Renal tubular VMP1 protects against acute kidney injury via modulating autophagy and autophagy-independent pathway

ABSTRACT

Macroautophagy/autophagy activation protects renal proximal tubular epithelial cells (PTECs) against acute kidney injury (AKI) induced by various challenges. The mechanism that regulates autophagy in PTECs, however, remains incompletely understood. Here, we report that VMP1 (vacuole membrane protein 1) plays an essential role in enabling PTECs to maintain high autophagic flow under AKI conditions. VMP1 in PTECs is strongly upregulated in AKI patients but not chronic kidney disease patients. The rapid elevation of VMP1 expression in PTECs during AKI is validated in mouse AKI models induced by cisplatin or ischemia-reperfusion injury (IRI). PTECs-specific vmp1-knockout mice (vmp1-cKO) display more severe renal injuries when challenged with cisplatin or IRI. In line with this, aging vmp1-cKO mice spontaneously develop defective calcium metabolism and display significant tubular damage. In contrast, adenovirus-mediated Vmp1 expression in renal tubular rescues IRI or cisplatin-induced renal tubular injury. Mechanistically, the level and distribution pattern of VMP1 are associated with the autophagy markers MAP1LC3/LC3 and SQSTM1, and VMP1 facilitates the formation of renal tubular cell autophagosomes. VMP1 deficiency also results in the accumulation of lipid droplets in renal tubular cells. Our studies thus reveal a critical role of VMP1 in protecting against AKI via facilitating tubular cell autophagic flux and lipid metabolism.

Introduction

Acute kidney injury (AKI) is a serious clinical syndrome characterized by a rapid loss of renal excretory function. With high morbidity, increasing healthcare costs, and mortality of about two million deaths per year [1,2], AKI is considered as an important contributor and accelerant in the development and progression of chronic kidney disease (CKD) [3–5]. Currently, no treatments can effectively reduce kidney injury or improve kidney recovery from AKI. The common causes of AKI include renal ischemia-reperfusion injury (IRI), sepsis, and exposure to nephrotoxins such as cisplatin. It has been known that renal tubular cell damage or death, the key pathological feature of AKI, is often followed by tubular cell regeneration and repair [6–8]. However, tubular repair after severe or repeated episodes of AKI tends to be incomplete or maladaptive, leading to chronic inflammation and renal interstitial fibrosis that contribute to the progression to CKD [6–8]. In particular, the kidney proximal tubular epithelial cells (PTECs) play a central role in the transition of AKI to CKD [6–9].

Macroautophagy/autophagy, a lysosomal degradation pathway that breaks down damaged cytoplasmic components via the formation of autophagosomes and autolysosomes [10–12], is known to occur in response to various stressful challenges, such as starvation conditions, hypoxia, endoplasmic reticulum stress or oxidative stress, organelle damage, and pathogen infections [13]. Alterations of autophagy have been demonstrated in both acute and chronic kidney diseases in experimental models and human patients [14–16]. In the acute injury phase of AKI, autophagy is induced in PTECs and acts as an intrinsic protective mechanism [17–19]. Compared with wild-type (WT) mice, the selective deletion of Atg5 or Atg7 in the proximal tubules of mice results in progressive kidney damage and increases tubular cell apoptosis and tubulointerstitial fibrosis [17,18]. In contrast, during kidney recovery from acute injury, autophagy needs to be deactivated for tubular cell proliferation and tubular repair, because tubular cells with persistent autophagy are defective in proliferation [20]. Therefore, the autophagy in PTECs must be tightly controlled in order to clear the damaged components and preserve self-tolerance of proximal tubules [21]. However, the regulatory mechanisms of autophagy and apoptosis in AKI are incompletely understood.

VMP1 (vacuole membrane protein 1) was first identified as a protein responsive to pancreatitis-induced stress [22]. As one of three essential autophagy genes conserved from worms to mammals [23], the role of VMP1 in triggering early steps of the autophagic pathway in mammalian cells was reported by different research groups [24,25]. Like autophagic flux, VMP1 can be induced by starvation or rapamycin treatment, and its overexpression triggers the formation of numerous vesicles that colocalize with MAP1LC3/LC3 (microtubule-associated protein 1 light chain 3), an autophagosome marker. The autophagy-promoting activity of VMP1 can be blocked by autophagy inhibitor 3-methyladenine [24]. VMP1 has been found to interact with BECN1/Beclin 1, a mammalian initiator of autophagy [26]. Expression of VMP1 with a mutated ATG domain fails to induce LC3 recruitment and interaction with BECN1. However, the expression of VMP1 in renal tubular cells and its potential role in regulating renal tubular injury during AKI remain elusive. In addition to regulating autophagy, recent studies have found that VMP1 displays non-autophagic functions in modulating protein trafficking, linking lipid metabolism and lipoprotein secretion to cell signaling and certain chronic diseases [27–30]. Mutation or loss of Vmp1 has been shown to accumulate LC3-positive autophagosomes and lipid droplets (LDs) [27] or neutral lipids within lipid bilayers of the ER membranes [29], consistent with defects in autophagy and lipid metabolism. The study by Wang et al. also demonstrates that depletion of Vmp1 in pancreatic acinar cells triggers spontaneous pancreatitis in mice via causing aberrant endoplasmic reticulum (ER) structures [31]. However, the autophagy-independent roles of VMP1 in modulating AKI development remain unclear.

In the present study, we demonstrate that VMP1 is predominantly expressed in renal tubular cells, not in other renal cell types, and its level is rapidly and transiently increased in response to AKI. Functionally, VMP1 protects PTECs against AKI through facilitating the autophagic flux and maintaining homeostasis of lipid metabolism in renal tubular cells.

Results

Rapid upregulation of renal tubular VMP1 in response to AKI

Activation of autophagic flux in response to AKI has been reported to play an important role in clearing the damaged components and preserving the self-tolerance of various renal cells [21]. To validate this, we analyzed gene expression profiles in GSE145085 and GSE30718 datasets related to human AKI from the Gene Expression Omnibus (GEO) database. As shown in Figure 1A, differentially expressed genes (DEGs) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses in the GSE145085 dataset indicated that the alterations of autophagy-related biology processes, particularly protein processing in endoplasmic reticulum and lysosome, are tightly associated with AKI. Analyzing 222 autophagy-related genes from the Human Autophagy Database and all DEGs in the GSE30718 dataset, we identified 21 AKI-DEGs that are related to autophagy. The heatmap depicted the differential expression of these AKI-DEGs between the AKI and control groups (Figure 1B). As shown, VMP1 was one of the most upregulated autophagic genes in the kidney of AKI patients compared to the healthy controls, arguing that the autophagic gene VMP1 may play a critical role in modulating kidney function in AKI. To explore the potential role of VMP1 in renal tissue damage under various stressful challenges, we compared its expression in kidney tissue from AKI or CKD patients with control subjects. In the experiment, renal PTECs were labeled with AQP1 antibody and glomerular cells were labeled with SYNPO (synaptopodin) antibody, respectively. As shown in Figure 1C, VMP1 was predominantly expressed in renal PTECs specifically, but not in other renal cells such as podocytes. In AKI patients, renal tubular VMP1 level was significantly higher than that in healthy control subjects. To our surprise, the level of renal tubular VMP1 in CKD patients was not markedly increased compared to that in normal renal tubules (Figure 1C and Figure S1A). We further scored VMP1 immunofluorescence staining of kidney tissue from patients with AKI and CKD and found a negative correlation between VMP1 level and the degree of renal tubular injury, suggesting that the progression of AKI to CKD is associated with VMP1 reduction in PTECs (Figure 1D). Data analyses also showed VMP1 upregulation in cisplatin-induced pluripotent stem cell-derived human kidney organoids and kidney tissue from AKI patients (Figure S1, B and C).

Figure 1.

VMP1 is rapidly and transiently induced in acute kidney injury. (A) KEGG enrichment analyses of the differentially expressed genes (DEGs) in the dataset GSE145085 revealed that the cellular autophagy process was strongly involved in AKI. (B) RNA sequencing analyses of the DEGs associated with autophagy from AKI dataset GSE30718. (C) Immunostaining of VMP1 in kidney tissue from AKI, CKD patients, and controls. AQP1 and SYNPO served as proximal renal tubule and glomerular protein markers, respectively. Scale bar: 200 μm. (D) VMP1 level (12 samples in each group) in renal tubule and glomerulus (top) and Pearson’s correlation of renal VMP1 level and tubular injury in patients (bottom). Data were analyzed by one-way ANOVA. (E) Immunofluorescence staining of VMP1 and AQP1 in the mouse cisplatin-induced AKI model. (F) Immunofluorescence staining of VMP1 and AQP1 in the mouse IRI-induced AKI model. In E and F, data from 6 mice per group were analyzed by unpaired t-test. Scale bar: 100 μm. (G) VMP1 level of mice treated with cisplatin analyzed by western blotting. (H) VMP1 level in mice treated with IRI analyzed by western blotting. In G and H, data from 3 biological replicates were analyzed using one-way ANOVA. Data were presented as group means ± SEM. * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001.

We next examined the alteration of renal tubular VMP1 expression in experimental AKI mouse models. Mouse models of AKI were induced by cisplatin treatment or unilateral renal IRI procedure, respectively. Immunofluorescence labeling showed that VMP1 expression was markedly increased in mouse renal tubules within one day after cisplatin treatment (Figure 1E) or IRI procedure (Figure 1F). We also monitored the dynamic change of VMP1 level in mouse renal tubules following kidney injury. As shown, cisplatin (Figure 1G) and IRI procedure (Figure 1H) both rapidly increased mouse renal tubular VMP1 expression. The result suggests that under AKI conditions, VMP1-associated autophagy is activated. Significant upregulation of Vmp1 in mouse kidney tissue after treatment with cisplatin, lipopolysaccharides (LPS), or IRI was validated by mRNA sequencing analyses in GSE240304 and GSE98622 datasets (Figure S1, D and E). However, VMP1 level in mouse renal tubules was gradually decreased after 3 days of cisplatin injection (Figure 1G) or one-week post-IRI procedure (Figure 1H), suggesting that upregulation of renal tubular VMP1 in AKI is a transient event. This observation is in line with the finding in CKD patients whose renal tubular VMP1 level is not higher than that in normal controls (Figure 1C). Given the physical and functional correlation between VMP1 and TMEM41B [32], we also explored the expression of TMEM41B in AKI and CKD. As shown in Figure S2 A-B, database analyses and immunofluorescence staining revealed no significant differences in TMEM41B expression within renal tissues of AKI or CKD patients. However, in cisplatin-induced AKI models, TMEM41B displayed a downward trend in renal tubules by 2-day post-induction (Figure S2, C, E and G). No notable change of TMEM41B expression was observed during the early phase of IRI-induced AKI (Figure S2, D and F), while TMEM41B expression was increased in tubular cells one-day post-IRI procedure (Figure S2H).

Renal proximal tubule-specific vmp1-knockout mice display severe tubular damage

To explore the function of VMP1 in renal tubules during AKI, we generated renal proximal tubule-specific vmp1-knockout mice (vmp1-cKO) using the Cre-LoxP system with CRISPR-Cas9 technology. Genotyping, immunofluorescence staining, and western blotting analysis all confirmed that the Vmp1 was specifically knocked out in the renal proximal tubules of vmp1-cKO mice (Figure 2A). Although there were no differences observed between WT and vmp1-cKO mice at birth and during early development, vmp1-cKO mice were found to weigh significantly less than their WT littermates after 4 months (Figure 2B, Figure S3A). At 4 months of age, vmp1-cKO mice exhibited reduced 24-hour urine output compared to the WT group (Figure 2C). Furthermore, urinary calcium levels in vmp1-cKO mice older than 4 months showed a continuous increase with advancing age (Figure 2D, Figure S3B). We examined mouse PTECs’ functions by monitoring the serum creatinine (Scr) and blood urea nitrogen (BUN) levels in both vmp1-cKO mice and WT littermates at 6 months post-birth. As shown in Figure 2D, serum calcium level in vmp1-cKO mice was significantly lower than that in their WT littermates, whereas levels of Scr and BUN in vmp1-cKO mice were markedly higher than those in WT littermates. These results indicated a defect in the kidney development of vmp1-cKO mice.

Figure 2.

Proximal renal tubule-specific vmp1-knockout (vmp1-cKO) mice displayed impaired renal tubule injury. (A) Genotype identification (upper left), western blotting analysis (upper right), and immunofluorescence labeling (lower panel) of VMP1 in WT and vmp1-cKO mouse renal tubules (6 mice per group). Scale bar: 50 μm. (B, C) Mouse body weight and 24-hour urine volume monitoring of WT and vmp1-cKO mice. (D) Ratio of urine calcium to urinary creatinine in WT and vmp1-cKO mice and levels of serum calcium, scr, and BUN in WT and vmp1-cKO mice. In B-D, data from 6 mice per group were analyzed with Mann Whitney test. (E) H&E and PAS staining in kidney tissue sections from WT and vmp1-cKO mice. Data from 6 mice per group were analyzed with one-way ANOVA. Scale bar: 100 μm. (F) TUNEL staining in kidney tissue sections from WT and vmp1-cKO mice. Scale bar: 100 μm. (G) TEM image of mitochondrial damage in renal tubule from WT and vmp1-cKO mice. At least 10 fields of view for each mouse were selected for statistics, and quantification of mitochondrial damage was based on the percentage of the number of damaged mitochondria to the total mitochondrial population. Scale bar: 5 μm. (H) BMD scanning of WT and vmp1-cKO mice. (I) Oil red O staining images of WT and vmp1-cKO mice. Scale bar: 100 μm. In F-I, data from 6 mice per group were analyzed with Mann Whitney test. Data were presented as group means ± SEM. ** p < 0.01, and **** p < 0.0001.

Renal tissue examination by hematoxylin-eosin (H&E) and periodic acid-schiff (PAS) staining confirmed markedly higher renal injury scores in vmp1-cKO mice compared to WT mice after 4 months (Figure 2E, Figure S3C). Terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining of 6-month mouse kidney sections also showed that Vmp1 deficiency resulted in significant apoptosis in renal tubules (Figure 2F). We further analyzed the kidney injury in 6-month-old vmp1-cKO mice and their WT littermates at an ultrastructural level. As shown in Figure 2G, transmission electron microscopy (TEM) showed considerable damage in the mitochondria of vmp1-cKO mice but not their WT littermates. Given that kidney PTECs are highly specialized for reabsorption [33], their damage can lead to a defect in the reabsorption of calcium and other substances. In line with this, bone mineral density (BMD) scans showed that vmp1-cKO mice displayed a rickety bone structure and severely reduced bone mineral density compared to their WT littermates (Figure 2H). We did notice that the phenotype observed in vmp1-cKO mice was more severe than that reported in atg5 or atg7 knockout mice, and this prompted us to postulate the involvement of autophagy-independent role of VMP1 in maintaining renal tubular function. As VMP1 has been shown to play a significant role in modulating lipid metabolism, such as the formation of LDs [28], we next performed Oil Red O staining of LDs in kidney tissue from WT and vmp1-cKO mice. As shown in Figure 2I, a considerable number of LDs were detected in the renal tubules of 6-month-old vmp1-cKO mice but not their WT littermates. Taken together, these results suggest that Vmp1 deficiency in renal tubules causes the loss of their protection against various stress challenges, leading to the gradual damage of renal tubules and the defect of renal reabsorption function.

Protective role of renal tubular VMP1 against AKI

To further explore the protective role of renal tubular VMP1 in AKI, we established AKI mouse models induced by cisplatin or IRI procedure with six-week-old vmp1-cKO mice and WT mice. As expected, vmp1-cKO mice at six weeks of age displayed no apparent defects in kidney morphology or function. However, when treated with cisplatin, vmp1-cKO mice displayed higher levels of Scr and BUN (Figure 3A). ADGRE1/F4/80 immunofluorescence staining of mouse kidney sections found that Vmp1 deficiency resulted in a higher level of infiltrated inflammatory macrophages (Figure 3B). Kidney tissue section staining (H&E and PAS) and TUNEL labeling showed that Vmp1-deficient renal tubules had more severe renal tubular injuries and apoptosis compared to their WT littermates following cisplatin treatment (Figure 3C). Similarly, significantly higher levels of Scr and BUN were detected in vmp1-cKO mice compared to their WT littermates following IRI procedure (Figure 3D). Kidney tissue section staining also showed that IRI procedure caused a higher level of infiltrated inflammatory macrophages (Figure 3E), as well as more severe renal tubular injuries and apoptosis (Figure 3F) in vmp1-cKO mouse renal tubules compared to those in their WT littermates. These results confirmed the protective role of renal tubular VMP1 in both cisplatin and IRI-induced AKI.

Figure 3.

Vmp1 deficiency aggravates renal tubular injury in cisplatin and IRI AKI models. (A) Scr and BUN levels in cisplatin-induced AKI mice. (B) Immunofluorescence staining of ADGRE1/F4/80 in cisplatin-induced AKI mouse kidneys. (C) H&E, Pas, and TUNEL staining in mouse kidneys with cisplatin-induced AKI. (D) Scr and BUN levels in mice with ongoing IRI AKI. (E) Immunofluorescence staining of ADGRE1 in mouse kidneys with IRI AKI. (F) H&E, Pas, and TUNEL staining in mouse kidneys with IRI AKI. Scale bars: 100 µm. At least 6 mice in each group were analyzed. Data were analyzed using the mann whitney test and presented as group means ± SEM. ** p < 0.01.

Loss of mouse renal tubular Vmp1 impairs autophagy and induces the formation of LDs

Given that VMP1 is essential for the formation of autophagosomes [34], which is positively regulated in PTECs during AKI (Figure S4), we constructed AKI models in C57BL/6J mice induced by cisplatin or IRI procedure. As shown by TEM images (Figure S4A-B), both AKI models promoted the formation of autophagosomes in renal PTECs. Immunofluorescence staining of mouse kidney tissues also showed the presence of LC3-labeled autophagosomes in PTECs (Figure S4C-F).

We next examined whether Vmp1 deficiency prevents autophagic degradation in PTECs. Immunoblot analyses showed elevated levels of LC3-II as well as SQSTM1 proteins in vmp1-cKO mice compared to WT littermates (Figure 4A,E), indicating impaired autophagic flux in PTECs of vmp1-cKO mice. In WT mouse PTECs treated with cisplatin or IRI, LC3 formed distinct puncta, which were separated from the ER labeled by anti-SEC61B/SEC61β antibody. This text is a description of Figure 4B and Figure 4F. Please correct it to: In contrast, LC3 was largely colocalized with the SEC61B-positive ER in vmp1-cKO PTECs treated with cisplatin or IRI (Figure 4B,F). In addition, we found that the number of autophagic vesicles formed in vmp1-cKO PTECs induced by cisplatin or IRI was significantly reduced (Figure 4C,G). Meanwhile, an accumulation of LDs in renal PTECs of vmp1-cKO mice, evidenced by Oil Red O staining, was observed (Figure 4D,H).

Figure 4.

Vmp1 deficiency induces autophagy flux blockade and lipid droplets accumulation in the proximal renal tubule in AKI mouse models. (A) Western blotting analyses in cisplatin-induced renal tubular tissues. (B) Immunofluorescence staining of LC3 and ER (SEC61B) in cisplatin-induced renal tubular tissue. Scale bars: 20 µm. (C) TEM image of proximal renal tubule from cisplatin-induced WT and vmp1-cKO mice. Scale bars: 5 µm. Autophagosomes and autolysosomes were indicated by blue and yellow arrowheads, respectively. (D) Oil red O staining images of cisplatin-induced WT and vmp1-cKO mice. Scale bars: 100 µm. (E) Western blotting analyses in IRI-treated renal tubular tissues. (F) Immunofluorescence staining of LC3 and ER in IRI-treated renal tubular tissue. Scale bars: 20 µm. (G) TEM image of proximal renal tubule from IRI-induced WT and vmp1-cKO mice. Scale bars: 5 µm. Autophagosomes and autolysosomes were indicated by blue and yellow arrowheads, respectively. (H) Oil red O staining images of IRI-induced WT and vmp1-cKO mice. Scale bars: 100 µm. Data from 6 mice per group were analyzed by Mann Whitney test and presented as group means ± SEM. ** p < 0.01.

VMP1 protects renal tubular cells via enhancing autophagic flow and reducing the accumulation of LDs

We next examined whether the protective function of VMP1 against AKI depends on VMP1-mediated autophagic flux or the accumulation of LDs. In these experiments, we isolated primary PTECs from C57BL6/J mice and treated cells overnight with cisplatin. As shown in Figure 5A, VMP1 protein level was markedly increased. A similar result was observed in human renal tubular epithelial HK-2 cells (Figure S5A). Subjecting HK-2 cells to rapamycin or starvation conditions also induced VMP1 expression (Figure S5, B and C). We silenced VMP1 in HK-2 cells via VMP1-specific siRNA (Figure S5D). As shown in Figure 5B, the depletion of VMP1 resulted in an abnormal accumulation of SQSTM1 and LC3. In contrast, VMP1 depletion did not affect the protein expression of ATG5 and ATG7; therefore, we speculated that the accumulation of LC3 was not caused by enhanced autophagosome synthesis but rather the failure of complete autophagic flux, which was unable to generate completely mature autophagic vesicles (Figure 5B). To validate this notion, intact natural autophagic vesicles were purified by FACS-sorting method [35], and the results showed that silencing of VMP1 interfered with autophagosome formation (Figure 5C-D). VMP1 has also been reported to be involved in the regulation of the unfolded protein response in the endoplasmic reticulum [36]. We next tested the impact of VMP1 in renal tubular cells integrated stress responses. As shown in Figure 5E, silencing of VMP1 indeed activated the EIF2AK3/PERK-EIF2A/EIF2α signaling pathway, leading to upregulation of the downstream ATF4 expression.

Figure 5.

VMP1 promotes autophagosome formation and protects renal tubular cells against cisplatin-induced apoptosis. (A) The protein level of VMP1 in mice PTECs treated with cisplatin for 12 h was analyzed by western blotting. Data from 3 mice per group were analyzed by one-way ANOVA. (B) Western blotting analyses in VMP1-knockdown HK-2 cells. (C) Western blotting analyses of autophagosomes isolated using LC3 antibody. Autophagic vesicles (AV) were represented with total lysate (TL). In B and C, data from 3 biological replicates were analyzed. (D) Quantification of fluorophore-labeled events in VMP1-knockdown and control HK-2 cells. The percentages represent the relative number of detected events in three independent experiments analyzed by unpaired t-test. (E) Western blotting analyses in VMP1 knockdown HK-2 cells. (F) Immunofluorescence staining of SQSTM1 and LC3 in VMP1-knockdown HK-2 cells treated with or without cisplatin for 12 h. Scale bars: 50 µm. (G) Immunofluorescence staining of LDs (BODIPY 493, green) and ER tracker (red). Scale bars: 10 µm. In F and G, data from three independent experiments (at least 20 cells in each experiment were counted) were analyzed by unpaired t-test. (H) KEGG pathway analyses of genes combined with VMP1. (I) Apoptosis evaluation of HK-2 cells subjected to cisplatin treatment for 12 h. Data from 3 biological replicates were analyzed by unpaired t-test. The results were presented as group means ± SEM. *p < 0.05, **p < 0.01, and *** p < 0.001.

In addition, a higher accumulation of LC3 and SQSTM1 proteins in the cytoplasm was observed in those lacking VMP1 compared to the control group with or without cisplatin (Figure 5F). Significantly more LDs were observed in HK-2 cells lacking VMP1 compared to control HK-2 cells with or without cisplatin treatment (Figure 5G). Furthermore, we transfected HK-2 cells with the FLAG-VMP1 plasmid and then performed RNA sequencing following the immunoprecipitation using FLAG antibody. Both KEGG pathway and Gene Ontology (GO) analyses showed that VMP1 was closely associated with the apoptotic process (Figure 5H, Figure S5E). Flow cytometry assay also showed that VMP1 depletion markedly enhanced the apoptosis in HK-2 cells treated with cisplatin (Figure 5I).

The observation in HK-2 cells was further validated in renal PTECs isolated from vmp1-cKO mice and their WT littermates. PTECs were treated with cisplatin, and the levels of reactive oxygen species (ROS) and apoptosis were examined, respectively. As shown in Figure 6A, cisplatin treatment generated higher ROS levels in vmp1-cKO PTECs than in WT PTECs. We performed western blotting and observed the accumulation of LC3 and SQSTM1 proteins in PTECs lacking Vmp1 (Figure 6B). Like in HK-2 cells, a considerable accumulation of LC3 and SQSTM1 proteins was observed in the cytoplasm of vmp1-cKO PTECs treated with or without cisplatin (Figure 6C). As increased SQSTM1 can lead to NFE2L2/NRF2 activation, as observed in pancreatic acinar cells of vmp1-cKO mice [31], we also examined the expression of NFE2L2 in PTECs after cisplatin induction. Both qPCR analysis and immunoblotting results showed that Vmp1 deficiency enhanced the upregulation of the NFE2L2 and its driven genes (Figure S6). Western blotting assay also indicated that Vmp1 deficiency activated the EIF2AK3/PERK-EIF2A signaling pathway (Figure 6D). Similarly, Vmp1 deficiency in PTECs enhanced the formation of LDs (Figure 6E,F) and also caused more apoptosis in cisplatin-treated renal PTECs (Figure 6G). These results collectively suggest that both autophagic flux formation and lipid metabolism are defective in VMP1-deficient PTECs.

Figure 6.

Functional analyses of primary renal tubular cells in vmp1-cKO mice. (A) ROS level of PTECs treated with cisplatin. 5 biological replicates per condition were analyzed. (B) Western blotting analyses of PTECs from WT and vmp1-cKO mice. (C) Immunofluorescence staining of LC3 and SQSTM1 in PTECs from WT and vmp1-cKO mice with or without cisplatin. Scale bars: 5 µm. (D) Western blotting analyses of PTECs from WT and vmp1-cKO mice. In B-D, 3 biological replicates per condition were analyzed. (E) Immunofluorescence staining of LDs (BODIPY 493, green) and ER tracker (red) in PTECs from WT and vmp1-cKO mice with or without cisplatin. Scale bars: 10 µm. (F) Quantitative analyses of LDs. Data from three independent experiments (at least 20 cells in each experiment were counted) were analyzed. (G) Apoptosis level of PTECs treated with cisplatin. 5 biological replicates per condition were analyzed. The results were analyzed by unpaired t-test and presented as group means ± SEM. **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Increasing Vmp1 expression in mouse renal tubules mitigates AKI

Given that renal tubular VMP1 plays a critical role in facilitating autophagic flux and LDs accumulation, we next tested whether enhancement of Vmp1 expression can protect renal tubular cells against AKI. In this experiment, AAV-mediated Vmp1 expression system was employed. We isolated PTECs and analyzed VMP1 levels by western blotting. As shown in Figure S7A, Vmp1-AAV strongly elevated VMP1 protein level. Six-week-old mice kidneys were in situ injected with Vmp1-AAV or Ctrl-AAV followed by cisplatin treatment (Figure 7A). Mice were sacrificed two weeks after cisplatin treatment, and plasma and kidney samples were collected for further assessment. As shown, Vmp1-AAV injection markedly reduced the levels of cisplatin-induced serum creatinine (Figure 7B) and BUN (Figure 7B) compared to those in the Ctrl-AAV injection group. H&E and PAS staining confirmed that Vmp1 overexpression was able to mitigate cisplatin-induced AKI (Figure 7C). Immunoblotting results showed that Vmp1-AAV was able to alleviate the cisplatin-induced impairment of autophagic flux compared with the control group (Figure 7D). As shown in Figure 7E-F an increase of Vmp1 level was able to reduce the accumulation of cisplatin-induced LDs as observed by both TEM and Oil Red O staining. Interestingly, in situ injection of Vmp1-AAV in WT mice without cisplatin treatment did not affect the number of autophagic vesicles and the accumulation of LDs (Figure S7B-C), arguing that the role of VMP1 in modulating mouse renal tubular cell autophagy and lipid metabolism may tightly link to stress conditions. A similar method of Vmp1-AAV or Ctrl-AAV injection was used in the IRI mouse model (Figure 7G). Vmp1-AAV administration also strongly reduced Scr and BUN levels in IRI mice compared with Ctrl-AAV (Figure 7H). H&E and PAS staining of mouse kidney tissue sections further confirmed that Vmp1-AAV treatment significantly reduced renal tubular injury in IRI mice (Figure 7I). The results of immunoblotting, TEM and Oil Red O staining all showed that Vmp1 overexpression largely restored autophagic flux and reduced the accumulation of LDs in PTECs from IRI-treated mice (Figure 7J-L). These results are in line with the notion that increasing Vmp1 expression in mouse renal tubules effectively mitigates AKI induced by cisplatin or IRI procedure.

Figure 7.

Increasing Vmp1 expression in mouse renal tubules attenuated tubular injury. (A) Schematic of experimental approach in cisplatin model with control or Vmp1-AAV infection. (B) Scr and BUN levels in mice with cisplatin-induced AKI. (C) H&E and PAS staining of renal tubules in cisplatin-treated mice with control or Vmp1-AAV infection. Scale bars: 100 µm. In B and C, data from 5 mice per group were analyzed using the Mann Whitney test. (D) Western blotting analyses of cisplatin-treated renal tubular tissues (3 mice per group). (E) TEM image of proximal renal tubule from cisplatin-treated mice. LDs were indicated by orange arrowheads. Scale bars: 5 µm. (F) Oil red O staining images of cisplatin-treated mice. Scale bars: 100 µm. In E and F, data from 6 mice per group were analyzed using the unpaired t-test. (G) Schematic of experimental approach in IRI model with control or Vmp1-AAV infection. (H) Scr and BUN levels in mice with IRI-induced AKI. (I) H&E and PAS staining of renal tubules in IRI-treated mice with control or Vmp1-AAV infection. Scale bars: 100 µm. In H and I, data from 5 mice per group were analyzed using the Mann Whitney test. (J) Western blotting analyses of IRI-treated renal tubular tissues (3 mice per group). (K) TEM image of proximal renal tubule from IRI-treated mice. LDs were indicated by orange arrowheads. Scale bars: 5 µm. (L) Oil red O staining images of IRI-treated mice. Scale bars: 100 µm. In K and L, data from 6 mice per group were analyzed using the unpaired t-test. Data were presented as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Discussion

In this study, we unveil the protective role of renal tubular VMP1 within the context of AKI. First, VMP1 expression in human and mouse renal tubules is rapidly increased in response to AKI challenge, and this upregulation of renal tubular VMP1 is transient and not detected in CKD patients. Second, 6-month-old renal PTECs-specific vmp1-cKO mice naturally develop defective calcium metabolism and display severe tubular damage. Challenged by cisplatin or IRI procedure, vmp1-cKO mice also show more severe renal injuries compared to their WT littermates. Supporting the protective role of VMP1, over-expressing Vmp1 in renal tubular cells via Vmp1-AAV strongly mitigates acute renal tubular injury induced by cisplatin or IRI procedure. The mechanistic study further demonstrates that VMP1 protects renal tubules during AKI via facilitating cellular autophagic flux and lipid metabolism.

As a highly conserved catabolic pathway in eukaryotic cells, autophagy contributes to the maintenance of homeostasis and function of the kidney. Although previous studies have shown that autophagy is activated in renal tubular cells to act as an intrinsic protective mechanism, the molecular basis that governs autophagic flux in renal tubular cells remains unclear. Through dataset analyses of kidney tissues from healthy or AKI patients, we demonstrate that autophagy is a key event in renal tubules during AKI. This observation is validated by mouse models of experimental AKI. Data analyses have further identified VMP1 as a major player in modulating renal tubular autophagic flux during AKI. This finding is in agreement with previous reports showing that VMP1, like other scramblases such as TMEM41B and ATG9, is involved in the autophagic processes, including the formation of autophagosomes and the subsequent fusion with lysosomes [37–39]. Double staining of both human and mouse kidney tissue sections indicates that the expression pattern of VMP1 is tightly associated with autophagy marker proteins such as LC3 and SQSTM1, and depletion of renal tubular cell VMP1 impairs the formation of autophagosomes and fusion of autophagosomes with lysosomes. Supporting this, although vmp1-cKO mice show no apparent phenotype compared with their WT littermates at birth, they gradually develop severe tubular injury when they are aging, evidenced by an impaired renal reabsorption function, such as an abnormal level of serum calcium. Challenged by cisplatin or IRI procedure, vmp1-cKO mice also show more severe acute renal injuries compared to their WT littermates. In contrast, maintaining a high level of Vmp1 in mouse renal tubules through renal in situ injection of Vmp1-AAV significantly mitigates kidney injuries in mice treated with cisplatin or IRI procedure. These results collectively demonstrate the protective function of renal tubular VMP1 against AKI via facilitating autophagic flux.

It is worth noting that the organ damage observed in vmp1-cKO AKI mouse models is more severe than that seen in atg5 or atg7 knockouts, arguing that VMP1 May have autophagy-independent functions that contribute to its protective role. Indeed, an elegant study by Jiang et al. showed that deletion of Vmp1 in mouse livers severely impairs hepatic lipoprotein secretion, resulting in rapid development of NASH features in mice [28]. Supporting broader cellular functions of VMP1, recent studies also reported non-autophagic roles of VMP1 in modulating lipid metabolism (e.g., LD accumulation) and organelle homeostasis [27,29–31]. In the present study, we found that VMP1 deficiency in renal PTECs not only impaired cellular autophagic flux but also induced the formation of LDs. Given that abnormal lipids accumulation in kidneys particularly mitochondria-enriched proximal renal tubular epithelial cells has been well-documented to significantly disrupt renal lipid metabolism, contributing to ischemia-reperfusion acute kidney injury and other renal dysfunctions [40], loss of Vmp1 in renal tubular cells may aggravate acute kidney injury induced by cisplatin treatment or IRI procedure through disrupting the cellular lipid metabolism. Taken together, our results suggest that both autophagic flux and lipid metabolism are defective in VMP1-deficient PTECs.

Our data show that renal tubular VMP1 is rapidly upregulated in AKI, but this induction is transient. The high level of VMP1 protein in mouse renal tubules shortly after AKI is gradually decreased along the course of kidney injury, and this is also in line with our observation that tubular VMP1 level is not increased in CKD patients. The rapid and transient induction of VMP1 May fit into its intrinsic protective mechanism in renal tubules during AKI. Altered VMP1 expression has recently been observed in the context of other diseases, including multiple sclerosis and Parkinson disease, in which VMP1 plays a role in modulating NLRP3 inflammasome activation and mitochondrial dysfunction [41]. The factor that upregulates VMP1 in the renal tubules during AKI, however, remains unknown. Our recent study has shown that inflammatory cytokines are involved in modulating protein expression of renal tubular protein PNPT1 [42], future study thus is required to explore whether AKI-associated cytokines play a role in regulation of VMP1. A previous study by Liu et al. showed that VMP1 protein expression, as well as its effects on myoblast proliferation, autophagy, and apoptosis, could be regulated by Mir124a [43]. However, it remains unknown whether Mir124a is involved in renal tubular VMP1 upregulation during AKI. In summary, this study reveals that renal tubular VMP1 mitigates AKI through facilitating autophagic flux in renal tubules and provides a novel intervention strategy based on renal tubular VMP1 for AKI patients.

Materials and methods

Human kidney biospecimens

All research protocols involving human biological samples in this study followed the requirements of the World Medical Association Declaration of Helsinki guidelines and were approved by the Ethics Committee of Jinling Hospital, School of Medicine, Nanjing University (2019NZKYKS–008–01). A written informed consent form was signed by each patient who provided the sample. We obtained renal biopsy samples from patients with AKI or chronic kidney disease, respectively, and used renal paracarcinoma tissues from renal clear cell carcinoma without tubular injury as the control.

Mouse strain and genotyping

All animal experiments in this study were performed in strict accordance with the Guidelines for the Care and Use of Laboratory Animals developed by the National Research Council and approved by the Nanjing University Laboratory Animal Welfare Ethics Review Committee (IACUC-2212002). Eight-week-old wild-type male C57BL/6J mice were purchased from Beijing Viton Lihua Laboratory Animal Technology Company. To directly assess the role of VMP1 in AKI-associated autophagy, we constructed renal proximal tubule-specific vmp1-knockout mice. Vmp1^flox transgenic mice (C57BL/6J genetic background) produced based on CRISPR-Cas9 technology were kindly provided by Dr Weidong Le (The First Affiliated Hospital of Dalian Medical University, Dalian, China). These mice were crossed with mice expressing Ggt-Cre (Jackson Laboratory 036,773), yielding Vmp1^flox/WT, Ggt-Cre^± mice. These mice crossed with each other to generate Vmp1^flox/WT, Ggt-Cre^± mice, and Vmp1^flox/flox, Ggt-Cre^−/− mice. Finally, proximal tubule-specific vmp1- knockout mice (cKO; gene type: vmp1^flox/flox, Ggt-Cre^±) were generated by crossbreeding Vmp1^flox/WT, Ggt-Cre^± mice, and Vmp1^flox/flox, Ggt-Cre^−/− mice. Vmp1^flox/flox, Ggt-Cre^−/−mice (WT) from the same breeding colony were used as controls. Genotypes were identified by polymerase chain reaction (PCR) and agarose gel electrophoresis using primers shown in Table S1. All mice were housed in a temperature-controlled room (22.0 ± 0.2°C) with a 12 h light-12 h dark cycle, had free access to water, and ad libitum consumption of standard rodent chow.

Mouse kidney disease models

To investigate the role of VMP1 in AKI, two models of kidney disease were constructed according to our previous study [44]. Briefly, the mice were injected intraperitoneally with cisplatin (Sigma-Aldrich, P4394) at a dose of 20 mg/kg or 15 mg/kg, and sacrificed at the designated time. The same volume of saline (Beyotime Biotechnology, ST341) was used as a vehicle control. Unilateral IRI-induced AKI was performed using a previously established method [44]. To overexpress Vmp1 in renal tubular cells, renal in situ injection of Vmp1-expressing adeno-associated virus (Vmp1-AAV) (OBiO Technology) was performed. Fifty µL Vmp1-AAV or control adeno-associated virus (Ctrl-AAV) (2 × 10¹² vg/mL) were injected into the mouse kidney during the establishment of the cisplatin or IRI AKI mouse model.

Collection of urine, blood, and kidney samples

The animals were kept in individual mouse metabolic cages (Tecniplast, 3600M021) to collect mouse urine. Mice were anesthetized with isoflurane (RWD Life Science, R510–22–10), blood was drawn through the retro-orbital venous sinus and collected in heparinized tubes (Thomas Scientific 367,871). Heparinized plasma was separated and used for biochemical assays. Levels of calcium (Sigma-Aldrich, MAK47), creatinine (Bioassay System, DICT-500), or blood urea nitrogen (Invitrogen, EIABUN) were assayed in urine or plasma, respectively. Mouse kidneys were isolated and sectioned for histological and immunohistochemical studies. The rest of the kidney tissue was snap-frozen in liquid nitrogen and stored at − 80°C for RNA or protein extraction.

Renal histology and immunohistochemical staining

Renal tissues were embedded in paraffin and stained with hematoxylin and eosin (H&E), periodic acid-Schiff (PAS), or TdT-mediated dUTP Nick-End Labeling (TUNEL) in 4-μm-thick sections and blocked in permanent sealer according to previously reported protocols [44]. All sections were captured using a whole slide scanning microscope (Olympus, VS200), and histological scoring of renal tubular injury and interstitial inflammation was performed in a double-blind method. To determine renal tubular injury, defined as loss of brush border, compensatory tubular dilation and detachment, tubular apoptosis, and cellular tubular pattern, a semi-quantitative scoring method was used. At least 12 randomly selected regions were scored under the microscope for each mouse tissue section, and the average score was calculated.

Oil red O staining

A stock solution of Oil Red O (Sigma-Aldrich, O0625) was prepared by dissolving 50 mg/mL in isopropanol (Sigma-Aldrich 34,863). For the working solution, 30 mL of stock stain was mixed with 20 mL of distilled water. Freshly cut 6 μm-thick frozen kidney sections were immersed in the Oil Red O working solution for 20 min at room temperature, followed by a 5-min rinse under running water. After hematoxylin counterstaining, sections were briefly washed.

Transmission electron microscopy

After obtaining kidney tissues, the tissues were immediately fixed in 2.5% glutaraldehyde (Absin, abs9277) for transmission electron microscopy (TEM) analyses. Sections were fitted on a copper grid (EMCN, ASH150) and photographed for analyses under a Hitachi 7500 transmission electron microscope (Japan). For autophagy level quantification, a minimum of 10 randomly TEM fields per experimental condition underwent systematic blinded analyses. Autophagosome structures were identified through ultrastructural examination of characteristic double-membrane topology, with dimensional parameters constrained to 300–900 nm in diameter.

Cell culture and in vitro interventions

HK-2 cell culture and primary renal tubular cell extraction were performed according to previously published methods [44]. Cellular autophagy was induced by using cisplatin, rapamycin (Sigma-Aldrich 553,211) or Earle’s Balanced Salt Solution (EBSS; Sigma-Aldrich, E2888). To silence VMP1 in HK-2 cells, we transfected VMP1 siRNA into HK-2 cells utilizing Lipofectamine RNAiMAX transfection reagent (Invitrogen 13,778,100). Sequences of siRNA are provided in Table S1. FuGENE HD transfection reagent (Promega, E2311) was used to overexpress the FLAG-VMP1 plasmid (Genscript 117,416) in HK-2 cells for RNA immunoprecipitation assay.

Isolation of autophagosomes

HK-2 cells were cultured overnight with 20 µM cisplatin and then exposed to 10 nM bafilomycin A₁ (APEXBIO, A8627) for 2 h. Cells were collected into PBS (Beyotime Biotechnology, C0221A) containing EDTA‐free protease inhibitors (Beyotime Biotechnology, P1045) with trypsin-EDTA (Thermo Scientific 15,400,054). Isolation of autophagic vesicles was performed according to the method reported previously [35]. Briefly, the collected cell precipitates were sonicated (Sonics materials, VCX 130) and subjected to ultracentrifugation (Beckman Coulter, OPTIMA XPN100) at 150,000 ×g at 4°C for 30 min. Autophagic vesicles were collected and incubated with LC3 primary antibody (MBL Life Science, PM036) for 30 min on ice before adding the corresponding fluorescent secondary antibody (Invitrogen, A-21206) for 1 h. Finally, positive signal sorting was performed by BD FACSAria II SORP (BD Biosciences, USA).

Immunofluorescence staining

Cells were treated as described, washed with PBS, and fixed with 4% paraformaldehyde (Beyotime Biotechnology, P0099) for 15 min at room temperature before permeabilization (Sigma-Aldrich, X100) and blocking (Invitrogen, R37624). Next, the cells were incubated overnight at 4°C using primary antibody and labeled for 1 h at room temperature using the corresponding fluorescent secondary antibody. Cell nuclei were labeled using DAPI (Thermo Scientific 62,248). Finally, images were recorded and analyzed with a laser confocal microscope (Zeiss, LSM 880). Fluorescence intensity was quantified based on average fluorescence intensity values using ImageJ software. The antibodies utilized are listed in Table 1.

Table 1.

Antibodies used in this study.

Antibodies	SOURCE	IDENTIFIER
Anti-VMP1	Cell Signaling Technology	12929S
Anti-VMP1	ABclonal	A15523
Anti-AQP1	Santa Cruz Biotechnology	sc -25,287
Anti-SYNPO	Santa Cruz Biotechnology	sc -21,537
Anti-GAPDH	Santa Cruz Biotechnology	sc -25,778
Anti-LC3B	Medical Biological Laboratories	PM036
Anti-LC3B	Novus Biologicals	NB100–2220
Anti-LC3B	Cell Signaling Technology	83506S
Anti-TMEM41B	Proteintech	29270–1-AP
Anti-ADGRE1/F4/80	Cell Signaling Technology	70076S
Anti-SQSTM1	Abcam	ab56416
Anti-SEC61B	Proteintech	15087–1-AP
Anti-BECN1	Cell Signaling Technology	3738S
Anti-ATG5	Novus Biologicals	NB110–53818
Anti-ATG7	Cell Signaling Technology	2631S
Anti-TOMM20	Cell Signaling Technology	42406S
Anti-GOLGA1	Absin	abs151924
Anti-LAMP1	Proteintech	21997–1-AP
Anti-TUBA/α-TUBULIN	Proteintech	11224–1-AP
Anti-p-EIF2AK3/PERK	Cell Signaling Technology	3179S
Anti-EIF2AK3/PERK	Proteintech	20582–1-AP
Anti-p-EIF2A	Cell Signaling Technology	3398S
Anti-ElF2A	Cell Signaling Technology	5324S
Anti-ATF4	Abcam	ab184909
Anti- p-ERN1/IRE1α	Abcam	Ab124945
Anti-ERN1/IRE1α	Cell Signaling Technology	3294T
Anti-ATF6	Cell Signaling Technology	65880T
Anti-NFE2L2/NRF2	Proteintech	16396–1-AP
Anti-GCLC	Proteintech	12601–1-AP
Anti-GCLM	Proteintech	14241–1-AP
Anti-HMOX1	Proteintech	10701–1-AP
Anti-NQO1	Proteintech	11451–1-AP
Goat anti-Rabbit IgG H&L (HRP)	Abcam	ab6721
Goat anti-Mouse IgG H&L (HRP)	Abcam	ab6789
Donkey anti-Rabbit, Alexa Fluor 488	Invitrogen	A-21206
Donkey anti-Mouse, Alexa Fluor 594	Invitrogen	A-21203
Donkey anti-Rabbit, Alexa Fluor 647	Invitrogen	A-31573
Donkey anti-Goat, Alexa Fluor 647	Invitrogen	A-21447

ER-Tracker and BODIPY co-staining

After cisplatin treatment, cells were washed with PBS and stained with 100 nM ER-Tracker red (Beyotime Biotechnology, C1041S) in pre-warmed buffer at 37°C for 30 min, followed by incubation with 10 μM BODIPY 493/503 (MedChemExpress, HY-D1614) under identical conditions. Nuclei were counterstained with Hoechst 33,342 (Beyotime Biotechnology, C1029) for 10 min. Fluorescent signals were visualized and recorded using a confocal laser scanning microscope (Zeiss, LSM 880) equipped with appropriate filter sets for multiplex detection.

Real-time PCR

RNA from HK-2 cells was extracted with TRIzol (Invitrogen, 15596018CN) and reverse-transcribed to cDNA (Vazyme, R323–01). Real-time quantitative PCR (Roche, Lightcycler 96 Instrument) was performed using ChamQ SYBR qPCR Master Mix (Vazyme, Q321–02). Primers used are listed in Table S1.

RIP sequencing

We transfected HK-2 cells with the FLAG-VMP1 plasmid, followed by immunoprecipitation with FLAG antibody (Cell Signaling Technology, 14793S), which was performed in an RNase-free environment with RNase inhibitors (Promega, N2111). We used fragmentation buffer (Invitrogen, AM8740) to break the enriched RNA into short fragments. The fragmented mRNAs were utilized as templates, and cDNAs were synthesized. cDNA libraries were sequenced by LC Biotech (China) on the Illumina NovaSeq 6000 platform. The sequencing data were submitted to the GEO database (GSE277267).

Immunoblotting

After cells or tissues were lysed on ice with RIPA buffer (Beyotime Biotechnology, P0013B) containing protease inhibitors and phosphatase inhibitors (Beyotime Biotechnology, P1046), proteins were collected, and the total amount of proteins was determined using BCA protein assay kits (Thermo Scientific 23,225). Equal amounts of proteins were separated on the gel by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and the proteins were blocked (Beyotime Biotechnology, P0023B) and incubated with antibodies after transferring them to PVDF membranes (Millipore, IPVH00010). Finally, the proteins were imaged in a chemiluminescent imaging system (Tanon, 5200). The antibodies used are listed in Table 1.

Measurement of ROS

ROS levels in cisplatin-induced renal tubular cells were detected using the ROS fluorescence assay kit (Invitrogen, EEA019). Briefly, after 20 µM cisplatin induction overnight in renal tubular cells, a 10 µM DCFH-DA working solution was added to the cultured cells and incubated at 37°C for 30 min. Cells were washed 3 times with serum-free medium (Gibco, C11330500CP), and the cell precipitates were collected and assayed by flow cytometry (Invitrogen, Attune NxT).

Apoptosis assay

Apoptosis level was determined using the FITC-Annexin V apoptosis detection kit (Invitrogen, V13242) according to the usage guidelines and assayed by flow cytometry (Invitrogen, Attune NxT).

Biological informatics analyses

Select the GEO database (https://www.ncbi.nlm.nih.gov/geo) to search and screen high-throughput sequencing datasets with the keyword “acute kidney injury” in Homo sapiens or mouse organisms. Through the GEO database, we identified GSE145085 and GSE30718 as AKI transcriptome datasets related to Homo sapiens organisms, while GSE240304 and GSE98622 were identified as renal cell datasets from acute kidney injury mouse models. We selected differentially expressed genes (DEGs) with a q-value less than 0.05 for analyses. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed in the DAVID database, and bubble plots were drawn using the ggplot2 package. Autophagy-related genes were obtained from the Human Autophagy Database (HADb; http://www.autophagy.lu/index.html).

Statistical analysis

All data are expressed as mean ± SEM. Comparisons between two independent groups are performed with two-tailed unpaired t-test. Comparisons between multiple groups are analyzed using one-way ANOVA, and post-hoc analyses are performed following the Tukey test. GraphPad Prism 8 software is used for all analyses. p < 0.05 indicates statistical significance.

Funding Statement

This work was supported in part by grants from National Key R&D Program of China (2018YFA0507100, K.Z.), National Natural Science Foundation of China (82170692, K.Z.), Basic Research Program of Jiangsu (BK20243061, K.Z.), and Jiangsu Graduate Student Research Innovation Program (KYCX24_0216, W.Y.).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All datasets produced were accessed in the GEO Series (accession number: GSE277267).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15548627.2025.2533306

Abbreviations

AKI

acute kidney injury

BUN

blood urea nitrogen

CKD

chronic kidney disease

DEGs

differentially expressed genes

GEO

Gene Expression Omnibus

H&E

hematoxylin-eosin

IRI

ischemia-reperfusion injury

KEGG

Kyoto Encyclopedia of Genes and Genomes

LD

lipid droplet

MAP1LC3/LC3

microtubule-associated protein 1 light chain 3

PAS

periodic acid-schiff

PTECs

proximal tubular epithelial cells

ROS

reactive oxygen species

Scr

serum creatinine

SQSTM1

sequestosome 1

TEM

transmission electron microscopy

TUNEL

terminal deoxynucleotidyl transferase dUTP nick-end labeling

VMP1

vacuole membrane protein 1