ERRα-NOS2-mediated citrulline metabolism attenuates tubular epithelial cell senescence in diabetic kidney disease

Abstract

Diabetic kidney disease (DKD) is a leading cause of end-stage renal disease, characterized by tubular epithelial cell (TECs) senescence, inflammation, and fibrosis. This study investigates the role of estrogen-related receptor alpha (ERRα) in regulating TECs senescence in DKD through nitric oxide synthase 2 (NOS2)-mediated citrulline metabolism. We demonstrate that ERRα expression is significantly downregulated in renal tubular cells of both diabetic mice and DKD patients, correlating with increased senescence markers and the senescence-associated secretory phenotype (SASP). Mechanistically, transcriptome and chromatin immunoprecipitation sequencing confirmed that ERRα regulates NOS2 transcription. TECs-specific knockout of ERRα led to reduced NOS2 expression and decreased citrulline levels, exacerbating TECs injury and senescence. In contrast, TECs-specific knock-in of ERRα alleviated TECs injury and senescence and restored citrulline metabolism. These findings indicate that ERRα plays a critical role in regulating NOS2-mediated citrulline metabolism, which is essential for maintaining kidney function and mitigating tubular senescence in DKD. Furthermore, overexpression of NOS2 and supplementation with citrulline ameliorated renal dysfunction and cellular senescence in diabetic mice, underscoring the importance of this metabolic axis. Modulating ERRα and NOS2 activity may present a potential therapeutic strategy to reduce kidney injury and slow the progression of DKD.

Graphical abstract

Highlights

• •ERRα expression is downregulated in TECs from both DKD patients and diabetic mice, correlating with increased cellular senescence.
• •ERRα regulates NOS2 expression and citrulline metabolism, impacting the senescence phenotype in renal tubular cells.
• •TECs-specific knockout of ERRα worsens kidney injury and senescence in diabetic mice, while its overexpression provides protective effects.
• •NOS2 overexpression and citrulline supplementation reduce TECs injury and senescence in diabetic kidney disease models.

1. Introduction

Diabetic kidney disease (DKD) is one of the most common and severe complications of diabetes mellitus, representing a leading cause of end-stage renal disease (ESRD) worldwide [1]. The pathogenesis of DKD is multifactorial, involving metabolic dysfunction, inflammation, oxidative stress, and fibrosis, which culminate in progressive renal damage [2]. Among the various cell types affected by DKD, renal tubular epithelial cells (TECs) play a crucial role in the initiation and progression of kidney injury [3]. TECs senescence, characterized by irreversible cell cycle arrest and the secretion of pro-inflammatory mediators, is a key feature of DKD and contributes significantly to kidney fibrosis and functional decline [4].

Recent studies have highlighted the importance of metabolic reprogramming in the pathogenesis of DKD, particularly with regard to amino acid metabolism [5]. One critical metabolic pathway that has received increasing attention is citrulline metabolism, which is involved in the urea cycle and nitric oxide (NO) production [6]. Citrulline is primarily synthesized from arginine by the enzyme nitric oxide synthase 2 (NOS2), and its levels are tightly regulated within renal cells [7]. Dysregulation of this metabolic pathway can lead to impaired kidney function and contribute to the progression of renal damage in DKD [8,9]. However, the molecular mechanisms that regulate citrulline metabolism and their impact on renal tubular cell senescence in the context of DKD remain poorly understood.

Estrogen-related receptor alpha (ERRα) is a nuclear hormone receptor that plays a central role in the regulation of energy metabolism, including mitochondrial function, fatty acid oxidation, and amino acid metabolism [10]. ERRα has been shown to regulate various metabolic processes in response to physiological and pathological stimuli [11]. In recent years, ERRα has garnered attention for its potential role in metabolic diseases, including diabetes and cardiovascular diseases, due to its involvement in energy homeostasis and mitochondrial function [[12], [13], [14]]. However, the role of ERRα in the context of DKD remains understudied.

In this study, we hypothesize that ERRα regulates NOS2-mediated citrulline metabolism, which in turn modulates renal tubular cell senescence in DKD. Using both in vitro and in vivo models of DKD, we aim to elucidate the regulatory role of ERRα in renal tubular cell aging and dysfunction. Specifically, we investigate how ERRα modulates NOS2 expression and citrulline metabolism in renal tubular cells and assess the impact of ERRα manipulation on kidney function, senescence, and inflammation in diabetic mouse models. By uncovering the molecular mechanisms through which ERRα regulates citrulline metabolism, this study aims to identify potential therapeutic targets for the prevention and treatment of DKD.

2. Methods

2.1. Human kidney biopsy samples and clinical index collection

The primary control kidney samples (n = 12) were obtained from individuals without diabetes or other known kidney diseases who had undergone tumor nephrectomy, and para-carcinoma cortical regions without obvious pathological changes were carefully selected. In addition, due to these tissues come from cancer patients who may have systemic alterations or subtle local changes due to nearby tumors or surgical stress, renal biopsy specimens from patients diagnosed with minimal change disease (MCD) (n = 12), a glomerular disorder characterized by near-normal proximal tubular morphology on light microscopy, were further analyzed as an additional reference group for almost normal tubular structure and function. The DKD kidney samples (n = 24) were obtained from patients who had been diagnosed with DKD both clinically and pathologically. These sample numbers were limited by the availability of suitable biopsy materials. Written informed consent was obtained from all patients for participation in the study and for publication of the data. The detailed clinical data for the two patient groups are presented in Supplementary Table 1. Notably, the DKD patients had substantially lower eGFR and higher urine albumin-to-creatinine ratio (UACR) than controls. Use of human kidney biopsy material was approved by the Medical Ethics Committee of Renmin Hospital of Wuhan University (WDRM‐20230310).

2.2. Animal models

All animal procedures followed the NIH Guidelines for the Care and Use of Laboratory Animals (Revised 2011) and received ethical clearance from the Animal Use Ethical Committee of Renmin Hospital of Wuhan University, China (WDRM‐20220205). Diabetes mouse model was induced in 8-week-old male C57BL/6J mice by a single intraperitoneal (i.p.) injection of streptozotocin (STZ, 60 mg/kg). After one week, mice were placed on a high-fat diet (HFD) for 12 weeks to induce DKD [15]. Successful diabetes induction included persistent hyperglycemia (fasting blood glucose >16.7 mmol/L and HbA1c >9.5% in STZ/HFD mice, compared to approximately 6 mmol/L and 4.4% in controls). In this study, diabetic mice exhibited marked proteinuria with significantly elevated UACR (mean > 40 mg/g, with most exceeding 100 mg/g). Extensive research confirms that an ACR >30 mg/g is typically regarded as indicative of microalbuminuria in mice [16,17]. As the data from this model surpassed this threshold, it indicates the presence of DKD.

As described previously [15], to generate kidney-specific ERRα knockout (ERRα^ptKO) and knock-in (ERRα^ptKI) mice, we used the CRISPR/Cas9 technology to delete or overexpress ERRα in kidney tubular cells. ERRα floxed allele was engineered with loxP sites flanking a critical exon of Esrra, and that crossing with Ggt1-Cre mice results in proximal tubule-specific deletion of ERRα (ERRα^ptKO, Fig. S1A) or activation of a Cre-dependent ERRα overexpression cassette inserted at the Rosa26 locus (ERRα^ptKI, Fig. S2A). The control group (ERRα^ctrl) consisted of littermate mice without the Cre recombinase transgene (Ggt1-Cre^-).

2.3. Intrarenal adeno-associated virus (AAV) delivery

Following a previously described method [15], the intrarenal administration of AAV effectively induced the overexpression of NOS2 in vivo. For this purpose, 1 × 10¹² genomic particles of AAV-CMV-NOS2 (AAV-NOS2) or AAV-CMV-null (AAV-ctrl) obtained from HANBIO Technology (Shanghai, China) were injected into the kidneys of different group of mice. The injection was performed at six distinct locations through in situ renal injection. We used an AAV serotype 9 (AAV9) vector with a CMV promoter. In addition, control AAV (AAV-ctrl) injections in diabetic mice did not produce significant renal injury or inflammation, indicating that the AAV vector and injection procedure alone had no detectable adverse effects.

2.4. Cell culture and treatments

Human kidney proximal tubular cells (HK-2) and primary TECs were obtained from from the Wuhan Procell Cell Bank and cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin. Cells were maintained in a humidified incubator at 37 °C with 5% CO₂.

To induce cellular senescence, TECs were treated with high glucose (HG, 40 mM) for 24 h [15]. For ERRα knockout, CRISPR/Cas9 technology was employed to knock out ERRα in HK-2 cells. Briefly, ERRα-specific guide RNAs (gRNAs) were designed and transfected into HK-2 cells using Lipofectamine CRISPRMAX (Thermo Fisher Scientific). For ERRα overexpression, lentiviral vectors carrying the ERRα gene were used to transduce HK-2 cells. Successful knockdown or overexpression was verified by Western blotting.

For NOS2 and ERRα regulation, HK-2 cells were transfected with lentiviral vectors expressing NOS2 (OE-NOS2), ERRα (OE- ERRα) or control vectors. Cells were then cultured under normal or HG conditions for further analysis.

2.5. Histological and biochemical analyses

Mice were euthanized by CO₂ asphyxiation, and kidneys were harvested, fixed in 4% paraformaldehyde, and embedded in paraffin. Kidney sections (5 μm) were stained with hematoxylin and eosin (HE) to assess renal histopathology, including tubular atrophy, interstitial fibrosis, and inflammation, and with Masson's trichrome to evaluate interstitial collagen deposition and fibrosis. To assess cellular senescence, β-galactosidase (β-gal) staining was performed using a commercial kit (Cell Signaling Technology). To assess oxidative stress, dihydroethidium (DHE) staining was performed on frozen kidney sections according to the manufacturer's instructions. For immunofluorescence staining, kidney sections were incubated with primary antibodies against p16 (1:100, GTX05147, GeneTex), p21 (1:100, GTX629543, GeneTex), GLB1 (1:100, GTX104360, GeneTex) and ERRα (1:100, GTX108166, GeneTex), followed by DAPI for nuclear staining. For ultrastructural analysis, small pieces of renal cortex (∼1 mm³) were fixed in 2.5% glutaraldehyde, post-fixed in osmium tetroxide, embedded in resin, and sectioned for transmission electron microscopy to assess mitochondrial morphology.

Renal function was assessed by measuring serum levels of blood urea nitrogen (BUN) and serum creatinine (Cr), as well as urine albumin-to-creatinine ratio (ACR).

2.6. Western blotting

For protein extraction, renal tissues and cultured cells were lysed in RIPA buffer (Thermo Fisher Scientific) containing protease and phosphatase inhibitors (Sigma-Aldrich). The protein concentration was determined using the BCA Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of protein (30–50 μg) were separated by SDS-PAGE and transferred onto PVDF membranes (Millipore). Membranes were blocked with 5% non-fat milk in TBS-T and incubated overnight with primary antibodies against ERRα (1:1000, GTX108166, GeneTex), NOS2 (1:1000, Abcam, #ab15323), p16 (1:1000, Cell Signaling Technology, #4824), p21 (1:1000, Cell Signaling Technology, #2947), and β-actin (1:5000, Sigma-Aldrich, #A5316). After washing, membranes were incubated with horseradish peroxidase-conjugated secondary antibodies and developed using an enhanced chemiluminescence detection system (Bio-Rad).

2.6.1. Enzyme-Linked immunosorbent assay (ELISA)

We specifically detected IL-1β and IL-6 as representative SASP markers in tubular cell isolates and cell culture supernatants in this study. IL-1β and IL-6 levels in HK-2 cells and renal tubular tissue were measured using commercially available ELISA kits according to the manufacturer's instructions. Briefly, HK-2 cells were cultured and treated in 6-well plates, and cell culture supernatants were collected and centrifuged to remove debris. For renal tubular tissue, kidney cortex samples were homogenized in ice-cold lysis buffer and centrifuged to obtain clear supernatants. Samples and standards were added to 96-well ELISA plates, incubated, washed, and developed with substrate solution. Absorbance was measured at 450 nm, and cytokine concentrations were calculated from standard curves and normalized to total protein content for tissue lysates.

2.7. Quantitative PCR (qPCR)

Total RNA was extracted from renal tissues or HK-2 cells using TRIzol reagent (Invitrogen), followed by reverse transcription using the iScript cDNA Synthesis Kit (Bio-Rad). qPCR was performed using SYBR Green Master Mix (Bio-Rad) on the CFX96 Real-Time PCR System (Bio-Rad). The following primers were used:

NOS2: Forward: 5′-GTTCTCAGCCCAACAATACAAGA-3′, Reverse: 5′-GTGGACGGGTCGATGTCAC-3′

ASS1: Forward: 5′-CTGATGGAGTACGCAAAGCA-3′, Reverse: 5′-CTCGAGAATGTCAGGGGTGT-3′

β-actin: Forward: 5′-AAGTGTGACGTTGACATCCG-3′, Reverse: 5′-GATCCACATCTGCTGGAAGG-3′

The expression levels of target genes were normalized to β-actin.

2.8. RNA sequencing and bioinformatic analysis

Total RNA was extracted from ERRα-ctrl and ERRα-KO HK-2 cells (n = 3 biological replicates per group) using TRIzol reagent (Invitrogen), and RNA quantity and integrity were assessed by NanoDrop spectrophotometry and agarose gel electrophoresis. Samples with an RNA integrity number (RIN) > 7.0 were used for library construction. Strand-specific mRNA libraries were prepared using a commercial library preparation kit according to the manufacturer's instructions, including poly(A)⁺ RNA enrichment, fragmentation, cDNA synthesis, adapter ligation and PCR amplification. The libraries were sequenced on an Illumina platform (paired-end, 150-bp reads), yielding at least 20 million raw reads per sample. After removal of adaptor sequences and low-quality reads, clean reads were aligned to the human reference genome (GRCh38) using HISAT2. Gene-level read counts were obtained with feature Counts, and differential expression analysis between ERRα-ctrl and ERRα-KO cells was performed using DESeq2. Genes with an adjusted p value (false discovery rate, FDR) < 0.05 and |log₂ fold change| > 1 were considered differentially expressed and used for subsequent volcano plot and pathway analyses.

2.9. Gene Set Enrichment Analysis (GSEA) and pathway analysis

To gain pathway-level insight into the transcriptional changes caused by ERRα loss, Gene Set Enrichment Analysis (GSEA v4.0.3) was performed using the MSigDB Hallmark and KEGG gene sets. Enrichment scores were calculated based on the ranked gene list from DESeq2, and pathways with FDR q value < 0.25 were considered significantly enriched. In parallel, KEGG pathway enrichment analysis was carried out on significantly downregulated genes using the clusterProfiler package to identify metabolic pathways affected by ERRα deletion. Particular attention was paid to amino acid metabolism–related pathways, including the urea cycle and arginine–proline metabolism, in which key enzymes of the citrulline/NO cycle (such as NOS2 and ASS1) were significantly downregulated. These analyses guided the subsequent targeted validation of citrulline metabolism, NOS2/ASS1 expression and ERRα–NOS2 promoter interactions.

2.10. Citrulline measurement assays

Citrulline levels were measured in both renal tissues and HK-2 cells using a commercial citrulline assay kit (ab273309, abcam) according to the manufacturer's instructions. The assays were performed in triplicates, and data were normalized to total protein concentration.

2.11. Luciferase reporter assay

Luciferase reporter assays were used to evaluate the transcriptional regulation of NOS2 by ERRα. The promoter region of NOS2 (approximately 1 kb upstream of the transcription start site) was cloned into a pGL3-luciferase reporter vector. The plasmids were transfected into HK-2 cells, and luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega). To confirm the specific binding of ERRα to the NOS2 promoter, site-directed mutagenesis was performed to introduce mutations in the ERRα-binding site within the NOS2 promoter.

2.12. Chromatin immunoprecipitation (ChIP) PCR

HK-2 cells were fixed with 1% formaldehyde, and chromatin was sheared by sonication. ChIP was performed using an anti-ERRα antibody (GTX108166, Genetex) and the ChIP-IT kit (Active Motif). After reverse cross-linking, PCR was used to detect ERRα binding at the NOS2 promoter. ChIP PCR confirmed that ERRα binds to a specific site in the NOS2 promoter.

2.13. Statistical analysis

Data are presented as mean ± standard deviation (SD) from at least three independent experiments. Statistical significance between groups was determined by unpaired t-tests for two groups or one-way ANOVA for multiple comparisons, followed by Tukey's post-hoc test. Pearson correlation coefficients (r) were calculated to evaluate associations between variables, with significance assessed by two-tailed tests (p < 0.05 considered significant). A p-value of <0.05 was considered statistically significant. All analyses were performed using GraphPad Prism software (version 9.0).

3. Results

3.1. Downregulation of ERRα in TECs and its association with senescence

To investigate the senescence changes in TECs of DKD patients, the renal biopsy samples were collected from 24 DKD patients and control samples from 12 patients undergoing nephrectomy for tumors or 12 MCD patients. Immunofluorescence co-staining of senescence marker (GLB1, p16, p21) and renal tubular marker LTL revealed significant senescence in TECs of DKD patients (Fig. 1A–C). Further co-staining of ERRα and LTL showed a marked downregulation of ERRα expression in TECs of DKD patients (Fig. 1D). Interestingly, the expression of ERRα was negatively correlated with the expression of GLB1 (r = −0.6145, p = 0.0014), suggesting a relationship between ERRα downregulation and TECs senescence (Fig. 1E). Additionally, we analyzed the expression of ERRα in TECs of diabetic mice. As previously described, diabetes was induced in mice by intraperitoneal injection of streptozotocin (STZ) combined with a high-fat diet (HFD) [18]. Immunofluorescence staining and Western blotting results showed that ERRα expression was downregulated in TECs of diabetic mice (Fig. 1F and G). Furthermore, high glucose (HG) stimulation significantly decreased the expression of ERRα in primary TECs (Fig. 1H). Taken together, these results indicate that ERRα expression is downregulated in TECs and is associated with cellular senescence in DKD.

Fig. 1.

Downregulation of ERRα in TECs and its association with senescence

(A) Representative images and quantification of GLB1 (red), LTL (green) and DAPI (blue)-immunofluorescent stained kidney sections from control (n = 12), MCD (n = 12) and DKD (n = 24) patients.

(B) Representative images and quantification of p16 (red), LTL (green) and DAPI (blue)-immunofluorescent stained kidney sections from control (n = 12), MCD (n = 12) and DKD (n = 24) patients

(C) Representative images and quantification of p21 (red), LTL (green) and DAPI (blue)-immunofluorescent stained kidney sections from control (n = 12), MCD (n = 12) and DKD (n = 24) patients

(D) Representative images and quantification of ERRα (red), LTL (green) and DAPI (blue)-immunofluorescent stained kidney sections from control (n = 12), MCD (n = 12) and DKD (n = 24) patients

(E) Correlation analysis of ERRα expression and GLB1 expression in DKD (n = 24) patients

(F) Representative images and quantification of ERRα (red), LTL (green) and DAPI (blue)-immunofluorescent stained kidney sections from Vehicle (n = 6) and STZ/HFD (n = 6) groups of mice

(G) Representative images and band density quantification of Western blotting using isolated renal tubules from Vehicle (n = 6) and STZ/HFD (n = 6) groups of mice to quantify protein expression of ERRα normalized to β-actin.

(H) Representative images and band density quantification of Western blotting using primary TECs from each mouse group to quantify protein expression of ERRα normalized to β-actin (n = 3)

n = 6 independent group mice. Data are expressed as mean ± SD. Statistical differences between two groups were assessed using t-test. For comparisons among three or more groups, one-way ANOVA followed by Tukey's post hoc test was used. ∗∗∗p < 0.001.

3.2. TECs-specific knockout of ERRα exacerbates TECs injury and senescence in diabetic mice

Although significant downregulation of ERRα was observed in the kidneys of DKD patients and diabetic mice, these findings do not establish a direct causal link between ERRα and TECs senescence. To address this, we utilized CRISPR/Cas9 technology to generate ERRα-flox mice, which were subsequently crossed with Ggt1-Cre mice to create TECs-specific ERRα knockout mice (ERRα^flox/flox, Ggt1-Cre⁺, ERRα^ptKO) and control mice (ERRα^flox/flox, Ggt1-Cre^-, ERRα^ctrl), as previously described (Fig. S1A–B) [15]. Importantly, ERRα^ptKO mice that were not rendered diabetic (vehicle-treated) showed no significant differences in renal function or histopathology compared to non-diabetic ERRα^ctrl mice, suggesting that ERRα deletion alone is not sufficient to cause kidney injury without diabetic stress. Subsequently, we induced diabetes in these mice by STZ injection and 12 weeks of HFD (Fig. 2A). Renal function analysis indicated that compared to ERRα^ctrl mice, ERRα^ptKO mice exhibited significantly elevated serum levels of blood urea nitrogen (BUN), serum creatinine (Cr), and albumin-to-creatinine ratio (ACR), suggesting that TECs-specific knockout of ERRα significantly worsened renal function in diabetic mice (Fig. 2B–D). Histopathological analysis (HE staining) revealed tubular atrophy, widespread interstitial inflammation, and infiltration in diabetic mice, which were further aggravated in ERRα^ptKO mice (Fig. 2E). Given that the ultimate pathological endpoint of DKD is renal interstitial fibrosis, we also assessed the degree of renal fibrosis in mice from each group using Masson's trichrome staining. The results demonstrated significant renal fibrosis in diabetic mice, which was further exacerbated in ERRα^ptKO mice (Fig. 2E). Moreover, β-galactosidase (β-gal) staining and GLB1 immunofluorescence analysis showed that ERRα^ptKO mice exhibited significantly increased senescence in TECs compared to control mice (Fig. 2E and F). Given that ERRα is known to regulate antioxidant genes (SOD2, GPX1, and catalase), we further analyzed its effects on oxidative stress and mitochondrial integrity. DHE staining and transmission electron microscopy (TEM) revealed that ERRα knockout significantly increased oxidative stress levels in TECs and exacerbated mitochondrial damage, including mitochondrial swelling and cristae disruption (Fig. 2G and H). Western blotting confirmed the upregulation of senescence markers (p16 and p21) in TECs of ERRα^ptKO diabetic mice (Fig. 2I and J). As well known, senescent cells secrete a group of specific cytokines and inflammatory mediators, known as the senescence-associated secretory phenotype (SASP) [19]. We further analyzed the SASP in these mice, and measurement of renal tubular inflammatory cytokines (IL-1β [20] and IL-6 [21]) revealed that knockout of ERRα significantly increased the SASP in diabetic mice (Fig. 2K and L). Taken together, these results indicate that kidney tubular epithelial cell-specific knockout of ERRα exacerbates renal tubular injury and senescence in diabetic mice.

Fig. 2.

TECs-specific knockout of ERRα exacerbates TECs injury and senescence in diabetic mice

(A) Schematic diagram of intraperitoneal injection of Streptozotocin (STZ) and 12 weeks treatment of high-fat diet (HFD) or normal saline (Vehicle) to ERRα^podKO and ERRα^ctrl mice (n = 6)

(B–D) Blood urea nitrogen (BUN), serum creatinine (Cr) and urine albumin-to-creatinine ratio (ACR) levels in each mouse group (n = 6)

(E–F) Representative HE, Masson's trichrome staining and β-gal staining images, quantification of tubular interstitial damage, Masson staining positive area and immunofluorescent staining of GLB1 (red), LTL (green), and DAPI (blue) in kidney sections from each mouse group (n = 6)

(G–H) Representative TEM images, DHE staining and quantification of proportion of damaged mitochondria, DHE staining positive area in kidney sections from each mouse group (n = 6).

(I–J) Representative images and band density quantification of Western blotting using isolated renal tubules from each mouse group (n = 6) to quantify protein expression of ERRα, p16 and p21 normalized to β-actin (n = 6)

(K–L) IL-1β and IL-6 contents in isolated renal tubules from each mouse group (n = 6).

n = 6 independent group mice. Data are expressed as mean ± SD. Statistical differences among three or more groups were evaluated using one-way ANOVA followed by Tukey's post hoc test. ns: not significant (p > 0.05), ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

3.3. TECs-specific knock-in of ERRα alleviates TECs injury and senescence in diabetic mice

To further investigate the role of ERRα in TECs senescence in DKD, we used CRISPR/Cas9 technology to generate ERRα-flox mice, which were then crossed with Ggt1-Cre mice to generate TECs-specific ERRα knock-in mice (ERRα^flox/flox, Ggt1-Cre⁺, ERRα^ptKI) and control mice (ERRα^flox/flox, Ggt1-Cre^-, ERRα^ctrl), as previously described (Fig. S2A–B) [15]. We then established the diabetic mouse model (Fig. 3A). Compared to ERRα^ctrl mice, ERRα^ptKI mice exhibited no significant differences in BUN, serum Cr, or ACR levels; however, they displayed significantly lower BUN, serum Cr, and ACR levels in the diabetic condition, indicating that TECs-specific overexpression of ERRα alleviates renal dysfunction in diabetic mice (Fig. 3B–D). Histopathological analysis, DHE staining and TEM results of renal tissues revealed that ERRα^ptKI mice showed reduced renal tubular damage, renal fibrosis, oxidative stress levels and mitochondrial damage in comparison to diabetic control mice (Fig. 3E–H). Additionally, β-gal staining and GLB1 immunofluorescence analysis revealed that ERRα^ptKI significantly reduced senescence in TECs in diabetic mice (Fig. 3E and F). Western blotting further showed a significant decrease in the expression of senescence markers (p16 and p21) in TECs of ERRα^ptKI diabetic mice (Fig. 3I and J). Measurement of renal tubular inflammatory cytokines (IL-1β and IL-6) indicated that ERRα overexpression significantly reduced the SASP in diabetic mice (Fig. 3K and L). These findings suggest that TECs-specific overexpression of ERRα ameliorates renal tubular injury and senescence in diabetic mice.

Fig. 3.

TECs-specific knock-in of ERRα exacerbates TECs injury and senescence in diabetic mice

(A) Schematic diagram of intraperitoneal injection of Streptozotocin (STZ) and 12 weeks treatment of high-fat diet (HFD) or normal saline (Vehicle) to ERRα^podKI and ERRα^ctrl mice (n = 6).

(B–D) Blood urea nitrogen (BUN), serum creatinine (Cr) and urine albumin-to-creatinine ratio (ACR) levels in each mouse group (n = 6).

(E–F) Representative HE and β-gal staining images, quantification of tubular interstitial damage, and immunofluorescent staining of GLB1 (red), LTL (green), and DAPI (blue) in kidney sections from each mouse group (n = 6).

(G–H) Representative TEM images, DHE staining and quantification of proportion of damaged mitochondria, DHE staining positive area in kidney sections from each mouse group (n = 6)

(I–J) Representative images and band density quantification of Western blotting using isolated renal tubules from each mouse group (n = 6) to quantify protein expression of ERRα, p16 and p21 normalized to β-actin

(K–L) IL-1β and IL-6 contents in isolated renal tubules from each mouse group (n = 6).

n = 6 independent group mice. Data are expressed as mean ± SD. Statistical differences among three or more groups were evaluated using one-way ANOVA followed by Tukey's post hoc test. ns: not significant (p > 0.05), ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

3.4. ERRα expression is associated with high glucose-induced senescence in HK-2 cells

To further investigate the role of ERRα in TECs senescence in vitro, we used human proximal tubular cells (HK-2 cells). We first generated ERRα knockout HK-2 cells using CRISPR/Cas9 technology. ERRα-KO cells proliferated and divided normally without affecting their growth. β-gal staining revealed that compared to control cells (ERRα-ctrl), ERRα-KO cells showed no obvious senescence under normal conditions but exhibited significantly enhanced senescence following HG stimulation (Fig. 4A). Additionally, measurement of inflammatory cytokines (IL-1β and IL-6) indicated that ERRα knockout exacerbated the SASP induced by HG in HK-2 cells (Fig. 4B and C). Western blotting showed that ERRα knockout further increased the expression of p16 and p21 in HG-stimulated HK-2 cells (Fig. 4D). To further confirm the protective role of ERRα, we overexpressed ERRα in HK-2 cells using lentiviral vectors. As expected, β-gal staining, inflammatory cytokine measurement, and Western blotting revealed that overexpression of ERRα significantly alleviated HG-induced cell senescence, SASP, and the expression of p16 and p21 (Fig. 4E–H). These results suggest that ERRα expression is closely associated with high glucose (HG)-induced senescence in HK-2 cells.

Fig. 4.

ERRα expression is associated with high glucose-induced senescence in HK-2 cells

(A) Representative β-gal staining images quantification of positive area from each HK-2 cell group (n = 3).

(B–C) IL-1β and IL-6 contents from each HK-2 cell group (n = 3).

(D) Representative images and band density quantification of Western blotting from each HK-2 cell group (n = 3) to quantify protein expression of p16 and p21 normalized to β-actin.

(E) Representative β-gal staining images quantification of positive area from each HK-2 cell group (n = 3)

(F–G) IL-1β and IL-6 contents from each HK-2 cell group (n = 3).

(H) Representative images and band density quantification of Western blotting from each HK-2 cell group (n = 3) to quantify protein expression of p16 and p21 normalized to β-actin

n = 3 cultures per group. HG, High gluose (40 mM, 24h); ERRα-KO, Knockout of ERRα in HK-2 cells using CRISPR/Cas9 technology. OE-ERRα, overexpression ERRα group. Data are expressed as mean ± SD. Statistical differences among three or more groups were evaluated using one-way ANOVA followed by Tukey's post hoc test. ns: not significant (p > 0.05), ∗∗p < 0.01, ∗∗∗p < 0.001.

3.5. ERRα regulates citrulline metabolism via NOS2 transcription in TECs

Building upon the observation that ERRα is downregulated in TECs and associated with senescence, we next explored the underlying molecular mechanisms. Transcriptome sequencing of ERRα-ctrl and ERRα-KO HK-2 cells revealed 677 upregulated genes and 1014 downregulated genes (Fig. 5A and B). Gene Set Enrichment Analysis (GSEA) showed significant changes in amino acid metabolism pathways (Fig. 5C), while Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis identified significant alterations in the urea cycle and arginine-proline metabolism pathways, suggesting that ERRα may regulate citrulline metabolism (Fig. 5D). To confirm this hypothesis, we measured citrulline levels in ERRα-ctrl and ERRα-KO cells and found that citrulline levels were significantly decreased in ERRα-KO cells (Fig. 5E). Similarly, HG stimulation in primary TECs significantly reduced citrulline levels (Fig. 5F). These results suggest that ERRα regulates citrulline metabolism in TECs. Further analysis of key enzymes involved in citrulline metabolism, namely ASS1 and NOS2. Consistent with RNA sequencing, qPCR results confirmed that ERRα knockout significantly reduced the mRNA levels of NOS2 but had no significant effect on ASS1 expression (Fig. 5G and H). Luciferase (LUC) assays showed that ERRα directly regulates NOS2 expression, as mutation of the ERRα binding sequence on the NOS2 promoter effectively reversed the transcriptional regulation (Fig. 5I and J). ChIP analysis confirmed ERRα binding at the NOS2 promoter (Fig. 5K–M) [22]. The LUC assays and mRNA results further confirmed that ERRα regulates NOS2 transcription through the sequence “CCAAGGTGGCA” (Fig. 5N and O). These results suggest that ERRα regulates NOS2 transcription and impacts citrulline metabolism in TECs.

Fig. 5.

ERRα regulates citrulline metabolism via NOS2 transcription in TECs

(A) ERRα-ctrl and ERRα-KO groups of HK-2 cells were used for transcriptome sequencing (n = 3)

(B) Volcano map showing differential genes in ERRα-ctrl and ERRα-KO groups of HK-2 cells (n = 3).

(C) Gene Set Enrichment Analysis (GSEA) pathway enrichment analysis enrichment analysis showing altered biosynthesis of amino acids in ERRα-ctrl and ERRα-KO groups of HK-2 cells (n = 3).

(D) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis enrichment analysis showing altered pathways in ERRα-ctrl and ERRα-KO groups of HK-2 cells (n = 3).

(E–F) Citrulline contents from each cell group (n = 3).

(G–H) Relative mRNA level of ASS1 and NOS2 from each cell group (n = 3)

(I–J) Relative luciferase (LUC) activity in each cell group (n = 3).

(K) The binding site of ACOX1 with ERRα in HK-2 cells using chromatin immunoprecipitation sequencing (ChIP-Seq) analysis (n = 3).

(L) The complementary sequence of ERRα targeted motif using JASPAR database.

(M) Diagram of the spatial binding pattern of ERRα protein and NOS2 gene.

(N) Relative luciferase (LUC) activity of NOS2 in each group (n = 3)

(O) Relative mRNA levels of NOS2 in each cell group (n = 3).

n = 3 cultures per group. HG, High gluose (40 mM, 24h); ERRα-KO, knockout of ERRα in HK-2 cells using CRISPR/Cas9 technology. si-ERRα, knockdown ERRα group. NOS2-MUT, transfection of NOS2 mutation site plasmid. Data are expressed as mean ± SD. Statistical differences among three or more groups were evaluated using one-way ANOVA followed by Tukey's post hoc test. ns: not significant (p > 0.05), ∗∗p < 0.01, ∗∗∗p < 0.001.

3.6. NOS2 is downregulated in DKD and associated with TECs senescence

Given that ERRα regulates NOS2 expression and citrulline metabolism, we further analyzed the expression of NOS2 in TECs in DKD. Analysis of NOS2 expression in aging kidneys from the ADEIP database (https://geneyun.net/ADEIP/) revealed a gradual decline in NOS2 expression with age (Fig. 6A). Immunofluorescence staining showed that NOS2 expression was reduced in TECs of DKD and negatively correlated with GLB1 expression (r = −0.5585, p = 0.0046) (Fig. 6B and C). As expected, NOS2 expression was lower in the kidneys of diabetic mice and HG-stimulated HK-2 cells. Following ERRα knockout, expression decreased further, whereas knock-in of ERRα restored NOS2 expression levels (Fig. 6D–G). To investigate the role of NOS2 in TECs senescence, we overexpressed NOS2 in HK-2 cells using lentivirus (OE-NOS2) (Fig. 6H). β-gal staining revealed that overexpression of NOS2 significantly reduced HG-induced senescence in HK-2 cells (Fig. 6I). These results suggest that NOS2 downregulation is associated with senescence in TECs of DKD.

Fig. 6.

NOS2 is downregulated in DKD and associated with TECs senescence

(A) NOS2 expression in the kidney of different age groups of the population using ADEIP database.

(B) Representative images and quantification of NOS2 (red), LTL (green) and DAPI (blue)-immunofluorescent stained kidney sections from control (n = 12), MCD (n = 12) and DKD (n = 24) patients.

(C) Correlation analysis of NOS2 expression and GLB1 expression in DKD (n = 24) patients

(D-E) Representative images and band density quantification of Western blotting using isolated renal tubules from each mouse group (n = 6) to quantify protein expression of NOS2 normalized to β-actin.

(F) Representative images and quantification of NOS2 (red), LTL (green) and DAPI (blue)-immunofluorescent stained kidney sections from each mouse group (n = 6).

(G) Representative images and band density quantification of Western blotting from each HK-2 cells group to quantify protein expression of NOS2 normalized to β-actin (n = 3)

(H) Representative images and band density quantification of Western blotting from each HK-2 cells group to quantify protein expression of NOS2 normalized to β-actin (n = 3)

(I) Representative β-gal staining images quantification of positive area from each HK-2 cell group (n = 3)

n = 3 cultures per group. HG, High gluose (40 mM, 24h); ERRα-KO, knockout of ERRα in HK-2 cells using CRISPR/Cas9 technology. si-ERRα, knockdown ERRα group. OE-NOS2, overexpression NOS2 group. Data are expressed as mean ± SD. Statistical differences among three or more groups were evaluated using one-way ANOVA followed by Tukey's post hoc test. ns: not significant (p > 0.05), ∗∗p < 0.01, ∗∗∗p < 0.001.

3.7. Overexpression of NOS2 alleviates TECs injury and senescence in diabetic mice

To further investigate the role of NOS2 in DKD, we employed adeno-associated viruses (AAV) to overexpress NOS2 in TECs (Fig. 7A). Renal function analysis demonstrated that NOS2 overexpression in TECs led to a significant reduction in BUN, serum Cr, and ACR levels in diabetic mice (Fig. 7B–D). These findings suggest that NOS2 overexpression may improve renal function in diabetic mice. HE/Masson's trichrome staining, DHE staining, TEM results and senescence analyses demonstrated that the overexpression of NOS2 ameliorated renal fibrosis, TECs injury, oxidative stress levels, mitochondrial damage and TECs senescence in diabetic mice (Fig. 7E–J). In addition, as expected, the expression of NOS2 also improved SASP in diabetic mice (Fig. 7K and L). In summary, these results suggest that the expression of NOS2 attenuated TECs injury and senescence in diabetic mice.

Fig. 7.

Overexpression of NOS2 Alleviates TECs Injury and Senescence in Diabetic Mice

(A) Schematic diagram of overexpression of NOS2 in TECs using adeno-associated viruses (AAV).

(B–D) Blood urea nitrogen (BUN), serum creatinine (Cr) and urine albumin-to-creatinine ratio (ACR) levels in each mouse group (n = 6).

(E–F) Representative HE and β-gal staining images, quantification of tubular interstitial damage, and immunofluorescent staining of GLB1 (red), LTL (green), and DAPI (blue) in kidney sections from each mouse group (n = 6).

(G–H) Representative TEM images, DHE staining and quantification of proportion of damaged mitochondria, DHE staining positive area in kidney sections from each mouse group (n = 6)

(I–J) Representative images and band density quantification of Western blotting using isolated renal tubules from each mouse group (n = 6) to quantify protein expression of ERRα, NOS2, p16 and p21 normalized to β-actin

(K–L) IL-1β and IL-6 contents in isolated renal tubules from each mice group (n = 6).

n = 6 independent group mice. Data are expressed as mean ± SD. Statistical differences among three or more groups were evaluated using one-way ANOVA followed by Tukey's post hoc test. ns: not significant (p > 0.05), ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

3.8. Citrulline Supplementation Alleviates TECs injury and senescence in diabetic mice

To investigate the in vivo biological effects of citrulline, we supplemented STZ/HFD mice with citrulline via drinking water. To determine the optimal dosage, we administered citrulline at different dose gradients (0.5 g/kg/day, 1 g/kg/day and 2 g/kg/day) [23]. Masson's trichrome and β-gal staining results revealed that 1 g/kg/day citrulline treatment in STZ/HFD mice significantly reduced renal fibrosis and tubular epithelial cell senescence, leading to its selection as the therapeutic concentration (Fig. 8A, Fig. S3A). Renal function analysis indicated that citrulline supplementation significantly improved renal function in diabetic mice (Fig. 8B–D). HE/Masson's trichrome staining, DHE staining, TEM results and senescence analyses demonstrated that citrulline supplementation ameliorated renal fibrosis, TECs injury, oxidative stress levels, mitochondrial damage and TECs senescence in diabetic mice (Fig. 8E–J). Moreover, Western blotting and inflammatory cytokine measurements revealed that citrulline supplementation reduced the expression of p16 and p21 and decreased the SASP in diabetic mice (Fig. 8K and L). These results suggest that citrulline supplementation mitigates renal tubular injury and senescence in diabetic mice. Notably, citrulline treatment did not significantly alter ERRα expression, indicating that its benefits likely stem from downstream metabolic effects rather than feedback on ERRα/NOS2.

Fig. 8.

Citrulline Supplementation Alleviates TECs Injury and Senescence in Diabetic Mice

(A) Schematic diagram of supplementation scheme of each group of mice with drinking water only or drinking waters containing citrulline (1 g/kg of body weight of mice) for 9 weeks.

(B–D) Blood urea nitrogen (BUN), serum creatinine (Cr) and urine albumin-to-creatinine ratio (ACR) levels in each mouse group (n = 6).

(E–F) Representative HE and β-gal staining images, quantification of tubular interstitial damage, and immunofluorescent staining of GLB1 (red), LTL (green), and DAPI (blue) in kidney sections from each mouse group (n = 6).

(G–H) Representative TEM images, DHE staining and quantification of proportion of damaged mitochondria, DHE staining positive area in kidney sections from each mouse group (n = 6).

(I–J) Representative images and band density quantification of Western blotting using isolated renal tubules from each mouse group (n = 6) to quantify protein expression of ERRα, NOS2, p16 and p21 normalized to β-actin.

(K–K) IL-1β and IL-6 contents in isolated renal tubules from each mice group (n = 6).

n = 6 independent group mice. Data are expressed as mean ± SD. Statistical differences among three or more groups were evaluated using one-way ANOVA followed by Tukey's post hoc test. ns: not significant (p > 0.05), ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

4. Discussion

In this study, we have demonstrated that ERRα plays a crucial role in regulating TECs senescence in DKD through the modulation of NOS2-mediated citrulline metabolism. Our findings show that ERRα expression is significantly downregulated in TECs in both human DKD biopsies and diabetic mouse models, and this downregulation is closely associated with kidney dysfunction, inflammation, and the progression of cellular senescence. Furthermore, we provide evidence that ERRα regulates NOS2 expression, which in turn impacts citrulline metabolism and modulates the senescence phenotype in renal tubular cells. This study not only identifies ERRα as a key regulator of renal tubular senescence but also implicates NOS2 and citrulline metabolism in the pathogenesis of DKD.

The downregulation of ERRα in renal tubular cells, observed in both human and mouse models of DKD, highlights the importance of metabolic dysregulation in the progression of kidney disease. ERRα, a nuclear receptor involved in regulating mitochondrial function, energy metabolism, and oxidative stress responses, has been shown to modulate a variety of pathways critical for cellular homeostasis [18,24,25]. In DKD, reduced ERRα expression appears to exacerbate the metabolic dysfunction seen in diabetic kidneys. In terms of mechanism, hyperglycemia and associated stressors likely contribute to ERRα downregulation in DKD. Indeed, we have previously shown that the E3 ubiquitin ligase RBBP6 can promote ERRα protein degradation under diabetic conditions [15]. In this study, we revealed that this downregulation contributed to TECs senescence, a key feature of DKD that accelerates kidney fibrosis and dysfunction [26]. This is particularly significant because senescent cells secrete pro-inflammatory cytokines and growth factors, collectively known as the senescence-associated secretory phenotype (SASP), which contribute to chronic inflammation and tissue damage [27,28]. Notably, consistent with our mouse experiment findings, increased p16-positive senescent cells have been reported in kidneys of patients with DKD, supporting the clinical relevance of targeting cellular senescence in DKD. In terms of mechanism, hyperglycemia and associated stressors likely contribute to ERRα downregulation in DKD. In addition to ERRα being regulated by ubiquitination, recent studies have also demonstrated that pro-inflammatory signaling pathways (such as the IL-6/STAT3 pathway mediated by HIF1α) can similarly suppress ERRα activity [25]. These factors may underlie the loss of ERRα in DKD, linking the metabolic and inflammatory milieu of DKD to ERRα suppression. By regulating ERRα, we propose that maintaining its expression may help mitigate the effects of TECs senescence in diabetic kidneys.

Moreover, our findings suggest that ERRα acts through the regulation of NOS2, a crucial enzyme involved in citrulline metabolism and nitric oxide (NO) production. Citrulline, synthesized from arginine by NOS2, plays an important role in the urea cycle and vascular homeostasis [29]. Dysregulated citrulline metabolism has been implicated in the progression of several renal pathologies, including DKD [[30], [31], [32]]. In our study, we observed that the knockout of ERRα in TECs resulted in a significant reduction in NOS2 expression and citrulline levels, which further exacerbated TECs senescence and kidney injury in diabetic mice. This suggests that ERRα not only maintains mitochondrial function and cellular metabolism but also plays a critical role in regulating the metabolic pathways that protect against kidney damage and cellular aging. Importantly, these results indicate that ERRα-mediated regulation of NOS2 and citrulline metabolism may be an essential mechanism in maintaining renal tubular cell function and preventing the pathological changes associated with DKD.

The importance of NOS2 in kidney function and aging is underscored by its involvement in inflammation and oxidative stress responses [33,34]. In DKD, aberrant NOS2 expression can lead to excessive NO production, contributing to endothelial dysfunction, glomerular hyperfiltration, and kidney injury. However, our findings suggest that ERRα′s regulation of NOS2 expression is protective, helping to maintain a balance in NO production and mitigate the adverse effects of HG conditions [35,36]. In diabetic mice, we observed that HG exposure significantly downregulated NOS2 expression, and this effect was further aggravated in the absence of ERRα. These findings suggest that ERRα plays a protective role in counteracting HG-induced metabolic disturbances, which are a hallmark of diabetes. By regulating NOS2 expression, ERRα appears to help preserve renal tubular function under diabetic conditions, preventing the exacerbation of TECs senescence and kidney damage.

The potential therapeutic implications of targeting ERRα and NOS2 in DKD are significant. Given the critical role of ERRα in regulating mitochondrial function and metabolic pathways, enhancing its activity could restore metabolic balance in TECs and prevent or reverse the progression of senescence [37,38]. Small molecules that activate ERRα or enhance its downstream signaling pathways could be explored as potential therapies for DKD [39,40]. Compounds such as resveratrol, berberine, and other synthetic potential ERRα activators have shown promise in preclinical studies for modulating metabolic dysfunction in diseases such as type 2 diabetes and cardiovascular disease [41,42]. Translating these findings into DKD treatment would require evaluating their efficacy in animal models and human clinical trials to ensure their safety and effectiveness.

Furthermore, targeting NOS2 represents another potential therapeutic strategy for DKD. NOS2 is an important regulator of citrulline metabolism and NO production, and its dysregulation contributes to kidney injury and senescence [43,44]. Strategies aimed at enhancing NOS2 expression or activity may help restore normal metabolic function in renal tubular cells, alleviating the effects of HG-induced kidney damage and aging [45]. However, the timing and dosage of such interventions must be carefully controlled, as excessive NOS2 activity could lead to detrimental effects, including increased oxidative stress and inflammation [46]. Therefore, developing targeted therapies that modulate NOS2 expression in a controlled manner is crucial for their success in clinical settings.

Although this study has provided valuable insights into the role of ERRα in regulating TECs senescence in DKD, several limitations remain. First, while our findings in mouse models offer strong evidence of the role of ERRα in DKD, further validation in human studies is necessary to confirm the clinical relevance of these results. Future studies using human DKD samples or organoid models could provide more direct evidence of ERRα′s role in human disease. Second, as ERRα is an estrogen-related nuclear receptor, its role in TEC aging may differ between sexes. All in vivo experiments in this study were performed in male mice, so future studies in female models are needed to define any sex-specific effects. Third, although our study focused on DKD, it remains to be determined whether other chronic kidney diseases exhibit similar downregulation of the ERRα-NOS2-citrulline axis and its implications, warranting further investigation. Fourth, this study utilized a small sample size of human subjects and exhibited potential selection bias. Firth, we focused on IL-1β and IL-6 as representative SASP factors due to their known roles in aging-associated inflammation, but did not evaluate other cytokines or chemokines (e.g., TNF-α, TGF-β, MCP-1). A more comprehensive inflammatory mediator profiling would provide additional insights. Finally, while we have focused on the regulation of NOS2 and citrulline metabolism, ERRα may influence other metabolic pathways that contribute to renal tubular function and senescence. Future research should explore these potential pathways, including those related to mitochondrial function, oxidative stress, and inflammatory responses, to gain a more comprehensive understanding of how ERRα modulates TECs health in DKD.

In conclusion, our study identifies ERRα as a potential regulator of TECs senescence in DKD, primarily through its modulation of NOS2-mediated citrulline metabolism. This novel mechanism underscores the importance of metabolic regulation in the pathogenesis of DKD and suggests that ERRα and NOS2 may serve as promising therapeutic targets for preventing or mitigating kidney damage in diabetic patients. Future research should aim to further investigate the therapeutic potential of targeting ERRα and NOS2 in clinical settings and explore additional metabolic pathways that contribute to the progression of DKD.

CRediT authorship contribution statement

Manli Hu: Formal analysis, Funding acquisition. Hongtu Hu: Formal analysis, Supervision, Writing – original draft. Jijia Hu: Methodology, Project administration. Zhaowei Chen: Project administration, Resources. Keju Yang: Investigation, Project administration. Zongwei Zhang: Project administration. Yanqin Fan: Software. Guohua Ding: Funding acquisition, Writing – original draft, Writing – review & editing. Wei Liang: Formal analysis, Funding acquisition, Writing – original draft, Writing – review & editing.

Declaration of competing interest

We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data generated and/or analyzed during this study are included in this published article and its supplementary information files. Additional data are available from the corresponding author upon reasonable request.