A novel Keap1-Nrf2 inhibitor 6K delays the progression of AKI-to-CKD by mitigating maladaptive repair

Abstract

Severe or recurrent acute kidney injury (AKI) is a critical risk factor for chronic kidney disease (CKD) progression, characterized by irreversible fibrosis and limited therapeutic options. Maladaptive repair in proximal tubular epithelial cells (PTECs) during AKI-to-CKD progression is crucial, with the oxidative stress (OS)-ferroptosis axis emerging as a potential therapeutic target. Nuclear factor erythroid 2-related factor 2 (Nrf2), a key regulator of redox balance and ferroptosis, is essential for cellular homeostasis, but its role in PTECs maladaptive repair is unclear. In this study, we employed Nrf2 knockout (KO) mice to establish an AKI-to-CKD model through bilateral ischemia-reperfusion injury (bIRI). Nrf2 deficiency significantly exacerbated renal fibrosis, OS markers (H₂O₂, NOX4, 8-OHdG), and ferroptosis indicators (4-HNE, MDA, ACSL4), while concurrently suppressing antioxidant enzyme activity (SOD, GPx) and the expression of maladaptive repair-related genes (Vcam1, Irf8, etc.). Pro-inflammatory cytokines (IL-1β, IL-6, TNF-α) were significantly elevated in KO mice. We found increased oxidative and inflammatory markers in CKD patients' serum and urine, highlighting the role of OS in disease progression. Treatment with 6K, a Kelch-like ECH-associated protein 1 (Keap1)-Nrf2 protein-protein interaction (PPI) inhibitor, markedly improved renal function, suppressed OS and ferroptosis, and upregulated adaptive repair-related genes and downregulated maladaptive repair-related genes in bIRI mice. Compound 6K exhibited a kidney-targeted distribution profile, with an area under the concentration-time curve (AUC) kidney/AUC blood ratio of 1.01, suggesting its potential for targeted treatment of kidney diseases. In human kidney-2 (HK-2) cells, 6K activated the Nrf2-GPX4 axis, thereby alleviated RSL3-induced suppression of GPX4 expression, reduced reactive oxygen species (ROS) accumulation and lipid peroxidation, and mitigated ferroptosis. Furthermore, 6K treatment significantly delayed fibrosis progression in bIRI and unilateral ureteral obstruction (UUO) models. In summary, our findings demonstrate that Nrf2 deficiency exacerbates AKI-CKD progression through redox imbalance and ferroptosis-mediated maladaptive repair. Targeting Keap1-Nrf2 with 6K protects against renal injury and fibrosis, highlighting its therapeutic potential for kidney diseases.

Graphical abstract

1. Introduction

Chronic kidney disease (CKD) is a significant global health burden, affecting over 10% of the population worldwide, with a prevalence of 8.2% in Chinese adults[1,2]. By 2040, CKD is predicted to rank among the top causes of years of life lost[3]. The increasing incidence is driven by traditional risk factors such as diabetes, hypertension, and aging, aggravated by environmental exposures like heavy metals[4]. Recently, a growing body of evidence identifies acute kidney injury (AKI) as a critical and independent risk factor for the development and progression of CKD[5,6].

The transition from AKI to CKD is caused by complex stress response, in which maladaptive repair of proximal tubular epithelial cells (PTECs) plays a pivotal role[7]. The pathological process is complex and often involves multiple mechanisms, such as oxidative stress (OS), organelle stress, mitochondrial damage, cell-cycle arrest, and DNA leakage from mitochondria/nucleus into cytosol, which mediate tubular inflammation[8,9]. Notably, the kidney's high metabolic demand and mitochondrial density render it particularly susceptible to OS, which not only initiates acute damage but also perpetuates the chronic inflammation and fibrosis that define CKD[10,11]. Consequently, targeting the molecular pathways that govern redox homeostasis represents a promising therapeutic strategy.

The transcription factor nuclear factor erythroid 2-related factor 2 (Nrf2) plays a pivotal role in mitigating OS by regulating the expression of antioxidant and detoxifying enzymes[12]. Under normal conditions, Nrf2 is ubiquitinated by the Kelch-like ECH-associated protein 1 (Keap1)-E3 ubiquitin ligase complex and degraded. However, under stress, modifications to Keap1 prevent Nrf2 degradation, allowing it to translocate to the nucleus and activate protective genes[13,14]. Nrf2 also modulates inflammatory responses, cellular metabolism, and cell survival, including the regulation of ferroptosis[13]. Importantly, the biological consequences of Keap1-Nrf2 signaling are highly context dependent. Appropriate activation of Nrf2 can protect against oxidative and inflammatory injury in chronic inflammatory and degenerative disorders, whereas persistent or aberrant Nrf2 activation in certain cancers may promote tumor cell survival and therapeutic resistance. Thus, dysregulation of the Keap1-Nrf2 pathway has been implicated in the onset and progression of diverse diseases, highlighting its multifaceted role in cellular homeostasis and disease pathogenesis[[15], [16], [17]].

Despite these context-dependent effects, activation of Nrf2 is predominantly cytoprotective in acute and chronic kidney injury models. In preclinical models, Nrf2 activation has been shown to protect against kidney injury, reduce inflammation, and attenuate fibrosis[18]. However, the therapeutic translation of Nrf2 activation remains challenging[19,20]. Several Nrf2 activators, including dimethyl fumarate (DMF)[21] and bardoxolone methyl, have demonstrated beneficial effects in inflammatory and oxidative stress-related diseases[22]. Nevertheless, these compounds primarily activate Nrf2 through electrophilic modification of reactive cysteine residues on Keap1, which may limit molecular selectivity because electrophilic reactivity is not necessarily restricted to Keap1[23]. Therefore, developing alternative strategies that directly modulate the Keap1-Nrf2 interaction represents an attractive approach to achieve more precise regulation of Nrf2 signaling. In previous work, we identified a series of non-covalent Keap1-Nrf2 protein-protein interaction (PPI) inhibitors featuring an azetidine-containing naphthalenesulfonamide scaffold, which directly disrupt the Keap1-Nrf2 interaction and promote Nrf2 activation without relying on electrophilic modification[18,24,25]. Among these, compound 6K exhibited high Keap1-binding affinity (K_D2 = 210 nM), favorable drug-like properties, and potent anti-inflammatory effects in vivo[26].

Building on this, the present study explores the potential of compound 6K in attenuating AKI injury and delaying AKI-to-CKD progression. We hypothesize that activation of the Keap1-Nrf2 pathway by compound 6K will attenuate OS, inflammation, and ferroptosis, thereby prevents maladaptive repair and ultimately slowing CKD progression.

2. Results

2.1. Nfe2l2 knockout accelerated the progression from AKI to CKD

To investigate the dynamics of Nrf2 during kidney injury, we first analyzed published single-cell sequencing datasets of ischemia-reperfusion injury (IRI) models. In both bilateral and unilateral IRI, Nfe2l2 expression transiently increased within the first 4-6 h of reperfusion but declined thereafter, suggesting that Nrf2 activation is an early yet unsustained protective response (Fig. 1A–B). Consistently, in our bilateral IRI (bIRI) model, Nrf2 protein levels rose at 4 h but diminished over days 1-14 (Fig. 1C–D).

Fig. 1.

Nfe2l2 deletion accelerated AKI-to-CKD progression. (A) Nfe2l2 expression in mouse IRI kidney profiled with sn-RNA-seq. (B) Nfe2l2 expression in mouse kidney uni-IRI profiled with sci-RNA-seq3. (C) Western blot analysis of cytoplasmic Nrf2 (Cyto-Nrf2) and nuclear Nrf2 (Nuc-Nrf2) protein expression in mice renal cortex at different reperfusion times following ischemia. (D) Quantification of C. (E) Kaplan-Meier survival curves of WT and KO mice following bIRI. (F-G) Scr levels (F) and BUN levels (G) at different reperfusion times following ischemia in WT and KO mice. (H) GFR of WT and KO mice after bilateral renal ischemia and various durations of reperfusion was dynamically monitored. (I) UACR at different reperfusion times following ischemia in WT and KO mice. (J-K) Relative expression levels of Ngal (J) and Kim1 (K) mRNA in renal cortex of WT and KO mice at different reperfusion times following ischemia. (L-M) Histopathological Masson staining images (L) and collage volume fraction quantification (M) of WT and KO mice kidney at different reperfusion times following ischemia. Scale bar: 200 μm. (N-O) Representative immunofluorescence images (N) and quantification (O) of α-SMA (red) and nuclei (blue) in WT and KO mice kidney tissue under bIRI of 14 days. Scale bar: 200 μm. Data were presented as mean ± SEM (n = 5-6) and analyzed using ANOVA analysis. *p < 0.05, **p < 0.01, ***p < 0.001 versus the sham-operated group with the same genotype. ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 versus the WT bIRI group.

To determine the functional role of Nrf2, we compared wild-type (WT) and Nfe2l2 knockout (KO) mice subjected to bIRI. KO mice showed reduced survival (Fig. 1E), more severe kidney dysfunction with higher serum creatinine (Scr), blood urea nitrogen (BUN), and urinary albumin-to-creatinine ratio (UACR) (Fig. 1F, G & I), together with a greater decline in glomerular filtration rate (GFR), indicating impaired recovery and maladaptive repair (Fig. 1H). Early injury markers kidney injury molecule 1​(Kim1) and neutrophil gelatinase-associated lipocalin​ (Ngal) were persistently higher in KO mice (Fig. 1J–K), and histology revealed aggravated tubular damage (Fig. S1). Fibrosis developed more rapidly in KO mice, as shown by increased collagen deposition, alpha-smooth muscle actin (α-SMA) upregulation, and epithelial cadherin (E-cadherin) loss (Fig. 1L–O, S2). Together, these data demonstrate that Nrf2 loss exacerbates acute injury and drives progression toward irreversible CKD.

2.2. Nfe2l2 knockout deteriorated OS and ferroptosis in kidney of bIRI mice

Given Nrf2's role in redox regulation, we examined OS in WT and KO kidneys after bIRI. Reactive oxygen species (ROS) accumulation, measured by hydrogen peroxide (H₂O₂) levels, was elevated in all IRI mice but significantly higher in KO mice (Fig. 2A–B). Antioxidant defenses including superoxide dismutase (SOD), manganese superoxide dismutase (Mn-SOD), glutathione peroxidase (GPx), and glutathione (GSH) were markedly suppressed, especially in KO mice (Fig. 2C–F). Similarly, Nrf2 targets NAD(P)H quinone dehydrogenase 1 (NQO1) and glutathione peroxidase 4 (GPX4) were downregulated, while pro-oxidant NADPH oxidase 4 (NOX4) was strongly induced (Fig. 2J–K).

Fig. 2.

Knockout of Nrf2 exacerbated OS and ferroptosis in AKI. (A) Serum H₂O₂ levels of WT and KO mice under 1_day bIRI. (B-F, H-I,L&N) H₂O₂ levels (B), SOD activity (C), Mn-SOD activity (D), GPx activity (E), GSH levels (F), mtDNA copy number (H), mitochondrial 8-OHdG levels (I), divalent iron concentration (L) and MDA levels (N) in renal cortex of WT and KO mice under bIRI of 1 d. (G) Electron microscopy images of mitochondria in WT and KO mice renal cortex under 1_d bIRI condition. Yellow arrows indicate swollen mitochondria with disrupted or damaged cristae structures. Scale bar: 100 nm. (J & K) Western blot analysis of NOX4, NQO1, GPX4, FTH and ACSL4 proteins (J) and relative levels (K) in WT and KO mice renal cortex subjected to bIRI of 1 d. (M) 4-HNE immunohistochemical staining image of WT and KO mice kidney subjected to bIRI of 1 d. Scale bar: 100 μm. (O-P) Relative levels of Ptgs2 (O) and Chac1 (P) mRNA expression in WT and KO mice renal cortex following bIRI of 1 d. (Q) Immunofluorescence images of HMGB1(red) and nuclei (blue) of WT and KO mice kidney tissues under bIRI of 1 d. Scale bar: 200 μm. Data were presented as mean ± SEM (n = 5-6) and analyzed using ANOVA analysis. *p < 0.05, **p < 0.01, ***p < 0.001 versus sham-operated group with the same genotype. ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 versus WT bIRI group.

Loss of Nrf2 also exacerbated mitochondrial injury. Compared with sham WT kidneys, bIRI induced mitochondrial swelling, disruption of cristae organization, and increased electron-lucent regions. These ultrastructural abnormalities were further aggravated in Nrf2-deficient mice, which exhibited more severe mitochondrial structural disruption and loss of mitochondrial integrity (Fig. 2G). Consistent with these ultrastructural changes, Nrf2 deficiency was associated with increased mitochondrial oxidative DNA damage, reflected by elevated 8-hydroxy-2'-deoxyguanosine (8-OHdG) levels, together with reduced mitochondrial DNA (mtDNA) copy number, indicating impaired mitochondrial genomic stability (Fig. 2H–I). Ferroptotic injury was further exacerbated, as evidenced by increased ferritin heavy chain (FTH) expression and ferrous iron (Fe²⁺) accumulation (Fig. 2J–L), elevated lipid peroxidation products, including 4-hydroxynonenal (4-HNE) and malondialdehyde (MDA) (Fig. 2M–N, S3A), and increased acyl-CoA synthetase long-chain family member 4 (ACSL4) expression (Fig. 2J–K). The mRNA levels of ferroptosis-related genes (prostaglandin-endoperoxide synthase 2 [Ptgs2], chaC glutathione-specific gamma-glutamylcyclotransferase 1 [Chac1]), along with the protein expression of high-mobility group box 1 (HMGB1) (Fig. 2Q–S3B) were also upregulated, linking redox imbalance to inflammatory signaling. These results confirm that Nrf2 deficiency promotes OS-driven ferroptosis, aggravating kidney injury.

2.3. Nrf2 deletion exacerbated maladaptive repair and inflammation in the progression of AKI-CKD

To assess how Nrf2 influences repair outcomes, we analyzed adaptive and maladaptive repair markers. Genes linked to adaptive repair (low-density lipoprotein receptor-related protein 2 [Lrp2], acyl-CoA synthetase medium-chain family member 2 [Acsm2], hepatocyte nuclear factor 1 homeobox B [Hnf1b]) were reduced in bIRI kidneys, while maladaptive markers (interferon regulatory factor 8 [Irf8], vascular cell adhesion molecule 1 [Vcam1], potassium voltage-gated channel interacting protein 4 [Kcnip4], SRY-box transcription factor 4 [Sox4], SRY-box transcription factor 9 [Sox9], and paired box 2 [Pax2]) were persistently upregulated, with stronger induction in KO mice (Fig. 3A–F, S4). Among these, VCAM1 served both as a maladaptive repair marker and an indicator of inflammation, and its expression was significantly higher in KO kidneys (Fig. 3G–S5).

Fig. 3.

Knockout of Nrf2 deteriorated maladaptive repair and inflammation in AKI-CKD progression. (A) Lrp2, Acsm2, Hnf1b, Irf8 and Vcam1 expression in mouse kidney uni-IRI profiled with sci-RNA-seq3. (B-F) Relative expression levels of Lrp2 (B), Acsm2 (C), Hnf1b (D), Irf8 (E) and Vcam1 (F) mRNA in WT and KO mice renal cortex at different reperfusion times following ischemia. (G) Immunofluorescence analysis of VCAM1 in WT and KO mice kidney subjected to bIRI of 1 day. Scale bar: 200 μm. (H) Serum IL-6 levels in WT and KO mice following bIRI of 1 day and 14 days. (I-K) Relative expression levels of Il1b (I), Il6 (J) and Tnf (K) mRNA in WT and KO mice renal cortex following bIRI of 1 day and 14 days. Data were presented as mean ± SEM (n = 5-6) and analyzed using ANOVA analysis. *p < 0.05, **p < 0.01, ***p < 0.001 versus sham-operated group with the same genotype. ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 versus WT bIRI group.

Concordantly, the mRNA expression levels of pro-inflammatory cytokines, including interleukin-1β (Il1b), interleukin-6 (Il6), and tumor necrosis factor-α (Tnf) were markedly induced in bIRI kidneys and further elevated in KO mice at both 1 day and 14 days of reperfusion, with IL-6 also increased in serum (Fig. 3H–K). These findings indicate that Nrf2 protects against maladaptive repair and sustained inflammation during the AKI-to-CKD transition.

2.4. Ferroptosis and maladaptive repair are strongly associated with CKD patients

To validate these mechanisms in humans, we analyzed samples from CKD patients (Supplementary Table S1-2). Compared with healthy controls, patients showed impaired kidney function (elevated Scr, BUN, reduced GFR) (Fig. 4A–C) and systemic redox imbalance, with higher serum H₂O₂, IL-6, and HMGB1 in serum and urine (Fig. 4D–G). Transcriptomic analysis further revealed reduced adaptive repair genes (IRF2, HNF1B, ACSM2A), increased maladaptive repair genes (IRF8, VCAM1), and enrichment of inflammatory and ferroptosis markers (TNF, IL1B, C-C motif chemokine ligand 2 [CCL2], HMGB1, CHAC1, FTH1) (Fig. 4H). These results support the clinical relevance of Nrf2-regulated redox imbalance, ferroptosis, and maladaptive repair in CKD progression.

Fig. 4.

CKD was strongly associated with ferroptosis and maladaptive repair in patients. (A-G) Serum BUN (A), GFR (B), Scr (C), serum H₂O₂ (D), serum IL-6 (E), serum HMGB1 (F) and urinary HMGB1 (G) levels in healthy volunteers (n = 34) and CKD patients (n = 25). Data were presented as mean ± SEM and analyzed using t-test. *p < 0.05, **p < 0.01, ***p < 0.001 versus healthy group. (H) LRP2, HNF1B, ACSM2, IRF8, VCAM1, TNF, IL1B, CCL2, GPX4, glutathione synthetase (GSS), HMGB1, CHAC1 and FTH mRNA expressions in CKD (n = 5) vs. normal kidney (n = 3) from validation set of Nakagawa CKD dataset.

2.5. 6K alleviates renal function impairment in AKI and exhibits preferential renal distribution in mice

We next evaluated the therapeutic efficacy and in vivo distribution profile of 6K (chemical formula: C₃₆H₄₃N₇O₈S₂; molecular weight: 765.90 g/mol) (Fig. 5A), a Keap1-Nrf2 protein-protein interaction inhibitor, in an AKI mouse model. Treatment with 6K significantly improved renal function, as evidenced by reduced serum Scr and BUN levels (Fig. 5B–C), decreased Kim1 and Ngal expression (Fig. 5D–E). Histological examination further confirmed the protective effects of 6K on renal structural integrity. Compared with sham mice, bIRI mice displayed prominent tubular pathological alterations, including tubular dilation, intraluminal cast formation, and tubular epithelial injury characterized by loss of brush border structures. Treatment with 6K markedly alleviated these pathological changes, with medium- and high-dose groups showing substantially reduced tubular damage (Fig. 5F and S6A).

Fig. 5.

Chemical structure, pharmacokinetics, preferential renal distribution, and acute renal protection of 6K (A) Chemical structure and molecular formula of compound 6K (previously synthesized and characterized in J. Med. Chem. 2022). (B-C, M) Scr (B), serum BUN (C) and serum H₂O₂ (M) levels of mice subjected to bIRI of 1 day with or without 6K treatments. (D-E) Relative expression levels of Ngal (D) and Kim1 (E) mRNA in mice renal cortex subjected to bIRI of 1 day with or without 6K treatments. (F) Histopathological HE staining images of mice renal cortex tissue under bIRI of 1 day with or without 6K treatments. Arrowheads indicate tubular dilation (yellow), intraluminal casts (green), and tubular epithelial damage (white). Scale bar: 200 μm. (G) Representative whole-body fluorescence images of mice after intraperitoneal administration of fluorescently labeled 6K (10 mg/kg). Fluorescence signals were monitored at indicated time points (0, 1, 5, 15, 30, 45, 60, 75, 90, 120, 180, and 360 min) using an in vivo fluorescence imaging system. A time-dependent increase in fluorescence signals was observed in the renal region after administration, followed by gradual signal attenuation over time. (H-I) Western blot analysis (H) and relative protein levels (I) of nuclear Nrf2 (Nuc-Nrf2), cytoplasmic Nrf2 (Cyto-Nrf2), Keap1, NQO1, NOX4, GPX4, ACSL4 and FTH protein levels in mice renal cortex subjected to bIRI of 1 day with or without 6K treatments. (J-L,N-O, R-S) GSH activity (J), GPx activity (K), kidney H₂O₂ levels (L), mtDNA copy number (N), mitochondrial 8-OHdG levels (O), MDA levels (R) and renal Fe²⁺ concentration (S) in mice renal cortex subjected to bIRI of 1 day with or without 6K treatments. (P) Electron microscopy images of mitochondria in kidney cortex of mice subjected to bIRI of 1 day with or without 6K treatments. Yellow arrows indicate swollen mitochondria with disrupted or damaged cristae structures. Scale bar: 100 nm. (Q) 4-HNE immunohistochemical staining image of mice kidney subjected to bIRI of 1 day with or without 6K treatments. Scale bar: 100 μm. (T) Immunofluorescence images of HMGB1 (red) and nuclei (blue) of mice kidney subjected to bIRI of 1 day with or without 6K treatments. Scale bar: 200 μm. For sham groups: mice were under sham operation and administered equivalent volume of solvent (sham) or intraperitoneally injected with 10 mg/kg of 6K (sham+6K). For bIRI group: mice were under bIRI operation of 1 d and administered equivalent volume of solvent (bIRI). For 6K treatment groups: mice were under bIRI operation of 1 d and intraperitoneally injected with 0.4 mg/kg (LD), 2 mg/kg (MD) and 10 mg/kg (HD) of 6K. Data were presented as mean ± SEM (n = 5-6) and analyzed using ANOVA analysis. *p < 0.05, **p < 0.01, ***p < 0.001 versus bIRI group.

Pharmacokinetic analysis revealed rapid systemic clearance of 6K, with a plasma elimination half-life (t_1/2) of 0.28 h following intraperitoneal administration. Tissue distribution analysis demonstrated preferential renal exposure compared with cardiac tissue. The kidney exhibited a higher maximum concentration (C_max) than the heart (2447 vs. 2146 ng/g) and showed prolonged retention, with a renal t_1/2 of 4.69 h. In addition, the kidney-to-plasma exposure ratio was 1.01 based on area under the concentration-time curve from time zero to infinity (AUC_0-∞), whereas the heart showed a lower exposure ratio (heart/plasma = 0.37) (Supplementary Table S3-4). These findings indicate favorable renal distribution and sustained kidney exposure of 6K after systemic administration.

To further visualize the in vivo distribution profile of 6K, fluorescence imaging was performed using fluorescently labeled 6K (BODIPY-FL conjugate, chemical formula: C₅₅H₅₅BF₂N₈O₁₂S₄; molecular weight: 1197.14 g/mol) (Fig. S7A) following intraperitoneal administration (10 mg/kg). Consistent with the pharmacokinetic findings, fluorescence signals were rapidly detected in the renal region after administration and gradually increased during the early distribution phase (Fig. 5G–S7B). Quantitative analysis of renal fluorescence intensity from three independent mice further confirmed the time-dependent renal localization pattern of labeled 6K. Together with pharmacokinetic analysis, in vivo imaging provides complementary quantitative and visual evidence that 6K preferentially accumulates in the kidney.

2.6. 6K restores Nrf2-dependent redox homeostasis and suppresses ferroptotic injury in AKI kidneys

Mechanistically, 6K exerted its protective effects by modulating the Keap1-Nrf2 antioxidant pathway. 6K disrupted the Keap1-Nrf2 interaction, promoted Nrf2 nuclear translocation, and activated downstream antioxidant responses without altering Keap1 expression (Fig. 5H–I). This enhanced Nrf2 signaling increased the expression of antioxidant enzymes, including NQO1 and GPX4, restored endogenous antioxidant capacity as indicated by recovered SOD, GPx, and GSH levels, and reduced ROS generation and accumulation through suppression of NOX4 expression and H₂O₂ production (Fig. 5H–I, 5J-M, S8). Consequently, mitochondrial oxidative damage was attenuated, reflected by decreased mitochondrial 8-OHdG accumulation and preservation of mtDNA copy number (Fig. 5N–O). Consistent with the improvement in redox homeostasis, ultrastructural analysis by transmission electron microscopy (TEM) demonstrated that bIRI-induced mitochondrial damage, characterized by mitochondrial swelling, disrupted cristae organization, and increased electron-lucent regions, was markedly attenuated by 6K treatment, particularly at medium and high doses (Fig. 5P). Restoration of antioxidant capacity by 6K further limited ferroptotic injury in AKI kidneys. Compared with bIRI mice, 6K treatment reduced Fe²⁺ accumulation (Fig. 5S), decreased lipid peroxidation products including 4-HNE (Fig. 5Q–S6B) and MDA (Fig. 5R), and suppressed ACSL4 expression (Fig. 5H–I). Furthermore, ferroptosis-associated genes, including Chac1, Ptgs2, and Acsl4, as well as HMGB1 expression, were significantly reduced following 6K treatment (Fig. 5T–S6C, S9). These results indicate that 6K-mediated Nrf2 activation restores redox balance, thereby limiting ferroptotic damage and preserving mitochondrial and renal integrity during AKI progression.

2.7. 6K activates the Nrf2-GPX4 antioxidant pathway and protects HK-2 cells against RSL3-induced ferroptotic injury

In vitro, immunofluorescence analysis demonstrated that 6K treatment induced a time-dependent increase in Nrf2 nuclear localization, reaching a peak at 4 h after treatment, whereas Keap1 fluorescence intensity remained largely unchanged (Fig. 6A, S10A). Interestingly, RSL3 stimulation itself triggered Nrf2 nuclear accumulation, suggesting an endogenous compensatory response to stress (Fig. 6B–S10B). However, this adaptive response was insufficient to prevent ferroptotic injury. RSL3 treatment resulted in a marked reduction in GPX4 expression at both protein (Fig. 6C–D) and mRNA levels (Fig. S10C), accompanied by depletion of intracellular GSH levels (Fig. 6E) and increased oxidative stress, as evidenced by elevated 8-OHdG and H₂O₂ (Fig. 6F–G). Compared with RSL3 treatment alone, 6K further enhanced Nrf2 nuclear localization (Fig. 6B–S10B), restored GPX4/GSH levels (Fig. 6C–E, S10C) and reduced 8-OHdG and H₂O₂ levels (Fig. 6F–G). Restoration of antioxidant capacity by 6K further suppressed ferroptotic responses, as demonstrated by reduced lipid peroxidation (Fig. 6H–S10D) and decreased expression of ferroptosis-associated genes, including PTGS2, CHAC1, and ACSL4 (Fig. 6I). Moreover, 6K preserved mitochondrial function by preventing RSL3-induced mitochondrial membrane depolarization and restoring adenosine triphosphate (ATP) production (Fig. 6J–K, S10E). Finally, 6K attenuated the induction of renal tubular injury markers KIM-1 and NGAL following ferroptotic stress (Fig. 6I), indicating protection of renal tubular epithelial cells.

Fig. 6.

6K activates the Nrf2-GPX4 antioxidant axis and alleviates oxidative stress and ferroptotic damage in RSL3-treated HK-2 cells. (A) Representative immunofluorescence images of in HK-2 cells treated with 10 μM 6K for 0, 1, 2, 4, 8 and 12 h. A: Nrf2 (red), Keap1 (green) and nuclei (blue). (B) Representative immunofluorescence images of Nrf2 (red), Keap1 (green) and nuclei (blue) in HK-2 cells treated under different treatments. Scale bar: 100 μm. (C-D) Western blot analysis (C) and relative protein levels (D) of GPX4 in HK-2 cells treated with RSL3 and different concentrations of 6K (5 and 10 μM). Data were presented as mean ± SEM (n = 3) and analyzed using ANOVA analysis. *p < 0.05, **p < 0.01, ***p < 0.001 versus RSL3 group. (E) Intracellular GSH levels determined by LC-MS/MS analysis in HK-2 cells after indicated treatments. (F-G) Quantification 8-OHdG (F) and H₂O₂ (G) in HK-2 cells treated with RSL3 and/or 6K. (I) Relative expression levels of KIM1, NGAL, CHAC1, PTGS2 and ACSL4 mRNA in HK-2 cells under different treatments. (H & J) Analysis of lipid peroxidation stained with C11-bodipy581/591 (H) and mitochondrial membrane potential stained with JC-1 (J) in HK-2 cells by flow cytometry under different treatments. (K) ATP levels in HK-2 cells under different treatments. Data were presented as mean ± SEM (n = 5-6) and analyzed using ANOVA analysis. *p < 0.05, **p < 0.01, ***p < 0.001 versus RSL3 group.

2.8. 6K alleviated maladaptive repair and inflammation in AKI kidney and retarded AKI-CKD progression in bIRI mice

Finally, we assessed long-term outcomes. In AKI mice, 6K promoted adaptive repair genes (Lrp2, Acsm2, Hnf1b) while suppressing maladaptive markers (Irf8, Pax2, Sox4, Sox9) (Fig. 7A–D, S11). It also reduced VCAM1 expression, lowered pro-inflammatory cytokines (IL-6, IL-1β, TNF-α, MCP1) mRNA expression in kidneys, and decreased serum IL-6 (Fig. 7E–G, S12-S13).

Fig. 7.

Effects of 6K on maladaptive repair and inflammation in the kidney of AKI mice. (A-D & G) Relative expression levels of Lrp2 (A), Acsm2 (B), Hnf1b (C), Irf8 (D) and Il6 (G) mRNA in mice renal cortex under bIRI of 1 day with or without 6K treatments. (E) Representative immunofluorescence image of VCAM1(red) and nuclei (blue) of mice kidney tissues subjected to bIRI of 1 day with or without 6K treatments. Scale bar: 200 μm. (F) Serum IL-6 levels in mice subjected to bIRI of 1 day with or without 6K treatments. Data were presented as mean ± SEM (n = 5-6) and analyzed using ANOVA analysis. *p < 0.05, **p < 0.01, ***p < 0.001 versus bIRI group. n.s. means no significance.

In bIRI mice, continuous 6K treatment for 7–14 days improved renal function (lower Scr, BUN) (Fig. 8A–B), alleviated tubular injury and fibrosis (Fig. 8C–S14-S15), reduced α-SMA, and restored E-cadherin. Similarly, in unilateral ureteral obstruction (UUO) models, 6K mitigated tubular dilation, suppressed fibrosis, and normalized epithelial-mesenchymal markers (Figs. S16–S17). Together, these results demonstrate that 6K not only alleviates acute injury but also slows maladaptive repair and fibrosis, thereby preventing AKI-to-CKD progression.

Fig. 8.

6K retarded renal function damage and fibrosis deposition in AKI-CKD progression. (A-B) Scr (A) and BUN (B) levels of mice under bIRI of 7 days and 14 days with or without 6K treatments. (C) Histopathological and immunofluorescence analysis of kidney tissues under bIRI of 7 days and 14 days with or without 6K treatments. Top two rows: HE staining images. Middle two rows: Masson's trichrome staining images. Bottom Three rows: representative immunofluorescence images of α-SMA (red) and nuclei (blue). Scal bar: 200 μm. Data were presented as mean ± SEM (n = 5-6) and analyzed using ANOVA analysis. *p < 0.05, **p < 0.01, ***p < 0.001 versus bIRI group. n.s. means no significance.

3. Discussion

In this paper, we identified that a novel Keap1-Nrf2 inhibitor 6K can attenuate the development of CKD after AKI and retard further progression. We demonstrated that deletion of Nfe2l2 in mice exacerbates OS, ferroptosis, and inflammation in AKI, which contributed to the maladaptive repair and promotes this AKI-CKD progression. 6K alleviates OS, ferroptosis, and maladaptive repair in AKI, and improves the renal function and reduces fibrosis deposition in the AKI-CKD progression.

To investigate the transition from AKI to CKD, we established a robust murine model of bIRI. This model successfully recapitulated the human disease progression, showing an initial severe decline in GFR with corresponding elevations in serum creatinine, BUN, and renal tubular injury biomarkers like KIM-1 and NGAL. Over 14 days, these acute changes evolved into hallmarks of CKD, including a sustained high UACR and significant interstitial fibrosis. Especially, our model captured the dual nature of the post-AKI kidney, characterized by a gradual improvement in GFR and a decline in serum creatinine after the initial peak. These results suggested partial recovery alongside a progression to maladaptive repair. However, the severity of the initial insult ultimately triggered a maladaptive cascade characterized by persistent inflammation, flattened tubular epithelium, capillary rarefaction, and the accumulation of α-SMA⁺ myofibroblasts. This profibrotic environment fosters irreversible tubular atrophy and nephron loss, driving the transition to CKD. This validated model provided a clinically relevant platform to demonstrate that Nrf2 knockout fatally accelerates this process, whereas treatment with compound 6K effectively attenuates it, underscoring its therapeutic potential.

OS is a key driver of the AKI-to-CKD transition. Our analysis of human (GSE66494) and single-cell mouse data (GSE190887, GSE139107) confirmed the altered expression of many inflammatory and antioxidant genes, including the downregulation of GPX4 and a dynamic, but ultimately insufficient, upregulation of Nrf2 in tubular cells post-injury. In our bIRI model, Nrf2 expression peaked transiently before declining, and Nrf2 knockout mice suffered from exacerbated injury, increased oxidative stress, and severe inflammation. This demonstrates that impaired Keap1-Nrf2 signaling is a critical factor accelerating the progression to CKD.

Although pharmacological activation of Nrf2 represents a promising therapeutic strategy, currently available Nrf2 activators have limitations associated with their mechanisms of action. For example, DMF, an approved Nrf2 activator for multiple sclerosis, and synthetic triterpenoids such as bardoxolone methyl activate Nrf2 through electrophilic modification of reactive cysteine residues in Keap1[23,27]. However, the electrophilic activity of these compounds is not restricted to Keap1. DMF, for instance, can induce succination of additional cysteine-containing proteins such as glyceraldehyde-3-phosphate dehydrogenase (GAPDH), thereby affecting cellular metabolism and immune responses beyond direct Nrf2 activation[28]. Similarly, highly reactive electrophilic compounds may modify multiple cellular targets, potentially reducing pathway selectivity. These considerations highlight the need for alternative strategies that selectively regulate the Keap1-Nrf2 axis.

In this study, we investigated compound 6K, a non-covalent Keap1-Nrf2 PPI inhibitor originally discovered via a fragment-based strategy, as a distinct therapeutic approach to activate endogenous antioxidant defense[25]. Unlike electrophilic Nrf2 activators that rely on covalent modification, 6K directly binds to the Kelch domain of Keap1 with high affinity (K_D = 0.21 μM) and sterically disrupts the Keap1-Nrf2 interaction, thereby promoting Nrf2 nuclear accumulation without altering total Keap1 abundance[26]. Importantly, pharmacokinetic and fluorescence imaging analyses demonstrated favorable renal exposure of 6K after systemic administration, characterized by a kidney-to-plasma AUC ratio of 1.01 and sustained renal retention. Although fluorescence imaging revealed temporal fluctuations in renal signals at intermediate time points, these changes should be interpreted cautiously because fluorescence intensity reflects the distribution of fluorescent species rather than the absolute concentration of parent 6K. Factors including tissue optical properties, animal positioning, and fluorophore-related characteristics may influence the measured signal. Therefore, the renal preferential distribution profile of 6K was interpreted based on the combined evidence from fluorescence imaging and LC-MS/MS pharmacokinetic analysis. Although the parent compound 6K exhibits direct Keap1-Nrf2 PPI inhibitory activity, the metabolic fate of 6K has not yet been comprehensively characterized. Therefore, the potential contribution of pharmacologically active metabolites to the magnitude or duration of the in vivo protective effects cannot be excluded. Future studies integrating metabolite profiling, structural identification, and pharmacological evaluation of major metabolites will be required to distinguish the respective contributions of parent 6K and its metabolites. These pharmacological properties provide a sound rationale for evaluating 6K as a targeted Nrf2 modulator for kidney diseases characterized by persistent oxidative stress and ferroptotic injury. By reinforcing the Nrf2-GPX4 antioxidant axis and improving tubular cell resilience, 6K effectively attenuated AKI-associated acute injury and delayed maladaptive repair and fibrosis progression.

OS and inflammation contribute to various forms of cell death, including ferroptosis, which is implicated in AKI-to-CKD progression[[29], [30], [31]]. Our bioinformatics analysis of CKD kidney tissue (GSE 66494) showed altered ferroptosis markers (ACSL4, GPX4, FTH1). Nrf2 deletion in IRI mice exacerbated severe ferroptosis, characterized by iron dysregulation, mitochondrial damage, and altered expression of key markers like GPX4 and ACSL4. Crucially, compound 6K activated the Nrf2-GPX4 axis, and potently rescued renal cells from ferroptosis, restoring mitochondrial integrity and normalizing iron metabolism. To further investigate the cellular mechanisms underlying the protective effects of 6K against ferroptotic injury, we established an RSL3-induced ferroptosis model in renal tubular epithelial cells. By enhancing Nrf2 activity, 6K reinforces the GPX4/GSH axis, thereby maintaining cellular redox homeostasis and limiting the propagation of oxidative damage.This improved antioxidant capacity may subsequently preserve mitochondrial function and protect renal tubular epithelial cells from ferroptotic injury.

AKI was once considered a self-limiting disease due to the kidney's self-repair capabilities, but studies now confirm that complete recovery is rare[32]. Maladaptive repair mechanisms, as indicated by increased HMGB1 levels in CKD patient serum and urine, contribute to poor recovery after AKI[33]. Our single-cell sequencing analysis revealed changes in maladaptive repair markers in the kidneys of mouse models as disease progressed. In the IRI model, maladaptive repair markers such as Sox4, Sox9, Vcam1, and Irf8 decreased initially but later increased, particularly in KO mice. 6K also protected the kidney away from a maladaptive repair program by suppressed the expression of maladaptive markers like Sox9 and VCAM1, resulting in a benificial recovery environment rather than pathological remodeling.

Ultimately, by suppressing these upstream processes, 6K effectively inhibited the progression to renal fibrosis, the irreversible hallmark of CKD and the AKI-to-CKD transition. In both post-ischemic and UUO-induced fibrosis models, 6K significantly reduced collagen deposition, decreased the expression of myofibroblast marker α-SMA, and restored the epithelial marker E-cadherin. These findings suggest that suppression of ferroptotic injury and maladaptive repair contributes to the anti-fibrotic effects of 6K.

It should be acknowledged that neither the whole-body Nrf2 knockout model nor systemic 6K administration cannot fully exclude extra-renal effects. Nonetheless, oxidative stress and inflammation in AKI and CKD are inherently systemic, with ROS and uremic toxins propagating beyond the kidney to aggravate multi-organ injury. Our findings of elevated H₂O₂, IL-6, and HMGB1 in serum from AKI mice and CKD patients further support this systemic nature. Thus, while systemic models have limitations, they remain clinically relevant for capturing the pathophysiology of AKI-to-CKD progression. Future work using kidney-specific knockout models or spatial transcriptomics will help refine the renal-specific contribution of Nrf2 signaling.

Our study indicates that pharmacological activation of the Keap1-Nrf2 pathway may provide a promising approach to prevent AKI-to-CKD transition. The preferential renal exposure of 6K and its ability to restore antioxidant defense and limit ferroptotic injury support its further development for kidney protection. Importantly, preliminary safety evaluation demonstrated that 14-day administration of 6K did not cause detectable pathological alterations in liver or heart tissues (Fig. S18A) and did not significantly affect serum markers of hepatic injury, suggesting favorable initial in vivo tolerability (Fig. S18B–C). Nevertheless, comprehensive multi-organ biodistribution studies and kidney-specific genetic models will be required to further define the organ-specific mechanisms and therapeutic potential of 6K. In addition, comprehensive metabolite profiling will be required to determine whether metabolites of 6K contribute to its in vivo pharmacological effects.

4. Materials and methods

4.1. Chemicals

Compound 6K and BODIPY-FL-labeled 6K was synthesized by Professor Chunlin Zhuang, following the previous protocols[25]. A solvent mixture composed of 5% dimethyl sulfoxide (DMSO) (Cat#D8371, Solarbio, Beijing, China), 5% Tween 80 (Cat#T8360, Solarbio, Beijing, China), and 10% polyethylene glycol 400 (PEG 400) (Cat#T19506, TargetMol, China) were used for 6K treatment in mice. RSL3 was purchased from Selleck Chemicals (Cat#S8155, Shanghai, China).

4.2. Animal experiments and ethics

Healthy 7∼8-week-old male C57BL/6J mice weighing approximately 18 g to 20 g in bIRI and UUO model were obtained from the Department of Laboratory Animal Science, Peking University Health Science Center (Beijing, China). All mice were housed under defined environmental conditions at 25 ± 2°C with 12-h light/dark cycle and were given free access to food and water. All animal care and experimental procedures complied with the Animals (Scientific procedures) Act 1986 and all procedures involving animals were confirmed to follow the Regulations for the Administration of Affairs Concerning Experimental Animals published by the State Science and Technology Commission of China. Nrf2 knockout (KO) mice (B6.129 × 1-Nfe2l2tm1Ywk/J) were obtained from the Jackson laboratory (Stock No. 017009).

For the bIRI mice model, the mice underwent a procedure of 30 min of bilateral renal ischemia at 37°C followed by reperfusion at various time points, including 4 h, 1 day, 7 days, and 14 days, to construct the process of AKI-CKD. No clamps are applied to the renal pedicle in the sham group. KO and WT mice were divided into four groups, 5-6 mice in each group: WT sham group, KO sham group, WT bIRI group, and KO bIRI group. The bIRI mouse model was induced in the WT bIRI group and the KO bIRI group. No clamps are applied to the renal pedicle in the WT sham group and KO sham group.

For bIRI mice with 6K treatments, eight-week-old male C57BL/6J mice were randomly divided into 6 groups (5-6 mice in each group): sham group, sham+6K group, bIRI group, LD (low dose) group, MD (middle dose) group and HD (high dose) group. Mice in the sham group were administered equivalent volume of solvent while in the sham + 6K group were intraperitoneally injected with 10 mg/kg of 6K before sham surgery. The bIRI model was induced after intraperitoneal injection of 0.4 mg/kg of 6K in mice of the LD group, 2 mg/kg in the MD group, 10 mg/kg in the HD group and solvent in the bIRI group. For the 7-day and 14-day bIRI models, mice were continuously treated with the corresponding doses of 6K or an equivalent volume of solvent once a day.

4.3. Non-invasive transcutaneous assessment of glomerular filtration rate (GFR)

Mice were anesthetized with 2% isoflurane (v/v) (Cat#SZKJ-821A, AIDISHENG, China) in 100% oxygen delivered via a calibrated vaporizer. Under light anesthesia, a part of the mouse's back was depilated. A transdermal reader-sensor device (MediBeacon, Mannheim, Germany) was attached to the skin in the dorsal region using a double-sided patch (MediBeacon, Mannheim, Germany). Fluorescein-labeled sinistrin tracer (FITC-S) (Cat#46950, Sigma Aldrich, USA) was injected via the tail vein at a dose of 7.5 mg/100 g body weight. Mice were kept in individual cages for 1 h. The elimination kinetics curve for FITC-S gives individual transcutaneous glomerular filtration rate (tGFR) values. Data were analyzed in 3 sections using MB Studio (MediBeacon, Germany). All animals displayed repeated measurements at different time points but always from the same sensor set.

4.4. In vivo fluorescence imaging and tissue distribution analysis

To evaluate the in vivo tissue disposition and real-time distribution kinetics of compound 6K, mice were intraperitoneally (i.p.) injected with BODIPY-FL-labeled 6K at a dose of 10 mg/kg (n = 3). Prior to imaging, mice were anesthetized with isoflurane (2% induction, 1.5% maintenance in 100% O₂) and placed in the imaging chamber of an in vivo optical imaging system. Serial whole-body fluorescence imaging was performed at designated time points (0, 1, 15, 30, 45, 60, 75, 90, 120, 180, and 360 min) post-administration. BODIPY-FL fluorescence signals were acquired using specific filter settings corresponding to its spectral characteristics (λ_ex = 505 nm, λ_em = 513 nm). For quantitative kinetic analysis, regions of interest (ROIs) encompassing the bilateral renal areas were defined using the system's software, and average radiant efficiency within the ROIs was measured across all time points. Data were normalized to baseline values to track the longitudinal renal accumulation and clearance kinetics of BODIPY-FL-labeled 6K.

4.5. Clinical samples and ethics

Blood and urine specimens were from 25 patients with stages 3-5 CKD and 34 control subjects that had been enrolled in the donation plan of Shanghai Fengxian District Central Hospital (ethics approval 2024-KY-09-01). All CKD patients met the clinical and the nephrological diagnostic criteria for CKD[4]. Control subjects had no history or signs of a kidney disease. Genders, ages, diagnosis, and other information are shown in SI Appendix, Table S1. Blood and urine samples were collected, frozen, and stored at -80°C.

4.6. Quantification of intracellular glutathione by LC-MS/MS

Intracellular GSH levels were quantified by liquid chromatography-tandem mass spectrometry (LC-MS/MS) according to a previously reported method with minor modifications[34]. HK-2 cells were cultured and incubated with 1 μM RSL3 for 2 h. After removing the RSL3, HK-2 were treated with different concentrations of 6K. After treatment, cells were washed with cold phosphate-buffered saline (PBS) and harvested. Cell pellets were collected and lysed in an appropriate volume of HPLC-grade methanol (Cat#CAEQ-4-00330, Anpel, China) containing an internal standard. The lysates were centrifuged at 4°C to remove insoluble debris, and the supernatants were collected for LC-MS/MS analysis. Chromatographic separation was performed using a suitable analytical column (Sepax Bio-ODS SP, 4.6 × 150 mm, 5 μm), and GSH was detected by multiple reaction monitoring (MRM) mode. The concentration of intracellular GSH was calculated based on calibration curves generated from standard solutions and normalized to protein concentration or cell number.

4.7. Statistical analyses

All results were presented as means ± SEM. Statistical analyses were performed using GraphPad Prism 8.0 software (GraphPad Software Inc., La Jolla, CA, USA). For normally distributed continuous variables, an unpaired two-tailed Student's t-test was used for statistical significance between two independent experimental groups, and one-way analysis of variance (ANOVA) followed by Tukey's post hoc test was used for multiple group comparison. Correlations between data on mice were analyzed using linear regression analysis and the sample Pearson's correlation coefficient. p values were determined using linear regression analysis. Immunofluorescent staining, immunohistochemical staining, and other fluorescent images were analyzed using open-source software (FIJI, VMTK, Stardist, MATLAB, DBSCAN). p < 0.05 was considered statistically significant.

CRediT authorship contribution statement

Yiming Wang: Writing – review & editing, Writing – original draft, Project administration, Investigation. Yi Sun: Writing – review & editing, Writing – original draft, Supervision, Investigation. Ting Zhou: Methodology. Cai Gao: Methodology, Investigation. Jihan Liu: Methodology, Investigation. Zhicong Chen: Methodology. PanShuang Qiao: Investigation. Guangying Shao: Methodology. Min Li: Supervision. Baoxue Yang: Supervision. Chunlin Zhuang: Supervision, Resources, Methodology. Hong Zhou: Supervision, Funding acquisition.

Ethics declaration

Written informed consent to take part in the study and to publish the article has been obtained from all participants or their legal representatives. The privacy rights of participants have been observed.

This study included organ or tissue donors. This study includes human biological material and consent was obtained by donors, or their next of kin or legal representatives, for use in this study and for publication of the article. The samples used in this research were not sourced from executed prisoners or prisoners of conscience.

This study was performed in compliance with relevant laws, regulatory frameworks and guidelines where the research took place. This study was approved by the Medical Ethics Committee of Shanghai Fengxian District Central Hospital. (Approval No. 2024-KY-09-01)

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article.

Data availability

No data was used for the research described in the article.