Keap1-independent Nrf2 regulation: A novel therapeutic target for treating kidney disease

Abstract

The transcription factor NF-E2-related factor 2 (Nrf2) is a master regulator of antioxidant responses in mammals, where it plays a critical role in detoxification, maintaining cellular homeostasis, combating inflammation and fibrosis, and slowing disease progression. Kelch-like ECH-associated protein 1 (Keap1), an adaptor subunit of Cullin 3-based E3 ubiquitin ligase, serves as a critical sensor of oxidative and electrophilic stress, regulating Nrf2 activity by sequestering it in the cytoplasm, leading to its proteasomal degradation and transcriptional repression. However, the clinical potential of targeting the Keap1-dependent Nrf2 regulatory pathway has been limited. This is evidenced by early postnatal lethality in Keap1 knockout mice, as well as significant adverse events after pharmacological blockade of Keap1 in human patients with Alport syndrome as well as in those with type 2 diabetes mellitus and chronic kidney disease. The exact underlying mechanisms remain elusive, but may involve non-specific and systemic activation of the Nrf2 antioxidant response in both injured and normal tissues. Beyond Keap1-dependent regulation, Nrf2 activity is modulated by Keap1-independent mechanisms, including transcriptional, epigenetic, and post-translational modifications. In particular, GSK3β has emerged as a critical convergence point for these diverse signaling pathways. Unlike Keap1-dependent regulation, GSK3β-mediated Keap1-independent Nrf2 regulation does not affect basal Nrf2 activity but modulates its response at a delayed/late phase of cellular stress. This allows fine-tuning of the inducibility, magnitude, and duration of the Nrf2 response specifically in stressed or injured tissues. As one of the most metabolically active organs, the kidney is a major source of production of reactive oxygen and nitrogen species and also a vulnerable organ to oxidative damage. Targeting the GSK3β-mediated Nrf2 regulatory pathway represents a promising new approach for the treatment of kidney disease.

1. Introduction

Under both physiological and pathological conditions, oxygen consumption by mammalian cells inevitably generates reactive oxygen and nitrogen species (RONS) as by-products of aerobic metabolism [1]. These highly reactive molecules encompass two primary categories: reactive oxygen species (ROS) and reactive nitrogen species. Under normal physiological or exercise conditions, the rate and magnitude of RONS formation in cells is balanced by the rate of oxidant elimination. At the physiological level, RONS are required for cellular homeostasis as they act as essential oxidation-reduction (redox) signaling messengers for normal cellular function. Maintaining a proper balance of RONS is vital for optimal cellular health, as both excessive and insufficient levels can be detrimental. Excessive RONS production can overwhelm the cell's antioxidant defenses, leading to oxidative stress [2]. As one of the most metabolically active organs, the kidney is a major source of RONS production and meanwhile an organ that is particularly vulnerable to oxidative damage [3]. Therefore, it is of paramount significance to explore innovative strategies to protect the kidney from oxidative injury.

NF-E2-related factor (Nrf2) is a cap-n-collar basic-region leucine zipper (bZIP) nuclear transcription factor that mediates the primary cellular defense against oxidative stress. Nrf2 activates a wide range of molecules to maintain redox homeostasis and allow adaptation to stress [4]. Under basal conditions, Nrf2 is sequestered in the cytoplasm and associated with an inhibitory protein called Kelch-like ECH-associated protein 1 (Keap1), which is an adaptor subunit of cullin 3-based E3 ubiquitin ligase, and a pivotal sensor of oxidative and electrophilic stress that is capable of facilitating Nrf2 ubiquitination, proteasomal proteolysis and transcriptional repression [5]. Upon activation by oxidative stimuli, Nrf2 dissociates from Keap1 and subsequently translocates to the nucleus, where Nrf2 binds to antioxidant response elements (AREs) and causes transcriptional activation of genes encoding cytoprotectants, such as antioxidant enzymes [6]. In addition, Keap1-independent regulatory pathways also play a key role in controlling the Nrf2 activity. Among these, glycogen synthase kinase 3β (GSK3β) has emerged as an integration point in the Keap1-independent Nrf2 antioxidant regulatory pathways [7]. GSK3β shuts down the self-protective antioxidant stress response after injury by promoting nuclear export and degradation of Nrf2. Here, we review recent literature on the role of Keap1-dependent versus Keap1-independent mechanisms in the regulation of the Nrf2 antioxidant response and compare their differences in controlling cellular and organ protective actions with a focus on the kidney disease scenario.

2. Biological functions of Nrf2

Nrf2 is a ubiquitously expressed and master regulator of an array of inducible cell defense systems involved in detoxification, antioxidation, anti-inflammation, and anti-fibrosis [4] (Fig. 1A). Nrf2 remains latent in cells until exposed to chemical or oxidative stress, which is a major cause and driver of many diseases [8]. When RONS production exceeds antioxidant capacity, the Nrf2 signaling cascade is triggered and modulates a variety of antioxidant proteins, drug-metabolizing enzymes, xenobiotic transporters, and other enzymes that catalyze multiple metabolic and biosynthetic pathways. Tissues that are rich in redox reactions, such as the kidney, are particularly susceptible to oxidative stress and RONS overload. Nrf2 expression is significantly higher in renal tubules compared to glomeruli in the kidney [9] (Fig. 1B). This differential expression pattern suggests an important biological role of Nrf2 in kidney health and disease.

Fig. 1.

Major biological functions of Nrf2 in the scenario of kidney disease. (A) Nrf2 has important functions including antioxidation, anti-inflammation, detoxification, and anti-fibrosis. (B) Bubble dot plots of the expression of Nrf2 (Nfe2l2), Keap1 and Gsk3β in various kidney cells based on the single nucleus RNA sequencing (snRNAseq) transcriptome of mouse kidney derived from the Wu Healthy Mouse Dataset that is publicly-available from Kidney Interactive Transcriptomics http://humphreyslab.com/SingleCell/. The size of each dot corresponds to the percentage of cells expressing indicated gene, while the color intensity indicates the average expression level. Abbreviations: Pod, podocyte; MC, mesangial cell; EC, endothelial cell; PT (S1–S2), proximal tubular (Segment 1–2); PT (S3), proximal tubular (Segment 3); LH (DL), loop of henle (descending limb); LH (AL), loop of henle (ascending limb); DCT, distal convoluted tubule; CNT, connecting tubule; CD-PC, collecting duct principal cell; IC-A, alpha intercalated cell; IC-B, beta intercalated cell; M∅, macrophage.

2.1. Key roles of Nrf2 in antioxidation and regulating redox status

Oxidative damage is a unifying pathophysiological process in almost all types of kidney disease, including acute kidney injury (AKI) and chronic kidney disease (CKD). A growing body of evidence suggests that Nrf2 plays a key role in modulating the antioxidant response and redox status upon kidney injury. For example, oxidative stress is a key mediator underlying ischemia-reperfusion injury (IRI)-induced AKI. In the early stages of unilateral IRI, Nrf2 knockout mice exhibited increased oxidative stress in outer medulla of the kidney compared to wild-type mice. Subsequently on day 14 of unilateral IRI, more severe tubular injury was observed in the IRI kidneys of Nrf2 knockout mice compared with wild-type mice [10]. In parallel, the expression of genes encoding antioxidant enzymes/proteins was significantly upregulated in the kidney of wild-type mice but not in Nrf2 knockout mice after renal IRI. Furthermore, IRI-induced renal dysfunction, renal histopathological injury, vascular permeability, and overall survival were significantly worsened in Nrf2 knockout mice. These detrimental effects of Nrf2 knockout could be ameliorated by treatment with the antioxidant N-acetyl-cysteine or glutathione [11]. To determine the cell-specific role of Nrf2 in regulating renal oxidative injury upon AKI, engineered mesenchymal stem cells overexpressing Nrf2 by recombinant adenovirus were transplanted into rats and conferred protection against cisplatin-induced kidney injury [12]. However, Nezu et al. [10] found that renal tubular-specific Nrf2 activation rather than myeloid-specific Nrf2 activation was critical for mitigating renal IRI.

Severe or recurrent AKI is more likely to transform to CKD, which is associated with persistent renal dysfunction, and progressive fibrosis and inflammation [13,14]. Knockout of Nrf2 exacerbated renal oxidative stress in mice with cisplatin-induced CKD. Specifically, cisplatin-induced increases in myeloperoxidase, an oxidative enzyme, and malondialdehyde, a by-product of unsaturated fatty acid peroxidation, and decreases in the antioxidant enzyme superoxide dismutase and the antioxidant glutathione were more pronounced in Nrf2 knockout mice [15]. Accumulation of the uremic toxin hippuric acid is positively correlated with CKD progression in patients. In cultured HK-2 cells, Nrf2 silencing downregulated the Nrf2-dependent antioxidant pathway, resulting in a significant increase in hippuric acid-evoked ROS and hydrogen peroxide generation [16]. Nrf2 also protects against diabetic kidney injury. In streptozotocin (STZ)-induced diabetic nephropathy, Nrf2 knockout mice exhibited higher ROS production and suffered more oxidative DNA damage and renal injury compared to wild-type mice [17]. Moreover, Nrf2 knockout exacerbated diabetes-induced oxidative or nitrosative damage in the kidney, marked by elevated levels of malondialdehyde, inducible nitric oxide synthase (iNOS), and 3-nitrotyrosine [18]. In type 1 diabetic Akita mice, Nrf2 knockout worsened diabetic kidney injury with more distended glomerular capillary loops and dilated distal renal tubules. This was accompanied by increased 8-hydroxydeoxyguanosine staining, and reduced renal expression of genes related to glutathione synthesis, as well as lower glutathione levels [19].

Proteinuria is an invariable clinical manifestation in patients with CKD and is by itself one of the few independent and modifiable risk factors for progression to end-stage renal failure and for cardiovascular morbidity and mortality [20]. As a cornerstone of the glomerular filtration barrier, podocytes play a pivotal role in controlling glomerular permselectivity, and preventing protein in the bloodstream from leaking into the urine [21]. Nrf2 is expressed in glomerular podocytes (Fig. 1B) and Nrf2 antioxidant defense has been found to be crucial for podocyte protection. In cultured human podocytes, Nrf2 silencing reduced the constitutive expression of podocyte homeostatic marker proteins, including nephrin and synaptopodin, and exacerbated high ambient glucose-elicited suppression of their expressions [22]. In addition, Nrf2 also mediates the podocyte protective effect of a number of pharmacological agents. For example, Nrf2 silencing abolished the protective effect of mitoquinone, a mitochondria-targeted antioxidant, on mitochondrial fission in cultured human podocytes injured by angiotensin II [23]. Nrf2 knockdown also abrogated the anti-oxidative and anti-apoptotic effect of β-cryptoxanthin, a natural carotenoid antioxidant, in podocytes exposed to high glucose [24]. In vivo in a mouse model of selective and inducible podocyte injury in NEP25 mice elicited by injection of the hCD25-targeted immunotoxin, Nrf2 knockout resulted in more severe glomerular injury [25]. In consistency in mouse model of STZ-induced diabetic nephropathy, Nrf2 knockout mice sustained more severe proteinuria and glomerulosclerosis than wild-type mice.

Nrf2 also regulates oxidative injury in other kidney diseases. In the Pkd1 knockout mouse model of autosomal dominant polycystic kidney disease (ADPKD), the additional loss of Nrf2 increased renal oxidative stress and accelerated cyst growth [26]. Nrf2 also plays a role in autoimmune kidney disease. For instance, Nrf2 deficiency led to autoimmune nephritis and shortened lifespan in aged female mice [27]. Furthermore, Nrf2 knockout sensitized oxidative DNA damage in a murine model of pristane-induced lupus nephritis model [28]. Nephrolithiasis, which involves oxidative stress, is also influenced by Nrf2, as Nrf2 knockdown exacerbated the induction of ROS in HK-2 cells exposed to calcium oxalate monohydrate, a major mineral component of kidney stones [29]. In addition, Nrf2 has been implicated in antioxidation in age-related kidney injury. Renal recovery after IRI was delayed in old mice compared to young mice, and this delay was further exacerbated by Nrf2 knockout. In vitro, Nrf2 silencing in senescent human proximal tubule epithelial cells increased oxidative stress and mitochondrial dysfunction [30].

2.2. Nrf2-mediated anti-inflammatory and anti-fibrotic effects

Apart from its role in antioxidant responses, Nrf2 modulates inflammation and fibrosis in the course of kidney disease. Following renal IRI, the proinflammatory cytokine granulocyte colony-stimulating factor was induced to a greater extent in the kidneys of mice lacking Nrf2 [11]. Likewise, renal fibrosis was greater in Nrf2 knockout mice when exposed to cisplatin [15]. Diabetic Akita mice lacking Nrf2 exhibited elevated markers of renal inflammation and increased macrophage infiltration [19]. In addition, in mice with high fat diet-induced type 2 diabetes, Nrf2 knockout completely abolished the renoprotective effects of sulforaphane against albuminuria, renal fibrosis, and renal inflammation [31]. In vitro, Nrf2 silencing exacerbated the fibrotic phenotypes in HK2 human proximal tubular cells exposed to the uremic toxin hippuric acid [16]. In mice with STZ-induced diabetic nephropathy, Nrf2 knockout increased the mRNA expressions of the profibrotic cytokine TGFβ1, and the extracellular matrix components fibronectin and collagen IV in the kidney [17]. Consistently, in another study, Nrf2 knockout in STZ-induced diabetic mice also enhanced the protein expressions of the profibrogenic factors like TGFβ1 and PAI-1 and the inflammatory mediators like ICAM-1 and VCAM-1 in the kidney [18].

In mouse models of ADPKD, Nrf2 knockout promoted the expression of proinflammatory cytokines such as MCP-1 and IL-6 [26]. In addition, in mice with pristane-induced lupus nephritis, Nrf2 knockout increased glomerular expressions of TGFβ1, fibronectin, and iNOS. The anti-inflammatory effect of Nrf2 appears to be operative in both kidney parenchymal cells and immune cells. Supporting this, in primary mesangial cells prepared from MRL/lpr lupus mice, Nrf2 silencing resulted in an increased expression of iNOS [28]. Furthermore, in cultured mouse macrophage RAW264.7, Nrf2 knockdown enhanced the lipopolysaccharide-induced inflammatory response and augmented the expression of inflammatory mediators, including iNOS, nitric oxide, and chemokine C–C motif ligand 2 [32]. Mechanistically, the anti-inflammatory/anti-fibrotic mechanism of Nrf2 may be mediated through the induction of antioxidant genes, which, in turn, effectively inhibit the transcription of proinflammatory and profibrogenic cytokines.

2.3. Nrf2-mediated detoxification in renal toxicity

As a master regulator of detoxification, Nrf2 plays a crucial role in protecting against cellular toxicity by detoxifying electrophiles and chemicals. In cultured rat kidney tubular epithelial cells (NRK-52E), cadmium exposure induced Nrf2 DNA binding activity in a time-dependent and dose-dependent manner, indicating a spontaneous, self-protective Nrf2 defense. Transient overexpression of Nrf2 through plasmid transfection conferred resistance to cadmium-induced apoptosis. Conversely, Nrf2 silencing sensitized NRK-52E cells to cadmium-elicited apoptotic death by reducing the expression of antioxidant enzymes [33]. In vivo, Nrf2 knockout consistently resulted in reduced renal expression of key antioxidant and phase II detoxification enzymes in mice, leading to exacerbated renal injury caused by subacute cadmium exposure [34]. In addition, Nrf2 silencing sensitized LLC-PK1 porcine renal proximal tubular cells to ochratoxin A-induced toxicity [35]. In contrast, adenovirus-mediated overexpression of Nrf2 inhibited the ability of ochratoxin A to generate ROS and induce TGFβ expression [36].

Cyclosporin A (CsA) is one of the most widely used immunosuppressants in organ transplantation and in the treatment of autoimmune disorders. Nevertheless, nephrotoxicity significantly limits its clinical use [37]. In vitro, Nrf2 knockdown worsened the CsA-induced impairment of viability in NRK-52E cells. In addition, Nrf2 inhibition in NRK-52E cells exacerbated CsA-induced epithelial-mesenchymal transition. Consistently, in vivo, Nrf2 knockout mice were more susceptible to CsA-elicited renal interstitial fibrosis compared to wild-type mice [38].

Nrf2-mediated detoxification has also been implicated in protecting against nephrotoxicity caused by other toxins. In mouse models of cisplatin nephrotoxicity, Nrf2 deficiency exacerbated renal damage [39]. Additionally, Nrf2 demonstrates a renoprotective effect in mice exposed to the renal carcinogen ferric nitrilotriacetate. Specifically, urinary excretion of N-acetyl-β-D-glucosaminidase, a marker of renal tubular injury, was significantly higher in Nrf2-null mice compared to wild-type mice following ferric nitrilotriacetate injury, concomitant with a worsened kidney dysfunction. This was accompanied by more severe histopathological changes in the kidney, including proximal tubular degeneration and necrosis [40]. Collectively, these findings highlight Nrf2's protective role against renal toxicity induced by various toxic substances.

3. The canonical Keap1-dependent regulation of Nrf2 response

Nrf2 activity is regulated by various cellular signaling pathways, which can be broadly classified into Keap1-dependent and Keap1-independent mechanisms. Keap1, a component of the cullin-3 (CUL3)-based E3 ubiquitin ligase complex, plays a key role in controlling the stability and accumulation of Nrf2. Recent advances in understanding the Keap1-Nrf2 system have highlighted its essential role in regulating cellular defense against oxidative stress [41].

3.1. The “hinge and latch” structure of the Keap1-Nrf2 system

Keap1 is composed of several functional domains: an N-terminal region (NTR), a Bric-à-Brac (BTB) domain, an intervening region (IVR), six Kelch repeats, and a C-terminal region (CTR) [42] (Fig. 2A). Within the DC domain of Keap1, the Kelch repeats directly bind to the N-terminal DLG motif of Neh2, while the Kelch domain of a second Keap1 molecule binds Nrf2 through the N-terminal EGTE motif of Neh2 [43] (Fig. 2B).

Fig. 2.

The canonical Keap1-dependent Nrf2 regulatory pathway. (A) Prediction of human Nrf2 and Keap1 protein structures using the AlphaFold server. Dark blue indicates greater than 90 % confidence, light blue indicates between 70 % and 90 % confidence, yellow indicates between 50 % and 70 % confidence, and orange indicates less than 50 % confidence. Nrf2 contains Neh1-7 domains. Keap1 contains a number of functional domains including N-terminal region (NTR), Bric-à-Brac (BTB), intervening region (IVR), Kelch repeats, and C-terminal region (CTR). (B) Prediction of the crystal structure of the Kelch-Neh2 complex in human tissue using the RCSB PDB website. The green structure is Kelch repeats of Keap1. The red structure is Neh2 of Nrf2. (C) Two Keap1 molecules and one Nrf2 molecule form a trimer. (D) The “hinge and latch” structure of Keap1-Nrf2 system. (E) Under basal conditions, Nrf2 undergoes ubiquitination and degradation. (F) Upon oxidative stress, Nrf2 translocates to the nucleus and binds to antioxidant responsive element (ARE), inducing the transcription of cell defense gene expression.

The Keap1-Nrf2 system operates as a two-component system: Keap1 functions as a sensor for electrophiles, while Nrf2 acts as an effector to coordinate the activation of cytoprotective genes [44]. It is proposed that two Keap1 molecules and one Nrf2 molecule form a trimer [45] (Fig. 2C). This trimer structure aligns with the hinge and latch hypothesis, where the EGTE motif, serves as the high-affinity “hinge”, and the DLG motif functions as the low-affinity “latch” (Fig. 2D). The interaction at the EGTE binding site allows for more dynamic movement, while the binding at the DLG site restricts Nrf2's detachment from Keap1. This dual binding accelerates the ubiquitination of lysine residues in the Neh2 domain, leading to the proteasomal degradation of Nrf2. Proteasomal degradation thus regulates the cell's response to inflammatory, hypoxic, oxidative, and xenobiotic stimuli, with Nrf2 being activated in response to these stimuli and deactivated upon their removal [7].

3.2. Ubiquitination and proteasomal degradation of Nrf2

Under basal conditions, Nrf2 is located in the cytoplasm, where it forms a trimeric complex with the Keap1 homodimer. In this inactive form, Nrf2 undergoes polyubiquitination by the CUL3 E3 ligase and is subsequently degraded by the 26S proteasome [5] (Fig. 2E). This process maintains Nrf2 at a very low basal level in the cytoplasm, preventing its constitutive activation of the antioxidant response [46]. Keap1 functions as an intrinsic repressor of Nrf2. Electrophiles or ROS modify reactive cysteine residues in Keap1, leading to the release of the DLG motif binding site (latch) of Nrf2 from Keap1 [47]. This disruption impairs the polyubiquitination process via the E3 ligase CUL3 and subsequently prevents its degradation by the 26S proteasome. Consequently, newly synthesized Nrf2 is able to escape the Keap1 complex. The accumulation of free Nrf2 in the cytoplasm then facilitates its translocation to the nucleus.

In the nucleus, Nrf2 heterodimerizes with one of the small musculoaponeurotic fibrosarcoma (sMaf) proteins. This Nrf2-sMaf heterodimer binds to CNC-sMaf binding elements (CsMBE), such as the consensus ARE [48] or the electrophile-responsive element (EpRE) [49], and strongly induces the transcription of various cell defense genes, including antioxidant proteins, xenobiotic transporters, ROS scavengers, and drug metabolism enzymes [50](Fig. 2F). Furthermore, the release of Nrf2 from Keap1 can also be triggered by kinase signaling, such as mitogen-activated protein kinases (MAPKs) or protein kinase C (PKC), even in the absence of modifying agents [51]. As a key regulatory pathway in oxidative stress, the Keap1-Nrf2 system is considered an attractive target for therapeutic intervention.

3.3. Autophagic degradation of Keap1

When autophagic flux is impaired and p62 accumulates, p62 sequesters Keap1 into autophagosomes, thereby preventing it from binding to Nrf2 and resulting in increased Nrf2 signaling [52]. Additionally, p62 contains an STGE motif that binds to the Kelch domain of Keap1 when the Keap1-Nrf2 complex is in an open conformation [53]. This binding disrupts the Keap1-Nrf2 interface. Phosphorylation of p62 enhances its affinity for Keap1, resulting in prolonged accumulation of Nrf2 [54] and increased Nrf2 activity [55] (Fig. 3A).

Fig. 3.

The Keap1-dependent and Keap1-independent regulatory pathways of Nrf2. (A) When autophagic flux is impaired and p62 accumulates, p62 sequesters Keap1 into autophagosomes, preventing it from binding to Nrf2 and thereby leading to enhanced Nrf2 signaling. (B) The ARE/XRE sequences are located near the promoters of Nrf2-regulated genes, including Nrf2 itself, enabling it to regulate its own transcription. The Nrf2-sMaf heterodimer binds to ARE, promoting the expression of various cytoprotective genes that combat oxidative stress. (C) NF-κB translocates to the nucleus, where it associates with CBP and competes with Nrf2, thereby inhibiting Nrf2-mediated transactivation. (D) P21 and caveolin-1 (Cav-1) serve as binding partners for Nrf2, stabilizing it and modulating its activity.

3.4. Epigenetic mechanisms

MicroRNAs (miRs) are short, single-stranded non-coding RNAs that bind to target mRNAs in the 3′UTR region after maturation, leading to mRNA degradation or inhibition of protein translation [56]. MiRs can suppress Keap1 signaling at multiple levels, thereby fine-tuning the regulation of Nrf2 activity (Fig. 4). For instance, miR-27a, miR-141, miR-144, and miR-432 inhibit the translation of Keap1 mRNA, facilitating increased Nrf2 activation in a Keap1-dependent manner [57].

Fig. 4.

MicroRNAs play a crucial role in fine-tuning the regulation of the Nrf2 signaling pathway. MiR-27a, miR-141, miR-144, and miR-432 repress the translation of Keap1 mRNA, thereby promoting Nrf2 activation. In contrast, miR-144, miR-155, miR-28, and miR-148b inhibit Nrf2 activity by repressing Nrf2 mRNA. Additionally, miR-181a, miR-193b, miR-424, and miR-125b are positively correlated with increased Nrf2 production and indirectly activate Nrf2. Notably, miR-200a inhibits Nrf2 ubiquitination, contributing to Nrf2 stabilization.

3.5. Post-translational mechanisms

Post-translational modifications play a crucial role in the Keap1-dependent Nrf2 regulation (Fig. 5). PKC, a family of serine/threonine kinases, can phosphorylate the Neh2 domain of Nrf2 at Ser40, disrupting its association with Keap1 and promoting Nrf2 translocation to the nucleus and ARE-mediated transcription [58]. Additionally, PERK phosphorylates Nrf2, as a direct PERK substrate, to activate the Nrf2-mediated antioxidant pathway [59], and the exact site of phosphorylation is presumed to be within the Neh2 domain, affecting the Nrf2-Keap1 interaction [59]. The deubiquitinase DUB3 stabilizes Nrf2 by reducing K48-linked ubiquitination and forming complexes with both Nrf2 and Keap1 [60].

Fig. 5.

Post-translational regulation of Nrf2. Nrf2 undergoes various post-translational modifications, including phosphorylation, ubiquitination, SUMOylation, methylation, glycation, and acetylation, mediated by a range of enzymes.

4. Pharmacological targeting of the canonical Keap1-dependent regulation of Nrf2

Given its crucial role in regulating Nrf2 activation, Keap1 is a highly enticing target for pharmacological intervention aimed at modulating the antioxidant response of Nrf2. Several chemical compounds have been successfully synthesized to acts as a "Keap1 blocker" by covalently binding to Keap1 protein, preventing it from binding to and degrading Nrf2, thereby leading to the activation of the Nrf2 pathway and an increase in antioxidant defenses within cells. The FDA has approved dimethyl fumarate, which oxidizes sulfhydryl groups of Keap1, finally activating Nrf2, for the treatment of relapsing forms of multiple sclerosis, and it is also being investigated for its potential application in the treatment of psoriasis. Research on the effect of dimethyl fumarate in kidney disease is emerging, with limited animal studies indicating promising results in AKI models [[61], [62], [63], [64]]. In addition, synthetic triterpenoids, like CDDO-imidazolide (CDDO-Im) and CDDO-methyl (CDDO-Me or bardoxolone methyl) can also effectively bind to Keap1 and thereby activates Nrf2. In particular, bardoxolone methyl, had undergone Phase 3 clinical trials in patients with type 2 diabetes and stage 4 CKD [65] as well as in Alport syndrome [66]. Additionally, certain dietary supplements, such as sulforaphane [67] extracted from broccoli sprouts, can activate Nrf2 by binding directly to Keap1 [68]. Curcumin from turmeric also has a Keap1 inhibitory activity [69], and carnosic acid can bind to the cysteine residues of Keap1, thereby activating Nrf2 [70].

5. The alternative Keap1-independent regulation of Nrf2 response

Keap1-mediated inhibition of Nrf2 occurs in the cytoplasm [71]. However, as a typical transcription factor, Nrf2 primarily exerts its biological functions in the nucleus [8]. Therefore, additional cellular processes that operate independently of Keap1 are crucial for regulating Nrf2 activity within the nuclear compartment.

5.1. Transcriptional mechanisms and autoregulation

Two core DNA sequences, the ARE and the xenobiotic-responsive element (XRE), are located in the promoter regions of cytoprotective genes targeted by Nrf2. Nrf2 binds directly to the ARE to activate genes involved in detoxification and antioxidant defense. The XRE is activated by the aryl hydrocarbon receptor, which induces the expression of phase I enzymes, such as cytochrome P450s, and phase II enzymes [72]. Both ARE and XRE are situated near the promoters of Nrf2 regulatory genes and Nrf2 itself [73], facilitating autoregulation of their transcription. Thus, Nrf2 can directly enhance its own expression by binding to its promoter region and forming a positive feedback loop, which significantly boosts cellular defense capabilities [74] (Fig. 3B). Under oxidative stress, NF-κB is released and translocates to the nucleus [75], where it competes with Nrf2 for the co-activator CBP, thereby inhibiting Nrf2 transactivation [76] (Fig. 3C)

5.2. Binding partners

P21 is a binding partner that interacts with Nrf2 at the DLG and ETGE motifs through its C-terminal KRR motif to stabilize Nrf2, resulting in an enhanced Nrf2 antioxidant response [77]. Another Nrf2 binding partner, caveolin-1, competes with Keap1 for binding to Nrf2, thereby contributing to the stabilization of Nrf2 [78] (Fig. 3D).

5.3. Epigenetic mechanisms

Epigenetic mechanisms, such as miRs, are able to modulate Nrf2 activity independently of Keap1 (Fig. 4). For example, increased miR-144 decreases Nrf2 expression and glutathione regeneration, thereby compromising oxidative stress tolerance in thalassemic erythroid cells [79]. MiR-155 was induced in bronchial epithelial cells exposed to arsenic and suppressed Nrf2 translation [80]. MiR-28 significantly reduced Nrf2 mRNA and protein levels in breast epithelial cells, although Keap1 protein expression and the Keap1/Nrf2 interaction remain unaffected [56]. In addition to targeting Keap1 mRNA, miR-200a can directly inhibit Nrf2 ubiquitination and protect it from degradation [81]. Furthermore, miR-181a, miR-193b, miR-424, and miR-125b are positively correlated with Nrf2 production and indirectly activate Nrf2 [57].

5.4. Post-translational mechanisms

Keap1-independent regulation of Nrf2 involves several post-translational modifications, including phosphorylation, ubiquitination, acetylation, SUMOylation, and methylation (Fig. 5). Among these, phosphorylation, particularly GSK3β-mediated phosphorylation of Nrf2, plays a significant role in regulating Nrf2 activity independent of Keap1(Fig. 6).

Fig. 6.

The GSK3β-mediated Keap1-independent regulation of Nrf2. (A) Prediction of GSK3β protein structures in human tissues using the AlphaFold server. Dark blue indicates greater than 90 % confidence, light blue indicates 70–90 % confidence, yellow indicates 50–70 % confidence, and orange indicates less than 50 % confidence. (B) In the nucleus, Nrf2 is phosphorylated by GSK3β, which enables its capture by β-TrCP, marking Nrf2 for nuclear export, ubiquitination, and subsequent degradation. (C) GSK3β can also indirectly phosphorylate Nrf2 through Fyn. (D) PI3K/Akt and MAPK pathways can phosphorylate GSK3β. Phosphorylated GSK3β loses its ability to phosphorylate Nrf2, resulting in the accumulation of Nrf2 in the nucleus.

5.4.1. GSK3β-mediated phosphorylation of Nrf2

Glycogen synthase kinase 3 (GSK3) is a serine/threonine protein kinase that plays a crucial role in various pathobiological processes, including glycogen biosynthesis, redox homeostasis, tissue injury, repair, and regeneration [82]. GSK3 was originally identified as a protein kinase capable of phosphorylating glycogen synthase and regulating glucose metabolism. GSK3 has two isoforms, GSK3α and GSK3β, which share extensive homology in their catalytic domain [83]. In the kidney, the β isoform is predominantly expressed. Unlike many other kinases, GSK3β is constitutively active. GSK3β is inactivated by phosphorylation at the Ser⁹ site and activated by phosphorylation at the Tyr²¹⁶ site. Nrf2 is a cognate substrate for GSK3β-mediated phosphorylation, harboring multiple GSK3β consensus motifs, including Ser³³⁵ and Ser³³⁸ in the Neh6 domain of Nrf2 [84]. As a convergence point for multiple Keap1-independent protective signaling pathways, GSK3β plays a key role in regulating Nrf2 phosphorylation and its activity [7] in the kidney [85].

Upon cellular injury, the interaction between Keap1 and Nrf2 is disrupted, leading to Nrf2 accumulation in the cytoplasm and subsequent translocation to the nucleus. Consequently, Nrf2 binds to an sMaf protein and initiates the transcription of antioxidant genes [50]. GSK3β, which is constitutively active, can directly phosphorylate Nrf2 in the Neh6 domain. This phosphorylation facilitates Nrf2's nuclear exit, ubiquitination, and subsequent proteasomal degradation [86] (Fig. 6B). Alternatively, GSK3β can phosphorylate Nrf2 indirectly through Fyn. After phosphorylation by GSK3β, Fyn could translocate into the nucleus and catalyze the phosphorylation of nuclear Nrf2, facilitating Nrf2 nuclear export (Fig. 6C) and degradation to switch off the Nrf2 antioxidant response [87,88]. This GSK3β-dictated nuclear exclusion and degradation of Nrf2 is pivotal in terminating the self-protective Nrf2 antioxidant stress response after injury [87].

Several cellular protective signaling pathways are simultaneously triggered after injury, including MAPK and PI3K/AKT signaling pathways (Fig. 6C), which can induce GSK3β Ser⁹ phosphorylation and thereby inhibit its activity [89]. This leads to reduced GSK3β phosphorylation of Nrf2 directly or indirectly through Fyn, which diminishes Nrf2 nuclear export and increases Nrf2 aggregation in the nucleus [83]. This results in a prolonged and enhanced Nrf2 response. In contrast, persistent injury can cause overactivation of GSK3β [90], which could promote Nrf2 nuclear export and subsequently dampen the Nrf2 defense response.

5.4.2. Post-translational mechanisms beyond GSK3β

Beyond GSK3β, Nrf2 is regulated by a variety of other post-translational modifications (Fig. 5). Casein kinase 2 (CK2) positively regulates Nrf2 activity by phosphorylating it at multiple sites within the Neh4 and Neh5 domains [91]. Inhibition of CK2 results in decreased accumulation, phosphorylation, and nuclear translocation of Nrf2 [92]. AMPK phosphorylates Nrf2 at Ser⁵⁵⁰ located within the Neh1 domain and prevents its nuclear export [93]. CDK5 phosphorylates Nrf2 at Thr³⁹⁵, Ser⁴³³, and Thr⁴³⁹, resulting in increased nuclear accumulation of Nrf2 [94]. MAP kinases target five phosphorylation sites on Nrf2: Ser²¹⁵, Ser⁴⁰⁸, Ser⁵⁵⁸, Thr⁵⁵⁹, and Ser⁵⁷⁷, although these MAPK-mediated phosphorylations appear to have limited contribution in modulating Nrf2 activity [95]. In addition, PKC isoforms can phosphorylate Nrf2 in the cytoplasm, thereby triggering Nrf2 signaling to the cell nucleus. Alternatively, PKC can translocate to the nucleus and act directly on Nrf2 [96].

In addition to TrCP-CUL1-dependent ubiquitination, Nrf2 can also be ubiquitinated by the E3 ligase Hrd1, leading to its degradation by the proteasome [97]. Acetylation and deacetylation are critical in regulating the nucleocytoplasmic shuttling of Nrf2. CREB binding protein (CBP) and p300, which share high homology, acetylate Nrf2 at Lys⁵⁸⁸ and Lys⁵⁹¹ in the Neh2 region, thereby enhancing the transcription of key antioxidant genes. Conversely, deacetylation causes Nrf2 to dissociate from the ARE, leading to transcriptional termination and nuclear export [98]. The Neh2 and Neh1 regions are particularly important for Nrf2 acetylation. SUMO-2-mediated sumoylation at Lys¹¹⁰ and Lys⁵³³ also regulates Nrf2, enhancing its stabilization and nuclear localization. Lys⁵³³ is located within the bZip region of the Neh1 domain, while Lys¹¹⁰ is adjacent to the activation domain in Neh4 [99]. In addition, methylation of Nrf2 by the protein arginine methyltransferase 1 (PRMT1) at Arg⁴³⁷ increased its transactivation and DNA-binding activity and protected against cell death [100]. These diverse post-translational modifications collectively affect the stability, localization, and activity of Nrf2.

6. Pharmacological targeting of GSK3β-mediated Keap1-independent regulation of Nrf2

Among the various Keap1-independent regulatory pathways, GSK3β-mediated regulation of Nrf2 has attracted considerable interest. This is largely due to the role of GSK3β as a practical and actionable drug target, given its involvement in the pathogenesis of numerous diseases, including kidney disease. GSK3β can be inhibited by a number of small molecule inhibitors, making it a promising therapeutic target. As a first-generation GSK3β inhibitor, lithium has been an FDA-approved mood stabilizer used as first-line therapy for affective disorders for over 50 years [101] and has been shown to enhance the Nrf2 response [102]. Long-term use of lithium, which is commonly prescribed to patients with psychiatric conditions, is generally well tolerated by most peripheral organs [103], making it a pragmatic therapeutic option for targeting GSK3β-mediated Keap1-independent regulation of Nrf2. In addition, pharmacological inhibition of GSK3β by VP3.15, a small molecule dual phosphodiesterase 7-GSK3 inhibitor, has been reported to enhance Nrf2 activity and provide protection against oxidative stress in the retina of diabetic mice [104]. Selective GSK3β inhibitors such as CHIR 98014 and CHIR 99021 have be shown to improve insulin action and glucose tolerance in insulin-resistant skeletal muscle in rodent models of type 2 diabetes [105]. Additionally, inhibition of GSK3β using small molecule inhibitors, such as SB216763, SB415286, as well as LiCl has been shown to reduce apoptosis following hyperosmotic stress [106]. Taken together, targeting GSK3β-mediated Keap1-independent regulation of Nrf2 represents a novel and promising approach for treating various kidney diseases.

7. Strategies to target the Keap1-dependent Nrf2 pathway in kidney disease

Numerous studies have explored strategies to regulate Nrf2 in renal diseases via genetic or pharmacological targeting of Keap1, with bardoxolone methyl (CDDO-Me) being tested in Phase 3 clinical trials.

7.1. Targeting the Keap1-dependent Nrf2 pathway in AKI

AKI is often associated with IRI, which accounts for approximately half of all hospitalized cases with AKI. In mouse models of IRI-AKI, Keap1 hypomorphic knockdown did not provide immediate renal tubular protection during the initial injury phase. The protective effects of Nrf2 only became apparent as the injury progressed [107]. Specifically, there was no significant difference in biochemical, functional or morphological protection of the renal tubules at the very early time point of 24 h post injury between wild-type and Keap1 hypomorphic mouse models. However, kidney function, measured by serum creatinine and blood urea nitrogen levels, improved in Keap1 hypomorphic mice at 3- and 10-days post-injury. Additionally, markers of renal tubular injury, such as kidney injury molecule-1 and neutrophil gelatinase-associated lipocalin, were reduced in these mice at 10 days [107]. In contrast, Nezu et al. [10] found that Keap1 knockdown reduced necrotic areas in the outer medulla of the kidney in the early stage of unilateral IRI. Furthermore, short-term administration of the Keap1 blockade CDDO during the early stage of injury was critical in protecting renal tubular cells from further damage. After 14 days of unilateral IRI, tubular injury was significantly reduced in Keap1 knockdown mice compared to wild-type mice. The improvement in tubular injury persisted in Keap1 knockdown mice until day 42. These findings suggest that early-phase Nrf2 hyperactivation may prevent tubular injury in renal IRI and attenuate the transformation to progressive CKD.

Bardoxolone methyl has shown promise in improving both renal function and histopathology in models of ischemic AKI. This improvement is thought to result from increased mRNA and protein expression of Nrf2 [108]. Similarly, sulforaphane activated Nrf2 and induced phase 2 antioxidant enzymes in HK2 cells and protected HK2 cells from hypoxia-reoxygenation injury, which models IRI in vitro [109]. In addition, in ferric nitrilotriacetate-injured mice, pretreatment with CDDO-Im prior to ferric nitrilotriacetate injection improved kidney function. This suggests that CDDO-Im protects against ferric nitrilotriacetate-induced nephrotoxicity. Overall, these studies suggest that targeting the Keap1-dependent Nrf2 pathway has therapeutic potential for the treatment of AKI, particularly by using pharmacological agents like bardoxolone methyl and sulforaphane.

7.2. Targeting the Keap1-dependent Nrf2 pathway in CKD

Keap1 is constitutively expressed in various kidney cells and highly enriched in podocytes (Fig. 1B). Keap1 inhibition has been shown to activate Nrf2 and attenuate glomerular disease in animal models. For example, podocyte-specific Keap1 knockout enhanced Nrf2 defense and protected against adriamycin-induced kidney injury in mice, as evidenced by reduced proteinuria, lower levels of the injury marker neutrophil gelatinase-associated lipocalin, and decreased fibrosis [110]. In addition, in the NEP25 mouse model of podocyte-specific injury induced by hCD25-targeted immunotoxin injection, Keap1-hypomorphic knockdown promoted Nrf2 activity and attenuated podocytopathy and glomerulosclerosis, despite no improvement in proteinuria [25]. Furthermore, type 1 diabetic Akita mice, which typically develop medullary hyaline casts in the kidneys at 40 weeks of age — signifying significant glomerular injury, proteinuria, and tubular damage — rarely showed these renal pathological changes when Keap1 was hypomorphically knocked down [19], suggesting a potential protective role for Keap1 inhibition in diabetic kidney injury. Apart from a direct podocyte protective effect, this improvement in diabetic kidney injury could be confounded by changes in the progression of diabetes. Indeed, in diabetic db/db mice with Keap1 hypomorphic knockdown, blood glucose levels decreased, attributable to improved insulin secretion and insulin sensitivity. Similarly, oral administration of CDDO-Im attenuated diabetes in db/db mice [111].

However, other studies challenge the notion that targeting Keap1-dependent Nrf2 pathway is protective in CKD. Paradoxically, genetic activation of Nrf2 through Keap1-hypomorphic knockdown or pharmacological blockade of Keap1 with CDDO-Im has been associated with increased proteinuria in CKD models. For instance, mice with Keap1 hypomorphic knockdown exhibited increased proteinuria in experimental glomerulopathy models evoked by adriamycin, angiotensin II, or protein overload. After these injuries, Keap1-hypomorphic mice developed increased albuminuria, glomerulosclerosis, nephrin disruption and shedding, podocyte injury, podocyte foot process effacement, and renal interstitial fibrosis [112]. To reconcile these conflicting findings, it is hypothesized that Nrf2 may have varying effects depending on the degree and mode of its activation, with systemic, non-specific Nrf2 activation potentially exacerbating CKD [112]. Indeed, Keap1-hypomorphic mice also exhibited elevated blood pressure, which could aggravate podocyte injury, glomerular damage and proteinuria. Similarly, while RTA 405, an analog of CDDO-Me, has been shown to increase the glomerular filtration rate by relaxing mesangial cell contraction in rats [113], RTA 405 treatment led to increased mortality, severe proteinuria, glomerulosclerosis, and renal tubular damage in Zucker diabetic fatty rats with overt type 2 diabetes [114].

In a mouse model of aristolochic acid-induced nephropathy, bardoxolone treatment increased renal Nrf2 expression and significantly ameliorated aristolochic acid-induced tubular necrosis and interstitial fibrosis [115]. In addition to bardoxolone, sulforaphane may offer an effective therapeutic approach to prevent the progression of renal fibrosis and CKD [16]. Paradoxically, neither sulforaphane nor genetic upregulation of Nrf2 activity protected against aldosterone-induced renal injury. This is likely due to the induction of GSK3β [116], a negative regulator of Nrf2 as discussed above, although the specific mechanism underlying GSK3β upregulation remains unclear.

7.3. Clinical trials targeting the Keap1-dependent Nrf2 pathway in kidney diseases

Since genetic targeting is not currently feasible in humans, Keap1 inhibitors have been explored in clinical trials as the primary approach to target Keap1-dependent Nrf2 regulation. The first-in-human trials of bardoxolone methyl demonstrated an increase in estimated glomerular filtration rate (eGFR) in patients with CKD [117]. The BEAM study later demonstrated that bardoxolone methyl did improve renal function in type 2 diabetic patients with CKD, although this improvement was associated with a significant increase in urinary albumin excretion [118]. However, the Phase 3 BEACON trial in patients with type 2 diabetes mellitus and stage 4 CKD further revealed that bardoxolone methyl did not reduce the risk of end-stage renal disease but was associated with a higher risk of cardiovascular complications particularly heart failure, resulting in the early termination of the study [65]. Following the cessation of the BEACON trial, a Phase 2/3 study known as CARDINAL was initiated to assess the efficacy and safety of bardoxolone methyl in patients with Alport syndrome. Results from CARDINAL indicated that bardoxolone methyl was effective in maintaining eGFR levels after 48 and 100 weeks’ treatment as compared to placebo [66]. However, the same number of patients in the bardoxolone methyl and placebo groups developed kidney failure, suggesting minimal renoprotective effects. Furthermore, over 90% of the patients exhibited an increase in transaminases, consistent with chronic liver toxicity [119]. In February 2022, the FDA rejected the new drug application for bardoxolone methyl with the proposed indication to slow CKD progression in patients with Alport syndrome. Collectively, bardoxolone methyl seems to possess negligible renoprotective effects and is associated with significant cardiovascular, liver, and renal toxicity in CKD patients.

8. Strategies to target the GSK3β-mediated Keap1-independent Nrf2 pathway in kidney disease

Overexpression or hyperactivity of GSK3β has been observed in various conditions associated with AKI and CKD, making it a potential therapeutic target for kidney disease.

8.1. Targeting the GSK3β-mediated Keap1-independent Nrf2 pathway in AKI

Bao et al. demonstrated that low-dose lithium, a GSK3β inhibitor, promoted renal tubular epithelial cell repopulation, enhanced renal repair, and accelerated renal function recovery after cisplatin- or IRI-induced AKI in mice [120]. Another GSK3β-specific inhibitor, SB216763, prevented contrast-induced AKI in rats by reducing oxidative stress and inflammation through activation of Nrf2 and inhibition of NF-κB [121]. Additionally, glucosamine has been shown to prevent contrast-induced AKI in rats by reducing apoptosis and oxidative stress, and this was associated with increased inhibitory phosphorylation of GSK3β and elevated Nrf2 levels [122], supporting the key role of GSK3β-regulated Nrf2 response in kidney injury. Moreover, targeting GSK3β through conditional knockout or weekly microdose lithium treatment in folic acid-injured mice potentiated the Nrf2 antioxidant response in the kidney and retarded the progression from AKI to CKD [123].

8.2. Targeting the GSK3β-mediated Keap1-independent Nrf2 pathway in CKD

In the kidney, GSK3β is widely expressed and highly enriched in podocytes (Fig. 1B). Burgeoning evidence suggests that GSK3β is hyperactive in injured or stressed podocytes in culture, as well as in glomerular podocytes in clinical and experimental CKD, correlating with the severity and progression of CKD and glomerular injury [124,125]. In cultured murine podocytes exposed to adriamycin, inhibition of overactivated GSK3β using the small molecule inhibitor SB216763 or by ectopic expression of a dominant negative GSK3β mutant potentiated the Nrf2 antioxidant response to adrimycin insults, resulting in protection against podocyte injury, as evidenced by preserved expression of podocyte homeostatic marker proteins, reduced apoptosis, and improved actin cytoskeleton integrity [85]. In contrast, overexpression of a constitutively active GSK3β mutant impaired the Nrf2 antioxidant response to adriamycin, leading to exacerbated podocyte injury [85]. The GSK3β-mediated Nrf2 antioxidant response is emerging as a novel therapeutic target for podocyte protection and treatment of glomerular diseases. Selective GSK3β knockout in mouse glomerular podocytes reduced proteinuria and attenuated glomerular and podocyte injury in multiple models of glomerulopathy, including protein overload nephropathy [126], adriamycin nephropathy, and nephrotoxic serum nephritis [85]. This protection was attributed to enhanced Nrf2 antioxidant responses [85]. GSK3β-modulated Nrf2 signaling also plays a critical role in diabetic podocytopathy. Knockdown of GSK3β enhanced Nrf2 antioxidant response, reduced oxidative stress and thereby attenuated podocyte injury and senescence in podocytes exposed to a milieu of type 2 diabetes. Conversely, overexpression of a constitutively active GSK3β mutant impaired the Nrf2 antioxidant response, increased oxidative stress, and exacerbated diabetic podocyte injury and senescence [127]. In murine models of STZ-induced diabetic kidney disease, β-hydroxybutyrate therapy inhibited GSK3β and thereby enhanced Nrf2 activation in glomerular podocytes, resulting in a reduction in podocyte senescence and injury and an amelioration of diabetic glomerulopathy and albuminuria [128]. In addition, microdose lithium was found to confer a protective effect on islet β-cells via targeting the GSK3β-regulated Nrf2 antioxidant response, thereby ameliorating type 1 diabetes and its associated renal impairment [129].

9. Keap1-dependent vs. GSK3β-Mediated Keap1-Independent Nrf2 regulatory pathways as therapeutic targets for kidney disease: Pros and Cons

In summary, Nrf2 regulation is mediated by two major mechanisms: 1) the primary Keap1-mediated regulation in the cytoplasm, and 2) the secondary GSK3β-mediated regulation of Nrf2 nuclear export [5,130]. Release from Keap1-based repression leads to a rapid initiation of the Nrf2 response, whereas GSK3β-dictated nuclear exclusion and degradation of Nrf2 is critical for shutting down the self-protective antioxidant [45].

9.1. Keap1-dependent Nrf2 regulatory pathway as therapeutic targets for kidney disease

As aforementioned, RONS are essential redox signaling messengers for normal cellular function. It is critical to maintain an optimal RONS level for cellular homeostasis (Fig. 7A). RONS levels that are either too high or too low can be detrimental. In support of this contention, despite the solid scientific premise that vitamin E reduces oxidative stress, its supplementation has not been successful in preventing chronic disease. For example, the HOPE trial demonstrated that vitamin E did not prevent cancer or cardiovascular events. Concerns have been raised about an unexpectedly high risk of heart failure in patients with vascular disease or diabetes [131]. Similarly, the MRC/BHF Heart Protection Study found that although vitamin E supplements did increase its blood levels, they did not reduce 5-year mortality or the incidence of vascular disease, cancer, or other key outcomes [132]. Furthermore, a meta-analysis highlighted potential harms associated with high-dose vitamin E supplements, including an increased risk of all-cause mortality [133]. In essence, vitamin E supplementation has failed to provide the expected benefits of reducing oxidative stress. The underlying pathomechanisms are unclear, but it is conceivable that long-term use of vitamin E results in a state of abnormally low oxidative state in all cells, which may disrupt the cellular signaling equilibrium due to low levels of redox signaling messengers. As such, optimal oxidative state is critical for physiological homeostasis and health.

Fig. 7.

Pharmacological targeting of Keap1-dependent versus GSK3β-mediated Keap1-independent regulatory pathways result in distinct modes of Nrf2 antioxidant response. (A) Schematic diagram illustrating the natural time-course of the Nrf2 antioxidant response in the diseased kidneys (light pink) to reactive oxygen and nitrogen species (RONS) insults in the absence of any interventions. The intensity of the pink color represents the magnitude of the Nrf2 response. (B) Pharmacological targeting of Keap1-dependent regulation of Nrf2, such as using Keap1 inhibitors like bardoxolone methyl, enhances the basal/constitutive Nrf2 antioxidant response systemically, rather than potentiating the stress-inducible Nrf2 antioxidant response in injured organs (depicted here in the diseased kidneys). This may lead to significant adverse effects. Note the elevated basal activity, with blunted amplitude of the Nrf2 antioxidant response to RONS insults in the diseased kidneys. (C) In contrast, therapeutic targeting of GSK3β does not affect Keap1 but allows for the fine-tuning of the inducibility, magnitude, and duration of the Nrf2 antioxidant response specifically in stressed or injured organs (depicted here in the diseased kidneys). Note the augmented amplitude and prolonged duration of the Nrf2 antioxidant response to RONS insults in the diseased kidneys, with minimal changes in basal levels.

Nrf2 is a master regulator of the antioxidant response. Therapeutic targeting of Keap1-dependent Nrf2 regulatory pathway may systemically potentiate the Nrf2 activity, leading to a systemic low-level oxidative state reminiscent of the effects seen with high-dose vitamin E supplementation. However, such systemic overactivation of Nrf2 antioxidant response may also result in potential adverse effects, similar to those observed with high-dose vitamin E. In the BEAM study, for example, increased albuminuria was noted despite improvements in eGFR. Similarly, in cynomolgus monkeys, bardoxolone methyl treatment increased albuminuria, and this effect was attributed to reduced renal expression of megalin, a protein crucial for reabsorbing filtered albumin in renal proximal tubules [134,135]. This downregulation of megalin, coupled with increased eGFR and single nephron GFR, contributed to the increased urinary albumin excretion [136]. This is concerning because increased albuminuria has been associated with renal inflammation and kidney disease progression, although the exact mechanism remains not fully understood [118].

In addition to increased albuminuria, the BEACON trial was halted early primarily due to an elevated risk of heart failure and cardiovascular events [65]. While the underlying pathomechnism is unknown, it may be related to a systemic low-level oxidative state caused by non-specific overactivation of Nrf2 antioxidant response, akin to the adverse effects of vitamin E supplementation. Furthermore, the increased albuminuria may play a role in this risk, because proteinuria is known to correlate strongly with cardiovascular events [137]. In fact, RTA 405, a compound tested in rats, also exacerbated diabetic nephropathy and caused additional adverse effects [114]. To exclude the possibility that RTA 405 contained impurities or degradation products, dh404, a variant of RTA 405, was selected for further validation. Unfortunately, dh404 did not show any beneficial effects on proteinuria, glomerulosclerosis and renal interstitial inflammation in diabetic rats [114]. Thus, there is a risk that bardoxolone may ultimately accelerate the decline in kidney function [138], though this has not been conclusively studied. In addition, even Keap1 knockout mice, which are supposed to have no off-target effects, died within three weeks of birth [139], again underscoring the detrimental effect of systemic enhancement of Nrf2.

It is arguable that an optimal level of Nrf2 activation may be beneficial and should be targeted to avoid adverse events in certain populations. Indeed, in contrast to the early postnatal lethality in Keap1 knockout mice, Keap1-hypomorphic knockdown did not cause mouse death [112]. Additionally, Keap1 heterozygotes extend the lifespan of Drosophila [140]. In the CARDINAL trial, which intentionally excluded patients with a history of heart failure or high baseline B-type natriuretic peptide levels and focused on younger patients, bardoxolone methyl maintained eGFR levels without serious cardiovascular events [66]. However, this finding may have masked the potential cardiovascular risks of bardoxolone methyl.

Keap1-dependent sequestration and degradation of Nrf2 in the cytoplasm is the primary mechanism regulating Nrf2 activity. Systemic inhibition of Keap1, whether through Keap1 inhibitors or Keap1 knockdown in mice, leads to enhanced Nrf2 activity across all organs, regardless of whether the organ is normal or injured. This approach boosts basal Nrf2 activation rather than stress/injury-induced activation, potentially resulting in significant deleterious effects (Fig. 7B). Therefore, the Keap1-dependent regulation may not be the optimal strategy for enhancing Nrf2 antioxidant defense. Developing alternative strategies that harness Keap1-independent pathways to activate the Nrf2 antioxidant response and promote cellular self-defense is crucial. GSK3β plays a pivotal role in the nuclear exclusion and degradation of Nrf2, which is necessary to terminate the self-protective antioxidant stress response following injury [85]. This mechanism enables Keap1-independent regulation to occur specifically in stressed or injured cells where Nrf2 activation is required, while leaving normal, healthy cells unaffected.

9.2. GSK3β-mediated Keap1-independent Nrf2 regulatory pathways as therapeutic targets for kidney disease

Optimally enhanced Nrf2 activity is critical for preventing kidney injury and delaying the progression of kidney disease. However, excessive induction of Nrf2 activity, especially when systemic, can be counterproductive [141]. GSK3β plays a key role by phosphorylating Nrf2 and excluding it from the nucleus, thereby shutting down the self-protective response after cellular damage and preventing Nrf2 overactivation [87]. Importantly, GSK3β-regulated Nrf2 activation and the transcription of Nrf2-targeted antioxidant genes occur selectively in stressed or injured cells and tissues rather than non-specifically throughout the entire system. Therapeutic targeting of the GSK3β-mediated Nrf2 regulatory pathway has been shown to potentiate the Nrf2 antioxidant response and protect multiple organs from injury. For example, patients with chronic hepatitis C, who received long-term treatment with lithium carbonate, a GSK3β inhibitor, due to concomitant psychiatric disorders, showed less liver injury. This was attributed to an enhanced Nrf2 response in the liver [142]. Additionally, GSK3β inhibition, either by gene targeting or pharmacological inhibitors, boosts Nrf2 antioxidant responses in podocytes following doxorubicin or nephrotoxic serum-induced injury, attenuating podocytopathy, glomerular injury, and proteinuria [85]. Notably, this regulation was independent of Keap1 [123].

These findings suggest that the GSK3β-regulated Nrf2 response offers a more targeted approach than Keap1-dependent Nrf2 regulation. Mechanistically, GSK3β regulates Nrf2 during the delayed/late phases of the oxidative stress response, and is effective only when Nrf2 is activated and accumulates in the nucleus. Unlike Keap1-dependent regulation, which systemically influences basal Nrf2 activity, GSK3β-directed Keap1-independent regulation of Nrf2 controls the inducibility, magnitude and duration of the Nrf2 response specifically in stressed or injured cells and tissues, without affecting healthy tissues (Fig. 7C). More importantly, recent evidence suggests that GSK3β becomes hyperactive upon injury, and this appears to be confined to stressed or injured tissues, without affecting healthy tissues [90,124,125,143,144]. Thus, it is conceivable that systemic pharmacological targeting of GSK3β may preferentially affect stressed or injured tissues with GSK3β hyperactivity. Therefore, targeting the GSK3β-mediated Nrf2 regulatory pathway represents a promising new strategy for the treatment of kidney diseases.

A potential concern for pharmacological targeting of GSK3β is the lack of a highly selective inhibitor at present. Most GSK3β inhibitors also exert some inhibitory effects on GSK3α [83]. Since GSK3 is in itself a master regulator of many biological processes beyond Nrf2 antioxidant response and phosphorylates a number of transcription factors — such as β-catenin, NF-κB, Jun, peroxisome proliferator-activated receptor, etc — many of which play important roles in kidney disease [82,83], there may be a concern that broad inhibition of GSK3 could impact these pathways. Indeed, Hurcombe et al. [145] demonstrated that podocyte-specific knockout of both GSK3α and GSK3β in embryonic or adult mice activated β-catenin and disrupted Hippo signaling in podocytes, leading to severe podocyte injury, glomerulosclerosis, and heavy proteinuria. Similarly, chronic high-dose lithium treatment in rats for six months caused inhibitory phosphorylation of both GSK3α and GSK3β, stabilizing β-catenin in podocytes and resulting in significant proteinuria and glomerulosclerosis [145]. However, by using compound knockouts of both GSK3α and GSK3β, Doble et al. [146] showed that genetic deletion of at least three of the four alleles of both GSK3 isoforms is required to cause an appreciable change in β-catenin activity. This suggests that there may be a therapeutic window during which chemical inhibitors can effectively block GSK3β signaling without impacting β-catenin levels in peripheral organs, such as the kidney. Our recent studies demonstrated that using micro-dose lithium, rather than using psychiatric high doses, may offer a viable approach. To this end, once-weekly low-dose lithium treatment has been shown to effectively inhibit GSK3β in the diseased kidney, enhance the Nrf2 response and reduce the transformation to CKD in mice with folic acid-induced AKI [123]. At this low dose, lithium had minimal effect on β-catenin signaling [147], yet was also able to prevent podocyte GSK3β hyperactivity in mice with lipopolysaccharide- or adriamycin-induced podocytopathy, resulting in protection against podocyte injury and albuminuria [147]. While the renoprotective effects of therapeutic targeting of GSK3β may be partially attributed to the inhibition of proinflammatory NF-κB activation [147], Nrf2, among the various biological pathways downstream of GSK3β, appears to play a central role in mediating kidney protection. Indeed, the beneficial effects of podocyte-specific knockout of GSK3β or GSK3β inhibitors in mouse models of adriamycin nephropathy or nephrotoxic serum nephritis were largely abolished by co-treatment with trigonelline, a selective Nrf2 antagonist [85]. While the precise impact of GSK3β inhibition on other downstream biological pathways in kidney disease is yet to be fully defined, recent evidence suggests that this is unlikely to pose a major concern. For example, once-weekly low-dose lithium administration for 3 or 6 months later in life effectively inhibited GSK3β in various kidney cells, preserved podocyte homeostasis and improved albuminuria in aged mice [148].

10. Summary

An imbalance between the scavenging and production of RONS in cells and tissues leads to oxidative stress. The kidney, as one of the most metabolically active organs, is both a major source of RONS production and a vulnerable target for oxidative injury. Nrf2, the master regulator of antioxidant defense, is primarily modulated by Keap1-dependent and Keap1-independent mechanisms. Targeting the Keap1-dependent Nrf2 regulatory pathway with Keap1 inhibitors or knockouts results in a low-level oxidative state throughout all organ systems. While this state may promote organ resistance to injury to some extent, it also leads to unintended deleterious effects. RONS, as essential redox signaling messengers, are critical for maintaining redox homeostasis and overall health when present at optimal levels [149]. Therefore, relying solely on the Keap1-dependent pathway to reinforce Nrf2 activity may not be ideal. The development of alternative strategies that exploit Keap1-independent pathways to activate the Nrf2 antioxidant response and promote cellular self-defense is crucial. In contrast to Keap1-dependent regulation, which affects basal Nrf2 activity systemically, GSK3β-directed Keap1-independent regulation of Nrf2 controls the inducibility, magnitude and duration of Nrf2 antioxidant response specifically in stressed or injured cells and tissues. This approach has shown promise as a therapeutic target for various kidney diseases, including AKI, CKD, and glomerular diseases. Thus, targeting the GSK3β-mediated Nrf2 regulatory pathway represents an actionable and pragmatic new strategy for treating kidney disease.

CRediT authorship contribution statement

Jiahui Zhang: Writing – original draft. Mingzhuo Zhang: Data curation, Writing – review & editing. Marc Tatar: Writing – review & editing. Rujun Gong: Writing – review & editing, Funding acquisition.

Disclosure

The authors declare that they have no conflict of interest. Figures were created with biorender.com and figdraw.com.

Declaration of competing interest

The authors declare that they have no conflict of interest.

Data availability

Data will be made available on request.