Activation of Stimulator of IFN Genes (STING) Causes Proteinuria and Contributes to Glomerular Diseases

Significance Statement

A signaling molecule that plays a role in the innate immune system, stimulator of IFN genes (STING), is a crucial regulator of the cyclic GMP-AMP synthase (cGAS)-STING pathway. This signaling pathway regulates inflammation and energy homeostasis under conditions of obesity, kidney fibrosis, and AKI, but its exact role in the pathogenesis of glomerular diseases remains unclear. The authors found that activation of STING in wild-type mice is sufficient to cause albuminuria and podocyte loss, and that the cGAS-STING signaling pathway is upregulated in mice with experimental diabetic kidney disease and Alport syndrome. They also demonstrated that either genetic or pharmacologic STING inhibition confers protection from kidney disease progression. These findings suggest that this signaling pathway plays an important role in mediating glomerular dysfunction.

Visual Abstract

Abstract

Background

The signaling molecule stimulator of IFN genes (STING) was identified as a crucial regulator of the DNA-sensing cyclic GMP-AMP synthase (cGAS)-STING pathway, and this signaling pathway regulates inflammation and energy homeostasis under conditions of obesity, kidney fibrosis, and AKI. However, the role of STING in causing CKD, including diabetic kidney disease (DKD) and Alport syndrome, is unknown.

Methods

To investigate whether STING activation contributes to the development and progression of glomerular diseases such as DKD and Alport syndrome, immortalized human and murine podocytes were differentiated for 14 days and treated with a STING-specific agonist. We used diabetic db/db mice, mice with experimental Alport syndrome, C57BL/6 mice, and STING knockout mice to assess the role of the STING signaling pathway in kidney failure.

Results

In vitro, murine and human podocytes express all of the components of the cGAS-STING pathway. In vivo, activation of STING renders C57BL/6 mice susceptible to albuminuria and podocyte loss. STING is activated at baseline in mice with experimental DKD and Alport syndrome. STING activation occurs in the glomerular but not the tubulointerstitial compartment in association with autophagic podocyte death in Alport syndrome mice and with apoptotic podocyte death in DKD mouse models. Genetic or pharmacologic inhibition of STING protects from progression of kidney disease in mice with DKD and Alport syndrome and increases lifespan in Alport syndrome mice.

Conclusion

The activation of the STING pathway acts as a mediator of disease progression in DKD and Alport syndrome. Targeting STING may offer a therapeutic option to treat glomerular diseases of metabolic and nonmetabolic origin or prevent their development, progression, or both.

CKD ultimately leading to kidney failure is a major health problem worldwide with no definitive cure. Primary glomerular disorders of metabolic and nonmetabolic origin are characterized by glomerulosclerosis and interstitial fibrosis and include many conditions with a variety of genetic and environmental causes. Diabetic kidney disease (DKD), a glomerular disorder of metabolic origin, is the most common single cause of ESKD in the United States¹ characterized by podocyte loss as a clinical feature and is an independent predictor of disease progression in patients with type 1 diabetes (T1D) and type 2 diabetes (T2D).^2–5 Alport syndrome (AS) is a genetic disorder caused by mutations in the gene coding for collagen type IV chains,^6,7 leading to structural abnormalities and dysfunction of the glomerular basement membrane, podocyte detachment, and ESKD.

Low-grade systemic and chronic inflammation are reported in both DKD and AS. Indeed, a recent study that examined the proteome demonstrated presence of a kidney risk inflammatory signature in patients with T1D or T2D, which consists of 17 circulating inflammatory proteins and is associated with ESKD progression.⁸ Similarly, in AS, glomerular damage due to the altered composition of the glomerular basement membrane further prompts a sustained inflammatory response leading to fibrosis^9,10 and resulting in irreversible loss of kidney function. Much less is known about the contribution of local inflammation to CKD progression. Interestingly, many studies report that podocytes exhibit features of immune cells, expressing various components of the adaptive^11–15 and innate^16–20 immune systems, which indicate that podocytes may have characteristics of nonhematopoietic professional antigen-presenting cells²¹ and may be involved in the local immune response and contribute to chronic inflammation and glomerular damage. We also recently reported an important role of glomerular-derived TNF in the pathogenesis of glomerular diseases.²² However, a significant gap of knowledge exists explaining the role of podocytes in inflammation in DKD and AS.

As an evolutionarily conserved system, the innate immune system relies on pattern recognition receptors, such as Toll-like receptors (TLRs), C-type lectin receptors, RIG-I–like receptors, intracellular Nod-like receptors, HIN-200 receptors, and the recently discovered cyclic GMP-AMP synthase (cGAS)-stimulator of IFN genes (STING) pathway: cGAS-STING. Interestingly, the predominant role of TLR2, TLR4, and pyrin domain containing 3 (NLRP3), as a downstream target of TLRs, was demonstrated in DKD, where the activation of proinflammatory responses were observed (reviewed by Panchapakesan and Pollock²³ and Tang and Yiu²⁴). Less is known about the role of innate immunity in the progression of AS, but a role of macrophage infiltration^25,26 and involvement of the glomerular endothelial cells,²⁷ but not of pattern recognition receptors, has been proposed.

More recent studies suggest a connection between the cGAS-STING pathway and a state of chronic inflammation. The cGAS-STING pathway regulates inflammation and energy homeostasis in obesity^28,29 in SLE,^30,31 in tubular epithelial cells in a mouse model of CKD,³² and in AKI.^33,34 STING is now recognized as a key player in insulin resistance and obesity-induced chronic low-grade inflammation.³⁵ In human podocytes a crosstalk between the activation of STING and increased expression of apolipoprotein 1 (APOL1) has been described, suggesting that STING activation may be a second hit leading to kidney disease development and progression in patients with lupus nephritis.³⁶ Although the crucial role of the cGAS-STING pathway in immune cells and the immune defense have been extensively investigated,^37–39 its function in nonimmune cells, such as podocytes, remains largely unknown. Moreover, whether STING contributes to the development and/or progression of glomerular diseases such as DKD and AS remains elusive. In this study, we tested the hypothesis that activation of STING causes podocyte and glomerular injury, contributing to kidney damage.

Terminally differentiated human and murine podocytes were utilized as a model system in vitro to investigate the expression of the cGAS-STING signaling pathway components and activity. For in vivo studies we used db/db mice as a model of T2D associated with DKD and Col4a3^−/− mice as an experimental model of kidney disease associated with AS to determine the presence and activity of the cGAS-STING pathway in the kidney. Here we report that both human and murine podocytes express all components of the cGAS-STING signaling pathway and that treatment with cyclic di-adenosine monophosphate (c-diAMP), an agonist of STING, leads to increased phosphorylation of STING, IRF3, and TBK1. Activation of STING in wild-type mice is associated with proteinuria and podocyte loss. In further support, we found that the cGAS-STING signaling pathway was also activated in kidney cortices and glomeruli of db/db mice and mice with AS and was associated with albuminuria. STING activation in all three mouse models occurs in the glomerular but not the tubulointerstitial compartment, and is not associated with changes in immune cell infiltration. Pharmacologic or genetic inhibition of STING was sufficient to ameliorate diabetes- or streptozotocin (STZ)-induced and AS-associated glomerular injury and kidney failure. Our findings demonstrate an important role of the cGAS-STING signaling pathway in mediating glomerular dysfunction and suggest that STING may represent a new potential target for treatment of patients with glomerular diseases of metabolic and nonmetabolic origin.

Methods

Reagents and Materials

RPMI 1640 media for human podocyte culture were obtained from Corning (Corning, NY; #10-040-CV). DMEM/F12 (1:1) media for murine podocyte culture (#11320-033), FBS (#26140-079), and penicillin/streptomycin (#15140-163) were purchased from Gibco | Thermo Fisher Scientific (Emeryville, CA). Insulin-transferrin-selenium (#25-800-CR) was obtained from Corning. cyclic di-adenosine monophosphate (c-diAMP; #SML1231-1UMO), 9-oxo-10(9H)-acridineacetic acid (CMA; #17927-250MG), C-176 (#SML2559-50MG), corn oil (#C8267-500ML), STZ (#S0130-500MG), and citrate buffer (#S4641-25G) were obtained from Sigma-Aldrich (St. Luice, St. Louis, MO). Amicon Ultra centrifugal filters to concentrate cytosolic fractions after subcellular fractionation (#UFC500324) were purchased from Millipore Sigma (Burlington, MD). RNeasy kit (#74106) was obtained from Qiagen (Germantown, MD). GoTaq green master mix (#M7123) was purchased from Promega (Madison, WI). qScript cDNA Supermix (#95048-500) and SYBR green ROX mix (#95073-012) were obtained from Quantabio (Beverley, MA). Pierce BCA Protein Assay kit (#23225) was purchased from Invitrogen | Thermo Fisher Scientific (Emeryville, CA). SDS-PAGE gels (#456-1094) and Trans-Blot Turbo Transfer Membranes (#1704156) for Western blot were purchased from Bio-Rad Laboratories (Hercules, CA). Albumin ELISA kit (#E90-134) was obtained from Bethyl Laboratories (Montgomery, TX) and creatinine kit (#0420-500) was obtained from Stanbio (Boerne, TX). Sieve strainers (#352350, 70 μm; #352360, 100 μm) were purchased from Falcon Biological (Kernville, CA). Hydrogen peroxidase blocking reagent (#GR3415498), DAB substrate kit (#GR3422585), and TUNEL assay kit (#ab66110) were obtained from Abcam (Branford, CT).

PCR Analysis

mRNA from human and murine podocytes and mRNA from murine kidney cortices and glomeruli was extracted using the RNeasy kit according to the manufacturer’s protocol. A total of 250 ng RNA was reverse transcribed using the high-capacity cDNA reverse transcription kit according to the manufacturer’s protocol. PCR was performed using the GoTaq green master mix at 60°C, 35 cycles for standard PCR. For RT-PCR the SYBR green ROX mix was used. Primer sequences used in the study are listed in Supplemental Table 1.

Protein Extraction and Western Blot Analysis

Podocytes or tissue from kidney cortices, glomeruli, or tubule fractions of mice were homogenized in ice-cold CHAPS buffer supplemented with protease and phosphatase inhibitor cocktails. Protein concentration was quantified using the Pierce BCA Protein Assay kit according to the manufacturer’s protocol. Samples were prepared in 4× Laemmli buffer, and 15 μg of protein was loaded onto 4%–20% SDS-PAGE gels followed by transfer onto polyvinylidene fluoride membranes. Membranes were blocked in 3% BSA for phosphoproteins and in 5% skim milk for total proteins at 4°C, overnight, followed by incubation with primary antibodies at 4°C, overnight. Blots were scanned using Azure Biosystems C600 imaging system (Dublin, CA). Antibodies used in the study and their respective dilution are reported in Supplemental Table 2. Uncropped blot images are shown in Supplemental Figure 1.

Cell Death Analysis

Apoptosis was assessed using the Caspase-3 ApoTox-Glo Triplex Assay (Promega) according to the manufacturer’s instructions. Briefly, differentiated podocytes were treated with 10 μM c-diAMP for 24 hours, and caspase-3 activity was determined after 2 hours at excitation 470 nm/emission 520 nm for the cytotoxicity, excitation 400 nm/emission 505 nm for viability and luminescence. Values were expressed as fold change to controls.

Autophagic cell death was assessed using expression of LC3B and Beclin-1 by Western blot in differentiated podocytes treated with 10 μM c-diAMP for 24 hours.

Animal Studies

All animal studies complied with the relevant ethical regulations and were performed in accordance with the National Institutes of Health Guidelines. The study protocol was approved by the Institutional Animal Care and Use Committee of the University of Miami, Miller School of Medicine.

The animals were housed in the animal facility of the Division of Veterinary Resources, University of Miami, Miller School of Medicine, on 12-hour light/dark cycles under controlled temperature (22°C±1°C) and were provided with water and Teklad Global 18% protein rodent chow diet with no restrictions.

c-diAMP Treatment of Mice

Eight-week-old male and female C57BL/6J mice (#000664; The Jackson Laboratory, Bar Harbor, ME) were randomly divided into two groups: (1) control (n=7) and (2) intraperitoneally injected with a single dose of c-diAMP, 25 μg/g (n=9). Spot morning urines were collected 6, 12, 24, 48, and 72 hours after c-diAMP injection. The animals were euthanized 72 hours after injection with c-diAMP, and kidneys were harvested and processed for in-depth phenotypic analysis.

C-176 Treatment of Mice

C-176 (750 nM), a specific inhibitor of STING, or 5% DMSO was dissolved in 100 μl of corn oil and injected intraperitoneally daily for 4 weeks.

Four-week-old male and female Col4a3 knockout (Col4a3^−/−) and wild-type (Col4a3^+/+) mice (129-Col4a3^tm1Dec/J, #002908; The Jackson Laboratory) were randomly divided into four groups: (1) Col4a3^+/+ DMSO (n=7), (2) Col4a3^+/+ C-176 (n=7), (3) Col4a3^−/− DMSO (n=7), and (4) Col4a3^−/− C-176 (n=7). For the mortality study, three groups of mice, Col4a3^+/+ (n=7), Col4a3^−/− DMSO (n=11), and Col4a3^−/− C-176 (n=14), were used.

Ten-week-old male and female diabetic (mice were considered diabetic when glucose levels were ≥250 mg/dl) db/db mice (B6.BKS(D)-Lepr<db>/J, #000697; The Jackson Laboratory) were used in the study. Animals were randomly divided into four groups: (1) db/+ DMSO (n=5), (2) db/+ C-176 (n=5), (3) db/db DMSO (n=7), and (4) db/db C-176 (n=8).

In both studies, spot morning urine was collected weekly. For db/db mice, random glycemia levels were also measured weekly. The animals were then euthanized, and kidneys were harvested and processed for in-depth phenotypic analysis. Blood samples were collected at euthanization and analyzed as described below.

Induction of Diabetes by Multiple Low Doses of STZ

Eight-week-old male and female C57BL/6J (#000664; The Jackson Laboratory) and MPYS^−/− STING knockout (#025805; The Jackson Laboratory) mice were given daily intraperitoneal injections of 40 mg/kg STZ dissolved in 0.1 M sodium citrate (pH 4.5) for 5 consecutive days. Animals were randomly divided into four groups: (1) B6 control (n=10), (2) B6 STZ (n=7), (3) STING^−/− (n=10), and (4) STING^−/− STZ (n=6). Blood glucose levels, body weight, and diabetes incidence were monitored weekly for the first 4 weeks. Mice were considered diabetic when glucose levels were ≥250 mg/dl for at least two consecutive measurements under nonfasting conditions. Body weight and random glycemia levels were then measured bi-weekly, and spot morning urines were collected every two weeks. Animals were euthanized 12 weeks after STZ injections, and kidneys were harvested and processed for in-depth phenotypic analysis. Blood samples were collected at euthanization and analyzed as described below.

Urine Sample Analysis

The urinary albumin content was measured by sandwich ELISA following the manufacturer’s protocol. The urinary creatinine was measured by an assay based on the Jaffe method, using albumin ELISA kit and creatinine kit. Values are expressed as microgram of albumin per milligram of creatinine.

Blood Sample Analysis

Blood samples were analyzed for lipid panel, aspartate aminotransferase, alanine transaminase, and BUN in the Comparative Laboratory Core Facility of the University of Miami. Serum creatinine was determined by tandem mass spectrometry at the University of Alabama at Birmingham–University of California San Diego O’Brien Core Center (University of Alabama, Birmingham, AL) as previously described.⁴⁰

Histology and Assessment of Mesangial Expansion

After perfusion of an animal with 1× PBS, the right kidney was removed for histologic analysis, and the left kidney was harvested for glomeruli isolation. Periodic acid–Schiff (PAS) and hematoxylin-eosin staining of paraffin-embedded kidney sections (4-μm-thick) was performed using a standard protocol. Histologic images were visualized using a light microscope (Olympus BX 41, Tokyo, Japan) at ×40 magnification and analyzed using ImageJ software.⁴¹ Sixty glomeruli per section were analyzed for mesangial expansion by semiquantitative analysis (scale 0–4) performed by two blinded independent investigators.^42,43

Glomeruli Isolation

The left kidney was harvested and mashed in 1× PBS buffer and processed through sieving steps using 100-μm and 70-μm Falcon sieve strainers. The final isolate was washed in 1× PBS and collected in tubes. Glomeruli were pelleted at 3000 × g at 4°C and used to extract mRNA as described above.

Transmission Electron Microscopy

For ultrastructural analyses, samples were fixed in 2% paraformaldehyde/2.5% glutaraldehyde (Polysciences Inc., Warrington, PA) in 100 mM sodium cacodylate buffer, pH 7.2 for 2 hours at room temperature and then overnight at 4°C. Samples were washed in sodium cacodylate buffer at room temperature and postfixed in 1% osmium tetroxide (Polysciences Inc.) for 1 hour. Samples were then rinsed extensively in dH₂0 before en bloc staining with 1% aqueous uranyl acetate (Ted Pella Inc., Redding, CA) for 1 hour. Following several rinses in dH₂0, samples were dehydrated in a graded series of ethanol and embedded in Eponate 12 resin (Ted Pella Inc.). Sections of length 95 nm were cut with a Leica Ultracut UCT ultramicrotome (Leica Microsystems Inc., Bannockburn, IL), stained with uranyl acetate and lead citrate, and viewed on a JEOL 1200 EX transmission electron microscope (JEOL USA Inc., Peabody, MA) equipped with an AMT 8-megapixel digital camera and AMT Image Capture Engine V602 software (Advanced Microscopy Techniques, Woburn, MA).

Wilms’ Tumor 1 Immunofluorescence Staining

To assess podocyte numbers in murine glomeruli, immunofluorescence labeling with anti–Wilms’ tumor 1 (WT1) was performed. WT1-positive nuclei were counted in ten consecutive glomerular cross sections per animal (n=3 animals per group reported) by two blinded investigators.⁴⁴ Fresh-cut 4-μm-thick kidney cortex tissue in OCT was used. No fixation agent was applied. Permeabilization was performed using 0.3% Triton X-100 in 1× PBS for 10 minutes at room temperature. A blocking step was performed using 5% BSA, 2.5% FBS in PBS for 1 hour at room temperature. Primary antibodies were used for 24 hours at room temperature. Alexa-conjugated secondary antibodies were used for 1 hour at room temperature. To detect nuclei, DAPI staining was applied in a 1:500 dilution for 20 minutes at room temperature. Images were acquired by laser scanning confocal microscopy using a Leica SP5 inverted microscope, 40× wet objective (Leica Microsystems CMS GmbH, Mannheim, Germany). Measuring of cell fluorescence was performed using ImageJ software.⁴⁵

STING Immunoperoxidase Staining

Immunoperoxidase was used to evaluate STING expression in fixed kidney sections from mice. Paraffin-embedded kidney sections (4-μm-thick) were incubated with hydrogen peroxidase blocking solution for 15 minutes at room temperature followed by protein block for 30 minutes at room temperature. Anti-STING antibodies (Novus Biologicals) were applied in 1:300 dilution overnight at room temperature followed by incubation with goat anti-rabbit IgG H&L antibodies. DAB substrate was added to the sections for 2 minutes at room temperature followed by counterstaining with hematoxylin (Sigma-Aldrich) for 7 minutes at room temperature. Histologic images were visualized using a light microscope (Olympus BX 41) at ×20 magnification. Negative control is shown in Supplemental Figure 2A.

Immunofluorescence Staining

Paraffin-embedded kidney sections (4-μm-thick) were deparaffinized in xylene, rehydrated by passing an ethanol serial gradient, and baked at 100°C for 10 minutes. Permeabilization was performed using 0.3% Triton X-100 in 1× PBS for 10 minutes at room temperature. A blocking step was performed using 5% BSA, 2.5% FBS in PBS for 1 hour at room temperature. Primary antibodies (MAC2 for macrophage infiltration and LC3B/synaptopodin for autophagy detection in glomeruli) were used for 24 hours at room temperature. Alexa-conjugated secondary antibodies were used for 1 hour at room temperature. To detect nuclei, DAPI staining was applied in a 1:500 dilution for 20 minutes at room temperature. Images were acquired by high-throughput VS120 Olympus slide scanner, ×4-×20 air lenses, four channels: DAPI, FITC, TRITC, and Cy5. Negative control for MAC2 is shown in Supplemental Figure 2, B and C. Positive control for MAC2 is shown in Supplemental Figure 2D.

TUNEL assay was carried out to examine apoptosis in kidney tissue according to the manufacturer’s protocol. Briefly, paraffin-embedded kidney sections (4-μm-thick) were deparaffinized in xylene, rehydrated by passing an ethanol serial gradient, and baked at 100°C for 10 minutes followed by incubation in 20 μg/ml proteinase K solution (Tris-HCl pH 8.0, 50 mM EDTA) for 5 minutes at room temperature. Next 100 μl of DNA labeling solution was applied to each slide, and slides were incubated for 1 hour at 37°C. The slides were then washed with PBS, followed by incubation with anti–BrdU-Red antibodies (2.5 μl per reaction) for 30 minutes at room temperature. DNA was counterstained using 7-AAD/RNase A staining buffer for 30 minutes at room temperature. Images were acquired by high-throughput VS120 Olympus slide scanner, ×4-×20 air lenses, four channels: DAPI, FITC, TRITC, and Cy5.

Statistical Analyses and Study Design

Data are expressed as mean±SD. A number of experiments ranging between three and five were utilized and are indicated for each distinct experiment. Data collection was randomized for all experiments. Experiments were blinded for imaging and data analyses. No sample size calculations were performed for in vitro studies. Sample sizes for in vitro studies were determined to be adequate based on the magnitude and consistency of measurable differences between groups. Minimal group sizes for in vivo studies were determined via power calculator using G Power (https://www.psychologie.hhu.de/arbeitsgruppen/allgemeine-psychologie-und-arbeitspsychologie/gpower.html) with an α of 0.05 and effect size (d) of 2 (using guidelines from Charan and Kantharia⁴⁶). Statistical analyses were performed using the GraphPad Prism, version 7.0 (GraphPad Software Inc.). Data were tested for normality using the Kolmogorov-Smirnov or Shapiro-Wilk test, and equality of variance was confirmed using the F-test. Two groups of data were compared using the one-tailed unpaired t test. Three and more groups of data were compared using one-way ANOVA or two-way ANOVA, followed by Tukey’s post hoc tests. P<0.05 was taken to indicate statistical significance.

Results

Expression of the Cytosolic DNA-Sensing Genes Is Increased in Kidneys of Mouse Models of DKD and AS

To determine whether the cytosolic DNA- or RNA-sensing pathways are affected in glomerular diseases, we performed quantitative real-time PCR analysis of kidney cortices from db/db (mouse model of DKD) and Col4a3^−/− (mouse model of AS) mice. Here, we analyzed mRNA expression of the RNA-sensing pathway molecules (RIG-I [retinoic acid-inducible gene I] and IFIH1 [IFN induced with helicase C domain 1]) and DNA-sensing pathway molecules (STING, ZBP1 [Z-DNA binding protein 1], AIM2 [absent in melanoma 2], DHX9 [DExH-Box helicase 9], DDX41 [DEAD-Box helicase 41], DDX60 [DExD/H-Box helicase 60], and MRE11A [double-strand repair protein MRE11 A]). Interestingly, among the tested molecules only Sting expression was significantly elevated in both (DKD and AS) mouse models, whereas Aim2 expression was increased in Col4a3^−/− mice only (Figure 1).

Figure 1.

mRNA expression heat map of genes related to the RNA- and DNA-sensing pathways in kidney cortices from mouse models of DKD and experimental AS. mRNA isolated from kidney cortices of control (Col4a3^+/+, n=3) and AS (Col4a3^−/−, n=3) mice and control (db/+, n=4) and diabetic (db/db, n=4) mice was used to obtain reverse transcription cDNA followed by quantitative real-time PCR. Upregulated mRNA expression is highlighted in yellow background color and downregulated mRNA expression is highlighted in blue background color. Significant changes are marked with an asterisk. Rig-I, retinoic acid-inducible gene I; Ifih1, IFN induced with helicase C domain; Sting, stimulator of IFN genes; Zbp1, Z-DNA binding protein 1; Aim2, absent in melanoma 2; Dhx9, DExH-Box helicase 9; Ddx41, DEAD-Box helicase 41; Ddx60, DExD/H-Box helicase 60; Mre11a, double-strand repair protein MRE11 A.

Additionally, reanalysis of microarray datasets obtained from the glomeruli of patients of the Pima Indians cohort, FSGS NEPTUNE cohort, and FSGS Miami cohort also demonstrated increased expression of the cGAS-STING signaling pathway (Supplemental Figure 3) and other inflammatory-related signaling pathways (Supplemental Data 1 and 2).

These results support the role of STING in mediating glomerular injury in CKD, such as DKD and AS.

Human and Murine Podocytes Express All of the Components of the cGAS-STING Signaling Pathway

We next confirmed that immortalized human and murine podocytes express cGAS, STING, TBK1, and IRF3 at the mRNA (Supplemental Figures 4A and 5A) and protein (Supplemental Figures 4B and 5B) levels under physiologic conditions.

As STING is a critical mediator for the production of IFNs, agonist of mouse STING-like c-diAMP elicits a potent immune response.⁴⁷ We therefore treated human and murine podocytes with c-diAMP and found, as expected, increased expression of cGAS, STING, TBK1, and IRF3, as well as IFN-β1, one of the downstream targets of STING, at mRNA (Supplemental Figures 4C and 5C) and protein (Supplemental Figures 4D, 4E, 5D, and 5E) levels. Notably, using the subcellular fractionation technique, we were able to confirm that in human podocytes activated STING is localized in the cytosol, whereas the inactive form of STING is localized in the endoplasmic reticulum (Supplemental Figure 4F).

c-diAMP–Mediated STING Activation in Mice Causes Glomerular Pathology and Proteinuria

c-diAMP, initially discovered as a regulator of import and export of potassium ions,^48,49 inhibits STING with low micromolar affinity, even in the presence of 1 mM guanosine derivatives (GTP) or ATP.⁵⁰ To further explore the role of STING as a potential contributor to chronic inflammation in kidney disease, C57BL/6 mice were injected intraperitoneally with c-diAMP (25 μg/g, once) or similar volume of vehicle (0.9% normal saline) as a control. The urine albumin-creatinine ratio (ACR) was significantly increased in c-diAMP–treated mice 48 hours after injection compared with control mice and persisted until 72 hours (time of euthanization) (Figure 2A), and no changes in body weight were observed (Figure 2B). PAS staining of kidney cortex sections was performed and revealed no changes in mesangial expansion in c-diAMP–treated mice compared with controls (Figure 2C). WT1 immunofluorescence staining (podocyte nuclear-specific antibody) to quantify podocyte number per glomerular cross section was performed. We found that c-diAMP–treated mice have significantly decreased numbers of podocytes per glomerular section compared with untreated controls (Figure 2D). Furthermore, using transmission electron microscopy, we observed that c-diAMP–treated mice experience a decrease in the number of foot processes per length of the glomerular basement membrane (Figure 2E). Additionally, we show that c-diAMP treatment leads to an increased expression of the components of the cGAS-STING pathway in kidney cortices at the protein level compared with untreated control mice (Supplemental Figure 6). Next, we aimed to distinguish whether the STING activation is restricted to the glomeruli or affects also the tubulointerstitial compartment in c-diAMP–treated mice. Using immunoperoxidase staining against STING, we found that glomeruli have higher STING expression in response to c-diAMP treatment compared with tubular epithelium (Figure 2F). Furthermore, the Western blot analysis of lysates from glomeruli and tubular fractions demonstrated higher STING phosphorylation in glomeruli, but not in tubular fraction from c-diAMP–treated mice (Figure 2G). Because activation of STING regulates IFN-mediated inflammation in many cell types, we next evaluated changes in immune cell infiltration in the kidneys of c-diAMP–treated mice. Macrophage content was determined by immunostaining using MAC2 antibodies against activated macrophages. No presence of activated macrophages in kidneys of c-diAMP–treated mice were noticed (Figure 2H). To evaluate a possible sex-related contribution into STING-mediated kidney damage, we divided male versus female where it was acceptable. No differences between males and females in the analyzed parameters were noticed (Figure 2).

Figure 2.

Activation of STING in mice leads to impaired renal function in wild-type mice. (A) Urine ACR of 8-week-old control (CTRL, n=7 animals) and C57BL/6 mice treated with c-diAMP (25 µg/g) for 72 hours (c-diAMP, n=9 animals). **P<0.01, two-tailed t test. (B) Bar graph analysis of body weight in CTRL and c-diAMP mice. (C) Representative PAS staining of kidney sections (4-µm-thick) and bar graph analysis (lower panel) of the scores for mesangial expansion in CTRL and c-diAMP mice. Original magnification ×40; scale bar, 30 µm. (D) Representative immunostaining for anti-WT1 (green) in 4-µm-thick OCT-frozen kidney sections and bar graph analysis (lower panel) for the number of podocytes per glomeruli in CTRL and c-diAMP mice. Original magnification ×40; scale bar 20 µm. *P<0.05, two-tailed t test. (E) Representative transmission electron microscopy (TEM) and bar graph analysis (lower panel) of foot process effacement (yellow arrow) in CTRL and c-diAMP mice. Original magnification ×10,000; scale bar, 500 nm. *P<0.05, two-tailed t test. (F) Representative immunoperoxidase staining for STING in 4-µm-thick kidney sections in CTRL and c-diAMP mice. Original magnification ×40; scale bar, 20 µm. (G) Representative Western blot (left panel) and bar graph analysis (right panel) of STING phosphorylation in tubules and glomeruli isolated from CTRL and c-diAMP mice. GAPDH served as a loading control. *P<0.05, two-tailed t test. (H) Representative immunostaining for macrophage infiltration (MAC2, red) in 4-µm-thick kidney sections in all four groups of mice (n=3 mice per group). Nuclei are stained with DAPI (blue). An empty channel color (green) was applied and resulted merged color (yellow) represents the levels of autofluorescence. Original magnification ×20; scale bar, 100 µm. For all bar graphs, blue dots correspond to male data and black dots correspond to female data.

To further confirm our hypothesis that STING activation contributes to kidney injury, we then used another antiviral small molecule, 9-oxo-10(9H)-acridineacetic acid (CMA, which binds directly to STING and triggers a type I IFN response through the TBK1/IRF3 pathway,⁵¹ but which is not a cyclic dinucleotide. C57BL/6 mice were intraperitoneally injected with DMSO (5%; CTRL) or CMA (224 mg/kg) and euthanized 72 hours after treatment. Like in c-diAMP–treated mice, mice treated with CMA developed increased ACR levels starting 48 hours after injection, which persisted 72 hours after injection (Supplemental Figure 7A), and no changes in body weight (Supplemental Figure 7B) or in glomerular histology as demonstrated by PAS staining (Supplemental Figure 7C and Supplemental Figure 7D) were found. Even CMA was previously reported to induce type I IFN expression in mice (using PBMCs),⁵¹ and our data demonstrate that CMA affects outcome to a lesser effect compared with c-diAMP.

Overall, these data highlight that STING activation by small molecule such as c-diAMP or CMA plays an important role in mediating kidney damage, primarily affecting the glomeruli.

STING Contributes to the Development of Glomerular Disease of Metabolic Origin

To validate our findings in the context of CKD, we used db/db mice as a model of T2D-associated DKD. Interestingly, Sting mRNA expression in both kidney cortices (Supplemental Figure 8A) and isolated glomeruli (Figure 3A) was elevated in db/db mice compared with db/+ controls. Additionally, we found that Sting mRNA expression in kidney cortices positively correlates with ACR in db/db mice (Figure 3B). Analysis of protein expression in kidney cortices revealed increased cGAS levels and phosphorylation of STING and IRF3 in db/db mice (Figure 3C). To investigate whether the observed STING activation in the kidney may be compartment specific, we evaluated expression of STING in glomeruli and tubules using immunoperoxidase staining of the kidney sections from db/+ and db/db mice. Our data suggest that STING expression is higher in glomeruli in db/db mice compared with tubular epithelium (Figure 3D). To further confirm these findings, we evaluated STING phosphorylation levels in isolated glomeruli and tubules from db/db mice. Unexpectedly, STING phosphorylation in tubules was significantly lower, whereas levels of phospho-STING in glomeruli were significantly higher (Figure 3E). Taken together, these data suggest that STING activation is restricted to glomeruli and may contribute to the progression of DKD.

Figure 3.

The cGAS-STING signaling pathway is upregulated in a mouse model of DKD. Sixteen-week-old control (db/+) and diabetic (db/db) mice were used in this study. (A) Quantitative real-time PCR analysis of the mRNA expression levels of cGas, Sting, Tbk1, and Irf3 in glomeruli isolated from db/+ and diabetic db/db mice. **P<0.01, ***P<0.001, two-tailed t test. (B) Regression analysis showing a correlation between mRNA expression levels of STING in kidney cortices and urine ACR in db/db mice. (C) Representative Western blot (left panel) and bar graph analysis (right panel) of STING, TBK1, and IRF3 phosphorylation and cGAS levels in kidney cortices of db/+ and db/db mice. n=4 mice per each group. *P<0.05, two-tailed t test. (D) Representative immunoperoxidase staining for STING in 4-µm-thick kidney sections in db/+ and db/db mice. Original magnification ×40; scale bar, 20 µm. (E) Representative Western blot (left panel) and bar graph analysis (right panel) of STING phosphorylation in tubules and glomeruli isolated from db/+ (n=3) and db/db (n=3) mice. GAPDH served as a loading control. *P<0.05, ***P<0.001, two-tailed t test. For all bar graphs, blue dots correspond to male data, black dots correspond to female data, and uncolored dots correspond to sex-mixed data.

To further elucidate the role of STING in glomerular diseases, we hypothesized that STING inhibition may have a protective effect in the development of DKD. In this study we used a potent and selective small molecule inhibitor of STING, C-176, which inhibits STING by covalently binding Cys91 and blocking activation-induced palmitoylation.⁵² The dose used here was based on a previous pharmacokinetics study, targeting a plasma concentration approximately 10× the IC₅₀ from the in vitro assay.⁵² This dose results in strong inhibition of IFN-γ production by 4hours after intraperitoneal injection in mice.

We performed a series of preventive experiments in hyperglycemic 10-week-old db/db mice and normoglycemic db/+ mice using C-176. Although no differences in glycemic levels (Figure 4A), liver function (Supplemental Table 3), and body weight (Figure 4B) were found between db/db mice treated with DMSO (vehicle) or C-176, ACR levels were significantly lower in db/db mice treated with C-176 (Figure 4C) compared with db/db DMSO mice. Additionally, BUN levels were similar between db/db DMSO and db/db C-176 groups (Figure 4D), but serum creatinine levels were significantly higher in db/db DMSO compared with db/+ mice but normalized in db/db mice treated with C-176 (Figure 4E). Next, to study the pathologic changes, we performed PAS staining to quantify mesangial expansion and WT1 immunofluorescence staining to estimate podocyte numbers per glomerular cross section. We found that db/db DMSO mice have a higher mesangial expansion score, whereas C-176 treatment prevented mesangial expansion (Figure 4F). Similarly, db/db DMSO mice are characterized by a decreased number of podocytes per glomerulus, a phenotype that was significantly improved in db/db C-176 mice (Figure 4G).

Figure 4.

Pharmacologic STING inhibition ameliorates kidney injury in a mouse model of DKD. For C-176–induced STING inhibition, four groups of 10-week-old mice were used: (1) db/+ intraperitoneally (ip) injected with 5% DMSO (n=5), (2) db/+ ip injected with 750 nM C-176 (n=5), (3) db/db ip injected with 5% DMSO (n=7), and (4) db/db ip injected with 750 nM C-176 (n=8). (A–C) Random glycemia levels (A), body weight (B), and urine ACR (C) in all four groups of mice. *P<0.05, two-tailed t test when comparing db/db DMSO and db/db C-176. ^$P<0.001, two-tailed t test when comparing db/+ DMSO and db/db DMSO; ^#P<0.001, two-tailed t test when comparng db/+ C-176 and db/db C-176. (D and E) Bar graph analysis of BUN (D) and serum creatinine levels (E) in all four groups of mice. **P<0.01, one-way ANOVA. (F) Representative PAS staining of kidney sections (4-µm-thick) and bar graph analysis (right panel) of the scores for mesangial expansion in all four groups of mice. Original magnification ×40; scale bar, 30 µm. *P<0.05, **P<0.01, one-way ANOVA. (G) Representative immunostaining for anti-WT1 (green) in 4-µm-thick OCT-frozen kidney sections and bar graph analysis (right panel) for the number of podocytes per glomeruli in all four groups of mice (n=3 mice per group). Original magnification ×40; scale bar, 30 µm. *P<0.05, ***P<0.001, one-way ANOVA. (H) Representative immunostaining for macrophage infiltration (MAC2, red) in 4-µm-thick kidney sections in all four groups of mice (n=3 mice per group). Nuclei are stained with DAPI (blue). An empty channel color (green) was applied and resulted merged color (yellow) represents the levels of autofluorescence. Original magnification ×20; scale bar, 100 µm. For all bar graphs, blue dots correspond to male data, black dots correspond to female data, and uncolored dots correspond to sex-mixed data.

Although DKD is not considered to be a primary immune-mediated form of kidney disease, several studies support involvement of many immune system components in DKD progression and even development. To evaluate whether STING-mediated IFN activation may cause changes in immune cell infiltration in the kidneys of db/db mice, we performed immunostaining for active macrophages using MAC2 antibodies. Although diabetic nephropathy monocyte and macrophage infiltration in kidney tissue is well known,^53–55 the presence of renal macrophage infiltration in DKD mouse models is not clear. In our study we failed to detect the presence of activated macrophages in kidneys of db/db mice compared with db/+ controls using MAC2 antibodies (Figure 4H) or F4/80 and CD11b antibodies (data not shown).

Next, we utilized STING knockout mice (STING^−/−) and C57BL/6 controls (CTRL) in a model of STZ-induced T1D (40 mg/kg for 5 consecutive days). We first confirmed that both CTRL and STING^−/− mice injected with STZ developed diabetes, based on the random glycemia levels (Figure 5A) and serologic analysis of liver function (Supplemental Table 4). No changes in body weight were noticed between the groups (Figure 5B). As expected, STZ-treated CTRL mice demonstrated significantly increased ACR levels starting 4 weeks afterthe establishment of diabetes, whereas in STZ-treated STING^−/− mice ACR levels remained unchanged (Figure 5C). Functional analysis of the kidney also demonstrated that BUN (Figure 5D) and serum creatinine (Figure 5E) levels remained unchanged in STZ-injected STING^−/− mice compared with STING^−/− control mice. Moreover, the effect of STZ-induced kidney damage was significantly attenuated in STING^−/− mice compared with CTRL mice based on the histologic analysis of mesangial expansion (Figure 5F) and WT1 immunofluorescence staining of the number of podocytes per glomerular cross section (Figure 5G). Given the fact that the importance of a sex difference contribution to DKD progression in both T1D and T2D is well recognized in the recent studies (reviewed by Shepard⁵⁶), we also evaluated the possible differences of renal outcomes between male and female in the mouse models used in this study. Although no significant changes were observed between male and female in T2D (Figures 3 and 4) or T1D mouse models (Figure 5), a tendency to have higher numbers for every evaluated parameter in males was presented across all of the experiments. Taken together, the data suggest that pharmacologic suppression or genetic deficiency of STING lowered kidney damage in mouse models of DKD (either T1D or T2D) and may be considered a potential therapeutic target to treat DKD.

Figure 5.

STING knockout mice are protected from DKD. For STZ-induced diabetes (STZ, 40 mg/kg) four groups of 8-week-old mice were utilized: (1) C57BL/6 mice intraperitoneally (ip) injected with citric buffer (vehicle) (CTRL), n=10; (2) C57BL/6 mice ip injected with STZ in citric buffer (CTRL STZ), n=7; (3) STING knockout mice ip injected with vehicle (STING^−/−), n=10; and (4) STING knockout mice ip injected with STZ (STING^−/− STZ), n=6. (A and B) Random glycemia levels (A) and body weight (B) in all four groups of mice. ***P<0.001, CTRL versus CTRL STZ; ^#P<0.001, STING^−/− versus STING^−/− STZ, two-tailed t test. (C) Urine ACR in all four groups of mice. *P<0.05, **P<0.01, ***P<0.001, CTRL STZ versus STING^−/− STZ, one-way ANOVA. (D and E) BUN levels (D) and serum creatinine levels (E) in all four groups of mice. **P<0.01, ***P<0.001, one-way ANOVA. (F) Representative PAS staining of kidney sections (4-µm-thick) and bar graph analysis (right panel) of the scores for mesangial expansion 3 months after induction of diabetes with STZ. Original magnification ×40; scale bar, 30 µm. *P<0.05, **P<0.01, one-way ANOVA. (G) Representative immunostaining for anti-WT1 (green) in 4-µm-thick OCT-frozen kidney sections and bar graph analysis (right panel) for the number of podocytes per glomeruli 3 months after induction of diabetes with STZ. Original magnification ×40; scale bar, 20 µm. *P<0.05, one-way ANOVA. For all bar graphs, blue dots correspond to male data, black dots correspond to female data, and uncolored dots correspond to sex-mixed data.

STING Pathway Activation Contributes to Kidney Damage in an Experimental Model of AS (Col4a3^−/−)

To investigate whether STING also contributes to glomerular diseases of nonmetabolic origin, we used mice with experimental AS, where a mutation in the gene coding for the collagen IV α3 chain causes the development of FSGS-like lesions in the kidney. Our data show that Col4a3^−/− mice have significantly increased expression of Sting, Tbk1, and Irf3 at mRNA levels in kidney cortices (Supplemental Figure 8B) and isolated glomeruli (Figure 6A), compared with control mice (Col4a3^+/+). Similar to db/db mice, Sting mRNA expression in kidney cortices positively correlates with ACR levels in Col4a3^−/− mice (Figure 6B). Analysis of the cGAS-STING signaling pathway expression at the protein level also revealed significantly increased levels of cGAS and phosphorylation of STING and TBK1 in Col4a3^−/− mice (Figure 6C). Interestingly, immunoperoxidase staining of the kidney sections against STING demonstrated that the STING activation is restricted to the glomeruli and not to tubular epithelium in mice with AS (Figure 6D). Moreover, this observation was confirmed by the Western blot analysis of lysates from glomeruli and tubular fractions collected from Col4a3^−/− mice (Figure 6E).

Figure 6.

The cGAS-STING signaling pathway is upregulated in the experimental AS. Eight-week-old control (Col4a3^+/+) and Col4a3^−/− mice, a mouse model for kidney disease associated with AS, were used in this study. (A) Quantitative real-time PCR analysis of the mRNA expression levels of cGas, Sting, Tbk1, and Irf3 in glomeruli isolated from Col4a3^+/+ and Col4a3^−/− mice. *P<0.05, **P<0.01, ***P<0.001, two-tailed t test. (B) Regression analysis showing a correlation between mRNA expression levels of STING in kidney cortices and urine ACR in Col4a3^−/− mice. (C) Representative Western blot (left panel) and bar graph analysis (right panel) of STING, TBK1, and IRF3 phosphorylation and cGAS levels in kidney cortices of Col4a3^+/+ and Col4a3^−/− mice. n=3 mice per group. *P<0.05, **P<0.01, ***P<0.001, two-tailed t test. (D) Representative immunoperoxidase staining for STING in 4-µm-thick kidney sections in Col4a3^+/+ and Col4a3^−/− mice. Original magnification ×40; scale bar, 20 µm. (E) Representative Western blot (left panel) and bar graph analysis (right panel) of STING phosphorylation in tubules and glomeruli isolated from Col4a3^+/+ (n=3) and Col4a3^−/− (n=3) mice. GAPDH served as a loading control. *P<0.05, two-tailed t test. For all bar graphs, blue dots correspond to male data, black dots correspond to female data, and uncolored dots correspond to sex-mixed data.

We next tested whether pharmacologic suppression of STING using C-176 protects from loss of kidney function in Col4a3^−/− mice. Four-week-old Col4a3^−/− and Col4a3^+/+ mice were injected with C-176 (750 nM) or vehicle (5% DMSO) daily for 4 weeks. Inhibition of STING in Col4a3^−/− mice resulted in significantly lower ACR levels starting 1 week after the treatment was initiated (Figure 7A). However, no body weight improvement in C-176–treated Col4a3^−/− mice was observed (Figure 7B). Expectedly, analysis of renal function demonstrated that C-176–treated Col4a3^−/− mice have significantly decreased BUN (Figure 7C) and serum creatinine (Figure 7D) levels. Histopathologic analysis of the kidney tissue revealed decreased mesangial expansion and fibrosis levels (Figure 7E) in C-176–treated Col4a3^−/− mice. Moreover, WT1 immunofluorescence staining of the number of podocytes per glomerular cross section showed that C-176–treated Col4a3^−/− mice have a higher number of podocytes compared with untreated Col4a3^−/− mice (Figure 7E). Macrophage content was determined by immunostaining using MAC2 antibodies against activated macrophages. No presence of activated macrophages in kidneys of Col4a3^−/− mice was noticed (Figure 7E). Moreover, using transmission electron microscopy, we were able to demonstrate that STING inhibition is beneficial in terms of preserving podocyte foot processes in C-176–treated Col4a3^−/− mice compared with untreated Col4a3^−/− mice (Figure 7F).

Figure 7.

Pharmacologic STING inhibition improves renal function and increases the lifespan of mice with experimental AS. For C-176–induced STING inhibition four groups of 4-week-old mice were used: (1) control mice injected with 5% DMSO (Col4a3^+/+ DMSO), n=5; (2) control mice injected with injected with C-176 (Col4a3^+/+ C-176), n=5; (3) AS mice injected with 5% DMSO (Col4a3^−/− DMSO), n=10; and (4) AS mice injected with C-176 (Col4a3^−/− C-176), n=10. (A and B) Urine ACR (A) and body weight (B) in all four groups of mice. *P<0.05, **P<0.01, Col4a3^−/− DMSO versus Col4a3^−/− C-176, two-tailed t test. (C and D) BUN levels (C) and serum creatinine levels (D) in all four groups of mice. **P<0.01, ***P<0.001, one-way ANOVA. (E) Representative PAS staining for mesangial expansion (left panel; original magnification ×40; scale bar, 30 µm), Picrosirius red (PSR) staining of kidney sections (4-µm-thick) for fibrosis levels (second from the left panel; original magnification ×40; scale ba, 30 µm), immunostaining for anti-WT1 in 4-µm-thick kidney sections for the number of podocytes per glomeruli (second from the right panel; original magnification ×20; scale bar, 30 µm), and immunostaining for macrophage infiltration (MAC2, red) in 4-µm-thick kidney sections (right panel; original magnification ×20; scale bar, 100 µm) in all four groups of mice (n=3 mice per group). Nuclei are stained with DAPI (blue). An empty channel color (green) was applied and resulted merged color (yellow) represents the levels of autofluorescence. (F) Representative transmission electron microscopy (TEM) and bar graph analysis of foot process effacement (yellow arrow) in Col4a3^−/− DMSO and Col4a3^−/− C-176 mice. Original magnification ×10,000; scale bar, 500 nm. *P<0.05, two-tailed t test. (G) Kaplan-Meier mortality curve showing lifespan in Col4a3^+/+ (n=7), Col4a3^−/− DMSO (n=11), and Col4a3^−/− C-176 mice (n=14). For all bar graphs, blue dots correspond to male data, black dots correspond to female data, and uncolored dots correspond to sex-mixed data.

Given the fact that Col4a3^−/− mice die around 8 weeks of age due to kidney failure, we performed a mortality study as a separate series of experiments in Col4a3^−/− mice. Four-week-old Col4a3^−/− mice were daily injected with C-176 (Col4a3^−/− C-176 group) of 5% DMSO (Col4a3^−/− DMSO group) for 4 weeks. As expected, STING inhibition in Col4a3^−/− mice resulted in a significantly extended lifespan with a median survival of 44 days in the Col4a3^−/− C-176 group versus 25 days in the Col4a3^−/− DMSO group (Figure 7G). No differences between male and female in every analyzed parameter were noticed (Figures 6 and 7).

STING Activation Leads to Podocyte Injury and Death

To delineate the type of cell death that may be associated with STING activation, differentiated human podocytes were treated with 10 μM c-diAMP for 24 hours. Apoptosis evaluation demonstrated higher levels of caspase-3 activity in c-diAMP–treated podocytes compared with control cells (Figure 8A). Interestingly, activation of autophagy in c-diAMP–treated cells was also found, as indicated by Western blot analysis against autophagy marker light chain isoform B (LC3B) and its downstream partner Beclin-1 (Figure 8, B and C). Although these data do not exclude the possibility that STING activation may contribute to other glomerular cells’ failure, they confirm a role of STING in mediating podocyte injury.

Figure 8.

STING activation in podocytes is associated with increased apoptosis and autophagy. (A) Caspase-3 activity levels in control (CTRL, n=6) and c-diAMP–treated (10 μM, n=6) podocytes. **P<0.01, two-tailed t test. (B and C) Representative Western blot (B) and bar graph analysis of LC3B (C, left panel) and Beclin-1 (C, right panel) expression in CTRL (n=3) and c-diAMP–treated (n=3) podocytes. *P<0.05, two-tailed t test. (D–F) Representative TUNEL BrdU-red staining (upper panel; magenta) and costaining for LC3B (green) and podocyte-specific marker synaptopodin (SYNPO, red; lower panel) in 4-µm-thick kidney sections in control db/+ and db/db mice treated with C-176 (D), control (CTRL) and STING knockout (STING^−/−) mice treated with STZ (E), and control (Col4a3^+/+) and AS (Col4a3^−/−) mice treated with C-176 (F). Nuclei were stained with DAPI (blue). BrdU: original magnification ×20; scale bar, 30 µm. SYNPO/LC3B: original magnification ×40; scale bar, 20 µm.

To validate our in vitro findings, we performed immunofluorescence staining of kidney sections from all of the mouse models recruited into this study. Thus, TUNEL assay demonstrated higher levels of apoptosis in glomeruli from mice with T2D-associated (Figure 8D, upper panel) and T1D-associated (Figure 8E, upper panel) DKD, but not in mice with experimental AS (Figure 8F, upper panel). Further, using LC3B as a marker of autophagy and synaptopodin as a podocyte-specific marker, we were able to demonstrate no changes in LC3B expression in glomeruli or tubule compartment in db/db (Figure 8D, lower panel) or STZ-treated (Figure 8E, lower panel) mice, whereas Col4a3^−/− mice had significantly increased LC3B levels in glomeruli only (Figure 8F, lower panel). Notably, C-176 treatment resulted in partial amelioration of autophagy in Col4a3^−/− mice. These data are consistent with the previously published report showing decreased mRNA levels of key autophagy genes in STZ and db/db mouse models of DKD.⁵⁷

In summary, pharmacologic STING inhibition in Col4a3^−/− mice, a model of experimental AS, results in improved kidney function, similar to what we observed in a mouse model of DKD, and may also be considered as a potential therapeutic strategy to treat kidney failure in patients with AS.

Discussion

Metabolic dysfunction, low-grade inflammation, and infiltration with immune cells were described in the kidneys of patients with DKD^58–61 and less frequently in the kidneys of patients with AS.^25–27 Despite the large body of literature demonstrating involvement of the immune system in the development and progression of glomerular diseases, the mechanism of immune activation remains poorly understood and no anti-inflammatory agents have been shown so far to confer renoprotection. STING is an adaptor protein involved in DNA-dependent activation of the innate immunity against viral or bacterial infection^62,63 and in autoinflammatory diseases such as vasculopathy⁶⁴ and SLE.⁶⁵ However, the importance of STING in the regulation of local inflammation in the kidney remains largely unstudied. Our own results suggest that among other possible RNA- and DNA-sensing pathways, only STING is upregulated in both mouse models of CKD. To our knowledge, this study identifies for the first time activation of the cGAS-STING signaling pathway in podocytes as a significant contributor to the pathogenesis of glomerular disease of either metabolic (DKD) or nonmetabolic (AS) origin, suggesting that STING targeting represents a potential therapeutic option to prevent kidney function loss in these conditions.

Activation of the cGAS-STING pathway has previously been shown in nonimmune cells such as mouse embryonic fibroblasts⁶⁶ and adipocytes.³⁵ Here, we first confirmed that human and murine podocytes express all components of the cGAS-STING signaling pathway either at mRNA or protein levels, which was associated with an increase in gene expression of STING, TBK1, and IRF3, along with increased phosphorylation after treatment with cyclic dinucleotides (c-diAMP). Activation of the cGAS-STING in podocytes also led to an increased expression of type I IFN, a “classic” downstream signal of cGAS-STING. These data confirm the notion that podocytes exhibit features of immune cells and support the idea that podocytes may also be considered as antigen-presenting–like cells.²¹ Moreover, increased caspase-3 activity and autophagy markers in c-diAMP–treated podocytes suggest that apoptosis and autophagic cell death may be associated with STING activation. Interestingly, increased autophagy levels in association with decreased podocyte-specific staining in glomeruli from AS mice but not db/db mice may suggest higher levels of autophagic podocyte death in glomerular disease of nonmetabolic origin. Other studies have shown decreased mRNA expression of autophagy-associated genes in db/db and STZ-treated mice⁵⁷ indicating that impaired autophagy in mouse models of DKD may also contribute to the disease progression. STING has been shown to regulate autophagy flux⁶⁷ via possible direct interaction with LC3B.⁶⁸ In contrast, the levels of apoptosis in glomeruli were higher in mice with T1D- and T2D-associated DKD, but not in mice with AS. These data are in accordance with previously published studies that showed increased apoptosis levels in renal tissue from db/db mice.^69,70 However, further investigation is required to delineate if other types of cell death via ferroptosis or endoplasmic reticulum stress are also involved in STING–associated podocyte injury in glomerular diseases. Interestingly, our data indicate increased Aim2 mRNA expression in kidney cortex from AS mice, which indicates that caspase-1–associated cell death may also take place.

Previous studies also revealed that expression of the components of the cGAS-STING signaling pathway are significantly upregulated by obesity in mice.^28,29 In humans, our data demonstrated that increased gene expression of cGAS and STING can be observed in glomeruli of patients with FSGS, whereas increased expression of the downstream mediator TBK1 is only observed in glomeruli of patients with TD2. The reason for this discrepancy remains obscure, but we speculate that this may be due to different mechanisms of cGAS-STING activation in FSGS and DKD, as STING induction may also occur via other DNA sensors such as DDX41, IFI16, and DAI or even through the RIG-I–dependent mitochondrial antiviral signaling protein activation and mitochondria stress-dependent release of mitochondrial DNA.⁷¹

Next, we show that the cGAS-STING signaling pathway is upregulated in murine models of CKD in the kidney cortex and glomeruli. Moreover, we demonstrated that activation of STING in wild-type C57BL/6 mice leads to kidney damage per se. These observations suggest an important role for activation of the cGAS-STING pathway in metabolic and nonmetabolic kidney diseases. Importantly, STING activation was restricted to glomeruli in c-diAMP–treated mice, suggesting that the tubular epithelium is less associated with observed proteinuria. Furthermore, glomerular accumulation of macrophages and T cells in early disease pathogenesis has been described in humans and in experimental models of DKD,^53,72 whereas the inhibition of inflammatory cell recruitment in mice was shown to protect from experimental DKD.^73,74 Interestingly, kidney cells, including podocytes, and endothelial and mesangial cells were shown to be equally capable of secreting proinflammatory cytokines.^19,75–77 In support of these findings, our results demonstrated that genetic or pharmacologic inhibition of STING is beneficial for kidney outcomes in animal models of CKD, which is in line with previously reported results in tubules from other groups.^32,34 Moreover, we found no macrophage infiltration in the kidneys of c-diAMP–treated, db/db and AS mouse models, suggesting that observed STING activation is related to the pathologic changes in kidney cells, primarily glomeruli. However, it is important to remember that the cGAS-STING signaling pathway is an important pathway of the innate immunity defense against infections and balancing the levels of STING inhibition would be required for the therapeutic benefit in treatment of CKD. In turn, it suggests that more studies are needed to confirm a safe dosage for STING inhibition in CKD patients without increasing susceptibility to infectious diseases.

Although we observed a significant effect of the cGAS-STING pathway on podocyte damage in glomerular diseases of both metabolic (DKD) and nonmetabolic (AS and FSGS) origin, the mechanism behind it remains unclear and needs further investigation. Some studies indicate that in the kidney, the cGAS-STING pathway may be activated by mitochondrial DNA leakage into cytosol.^32,34 Indeed, STING knockout or its pharmacologic inhibition did not completely abolish kidney injury in our mouse models of CKD, which might be partially explained by the continued presence of mitochondrial dysfunction and energy deficit, as we previously reported in mouse models of DKD.⁷⁸ Thus, it will be important to assess if mitochondrial DNA leakage also occurs in podocytes and contributes to the activation of the cGAS-STING pathway and the degree of that leak, if any. On the other hand, negative regulation of STING signaling is also essential for the prevention of chronic inflammation. Interestingly, an alternative mechanism of STING downregulation has been proposed via phosphorylation of cGAS by protein kinase B (AKT),⁷⁹ which plays a significant role in podocyte insulin signaling,⁸⁰ as we previously demonstrated that AKT phosphorylation is decreased in kidneys of diabetic db/db mice.⁸¹ Thus, it is also possible that STING activation in the kidneys of db/db mice is affected by a negative feedback loop. Importantly, to investigate the intrinsic effects of STING inhibition on podocytes survival in glomerular diseases such as DKD and FSGS, podocyte-specific STING knockout db/db and AS mice should be used, which presents a limitation of this study.

In conclusion, our data suggest that activation of STING acts as a mediator of glomerular disease progression independently if the disease is of metabolic or nonmetabolic origin. Thus, targeting of STING may represent another therapeutic option to treat or even prevent glomerular disease development and/or progression.

Disclosures

G. Burke, A. Fornoni, and S. Merscher are inventors on pending or issued patents (US10,183,038 and US10,052,345) aimed to diagnose or treat proteinuric kidney diseases; they stand to gain royalties from their future commercialization. G. Burke reports consultancy fees and honoraria from CareDx; and is associate editor (unpaid) for Transplantation. S. Eddy reports research funding from Angion Biomedica, AstraZeneca, Certa, Eli Lilly, Gilead Sciences, Ionis, Janssen, Moderna, and Novo Nordisk. A.Fornoni is chief scientific officer of L&F Health LLC, holds equity interests in L&F Research and is the inventor of assets developed by ZyVersa Therapeutics (ZyVersa has licensed worldwide rights to develop and commercialize hydroxypropyl-β-cyclodextrin for treatment of kidney disease from L&F Research); and holds equity in River 3 Renal Corporation. A. Fornoni reports consultancy fees from Dimerix, Horizon, Kaneka, and ONO; is a shareholder of UpToDate; reports research funding from Aurinia Pharmaceuticals, Boehringer Ingelheim, and Kyowa Kirin; reports patents or royalties for two patents (patent on the use of cyclodextrin for the treatment of kidney diseases and patent on the use of small molecule inducers of cholesterol efflux); reports advisory or leadership role for Journal of Clinical Investigation and Kidney International; is on the speakers bureau for American Physician Scientists Association physician-scientist seminar series, Dalian University, Drexel University, Duke University, European Renal Association–European Dialysis and Transplantation Association, International Podocyte Conference, Massachusetts General Hospital/Brigham and Women’s Hospital combined Renal Grand Round, University of California Irvine, University of California Los Angeles, University of Heidelberg, University of Southern California, and World Congress of Nephrology; and is the inventor of five pending US patents and one published patent. M. Ito reports interests or relationships with Manpei Suzuki Diabetes Foundation. M. Kretzler reports consultancy fees as an employee of University of Michigan from Astellas, Boehringer Ingelheim, Certa, Janssen, Novo Nordisk, and Poxel,; reports research funding at University of Michigan in sponsored research projects as principal investigator from amfAR, Angion, AstraZeneca, Boehringer Ingelheim, Certa, Chan Zuckerberg Initiative, Chinook, Eli Lilly, Elpidera, Gilead Sciences, Goldfinch, Ionis, Janssen, JDRF, National Institutes of Health (NIH), Novo Nordisk, Regeneron, RenalytixAI, and Travere; is on the editorial boards for JASN, Kidney International, and Kidney Diseases; and is on the advisory board for NephCure Kidney International. S. Merscher holds equity interest in L&F Research. S. Merscher reports consultancy fees from Kintai Therapeutics (expired October 2020); reports research funding from Aurinia Pharmaceuticals and Boehringer Ingelheim; is co-inventor of several pending and issued patents, and has indirect equity interest in and potential royalty from ZyVersa Therapeutics by virtue of assignment and licensure of a patent estate; and is review editor, guest topic editor, and associate editor for Frontiers in Medicine (section nephrology). A. Mitrofanova is a member of American Society of Nephrology and a member of American Heart Association. All remaining authors have nothing to disclose.

Funding

A. Mitrofanova and G. Burke are supported by Chernowitz Medical Research Foundation (GR016291). A. Mitrofanova is supported by Carl W. Gottschalk Research Scholar Grant, American Society of Nephrology (GR018262).

Author Contributions

A. Mitrofanova conceptualized the study, was responsible for project administration, supervision, and visualization, and wrote the original draft; A. Fornoni and A. Sloan were responsible for software; A. Fornoni and A.Mitrofanova were responsible for funding acquisition; M. Boulina, S. Eddy, M. Kretzler, S. Mallela, S. Merscher, and A. Mitrofanova were responsible for data curation; M. Boulina, S. Mallela, S. Merscher, A. Mitrofanova, J. Molina David, J. Santos, and Y. Zuo were responsible for formal analysis; M. Boulina, A. Fontanella, M. Ito, A. Mitrofanova, J. Molina David, and M. Tolerico were responsible for methodology; M. Boulina, S. Eddy, A. Fornoni, M. Kretzler, R. Nelson, and A. Sloan were responsible for resources; S. Eddy, M. Kretzler, S. Merscher, and A. Mitrofanova, and J. Pressly were responsible for validation; A. Fontanella, M. Ge, M. Gurumani, W. Issa, M. Ito, J.-J. Kim, A. Mitrofanova, J.Pressly, and M. Tolerico were responsible for investigation; and G. Burke, A. Fontanella, A. Fornoni, M. Ge, J.-J. Kim, S. Mallela, S. Merscher, A. Mitrofanova, R. Nelson, J. Santos, A. Sloan, M. Tolerico, and Y. Zuo reviewed and edited the manuscript.

Data Sharing Statement

Data from the University of Miami microarray study (Miami cohort) have been released for free and open accessibility.⁸² The complete dataset from the NEPTUNE study cannot be shared publicly. Complete datasets of patients enrolled in the NEPTUNE study are considered sensitive information, as they provide genomic datasets from patients with an orphan disease with significant risk of reidentification. No individual-level genetic data can be publicly released from the Pima Indians study. All other relevant data are available from the corresponding author upon a reasonable request.

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2021101286/-/DCSupplemental.

Supplemental Methods. Details on methods used for cell culture and treatment, subcellular fractionation, CMA treatment of mice, immunocytochemistry, patient data, microarray analysis and pathway analysis are provided in Supplemental Material.

Supplemental Figure 1. Uncropped Western blot pictures related to Figures 2–4 and 7.

Supplemental Figure 2. Negative and positive controls for immunostaining related to STING detection and macrophage activation.

Supplemental Figure 3. Increased expression of genes of the cGAS-STING signaling pathway in DKD sera treated podocytes and glomeruli from patients with DKD and FSGS.

Supplemental Figure 4. Human podocytes express all of the components of the cGAS-STING signaling pathway and this pathway could be activated by small molecule STING agonist.

Supplemental Figure 5. Murine podocytes express all of the components of the cGAS-STING signaling pathway and this pathway could be activated by small molecule STING agonist.

Supplemental Figure 6. c-diAMP treatment leads to increased expression of the components of the cGAS-STING pathway.

Supplemental Figure 7. CMA treatment results in mild kidney damage in wild-type C57BL/6 mice.

Supplemental Figure 8. The cGAS-STING signaling pathway is upregulated in the kidney cortices of the animal models of DKD and AS.

Supplemental Table 1. List of primers used in the study.

Supplemental Table 2. List of antibodies used in the study.

Supplemental Table 3. Serologic analysis of control and STING^−/− mice with STZ-induced diabetes.

Supplemental Table 4. Serologic analysis of db/+ and db/db mice treated with C-176.

Supplemental Data 1. Pathways enrichment analysis performed using DAVID 6.8 database in FSGS Miami cohort, related to Figure 1 (an Excel file).

Supplemental Data 2. Pathways enrichment analysis performed using DAVID 6.8 database in Pima Indians cohort, related to Figure 1 (an Excel file).