Inhibition of STING modulates dendritic cell responses and ameliorates podocyte injury in SLE

Abstract

Background

SLE features aberrant type I interferon production, with the stimulator of interferon genes (STING) pathway acting as a key upstream driver. However, its expression patterns in patients with SLE and contributions to clinical phenotypes and tissue injury remain insufficiently defined.

Methods

The proportion of STING-positive cells in peripheral blood mononuclear cells (PBMCs) from 28 patients with SLE and 15 healthy controls was analysed alongside clinical and serological features. Glomerular STING expression and podocyte injury were assessed in kidney biopsies from 39 patients with lupus nephritis (LN) and circulating and renal STING expression was compared in a paired cohort. The therapeutic effects of STING inhibitor C176 were evaluated in MRL/lpr mice by examining kidney pathology, proteinuria, function, immune cell subsets and podocyte injury. STING-mediated regulation of dendritic cell (DC) activation and plasmacytoid DC (pDC) differentiation was further investigated in vivo and in vitro.

Results

STING expression was significantly elevated in PBMCs from patients with SLE and kidney tissues from patients with LN, particularly in proliferative LN and correlated with glomerular proliferation and decreased synaptopodin expression, suggesting an association with podocyte injury. STING activation promoted DC activation and pDC differentiation. Furthermore, conditioned media from STING-activated DCs induced podocyte apoptosis and reduced synaptopodin expression, effects attenuated by STING inhibition, indicating that STING activation in DCs promotes podocyte injury and disrupts podocyte homeostasis in a paracrine manner. C176 treatment in MRL/lpr mice significantly alleviated splenomegaly, proteinuria, kidney inflammation and renal dysfunction, while restoring podocyte integrity and reshaping immune responses.

Conclusion

Our findings suggest a potential role for STING in the interplay between immune dysregulation and podocyte injury in lupus. Pharmacological STING inhibition modulates DC responses and partially protects podocytes and renal structure, supporting STING as a promising therapeutic target for SLE.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Stimulator of interferon genes (STING) is a key upstream activator of interferon-I signalling, yet its specific expression patterns in patients with SLE and its contribution to lupus nephritis (LN) pathogenesis remain poorly understood.

WHAT THIS STUDY ADDS

• STING expression is abnormally upregulated in peripheral blood from patients with SLE and in kidney tissues from patients with LN. Increased STING expression is associated with disease activity, proliferative glomerular lesions and podocyte injury.

• Common immunosuppressive therapies, including glucocorticoids, mycophenolate mofetil, hydroxychloroquine and B cell-targeted biologics, are associated with reduced renal STING expression, suggesting the potential modulation of STING expression by current treatment strategies.

• STING activation in dendritic cells (DCs) may promote podocyte injury through altered DC responses and soluble mediators, while pharmacological STING inhibition ameliorates LN in MRL/lpr mice, highlighting STING as a potential link between immune dysregulation and renal injury.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Our study highlights the STING pathway as a potential therapeutic target in LN. The observed association between standard therapies and reduced STING expression underscores the need for future clinical studies stratified by medication status to evaluate STING as a biomarker. These findings support further evaluation of STING-targeted interventions as a strategy to concurrently control systemic inflammation and protect renal structure in LN.

Introduction

SLE is a prototypical autoimmune disease characterised by multisystem involvement, autoantibody production and chronic inflammation.¹ Among its diverse clinical manifestations, lupus nephritis (LN) is one of the most common and severe complications. Epidemiological studies indicate that LN increases the risk of progression to end-stage kidney disease by up to 45-fold and triples all-cause mortality, particularly among young patients.^{2 3} The pathogenesis of LN involves a complex interplay of genetic susceptibility, gut microbiota dysbiosis, neutrophil extracellular trap formation, defective autophagy and abnormal immunometabolism.^4–6 Meanwhile, podocyte injury is closely associated with proteinuria and disease progression, indicating that it represents a key pathological feature contributing to LN progression.^{7 8}

Type I interferon (IFN-I) signalling is a central pathogenic axis in LN. Transcriptomic studies have identified extensive activation of IFN-related gene signatures in patients with active LN.⁹ Experimental studies have further demonstrated that IFN-I accelerates glomerulonephritis and kidney injury, whereas its inhibition ameliorates histopathological damage and improves survival.¹⁰ These findings establish the central role of the IFN-I signalling axis and its upstream regulatory pathways in the pathogenesis of LN.

The cGAS-stimulator of interferon genes (STING) pathway is a key cytosolic DNA-sensing mechanism that drives IFN-I signalling. STING, an adaptor protein localised to the endoplasmic reticulum, acts as a central hub linking DNA sensing to downstream immune activation.¹¹ on recognising endogenous or pathogen-derived double-stranded DNA (dsDNA), cGAS catalyses the production of cyclic GMP-AMP (cGAMP), which binds to and activates STING. Activated STING subsequently recruits TANK-binding kinase 1 (TBK1) and interferon regulatory factor 3, leading to the transcriptional induction of IFN-I and inflammatory cytokines.^12–14

Given the pivotal role of STING in IFN-I signalling, multiple classes of small-molecule STING inhibitors have been developed to therapeutically target this pathway. These include agents that block cGAMP-STING binding, inhibit STING palmitoylation or disrupt its trafficking from the endoplasmic reticulum to the Golgi apparatus.^15–17 Among them, the nitrofuran derivative C176 is a potent and selective STING inhibitor that prevents STING palmitoylation, a critical post-translational modification required for TBK1 recruitment and signalling complex assembly.^{18 19} In Trex1^−/− mice, which develop severe multiorgan inflammation due to persistent activation of the cGAS-STING pathway and recapitulate key pathogenic features of human Aicardi-Goutières syndrome, C176 effectively attenuates systemic IFN-I responses and inflammatory phenotypes, highlighting its therapeutic potential in STING-driven autoinflammatory diseases.¹⁶

Despite these advances, the role of STING in autoimmunity appears context-dependent and remains controversial. In lupus-prone MRL/lpr and pristane-induced mouse models, loss of STING exacerbates disease severity, manifested by splenomegaly, lymphadenopathy, increased autoantibody production and aggravated proteinuria.^{20 21} In contrast, in 129/B6.Fcgr2b^−/− mice, a lupus-prone model with an IFN-dependent autoimmune phenotype, upregulation of STING expression has been observed and genetic ablation of STING (Fcgr2b^−/−.Sting^gt/gt) significantly reduces serum anti-dsDNA and ANA levels and ameliorates kidney pathology.²² These conflicting findings from different lupus mouse models underscore the complexity of STING signalling in SLE. Given that MRL/lpr mice closely recapitulate key features of human proliferative LN, they represent a suitable model for LN research. Moreover, prior studies have largely focused on genetic deletion of STING, whereas the effects of pharmacological STING inhibition remain insufficiently explored in this model. Thus, there is an urgent need to evaluate the therapeutic potential of pharmacological STING inhibition in SLE and to elucidate the mechanisms by which it modulates disease progression.

In this study, we investigated the expression and potential pathological role of STING in patients with SLE and LN, and evaluated the therapeutic efficacy of STING inhibition in MRL/lpr mice. We found that STING expression was elevated in both peripheral blood mononuclear cells (PBMCs) and kidney tissues of patients with LN. Notably, renal STING expression correlated with disease activity and histopathological indices. Furthermore, we explored whether STING activation in dendritic cells (DCs) could contribute to podocyte injury and evaluated whether pharmacological STING inhibition could ameliorate LN. Our findings identify STING as a potential mediator linking innate immune activation to podocyte injury and support STING inhibition as a promising therapeutic strategy for LN.

Materials and methods

Study population

A total of 28 hospitalised patients with SLE were enrolled in this study (admitted to the Department of Rheumatology and Immunology, Nanjing Drum Tower Hospital, the Affiliated Hospital of Medical School, Nanjing University, between April and May 2020), including 6 male and 22 female patients. 15 age-matched and sex-matched healthy controls were also included (3 male and 12 female participants), with a median age of 36 years (IQR, 27–47 years). The patients met the 2019 European Alliance of Associations for Rheumatology/American College of Rheumatology (EULAR/ACR) classification criteria for SLE.²³ Patients with malignancy, infection or other autoimmune diseases (except secondary Sjögren’s syndrome) were excluded.

For renal tissue analysis, archived kidney biopsy specimens from a total of 39 patients with LN were obtained from the pathology database between April 2021 and June 2026. These included 24 patients enrolled between April 2021 and July 2023 and an additional 15 patients enrolled between April and June 2026. The cohort consisted of 5 male and 34 female patients. Histological classification was performed according to the 2018 International Society of Nephrology/Renal Pathology Society criteria.²⁴ Seven samples of normal kidney tissue adjacent to kidney carcinoma were used as controls. The control cohort included three males and eight females, with a median age of 37 years (IQR, 31–42 years).

To further evaluate the association between peripheral and renal STING expression within the same individuals, peripheral blood samples were additionally collected from the 15 patients enrolled between April and June 2026. The 15 patients in the paired cohort represented a subset of the 39 patients with LN included in the renal tissue analysis. STING expression in PBMCs was analysed by flow cytometry, and STING expression in renal tissues was assessed by multiplex immunohistochemistry. The renal biopsy specimens from this paired cohort were included in the overall renal tissue analysis described above.

Animal experiments

Female MRL/lpr mice and age-matched MRL/Mpj mice (18 weeks old) were purchased from Sibeifu Biotechnology (China). All animals were housed under specific pathogen-free conditions with free access to food and water. Two independent intervention protocols were conducted to evaluate the therapeutic efficacy of STING inhibition in lupus-like disease.

For the dose-response study, MRL/lpr mice were randomly assigned to experimental groups using a random number table prior to treatment initiation to minimise allocation bias. Mice received either vehicle or C176 (MedChemExpress, USA) at doses of 187.5 or 375 nmol per injection, administered intraperitoneally two times per day from 18 to 20 weeks of age (n=7 per group). For therapeutic evaluation, 18-week-old female MRL/lpr mice (n=10) and age-matched MRL/Mpj controls (n=5) were purchased from Sibeifu Biotechnology (China). MRL/lpr mice were randomly assigned, using a random number table, to receive either vehicle or C176 (MedChemExpress, USA) at 375 nmol per injection two times per day from 18 to 22 weeks of age. 24-hour urine was collected in metabolic cages before treatment and prior to sacrifice. Following euthanasia at 22 weeks, serum, spleen and kidneys were collected for subsequent biochemical and histological analyses.

Flow cytometry

PBMCs from human subjects were isolated by Ficoll-Paque density gradient centrifugation and stained with PE anti-human STING (BD Biosciences, USA). For animal experiments, splenocytes (1×10⁶ cells) were stained with the following fluorochrome-conjugated antibodies: Ghost Dye Violet 510 Fixable Viability Dye (Thermo Fisher Scientific), BV711 anti-mouse-CD45, AF700/PE anti-mouse-MHC-II (I-A/I-E), FITC anti-mouse-Lin, APC/Fire 750/PE anti-mouse-CD11c, PE-Cy7 anti-mouse-IFNγ, APC/PE anti-mouse-Siglec-H, APC anti-mouse-CD80, BV605 anti-mouse-PDCA1, PerCP-Cy5.5/BV421 anti-mouse-CD11b, APC anti-mouse-CD4, APC anti-mouse-CD138, BV785/PE-Cy7 anti-mouse-B220, FITC anti-mouse-Ly6G, APC anti-mouse-Ly6C, PE-Cy7 anti-mouse-F4/80, FITC anti-mouse-CD19 and FITC anti-mouse-CD3 (all from BioLegend, USA).

Prior to the detection of IFN-γ and interleukin (IL)-17A, cells were stimulated with a cytokine activation cocktail (BioLegend) for 4 hours at 37°C in a humidified atmosphere containing 5% CO₂. For intracellular staining (including IFN-γ, IL-17A and STING), cells were fixed and permeabilised using a fixation/permeabilisation kit according to the manufacturer’s instructions. Fluorochrome-conjugated monoclonal antibodies were added at the recommended concentrations and incubated for 30 min at 4°C in the dark. After staining, cells were washed twice with staining buffer and analysed on a BD Fortessa flow cytometer (BD Biosciences). Data were compensated and analysed using FlowJo software (V.10.8.1, Tree Star). The percentage of STING-positive cells was quantified based on a standardised gating strategy.

Multiplex immunohistochemistry staining

Paraffin-embedded kidney sections (3 µm) were deparaffinised, rehydrated and subjected to antigen retrieval in AR buffer (Melady, China). After blocking with goat serum (Zhongshan Golden Bridge, China), sections were incubated overnight at 4°C with anti-CD11c (Abcam, 1:100), anti-Synaptopodin (Abcam, 1:300) and anti-STING (Abcam, 1:4000), followed by horseradish peroxidase (HRP)-conjugated secondary antibody (Zhongshan Golden Bridge, China). Signals were detected using tyramide signal amplification (TSA) reagents (Melady, China) and nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) antifade mounting medium (Beyotime, China). Images were visualised with SlideViewer software and quantitatively analysed using Image-Pro plus.

H&E and periodic acid-Schiff staining

Paraffin-embedded kidney sections (3 µm) were deparaffinised, rehydrated and stained with H&E or periodic acid-Schiff (PAS) according to standard protocols. The sections were then dehydrated, cleared and mounted. Histopathological changes were observed under a light microscope (Olympus).

Kidney histopathological scoring in mice

Kidney injury was semiquantitatively evaluated on H&E-stained and PAS-stained kidney sections according to previously described methods. 20 glomerular cross-sections (GCSs) per mouse were assessed for glomerular lesions and scored as follows: 0, normal glomeruli (35–40 cells/GCS); 1, mild mesangial proliferation (41–50 cells/GCS) or focal exudation; 2, moderate mesangial or segmental proliferation (51–60 cells/GCS) with hyaline change or moderate exudation; and 3, severe glomerulosclerosis (>60 cells/GCS), fibrinoid necrosis, crescent formation or marked exudation. Tubulointerstitial injury was evaluated in ten randomly selected high-power fields and scored as: 0, normal; 1, mild focal interstitial inflammation or tubular vacuolisation; 2, multifocal inflammatory infiltration with interstitial oedema or tubular dilation; and 3, diffuse interstitial fibrosis, tubular atrophy or necrosis, and cast formation. Perivascular inflammation was graded according to the number of inflammatory cell layers surrounding the vessels: 0, none; 1, <5 layers; 2, 5–10 layers; and 3, >10 layers with vascular wall disruption. Tissue sections and staining images were anonymised using randomly assigned identifiers prior to analysis. Histological scoring and quantitative image analyses were conducted in a blinded manner with respect to group allocation, and group identities were not revealed until completion of the analyses. The mean values were used for statistical analysis.

In vitro DCs and plasmacytoid dendritic cells stimulation assays

Bone marrow cells isolated from the femurs and tibiae of MRL/lpr mice were prepared into single-cell suspensions and cultured in the presence of granulocyte-macrophage colony-stimulating factor and IL-4 to generate bone marrow-derived dendritic cells (BMDCs). For DC activation experiments, BMDCs were seeded into 96-well plates at a density of 5×10⁵ cells/well and treated with dimethyl sulfoxide (DMSO) (vehicle control; Sigma-Aldrich), 5,6-Dimethylxantheonone-4-acetic acid (DMXAA) (STING agonist; MedChemExpress, USA), C176 (STING inhibitor; MedChemExpress, USA) or a combination of DMXAA and C176 for 24 or 48 hours. Single-cell suspensions prepared from the bone marrow or spleen of MRL/lpr mice were used for plasmacytoid DC (pDC) phenotype analysis. Following treatment, DC activation markers and pDC-associated markers were assessed by flow cytometry.

Conditioned medium preparation and podocyte apoptosis assay

For assessment of podocyte apoptosis, podocytes were treated with conditioned media collected from differently treated BMDCs as described above. After treatment, podocytes were collected, washed twice with cold phosphate-buffered saline and stained with Annexin V and 7-aminoactinomycin D using an apoptosis detection kit (Vazyme, China) according to the manufacturer’s instructions.

Western blotting

Podocytes treated with conditioned media from differently treated BMDCs were lysed in RIPA buffer. Equal amounts of protein were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride (PVDF) membranes. After incubation with primary and secondary antibodies, protein expression was detected using an enhanced chemiluminescence detection system. The anti-synaptopodin antibody (Abcam, UK) and β-actin antibody (ABclonal, China) were used, with β-actin as the loading control.

Transmission electron microscopy

Mouse kidneys were perfusion-fixed in 2.5% glutaraldehyde for ultrastructural examination of glomerular podocytes. After fixation, tissues were treated with 1% osmium tetroxide, dehydrated in a graded acetone series and embedded in Epon. Ultrathin sections (70 nm) were generated with an ultramicrotome, stained with uranyl acetate and lead citrate, and placed on copper grids. Samples were examined using a transmission electron microscope operated at 80 kV (Hitachi, Japan).

For ultrastructural analysis, five glomeruli were randomly selected from each kidney sample. Transmission electron microscopy (TEM) images were acquired at different magnifications to evaluate podocyte ultrastructure and foot process morphology. Quantitative measurements of podocyte foot process width were performed blinded to group allocation, and group identities were revealed only after completion of the analyses. The mean values from each sample were used for statistical analysis.

ELISA

Serum levels of ANA, anti-dsDNA antibodies and IFN-α were measured using the mouse ANA ELISA Kit (Alpha Diagnostics, USA), mouse anti-dsDNA antibody ELISA Kit (CUSABIO, China) and mouse IFN-α ELISA Kit (CUSABIO, China), respectively, following the manufacturer’s protocols.

Detection of serum creatinine and blood urea nitrogen

Serum creatinine and blood urea nitrogen levels were measured using enzymatic methods, with serum creatinine determined by the sarcosine oxidase method and blood urea nitrogen assessed using a urease-glutamate dehydrogenase-coupled enzymatic assay, according to the manufacturer’s instructions (Nanjing Jiancheng Bioengineering Institute, China).

Statistical analyses

All statistical analyses were performed using SPSS V.24.0 (IBM, USA) and GraphPad Prism V.9. Continuous variables were presented as mean±SD or median (IQR), and categorical variables were expressed as percentages. For group comparisons, normally distributed data with equal variances were analysed using a two-tailed unpaired t-test (for two groups) or one-way analysis of variance (for multiple groups), whereas non-normally distributed data were analysed using the Kruskal-Wallis test. Categorical variables were compared using Pearson’s χ² test or the continuity-corrected chi-square test. Correlations between continuous variables were assessed using Pearson’s correlation analysis, while Spearman’s rank correlation was used for non-normally distributed or ordinal variables. For analyses involving multiple comparisons within predefined panels, p values were adjusted using the Benjamini-Hochberg (BH) procedure to control the false discovery rate. Adjusted p values were reported where applicable. A two-sided p value <0.05 was considered statistically significant.

Results

STING-positive cells were elevated in PBMCs of patients with SLE and correlated with disease activity

To investigate the expression of STING in patients with SLE, we first examined the proportion of STING-positive cells in PBMCs. The results showed that the proportion of STING-positive cells was significantly higher in PBMCs from patients with SLE than in those from healthy controls (mean difference vs controls, 9.33; 95% CI 2.81 to 15.85; adjusted p=0.006) (figure 1A), indicating increased STING expression in SLE. The clinical characteristics and medication profiles of all enrolled patients were summarised in tables 1 and 2. We then analysed the correlations between the proportion of STING-positive cells and various clinical parameters. The analysis revealed that the proportion of STING-positive cells was significantly correlated with the SLE Disease Activity Index (SLEDAI) score (figure 1B), renal-SLEDAI (rSLEDAI) score (figure 1C), erythrocyte sedimentation rate (figure 1G), Systemic Lupus International Collaborating Clinics/ACR Damage Index (figure 1H), serum IgM (figure 1I), D-dimer (figure 1J), serum uric acid (figure 1K) and 24-hour urinary protein excretion (figure 1L). In contrast, no significant correlations were observed between STING expression and complement C3 (figure 1D), complement C4 (figure 1E) or anti-dsDNA antibody levels (figure 1F). Collectively, these findings indicate that STING upregulation is associated with multiple clinical indicators of disease activity and organ damage in SLE, particularly renal involvement, although not all conventional biomarkers showed significant correlations, suggesting a potential role in SLE pathogenesis.

Figure 1. STING-positive cells were elevated in PBMCs of patients with SLE and correlated with disease activity. (A) Representative flow cytometry analysis of the proportion of STING-positive cells in PBMCs from healthy donors (n = 15) and patients with SLE (n = 28). (B–L) The correlation between STING-positive cells proportion and clinical parameters, including SLEDAI score (B), rSLEDAI score (C), complement C3 (D), complement C4 (E), anti-dsDNA antibody levels (F), ESR (G), SDI (H), serum IgM levels (I), D-dimer levels (J), serum uric acid (K) and 24-hour urinary protein (L). Data are shown as mean±SD. Pearson’s correlation was applied to normally distributed continuous variables, and Spearman’s rank correlation to non-normal or ordinal variables. **p<0.01. dsDNA, double-stranded DNA; ESR, erythrocyte sedimentation rate; HC, healthy controls; PBMCs, peripheral blood mononuclear cells; rSLEDAI, renal-SLEDAI; SDI, Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SLEDAI, SLE Disease Activity Index; STING, stimulator of interferon genes.

Table 1. Clinical features of patients with SLE.

Parameter	Mean/median	SD/IQR	Reference range
Age (years) (median, IQR)	39.61	28.00, 48.75
SLEDAI	10.46	6.52
SDI	1.93	1.65
Platelet count (×10⁹/L)	146.11	98.60	125–350
Haemoglobin (g/L)	84.71	25.90	115–150
Creatinine (µmol/L)	112.14	119.64	44–106
Blood urea nitrogen (mmol/L)	10.81	8.34	2.9–7.5
Uric acid (μmol/L)	439.82	134.36	90–420
eGFR (mL/min/1.73 m²)	95.64	57.84	>90
Albumin (g/L)	31.10	6.03	40–55
D-dimer (mg/L)	3.04	3.89	0–0.5
24-hour urine protein (g)	2.80	3.79	0–0.15
ESR (mm/hour)	55.35	38.83	0–20
Anti-dsDNA antibody (IU/mL)	337.52	404.31	0–100
Complement C3 (g/L)	0.61	0.28	0.8–1.6
Complement C4 (g/L)	0.11	0.09	0.2–0.4
IgA (g/L)	2.51	1.09	0.7–3.3
IgG (g/L)	12.83	4.75	8–16
IgM (g/L)	0.95	0.60	0.5–2.2

dsDNA, double-stranded DNA; eGFR, estimated glomerular filtration rate; ESR, erythrocyte sedimentation rate; SDI, Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SLEDAI, SLE Disease Activity Index.

Table 2. Summary of medication use in patients with SLE.

Medication	Users (%)	Non-users (%)
Glucocorticoids	23 (82.14)	5 (17.86)
Hydroxychloroquine	14 (50.00)	14 (50.00)
Leflunomide	3 (10.71)	25 (89.29)
Cyclophosphamide	5 (17.86)	23 (82.14)
Mycophenolate mofetil	4 (14.29)	24 (85.71)
Tripterygium Wilfordii	2 (7.14)	26 (92.86)
Total Glucosides of Peony	1 (3.57)	27 (96.43)
Intravenous immunoglobulin	3 (10.71)	25 (89.29)
Thalidomide	2 (7.14)	26 (92.86)

Increased renal STING expression correlated with glomerular proliferation and podocyte injury in LN

Given the observed association between STING expression in PBMCs and renal injury, we next investigated the expression of STING within kidney biopsy specimens from patients with LN. Clinical characteristics and medication profiles of these patients were summarised in tables 3 and 4. Renal STING expression was significantly elevated in class II (median difference vs controls, 1.19; 95% CI 0.97 to 1.55; BH-adjusted p=0.02), III (median difference vs controls, 1.14; 95% CI 0.62 to 1.63; BH-adjusted p=0.01), IV (median difference vs controls, 1.16; 95% CI 0.38 to 1.88; BH-adjusted p=0.006) and IV+V (median difference vs controls, 1.29; 95% CI 0.97 to 1.73; BH-adjusted p=0.0003) LN compared with normal kidney tissues adjacent to renal carcinoma, but not in class V LN. Furthermore, class IV+V LN (median difference vs class V, 0.82; 95% CI 0.28 to 1.39; BH-adjusted p=0.03) showed higher STING expression than class V LN, whereas no significant differences were observed among the other LN classes (figure 2A,B). Furthermore, renal STING expression was positively correlated with peripheral blood STING expression in the paired cohort (figure 2E), suggesting a potential link between systemic STING activation and renal STING upregulation in LN.

Table 3. Characteristics of patients with lupus nephritis.

Parameter	Mean/median	SD/IQR	Reference range
Age (years) (median, IQR)	31.00	24.00, 44.00
SLEDAI	13.82	5.574
Platelet count (×10⁹/L)	205.23	70.23	125–350
Haemoglobin (g/L)	107.64	17.28	115–150
Creatinine (µmol/L)	72.41	48.48	44–106
Blood urea nitrogen (mmol/L)	7.47	4.44	2.9–7.5
Uric acid (μmol/L)	381.38	139.37	90–420
eGFR (mL/min/1.73 m²)	111.27	37.21	>90
Albumin (g/L)	29.63	5.52	40–55
D-dimer (mg/L)	1.67	2.18	0–0.5
24-hour urine protein (g)	3.60	2.40	0–0.15
ESR (mm/hour)	37.45	25.04	0–20
Anti-dsDNA Antibody (IU/mL)	420.66	350.97	0–100
Complement C3 (g/L)	0.63	0.29	0.8–1.6
Complement C4 (g/L)	0.13	0.09	0.2–0.4
IgA (g/L)	2.55	0.97	0.7–3.3
IgG (g/L)	11.73	5.78	8–16
IgM (g/L)	1.02	0.77	0.5–2.2

dsDNA, double-stranded DNA; eGFR, estimated glomerular filtration rate; ESR, erythrocyte sedimentation rate; SLEDAI, SLE Disease Activity Index.

Table 4. Summary of medication use in patients with lupus nephritis.

Medication	Users (%)	Non-users (%)
Glucocorticoids	23 (58.97)	16 (41.03)
Hydroxychloroquine	20 (51.28)	19 (48.72)
Leflunomide	6 (15.38)	33 (84.62)
Cyclophosphamide	7 (17.95)	32 (82.05)
Mycophenolate mofetil	15 (38.46)	24 (61.54)
Tacrolimus	12 (30.77)	27 (69.23)
Tripterygium Wilfordii	1 (2.56)	38 (97.44)
Belimumab/Telitacicept	5 (12.82)	34 (87.18)

Figure 2. Increased renal STING expression correlated with glomerular proliferation and podocyte injury in LN. (A) Representative mIHC staining of STING (green) and synaptopodin (red) in renal paracarcinoma tissues (n=11) and renal tissues from patients with different classes of LN, including class II (n=3), class III (n=8), class IV (n=9), class IV+V (n=11) and class V (n=8). Nuclei were counterstained with DAPI (blue), original magnification, ×400, scale bar: 100 μm. (B) Quantification of renal STING expression across different LN classes compared with renal paracarcinoma tissues. Data are presented as relative expression levels normalised to the mean value of the control group. (C) Quantification of synaptopodin expression across different LN classes and renal paracarcinoma tissues. Data are presented as relative expression levels normalised to the mean value of the control group. (D) Correlation between relative expression levels of renal STING and synaptopodin. (E) Correlation between the percentage of STING-positive cells in PBMCs and renal STING expression levels (n = 15). (F–O) Correlation analyses between renal STING expression and clinical or pathological parameters, including rSLEDAI score (F), complement C3 (G), complement C4 (H), anti-dsDNA antibody levels (I), glomerular cellular proliferation (J), neutrophilic infiltration (K), activity index (L), chronicity index (M), 24-hour urinary protein (N) and immunosuppressive therapy (O). Data are shown as mean±SD. Pearson’s correlation was used for normally distributed continuous variables, while Spearman’s rank correlation was applied to non-normal or ordinal variables. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. BCTAs, B-cell targeted therapies; CTX, cyclophosphamide; DAPI, 4′,6-diamidino-2-phenylindole; dsDNA, double-stranded DNA; GCs, glucocorticoids; HCQ, hydroxychloroquine; LEF, leflunomide; LN, lupus nephritis; mIHC, multiplex immunohistochemistry; MMF, mycophenolate mofetil; PBMCs, peripheral blood mononuclear cells; rSLEDAI, renal-SLE Disease Activity Index; STING, stimulator of interferon genes; TAC, tacrolimus; TwHF, Tripterygium wilfordii.

Notably, renal STING expression was significantly correlated with rSLEDAI scores (figure 2F), serum complement C3/C4 levels, anti-dsDNA antibody levels (figure 2G–I), glomerular cell proliferation scores (figure 2J), LN activity index (figure 2L) and 24-hour urinary protein excretion (figure 2N). In addition, correlation analyses revealed that higher glucocorticoid doses and exposure to mycophenolate mofetil (MMF), hydroxychloroquine (HCQ) and B-cell targeted biologics (including belimumab and telitacicept) were negatively associated with renal STING expression (figure 2O), suggesting that these therapy strategies may influence STING expression in vivo. However, STING expression was not significantly associated with glomerular neutrophil infiltration (figure 2K) or chronicity index (figure 2M). Notably, the highest STING expression was observed in class IV LN. Coupled with its positive correlation with glomerular cell proliferation, this pattern may suggest that STING may be involved in the pathological processes underlying proliferative lesions in LN.

Using synaptopodin as a podocyte marker, we observed a significant decrease of synaptopodin expression in kidney tissues from patients with LN compared with normal kidney tissues (figure 2C). Notably, consistent with previous reports, synaptopodin expression was higher in pure membranous class V LN than in class IV and class IV+V LN,²⁵ suggesting more severe podocyte injury in proliferative LN. Further correlation analysis revealed that renal STING expression was negatively associated with synaptopodin levels (figure 2D), indicating that STING may be involved in podocyte injury. In summary, these findings suggest that upregulation of STING in the kidney is associated not only with glomerular proliferation in proliferative LN but also with podocyte injury, supporting a potential pathogenic role of STING in LN progression.

STING activation in DCs induced podocyte injury and promoted pDC differentiation

Considering that STING is a key molecule in the IFN-I pathway and that DCs are a major source of IFN-I, we firstly explored whether STING activation was sufficient to trigger DC activation in a STING-dependent manner. Bone marrow cells from MRL/lpr mice were treated with the STING agonist DMXAA, the selective STING inhibitor C176 or DMSO as a vehicle control. The expression of DC activation markers (major histocompatibility complex (MHC)-II, CD80 and CD86) was determined by flow cytometry 24 hours after treatment. We found that DMXAA-mediated STING activation markedly upregulated the expression of MHC-II, CD80 and CD86 relative to the control group (figure 3A–D). Importantly, cotreatment with C176 effectively abrogated DMXAA-induced upregulation of these activation markers, demonstrating that DC activation was mediated specifically through STING signalling.

Figure 3. STING activation in DCs induced podocyte injury and promoted pDC differentiation. (A–D) Expression of DC activation markers MHC-II (A), CD80 (B) and CD86 (C) after 24 hours of treatment (n = 5 per group). Quantification of MHC-II⁺, CD80⁺ and CD86⁺ cells among CD11c⁺ BMDCs following different treatments (D). (E) Experimental scheme for evaluating the effect of BMDC-derived conditioned media on podocyte injury. GM-CSF-induced BMDCs were treated with DMSO, C176, DMXAA or C176 combined with DMXAA for 48 hours. The collected conditioned media were then used to stimulate podocytes for 24 hours, followed by assessment of podocyte apoptosis and synaptopodin expression. (F) Podocyte apoptosis induced by conditioned media derived from differently treated BMDCs was evaluated by Annexin V/7-AAD staining. Representative flow cytometry plots and quantification of apoptotic podocytes are shown (n=5 per group). (G) Western blot analysis of synaptopodin expression in podocytes treated with conditioned media derived from differently treated BMDCs. β-actin was used as the loading control. A representative blot from three independent biological experiments is shown. (H–I) Proportions of PDCA1⁺ pDCs in BM (H) and spleen (I) after 48 hours of treatment (n=5 per group). Data are presented as mean±SD. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. 7-AAD, 7-aminoactinomycin D; BM, bone marrow; BMDCs, bone marrow-derived dendritic cells; DC, dendritic cell; DMSO, dimethyl sulfoxide; DMXAA, 5,6-dimethylxanthenone-4-acetic acid; GM-CSF, granulocyte-macrophage colony-stimulating factor; MHC, major histocompatibility complex; pDC, plasmacytoid dendritic cell; STING, stimulator of interferon genes.

Based on the correlation between renal STING expression and podocyte injury in patients with LN, we next examined whether STING activation in DCs could indirectly affect podocyte homeostasis. Conditioned media were collected from MRL/lpr-derived BMDCs treated with DMXAA and/or C176, and subsequently applied to cultured podocytes (figure 3E). Functional experiments showed that conditioned media from STING-activated DCs significantly induced podocyte apoptosis and reduced synaptopodin expression (figure 3F,G; online supplemental figure 1), whereas C176 treatment could reverse these effects. These results indicate that STING activation in DCs may promote podocyte injury in a paracrine manner.

In addition to its effects on DC activation, STING may also regulate DC differentiation. Given the well-recognised pathogenic role of pDCs in lupus progression, we further investigated whether STING signalling modulates pDC differentiation. In the pDC differentiation assay, we observed that, compared with the DMSO vehicle control, the STING agonist DMXAA treatment significantly increased the proportion of PDCA1⁺ pDCs in both bone marrow and splenic cultures. Conversely, C176 treatment efficiently abolished this inductive effect (figure 3H,I), indicating that STING activation promotes pDC differentiation.

Taken together, these findings suggest that STING signalling influences DC-mediated immune responses by promoting DC activation and pDC differentiation, which may contribute to the pathogenesis of LN.

STING inhibition ameliorated lupus manifestations in MRL/lpr mice

Given the regulatory role of STING in immune responses and its potential pathogenic contribution in lupus, we next evaluated whether STING blockade could confer therapeutic benefits in vivo. To optimise the intervention conditions for subsequent therapeutic assessment, we initiated a 2-week dose-exploration study in MRL/lpr mice using two dosing regimens of the STING inhibitor C176 (187.5 or 375 nmol, two times per day; online supplemental figure 2A). This short-term treatment partially reduced proteinuria (online supplemental figure 2D), supporting a potential renoprotective effect of STING blockade. However, the ANA and anti-dsDNA antibody levels showed no significant changes (online supplemental figure 2B and C). These findings indicated that while C176 exerted early benefits on renal function, a longer intervention period might be necessary to achieve systemic immunomodulation and attenuate autoantibody production. Therefore, we proceeded with a 4-week treatment regimen to further assess the therapeutic potential of STING inhibition in vivo. 18-week-old female MRL/lpr mice received C176 (375 nmol, two times per day) or vehicle, with age-matched MRL/Mpj mice included as healthy controls (figure 4A). Compared with the vehicle control group, C176 treatment significantly alleviated splenomegaly in MRL/lpr mice (mean difference vs controls, −4.69; 95% CI −8.55 to −0.83; BH-adjusted p=0.01) (figure 4B,C), without causing body weight fluctuations or drug-related mortality (figure 4D), indicating good tolerability. More importantly, mice in the C176-treated group exhibited significantly reduced serum levels of ANA (mean difference vs controls, −165.80; 95% CI −300.2 to −31.34; BH-adjusted p=0.01) (figure 4E), anti-dsDNA antibody (mean difference vs controls, −70.48; 95% CI −133.30 to −7.69; BH-adjusted p=0.02) (figure 4F) and IFN-α (mean difference vs controls, −118.5; 95% CI −192.20 to −44.80; BH-adjusted p=0.002) (figure 4G) compared with controls.

Figure 4. STING inhibition ameliorated lupus manifestations in MRL/lpr mice. (A) Schematic of the experimental design. 18-week-old female MRL/lpr mice were treated for 4 weeks with the STING inhibitor C176 (375 nmol per dose, two times per day, i.p.) or vehicle control, with age-matched MRL/Mpj mice serving as healthy controls. (B–C) Representative images and quantitative analysis of the spleen, with spleen/body weight ratio used to quantify spleen size in mice. (D) Representative weekly changes in body weight (%) of mice in each group throughout the experiment. (E–G) Serum levels of ANA (E), anti-dsDNA antibodies (F), IFN-α (G) in each group. Data are presented as mean±SD (n=5 per group). *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. dsDNA, double-stranded DNA; IFN, interferon; i.p., intraperitoneally; STING, stimulator of interferon genes.

Building on the observed improvements in systemic immune parameters, we further investigated the effects of C176 on local renal pathology. Histological analysis of kidney tissues, including H&E and PAS staining combined with semi-quantitative scoring, demonstrated that C176 treatment markedly alleviated glomerular structural damage (median difference vs controls, −1.00; 95% CI −2.00 to −0.00; adjusted p=0.04), perivascular inflammatory infiltration (median difference vs controls, −1.00; 95% CI −2.00 to −1.00; adjusted p=0.02) and tubulointerstitial lesions (median difference vs controls, −1.00; 95% CI −2.00 to −0.00; adjusted p=0.02) (figure 5A,B). Correspondingly, 24-hour urinary protein excretion was significantly reduced in the C176-treated group (mean difference vs 18 weeks, −0.50; 95% CI −0.99 to −0.004; adjusted p=0.049) (figure 5C), accompanied by decreased serum creatinine (mean difference vs control, −53.55; 95% CI −102.20 to −4.86; BH-adjusted p=0.02) (figure 5D) and blood urea nitrogen levels (mean difference vs control, −3.70; 95% CI −6.50 to −0.89; adjusted p=0.01) (figure 5E), further supporting its renoprotective effects. In addition to improving renal inflammation and proteinuria, we evaluated the effects of C176 on podocyte injury. Multiplex immunohistochemical analysis showed that C176 treatment markedly reduced STING expression in the kidneys of MRL/lpr mice (figure 5F,G), while significantly increasing the podocyte marker synaptopodin (figure 5F,H). TEM further revealed that C176 effectively ameliorated podocyte foot process effacement in MRL/lpr mice (mean difference vs control, −182.50; 95% CI −334.60 to −30.50; adjusted p=0.02) (figure 5I,J). These findings suggested that the renoprotective effects of STING inhibition were at least partially mediated through the alleviation of podocyte injury. Taken together, STING inhibition markedly ameliorated multiple lupus-associated phenotypes in MRL/lpr mice, not only reducing splenomegaly and autoantibody production but also effectively alleviating renal involvement, as evidenced by decreased proteinuria and podocyte injury.

Figure 5. STING inhibitor alleviated renal inflammation and podocyte injury in MRL/lpr mice. (A–B) Representative H&E and PAS staining of kidney sections in MRL/lpr mice (A) and semiquantitative scoring of renal glomerular, tubulointerstitial and perivascular pathology (B). Original magnification, ×400, scale bar: 100 μm. (C) Quantification of 24-hour urinary protein in mice at 18 and 22 weeks of age. (D–E) Serum levels of creatinine (D) and blood urea nitrogen (E) in each group. (F) Representative mIHC staining of STING (green) and synaptopodin (red) in kidney sections from MRL/Mpj controls, vehicle-treated MRL/lpr mice and C176-treated MRL/lpr mice. Original magnification, ×400, scale bar: 100 μm. (G–H) Quantification of STING (G) and synaptopodin (H) expression (IOD per mm²) in kidney. (I) TEM of renal tissues showing ultrastructural features of the GBM, podocyte foot processes (po) and slit diaphragm (red asterisk), ×6000 (scale bar: 1 μm) and ×30000 (scale bar: 200 nm). Podocyte foot process effacement was observed in glomeruli from MRL/lpr mice. (J) Quantitative analysis of podocyte foot process width in mice. Data are presented as mean±SD (n=5 per group). *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. DAPI, 4′,6-diamidino-2-phenylindole; GBM, glomerular basement membrane; IOD, integrated optical density; mIHC, multiplex immunohistochemistry; PAS, periodic acid-Schiff; STING, stimulator of interferon genes; TEM, transmission electron microscopy.

STING inhibition restricted DC expansion and modulated pDC differentiation in MRL/lpr mice

To elucidate the immunological mechanisms underlying the therapeutic effects of STING inhibition in lupus, we systematically analysed the changes in splenic immune cell subsets in MRL/lpr mice. Flow cytometric analysis revealed that treatment with the STING inhibitor C176 markedly reduced the absolute number of DCs (mean difference vs control, −531.40; 95% CI −840.70 to −222.20; BH-adjusted p=0.002) (figure 6F,G), whereas no significant changes were observed in monocytes (figure 6A,B), neutrophils (figure 6A,C) or macrophages (figure 6D,E). Building on our earlier findings that STING regulated DC activation and pDC differentiation, we next examined the role of STING signalling in shaping pDC populations in lupus-prone mice. To identify pDCs, we used two established surface markers, PDCA1 and Siglec-H. In MRL/lpr mice, both PDCA1⁺ pDCs (figure 6H,I) and Siglec-H⁺ pDCs (figure 6J,K) were markedly expanded in the spleen compared with age-matched MRL/Mpj controls, and C176 treatment significantly reduced the abundance of both populations. Notably, despite the lack of statistical significance, C176 treatment was associated with a consistent downward trend in the absolute numbers of both PDCA1⁺ and Siglec-H⁺ pDCs compared with controls (figure 6I,K), which may be attributable to increased intergroup variability. Besides, we also found that treatment with the STING inhibitor significantly reduced renal infiltration of CD11c⁺ cells compared with the vehicle control group (figure 6L). In summary, these findings indicated that STING inhibition reshaped the DC compartment in lupus by restraining DC expansion and modulating pDC populations, thereby potentially limiting downstream autoimmune activation.

Figure 6. STING inhibition restricted DC expansion and modulated pDC subsets in MRL/lpr mice. (A–C) Flow cytometric analysis of splenic monocytes and neutrophils (A), with panels B, C showing percentages and absolute numbers of each population. (D–G) Analysis of splenic macrophages (D) and DCs (F), with panels E, G showing percentages and absolute numbers. (H–K) Analysis of splenic pDC subsets, including PDCA1⁺ pDCs (H) and Siglec-H⁺ pDCs (J), with panels I, K showing percentages and absolute numbers. (L) Representative mIHC staining of renal sections showing CD11c⁺ cells infiltration (red) and nuclei (DAPI, blue) in MRL/Mpj controls, MRL/lpr vehicle controls and C176-treated groups. Original magnification, ×400, scale bar: 100 μm. Data are presented as mean±SD (n=5 per group). *p<0.05, **p<0.01, ****p<0.0001. DAPI, 4′,6-diamidino-2-phenylindole; DCs, dendritic cells; mIHC, multiplex immunohistochemistry; pDC, plasmacytoid dendritic cell; STING, stimulator of interferon genes.

STING inhibition attenuated the expansion of Th1, Th17 and plasma cells in MRL/lpr mice

Given the critical role of DCs in directing T and B cell responses, we hypothesised that STING inhibition would impact downstream adaptive immunity. Flow cytometric analysis revealed that treatment with C176 markedly reduced both the frequency (mean difference vs control, −10.67; 95% CI −18.57 to −2.76; BH-adjusted p=0.01) and absolute number (mean difference vs control, −662.10; 95% CI −895.40 to −428.80; BH-adjusted p<0.0001) of T helper cell (Th)1 cells (figure 7A,B), accompanied by significant decreases in Th17 cells (figure 7C,D) and plasma cells (figure 7E,F). In summary, these findings demonstrated that STING inhibition exerted broad immunomodulatory effects in lupus. By restraining DC expansion and modulating pDC populations, it limited downstream pathogenic Th1, Th17 and plasma cell responses, collectively contributing to the amelioration of lupus-like phenotypes in MRL/lpr mice.

Figure 7. STING inhibition attenuated the expansion of Th1, Th17 and plasma cells in MRL/lpr mice. (A–F) Flow cytometric analysis of splenic Th1 cells (A), Th17 cells (C) and plasma cells (E), with panels B, D and F showing the corresponding percentages and absolute numbers of each population (n=5 per group). Data are presented as mean±SD. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. IFN, interferon; IL, interleukin; STING, stimulator of interferon genes; Th, T helper cell.

Discussion

Our study systematically investigated the role of the STING signalling pathway in the pathogenesis of SLE by integrating clinical observations with experimental evidence. We observed markedly elevated STING expression in PBMCs from patients with SLE and in renal tissues from patients with LN, with levels correlating with disease activity, complement consumption and organ damage markers. Inhibition of STING significantly reduced proteinuria, ameliorated renal histopathology and function, and partially restored podocyte structural integrity, accompanied by reduced DC activation, modulation of pDC populations and suppression of pathogenic immune cell expansion. Furthermore, in vitro studies showed that conditioned media from STING-activated DCs promoted podocyte apoptosis and decreased synaptopodin expression, while STING inhibition alleviated these alterations, suggesting that DC-intrinsic STING activation may contribute to podocyte injury through paracrine mechanisms. Collectively, our findings identified STING as a potential mediator connecting immune dysregulation with podocyte structural damage in LN.

Previous studies have further elucidated the pathogenic role of STING signalling in LN. Chen et al reported that STING expression was significantly increased in the renal tissues of patients with LN and in MRL/lpr mice.²⁶ Mechanistically, STING aggravated renal injury by inducing ferroptosis and inflammatory responses through activation of the TBK1/NF-κB signalling pathway, whereas knockdown of STING alleviated lipid peroxidation, reduced inflammatory cytokine production and improved renal function. Consistent with these observations, we also observed upregulated STING expression in LN. However, our study extended these observations by examining STING expression across different histopathological classes of LN, revealing higher STING levels in proliferative forms of LN. The mechanistic focus differs between the two studies. Chen et al primarily emphasised STING-driven ferroptosis-related renal injury, whereas our study focused on the immunological consequences of STING activation, particularly its impact on DC responses and downstream immune dysregulation. Moreover, we found that pharmacological inhibition of STING using the small-molecule inhibitor C176 modulated DC activation, reduced pathogenic immune cell responses and ameliorated podocyte injury in MRL/lpr mice.

Accumulating evidence indicates that podocyte injury plays an important role in LN progression. Loss of podocyte-associated markers, including synaptopodin, nephrin and GLEPP1, is frequently observed in proliferative LN and is indicative of impaired glomerular filtration barrier integrity.²⁷ Multiple converging mechanisms contribute to the destabilisation of podocyte architecture, including immune complex-mediated complement activation, innate immune sensing, oxidative stress and inflammatory responses, ultimately leading to disruption of slit diaphragm proteins and glomerular filtration barrier dysfunction.^28–32 Within this framework, our study provides additional evidence linking innate immune sensing pathways to podocyte structural injury in LN. We observed a marked reduction in synaptopodin expression accompanied by a significant increase in STING expression in proliferative LN, with renal STING levels showing a significant inverse correlation with synaptopodin expression. Moreover, renal STING expression was significantly associated with glomerular cell proliferation scores, supporting a link between increased STING expression and proliferative glomerular injury. To further explore the mechanism by which STING activation in immune cells contributes to podocyte injury, we first investigated the impact of DC-associated STING signalling on podocyte homeostasis. Conditioned media derived from STING-activated DCs induced podocyte apoptosis and reduced synaptopodin expression, whereas STING inhibition attenuated these effects. These findings suggest that STING activation in DCs may contribute to podocyte injury through DC-derived soluble factors, providing a potential mechanistic link between immune activation and structural renal damage in LN. Furthermore, in vivo inhibition of STING partially restored podocyte integrity in lupus-prone mice, supporting a role for STING signalling in the progression of podocyte injury. In addition to immune cell-mediated effects, intrinsic activation of the cGAS-STING pathway within podocytes has also been implicated in podocyte dysfunction. Consistent with previous studies, SMURF1-YY1 signalling and APOL1 risk variants can engage the cGAS-STING pathway to induce podocyte apoptosis or autophagy defects.^{33 34} However, unlike these studies that primarily focus on intrinsic podocyte signalling, our work provides evidence that STING may contribute to LN progression through systemic immune remodelling and immune cell-mediated effects on podocyte homeostasis, based on human biopsy data and in vivo pharmacological intervention.

Although renal STING expression was generally elevated in patients with LN, its distribution varied across pathological classes, with higher levels observed predominantly in proliferative lesions rather than pure membranous LN. This pattern suggested that STING expression might be preferentially associated with inflammatory proliferative injury. Given the prominent inflammatory and immune activation features of proliferative LN, enhanced STING signalling may participate in innate immune-mediated renal injury. However, the heterogeneity of STING expression among LN subtypes may also reflect differences in disease activity, pathological features and treatment exposure. Notably, treatment with glucocorticoids, MMF, HCQ and B cell-targeted biologics was negatively associated with renal STING expression, suggesting that commonly used immunosuppressive therapies may modulate STING pathway activity in vivo. This observation is consistent with their reported effects on IFN-I related pathways,^35–38 which are closely linked to STING signalling. However, given the observational design and limited sample size, residual confounding, particularly from heterogeneity in treatment regimens and disease stage, cannot be excluded. Future studies with larger, well-stratified cohorts and longitudinal sampling are needed to clarify the relationship between STING expression, therapeutic exposure and histopathological subtypes of LN.

Given the central role of STING in IFN-I signalling and the importance of pDCs as IFN-I producers, both the number and functional state of pDCs are critical for SLE pathogenesis.^39–41 In our study, STING activation enhanced the activation phenotype of DCs and increased PDCA1⁺ pDC populations in vitro, whereas pharmacological inhibition of STING reversed these effects. Consistently, MRL/lpr mice exhibited expansion of splenic PDCA1⁺ and Siglec-H⁺ pDC populations accompanied by elevated IFN-α production, which was attenuated by C176 treatment. These findings indicated that STING signalling may contribute to the remodelling of the DC compartment and may amplify pDC-associated IFN-I responses in lupus.

Moreover, STING activation may also influence broader immune networks in LN. Our findings indicated that STING inhibition may limit the differentiation of pathogenic T cells including Th1 and Th17 and reduced plasma cell numbers, through constraining DC expansion and regulating pDC differentiation. These effects may be mediated by reduced IFN-I production and altered immune cell activation, including changes in DC responses and downstream T-cell and B-cell immunity.⁴² In addition, STING inhibition also decreased renal CD11c⁺ cell infiltration in MRL/lpr mice, indicating a direct modulation of the renal immune microenvironment that likely contributes to reduced proteinuria and podocyte injury.

As noted earlier, the functional role of STING in lupus pathogenesis remains controversial, with studies in different lupus-prone mouse strains reporting conflicting results, suggesting that its effects may depend on genetic background and experimental context. In this study, we observed that inhibition of STING with the selective inhibitor C176 effectively alleviated lupus-like pathology in MRL/lpr mice. This observation differs from previous reports based on STING knockout models. The discrepancy between pharmacological inhibition and genetic ablation suggests that STING may play a more complex, dualistic role in the development of SLE. Selective blockade of aberrantly activated STING signalling may suppress excessive production of IFN-I and other downstream inflammatory mediators, thereby mitigating tissue injury. By contrast, germline STING deficiency eliminates basal signalling and may provoke compensatory inflammatory pathways or impair homeostatic responses. Definitive resolution of these differences will require comparative studies using inducible or cell-type-specific STING deletion, controlled pharmacological regimens and cross-strain validations. Together, our findings suggest that selective pharmacological modulation of STING, rather than complete genetic ablation, may represent a more clinically relevant strategy to rebalance immune activation while preserving physiological immune function in SLE.

Several limitations of our study should be acknowledged. First, although our data demonstrate that STING activation in DCs promotes podocyte injury through soluble mediators and influences pDC-related phenotypes, the specific inflammatory factors involved and the functional contribution of the pDC–IFN-I axis require further investigation. Second, despite the integration of human samples, MRL/lpr mice and in vitro models, confirmation in additional lupus models and clinically diverse patient cohorts will be important to further establish the generalisability of these findings. Third, C176 effectively inhibits murine STING and demonstrated therapeutic effects in the MRL/lpr model, however, its translational applicability to human SLE requires further validation. Evaluation of human-STING-targeting compounds, such as H-151 or SN-011, together with further assessment of safety, dosing strategies and cell-specific mechanisms, will be required before clinical application. In particular, future studies should clarify the relative contribution of immune cell-derived and podocyte-intrinsic STING signalling to LN pathogenesis.

In conclusion, our study demonstrated that increased STING expression may contribute to systemic immune dysregulation including renal injury, through induction of IFN-I responses, alteration of immune cell composition and promotion of podocyte injury. Inhibition of STING restored immune homeostasis and alleviated renal damage in the MRL/lpr model, underscoring its potential as a novel therapeutic target for SLE. Our work further supports the involvement of STING in LN and provides a basis for future studies exploring STING-directed interventions.

Data availability statement

Data are available upon reasonable request.