a spherical approximation formula (GV = 1.2545(GA)^1.5).

Renal cortical tissue (∼1 mm³) from diabetic and nondiabetic WT and REDD1 KO mice was fixed with 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 mol/L phosphate buffer (pH 7.4), followed by 1% osmium tetroxide in 0.1 mol/L phosphate buffer (pH 7.4) for 1 h. Samples were dehydrated in a graduated series of ethanol and acetone and embedded in LX 112 (Ladd Research Industries, Williston, VT). Thin sections (65 nm) were stained with uranyl acetate and lead citrate and viewed in a JEOL JEM-1400 Transmission Electron Microscope (JEOL USA) located at the Penn State College of Medicine Transmission Electron Microscopy Core (Research Resource Identifier [RRID] SCR_021200).

REDD1 mRNA was detected with the RNAscope 2.5 HD Assay-RED detection kit and the RNAscope probe Mm-Ddit4 (Advanced Cellular Diagnostics) targeting nucleotides 133–1480 of NM_029083.2. RNA in situ hybridization (ISH) was carried out on 6 μm FFPE kidney sections according to the manufacturer’s protocol. Sections were deparaffinized and pretreated with heat and protease before hybridization of REDD1 mRNA probe. Preamplifier, amplifier, and alkaline phosphatase–labeled oligonucleotides were sequentially hybridized followed by application of a chromogenic substrate to produce red punctate dots. Tissue was counterstained with Mayer hematoxylin, and micrographs were captured with an Olympus BX51 brightfield microscope.

Apoptosis was measured with the DeadEnd fluorometric terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) system assay kit (Promega) according to the manufacturer’s instructions. Prior to TUNEL staining, FFPE kidney sections were rehydrated and immunofluorescence staining for WT-1 was carried out. Sections were then fixed in 4% paraformaldehyde and permeabilized with 20 µg/mL Proteinase K and then incubated with TUNEL reaction mixture for 60 min at 37°C in a humidified dark chamber. Kidney sections were subsequently washed with 2× saline-sodium citrate buffer (0.3 mol/L sodium chloride, 0.03 mol/L sodium citrate) and PBS and mounted with VECTASHIELD + DAPI (Vector Laboratories). Images were captured with the Leica SP8 confocal laser microscope with frame-stack sequential scanning.

Conditionally immortalized human podocytes (CIHP-1) were purchased from Dr. Moin Saleem’s laboratory at the University of Bristol (Bristol, U.K.), proliferated at 33°C in 5% CO₂, and then differentiated for 10 days at 37°C in 5% CO₂ (25). Cells were cultured in RPMI-1640 media (Sigma-Aldrich) supplemented with 10% FBS and 1% penicillin/streptomycin. CRISPR/Cas9 genome editing was used to generate a stable CIHP-1 cell line deficient in REDD1 (REDD1 KO) as previously described (26). Cells were exposed to culture medium containing either 30 mmol/L glucose or 5 mmol/L glucose supplemented with 25 mmol/L mannitol as an osmotic control. In specific studies, cells were exposed to medium containing the GSK3 inhibitor VP3.15 (1 μmol/L).

Renal lysates were centrifuged at 1,500g for 3 min, and the supernatant was exposed to 10 μmol/L 2′,7′-dichlorofluorescein (DCF). Fluorescence (excitation/emission = 504/529 nm) was measured with a SpectraMax M5 plate reader (Molecular Devices, San Jose, CA). Intracellular ROS were assessed in CIHP-1 cultures with a DCFDA Cellular Reactive Oxygen Species Detection Assay Kit (Abcam, Cambridge, U.K.) according to the manufacturer’s instructions. Cellular ROS were imaged by confocal microscopy.

Differentiated CIHP-1 WT and REDD1 KO cells were cotransfected with the pRL-CMV Renilla luciferase (Promega) and antioxidant response element (ARE)-firefly luciferase (kindly provided by Dr. Jiyang Cai, University of Texas Medical Branch) plasmids combined in a 10:1 ratio using jetPRIME (Polyplus transfection). After 24 h, transfection media were removed, and cells were exposed to hyperglycemic conditions for 48 h. Luciferase activity was measured on a FlexStation 3 (Molecular Devices) with a Dual-Luciferase Assay Kit (Promega).

WT and REDD1 KO cells were plated at a density of 10⁴ cells per well in a 96-well plate in 200 μL culture medium and allowed to adhere overnight at 33°C. Cells were then differentiated for 10 days followed by exposure to hyperglycemic conditions for 0–48 h. Cell viability was determined with the MTT Cell Proliferation Assay Kit (Cayman Chemical, Ann Arbor, MI) following the manufacturer’s instructions. Briefly, 10 μL of a 5 mg/mL 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) solution was added to each well. After 4 h incubation, the medium was carefully aspirated, the purple formazan crystals were solubilized, and optical density was measured at 570 nm.

Total protein was extracted from cells in culture and tissue that was flash frozen in liquid nitrogen. Approximately 20 mg of renal cortical tissue was homogenized in 200 µL of lysis buffer (50 mmol/L Tris [pH 7.4], 150 mmol/L NaCl, 1% NP-40, 0.5% Na-deoxycholate, 1% protease-phosphatase inhibitors) and protein concentrations were estimated with DC protein assay (Bio-Rad Laboratories). Protein from cell lysates or kidney homogenates were combined with Laemmli buffer, boiled, and fractionated with use of Criterion precast 4–20% gels (Bio-Rad Laboratories). Proteins were transferred to polyvinylidene fluoride membrane, reversibly stained for assessment of protein loading, blocked in 5% milk in Tris-buffered solution Tween 20, and evaluated with the appropriate antibodies (Supplementary Table 1).

Total RNA was extracted with TRIzol (Invitrogen, Waltham, MA). RNA (1 μg) was reverse transcribed with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Waltham, MA) and subjected to quantitative real-time PCR (QuantStudio 12K Flex Real-Time PCR System; Thermo Fisher Scientific [RRID SCR_021098]) using QuantiTect SYBR Green Master Mix (QIAGEN, Germantown, MD). Primer sequences are listed in Supplementary Table 2. Mean cycle threshold (CT) values were determined for control and experimental samples. Changes in mRNA expression were normalized to GAPDH mRNA expression with the 2^−ΔΔCT method.

Data are expressed as means ± SD. Statistical analysis of data with more than two groups were analyzed with two-way ANOVA, and pairwise comparisons were made with the Tukey test for multiple comparisons. Difference between two-groups was determined with unpaired Student t test. The relationships between REDD1 expression and blood glucose levels were tested with Spearman correlation analysis. Significance was defined as P < 0.05 for all analyses.

The data sets and resources generated during this study are available from the corresponding author on reasonable request.

After 16 weeks of STZ diabetes, REDD1 mRNA abundance was increased in kidney tissue homogenates in coordination with fasting blood glucose concentrations (Fig. 1A). A corresponding increase in REDD1 protein expression was observed in kidney tissue homogenates from diabetic mice as compared with nondiabetic controls (Fig. 1B). Negligible REDD1 mRNA was detected in the glomeruli of nondiabetic mice by ISH, whereas enhanced staining was observed throughout the kidney cortex of diabetic mice (Fig. 1C). For evaluation of a role for REDD1 in diabetes-induced renal damage, mice with a germline disruption of REDD1 were administered STZ. Following 16 weeks of diabetes, diabetic and nondiabetic REDD1 KO mice exhibited fasting blood glucose concentrations similar to those observed in diabetic and nondiabetic WT mice (Fig. 2A). Diabetes enhanced REDD1 mRNA (Fig. 2B) and protein (Fig. 2C) expression in the kidney of WT mice. Diabetic WT and REDD1 KO mice had reduced body weights as compared with nondiabetic mice after 16 weeks of STZ diabetes. However, there was no difference in body weight with WT versus REDD1 KO mice (Fig. 2D). In WT mice, diabetes increased kidney weight (Fig. 2E) and kidney–to–body weight ratio (Fig. 2F). As compared with diabetic WT mice, diabetic REDD1 KO mice exhibited reduced kidney weight and kidney–to–body weight ratio. In support of renal hypertrophy, rpS6 phosphorylation was enhanced in STZ diabetic WT mice (Supplementary Fig. 1); however, diabetes did not alter rpS6 phosphorylation in REDD1 KO mice. Urine volumes collected from diabetic WT and REDD1 KO mice were significantly higher as compared with their respective nondiabetic controls (Fig. 2G). Urinalysis of diabetic WT mice showed an increased urinary protein band at 60 kDa (Fig. 2H) and elevated urinary albumin (Fig. 2I). Diabetic mice also exhibited a reduction in urinary creatine concentration (Fig. 2J) and increased urinary albumin–to–creatinine ratio (ACR) (Fig. 2K). REDD1 deletion prevented proteinuria and normalized urinary ACR in diabetic mice.

Glomerular structure (Fig. 3A, panels a–d) and mesangial proliferation (Fig. 3A, panels e–h) were examined with H-E and PAS staining, respectively. Histologic examination revealed glomerular hypertrophy, basement membrane thickening, and mesangial proliferation in diabetic WT mice as compared with nondiabetic controls. Increased GV (Fig. 3B) and compressed Bowman capsule were observed in diabetic WT mice. Mesangial expansion was also observed within glomeruli of diabetic WT mice (Fig. 3C). Features of renal pathology were significantly attenuated or absent in diabetic REDD1 KO mice. In particular, GV and mesangial index were reduced in diabetic REDD1 KO mice as compared with diabetic WT mice. REDD1 deletion also decreased matrix deposition in diabetic mice, evidenced by reduced fibronectin expression (Fig. 3A, panels i–l, and Fig. 3D). Expression of mRNAs encoding profibrotic markers TGF-β1, α-SMA, fibronectin, and collagen 1A1 were enhanced in the kidney of diabetic WT mice as compared with nondiabetic controls, whereas the mRNA encoding the tight junction protein E-cadherin was reduced (Fig. 3E). Fibrotic markers were reduced, and E-cadherin expression was increased, in the kidney of diabetic REDD1 KO mice as compared with diabetic WT mice.

Hyperglycemia-induced podocyte loss has previously been reported in cell cultures and rodent models of diabetes (12,27). WT-1 immunolabeling was used to identify podocytes. WT-1 is essential for renal development as well as maintaining the differentiated state of podocytes in the adult kidney (28). A decrease in WT-1–positive cells was observed in glomeruli from diabetic WT mice (Fig. 4A). However, REDD1 deletion prevented reduced WT-1 expression in diabetic mice (Fig. 4B). A role for REDD1 in podocytopenia was also supported by reduced expression of the podocyte marker podocin in kidney homogenates from diabetic WT mice but not diabetic REDD1 KO mice (Fig. 4C). Normal podocyte architecture with filtration slits and evenly spaced slit diaphragms were observed in electron micrographs of renal tissue from nondiabetic mice (Fig. 4D). By contrast, glomerular basement membrane thickening and foot process effacement with distancing of slit pores were observed diabetic WT mice. In contrast with diabetic WT mice, podocyte architecture was preserved in diabetic REDD1 KO mice. For investigation of a role for REDD1 in human podocytes, CIHP cells were exposed to hyperglycemic culture conditions, which increased both REDD1 mRNA (Fig. 4E) and protein (Fig. 4F) expression. Metabolic viability assessment using MTT showed reduced viability of WT podocytes after 24–48 h of exposure to hyperglycemic conditions (Fig. 4G). However, hyperglycemic conditions did not reduce viability of REDD1-deficient podocytes. Hyperglycemic conditions also increased cleaved caspase 3 expression in WT, but not in REDD1 KO, podocytes (Fig. 4H).

Increased ROS levels were observed in kidney homogenates from diabetic mice (Fig. 5A). Protein oxidation identified by nitrotyrosine immunolabeling was also elevated in the kidney of diabetic mice (Fig. 5B and C). Both ROS levels and nitrotyrosine labeling were reduced in the kidney of diabetic REDD1 KO mice as compared with diabetic WT mice. Exposure to hyperglycemic culture media also enhanced ROS levels in WT podocytes; however, high glucose–induced ROS were not observed in REDD1 KO podocytes (Fig. 5D and E). To assess the impact of REDD1 on the Nrf2-mediated antioxidant response, an Nrf2 ARE-luciferase reporter was expressed in podocytes. Exposure to hyperglycemic conditions reduced Nrf2 activity in WT cells (Fig. 5F). REDD1 KO podocytes exhibited enhanced Nrf2 activity as compared with WT podocytes, and the suppressive effect of hyperglycemic conditions on Nrf2 required REDD1. Examination of Nrf2 target mRNA transcript (i.e., GCLC, GCLM, HO-1, and NQO1) expression was consistent with constitutive activation of Nrf2 in the absence of REDD1 (Fig. 5G).

We previously demonstrated that REDD1-dependent GSK3β activation contributes to Nrf2 suppression in the retina of diabetic mice (18). In the kidney of diabetic WT mice, we observed a reduction in the inhibitory phosphorylation of GSK3β at Ser9 (Fig. 6A–B). The suppressive effect of diabetes on GSK3β phosphorylation was absent in REDD1 KO mice, and GSK3β phosphorylation was increased in diabetic REDD1 KO mice as compared with diabetic WT mice. Similarly, GSK3β phosphorylation was reduced in WT podocytes on exposure to hyperglycemic conditions, and REDD1 was required for this effect (Fig. 6C). Together, the observations support REDD1-dependent activation of GSK3β in response to diabetes/hyperglycemic conditions. For examination of the role of GSK3β, GSK3 activity was inhibited using VP3.15. In podocytes exposed to hyperglycemic conditions, VP3.15 reduced glycogen synthase phosphorylation (Supplementary Fig. 2A) and attenuated ROS levels (Fig. 6D). Moreover, GSK3β inhibition attenuated hyperglycemia-induced loss of podocyte viability (Fig. 6E). For determination of the benefits of GSK3β suppression on renal defects, 13 weeks after STZ administration diabetic mice were administered VP3.15 daily for 3 weeks (Supplementary Fig. 2B). VP3.15 did not alter body weight (Supplementary Fig. 2C). Efficacy of the treatment was confirmed by reduced glycogen synthase phosphorylation in renal tissue lysates of diabetic mice (Supplementary Fig. 2D). In diabetic mice receiving VP3.15, renal hypertrophy (Supplementary Fig. 2E and F) and urinary proteins (Supplementary Fig. 2G) were reduced. Evidence for improved renal function was suggested by a trend toward lower urine ACR in diabetic mice receiving VP3.15 as compared with DMSO vehicle (P = 0.1276) (Supplementary Fig. 1H–J). Importantly, GSK3β inhibition prevented diabetes-induced ROS in kidney homogenates (Fig. 6F). GSK3β inhibition also prevented diabetes-induced podocyte loss (Fig. 6G and H) and attenuated podocyte apoptosis in diabetic mice, as assessed according to colocalization of WT-1– and TUNEL-positive cells (Fig. 6G and I and Supplementary Fig. 3).

Herein, we examined a role for REDD1 in the development of renal complications in a well-characterized preclinical model of type 1 diabetes. REDD1 induction was observed in the kidney of STZ diabetic mice, as well as in human podocytes exposed to hyperglycemic conditions. Diabetes-induced mesangial expansion, matrix deposition, and glomerular ultrastructural alternations were relieved in REDD1 KO mice. Single-cell sequencing supports that REDD1 expression is not localized to a particular cell type within the kidney (29). REDD1 was initially identified as a stress response gene that was transcriptionally induced in response to hypoxia and oxidative stress (30) but subsequently shown to be elevated in response to cellular stressors including nutrient deprivation and endoplasmic reticular stress (31,32). Increased REDD1 mRNA expression in the renal cortex of diabetic mice supports transcriptional upregulation of REDD1 with STZ diabetes. The studies here also provide evidence that REDD1 acts to promote renal oxidative stress via GSK3β-dependent suppression of the Nrf2 antioxidant response. Overall, the findings support a model wherein diabetes-induced renal REDD1 causes glomerular podocyte damage through disrupted redox homeostasis (Fig. 7).

Microalbuminuria is an important indicator of DN progression and often caused by the dysfunction of the glomerular filtration barrier (33). Podocytes are a core component of this filtration barrier and play a central role in maintaining glomerular structural and functional integrity. Podocyte injury is a key factor in DN progression, with foot process effacement, detachment from the glomerular basement membrane, and podocyte apoptosis often observed in the earliest stages of disease (34,35). Disrupted glomerular architecture, foot process effacement, distancing of slit pores, and podocytopenia were observed in diabetic mice. Remarkably, markers of podocyte dysfunction were absent or reduced in the kidney of diabetic REDD1 KO mice and renal barrier function was preserved.

Hyperglycemia is a major determinant of podocytopenia and DN progression (12,36). Podocyte cultures exposed to conditions including elevated glucose concentrations undergo apoptosis due to activation of p38 MAPK and classical caspase 3 signaling (12). A role for REDD1 in high glucose–induced caspase activation and cell death was previously demonstrated in R28 neuronal precursor cells (15). In support of that observation, we found that REDD1 deletion prevented an increase in cleaved caspase 3 expression and reduced viability in human podocyte cultures exposed to hyperglycemic conditions. The principle pathways responsible for hyperglycemia-induced tissue damage are all linked to the overproduction of ROS (37). An imbalance between ROS production and the cellular antioxidant response leads to development of oxidative stress, which is a key driver of DN pathology (38–40). Combating oxidative stress with antioxidant adjuvant therapies has been successful in delaying if not preventing the development of DN pathologies in preclinical rodent models (41,42). In clinical studies, antioxidant supplementation consistently improves albuminuria (43). Enhanced REDD1 expression is associated with multiple disease models in which oxidative stress is a driver of pathology (44). In the retina of diabetic mice and in retinal cell cultures exposed to hyperglycemic conditions, REDD1 contributes to the development of oxidative stress (15,17). Moreover, multiple reports have implicated hyperglycemia-induced oxidative stress as a trigger for cell death mechanisms including apoptosis (12,15,36,45). In the current study, REDD1 deletion blunted oxidative stress in the kidney of diabetic mice and in human podocytes exposed to hyperglycemic conditions.

The transcription factor Nrf2 regulates expression of >200 antioxidant response genes by binding to an ARE enhancer region of their promoters (45,46). Nrf2 is classically regulated by the adaptor protein Kelch-like ECH-associated protein 1 (Keap1), which facilitates proteasomal degradation of the transcription factor. When ROS accumulate, cysteine residues on Keap1 become oxidized and Nrf2 is allowed to translocate into the nucleus to promote the production of antioxidants (47). In addition, GSK3β directly phosphorylates Nrf2 to promote its nuclear exclusion and degradation by a protein cluster containing the ubiquitin ligase adapter β-transducin repeat–containing protein (β-TrCP) (48). In human podocytes exposed to hyperglycemic conditions, ROS levels were enhanced in coordination with suppression of Nrf2 activity. REDD1 deletion prevented the suppressive effect of hyperglycemic conditions on Nrf2 activity, and also podocytes lacking REDD1 exhibited chronic hyperactivation of Nrf2. Previous work from our laboratory provides evidence that REDD1 acts to suppress Nrf2 activity by promoting GSK3-dependent proteasomal degradation of the transcription factor (18).

Diabetes-induced GSK3β activation is observed in various tissues including the kidney (21). In the glomerulus, GSK3β is an evolutionarily conserved, redox-sensitive serine/threonine protein kinase that is predominantly expressed in podocytes (49,50). GSK3β expression in podocytes serves as a critical regulator of kidney function, injury, and repair (49,50). REDD1 promotes GSK3β activation via suppression of the protein kinase Akt. More specifically, REDD1 mediates dephosphorylation of Akt at Thr308 via recruitment of the protein phosphatase 2A, resulting in attenuated Akt activation (20). Akt phosphorylates the N-terminus of GSK3β at Ser9, which reduces GSK3β kinase activity by obstructing substrate recognition (51). In the retina of diabetic mice, REDD1 protein expression is enhanced and necessary for attenuated GSK3β phosphorylation (17). Data presented here extend from these studies in demonstrating that REDD1 was necessary for suppression of GSK3β phosphorylation in both the kidney of diabetic mice and in podocytes exposed to hyperglycemic conditions. GSK3 inhibitors have been successfully used to prevent deficits in renal function and cardiac remodeling in rodent models of diabetes and diabetic ischemia reperfusion injury (22,52,53). Herein, benefits of REDD1 deletion in the prevention of DN pathology were recapitulated by GSK3 inhibition. We found that administration of a GSK3 inhibitor to diabetic mice reduced kidney hypertrophy, normalized ROS levels, and attenuated podocyte loss.

Mechanisms underlying compensatory kidney hypertrophy are not well characterized but are thought to involve hyperactivation of the protein kinase mTOR (mechanistic target of rapamycin) in complex 1 (mTORC1) (54). Activation of mTORC1 enhances phosphorylation of rpS6, promoting cell growth and proliferation in response to nutrients and growth factors. Rheb-GTP is an obligate activator of mTORC1, and insulin binding to its receptor promotes Rheb-GTP loading (55,56). REDD1 acts as a dominant regulator of mTORC1 signaling and suppresses mTORC1 kinase activity by reducing Rheb-GTP (20). In support of the prior report (57), we observed enhanced phosphorylation of rpS6 in the kidney of STZ diabetic mice, which is consistent with activation of mTORC1. In nondiabetic REDD1 KO mice, rpS6 phosphorylation in the kidney was enhanced as compared with nondiabetic WT mice, and there was no change with STZ diabetes. Consequently, rpS6 phosphorylation was similar in diabetic WT and diabetic REDD1 KO mice. The findings suggest that REDD1 expression may contribute to rpS6 phosphorylation in the kidney of nondiabetic mice but do not support REDD1-dependent mTORC1 suppression in the context of diabetes. Upregulation of mTORC1 signaling in the context of diabetes is paradoxical, as both insulin deficiency and increased REDD1 expression should act counter to its activation. One possibility is that diabetes promotes mTORC1 activation despite a relative reduction in Rheb-GTP. This could potentially be accomplished by enhanced lysosomal localization of mTORC1, as is seen with amino acid signaling (58,59).

Overall, the findings provide new insight into the mechanism whereby type 1 diabetes and hyperglycemic conditions contribute to the development of renal complications. At present, whether REDD1 plays a similar role in the development of renal damage in type 2 diabetes is not known. However, REDD1 expression is upregulated in skeletal muscle of both patients with type 2 diabetes (60) and rodent models of obesity (61), and evidence supports a role for the protein in development of insulin resistance. In the kidney of STZ diabetic mice, REDD1 expression was increased and GSK3β phosphorylation was decreased concomitant with podocytopenia and renal injury. However, REDD1 deletion reduced oxidative stress in the kidney of diabetic mice and prevented both podocyte loss and the resulting glomerular filtration defects. The data are consistent with a mechanism wherein a REDD1/GSK3β signaling axis acts to negatively regulate redox homeostasis in the podocytes resulting in podocytopenia and albuminuria. These proof-of-concept studies support the possibility that therapeutic intervention to repress REDD1 expression in the kidney may offer hope for maintaining renal function in patients with diabetes.

Acknowledgments. The authors thank Dr. Elena Feinstein (Quark Pharmaceuticals) for permission to use the REDD1 KO mice. The preparation of histology sections and H-E staining of tissue sections, carried out by Ellen Mullady and Gretchen Snavely, Department of Comparative Medicine, Penn State College of Medicine, are greatly appreciated.

Funding. This research was supported by the American Diabetes Association Pathway to Stop Diabetes grant 1-14-INI-04, National Institutes of Health grants R01 EY029702 and R01 EY032879 (to M.D.D.), and a Children’s Miracle Network Hospitals Trainee Research Grant (to S.S.).

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

Author Contributions. S.S., S.R.K, and M.D.D. contributed to study conceptualization. S.S., E.I.Y., A.L.T., and M.D.D. contributed to data curation. S.S., E.I.Y., A.L.T., and M.D.D. contributed to formal analysis. S.S. and M.D.D. acquired funding. S.S., E.I.Y., A.L.T., W.P.M., and H.C. performed the investigation. S.S. and M.D.D. visualized the data. S.S., H.C., K.H., S.R.K., and M.D.D. contributed to development of study methodology. S.S. and M.D.D. contributed to writing the manuscript. S.S., E.I.Y., A.L.T., W.P.M., K.H., S.R.K., and M.D.D. reviewed and edited the manuscript. K.H., S.R.K., and M.D.D. provided resources. S.R.K. and M.D.D. supervised the study. M.D.D. 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.

Prior Presentation. Parts of this study were presented in abstract form at the 2022 Experimental Biology annual meeting, Philadelphia, PA, 2–5 April 2022.