β-Arrestin-2 enhances endoplasmic reticulum stress-induced glomerular endothelial cell injury by activating transcription factor 6 in diabetic nephropathy

Abstract

BACKGROUND

Glomerular endothelial cell (GENC) injury is a characteristic of early-stage diabetic nephropathy (DN), and the investigation of potential therapeutic targets for preventing GENC injury is of clinical importance.

AIM

To investigate the role of β-arrestin-2 in GENCs under DN conditions.

METHODS

Eight-week-old C57BL/6J mice were intraperitoneally injected with streptozotocin to induce DN. GENCs were transfected with plasmids containing siRNA-β-arrestin-2, shRNA-activating transcription factor 6 (ATF6), pCDNA-β-arrestin-2, or pCDNA-ATF6. Additionally, adeno-associated virus (AAV) containing shRNA-β-arrestin-2 was administered via a tail vein injection in DN mice.

RESULTS

The upregulation of β-arrestin-2 was observed in patients with DN as well as in GENCs from DN mice. Knockdown of β-arrestin-2 reduced apoptosis in high glucose-treated GENCs, which was reversed by the overexpression of ATF6. Moreover, overexpression of β-arrestin-2 Led to the activation of endoplasmic reticulum (ER) stress and the apoptosis of GENCs which could be mitigated by silencing of ATF6. Furthermore, knockdown of β-arrestin-2 by the administration of AAV-shRNA-β-arrestin-2 alleviated renal injury in DN mice.

CONCLUSION

Knockdown of β-arrestin-2 prevents GENC apoptosis by inhibiting ATF6-mediated ER stress in vivo and in vitro. Consequently, β-arrestin-2 may represent a promising therapeutic target for the clinical management of patients with DN.

Key Words: Diabetic nephropathy; Glomerular endothelial cell; β-Arrestin-2; Activating transcription factor 6; Endoplasmic reticulum stress

Core Tip: This study investigates the role of β-arrestin-2 in glomerular endothelial cells (GENCs) in the context of diabetic nephropathy (DN). Under DN conditions, upregulated β-arrestin-2 increases the expression of activating transcription factor 6 and facilitates its translocation into the nucleus. This activation of endoplasmic reticulum stress induces apoptosis of GENCs, exacerbating renal damage. Silencing β-arrestin-2 alleviates renal injury and GENC apoptosis both in vivo and in vitro, suggesting that β-arrestin-2 could be a promising therapeutic target for the clinical management of patients with DN.

INTRODUCTION

Diabetic nephropathy (DN) is a microvascular complication of diabetes and approximately 30% of patients with diabetes develop DN, which is the leading cause of end-stage renal failure worldwide[1]. Lowering blood pressure and tight glycemic control are common clinical treatments for patients with DN; however, these strategies do not prevent DN progression[2]. As an early manifestation of DN, microalbuminuria, which indicates damage to the glomerular filtration barrier (GFB), is one of the main causes of DN deterioration. Podocytes are a component of the GFB, and previous studies have demonstrated that the inhibition of podocyte damage can protect against DN. Glomerular endothelial cells (GENCs) are also a crucial component of the GFB; however, studies indicating their role in DN remain limited. Increasing evidence suggests that GENC damage occurs before podocyte injury[3-5]. In mouse models, targeting specific genes that cause endothelial damage exacerbates the development of DN[6-8]. Injury and apoptosis of GENCs are characteristic features of the early stages of DN. However, mechanisms underlying GENC injury and apoptosis in DN remain unclear. Therefore, identifying key molecules and mechanisms involved in GENC injury may provide clues for developing new therapeutic strategies for DN in clinical practice.

Arrestins are small molecular proteins with multiple functions. β-Arrestin-1 and -2 are two subtypes of arrestins that are widely expressed in various mammalian tissues, and they act as negative regulators of G protein-coupled receptors and function as scaffold proteins that interact with different signaling molecules. β-Arrestins are closely related to the development of many diseases. In adriamycin induced nephropathy, β-arrestin-1 activates the endothelin-A receptor signaling pathway to promote podocyte injury and apoptosis[9]. According to a previous study, high glucose (HG) levels induced the upregulation of β-arrestin-1 and β-arrestin-2, which negatively regulated the conjugation of ATG5–ATG12 and suppressed autophagy in podocytes and induced podocyte apoptosis in DN[10]. However, the role of β-arrestin-1/2 in GENCs during DN remains unclear.

MATERIALS AND METHODS

Animal studies

Eight-week-old wild-type male C57BL/6 mice were selected to establish a DN model by intraperitoneal injection of streptozotocin (STZ), as previously described[10-12]. The details of the methods are shown in the Supplementary Materials. The mice were sacrificed, and blood, urine, and kidney samples were collected for subsequent experimentation (eight mice of each group; Supplementary Table 1).

Human renal biopsies

All renal biopsy tissues were obtained from the Department of Pathology at The Second Hospital of Shandong University, China and clinical samples were collected as previously described from five subjects with diabetes and biopsy-proven DN and five normal control subjects[11,12] (Supplementary Table 2). Normal control renal tissues were obtained from the cortex of the normal pole opposite the tumorous pole which otherwise had no clinical or histologic renal disease. The patients had a final histologic diagnosis of DN with nodular diabetic glomerulosclerosis and no other renal pathology.

Ethics

The investigations were conducted in accordance with the principles of the Declaration of Helsinki. All patients (or their legal guardians) signed an informed consent form. All the human study participants did not receive compensation. All experimental protocols for the animal studies were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. This study was approved by the Ethics Committee of Second Hospital of Shandong University, No. KYLL-2020 (LW)-072.

Cell culture and treatments

Rat GENCs were kindly donated by Professor Yi Fan at Shandong University and cultured in RPMI 1640 medium containing 5.5 mmol/L glucose and 10% fetal bovine serum (FBS; Gibco, Grand Island, NY, United States). Since chronic low-grade inflammation is a key factor in DN pathogenesis, diverse components of the immune system participate in the initiation and progression of DN, including adhesion molecules, chemokines, and pro-inflammatory cytokines[13-16]. Therefore, the GENCs were treated with different stimuli.

RNA interference and overexpression of β-arrestin-2 and activating transcription factor 6

Small interference RNA (siRNA) against β-arrestin-2 (5′-CCUACAGGGUCAAGGUGAATT-3′) and the negative control (5′-UUCUCCGCGUGUCACGUTT-3′) were synthesized by BioSune (Jinan, China). The siRNA for activating transcription factor 6 (ATF6) was synthesized and constructed into pGPU6/GFP/Neo to obtain shRNA-ATF6 (5’-GAGTGAGCTGCAGGTGTATTA-3’) by Biomics Biotechnologies Co., Ltd. (Nantong, Jiangsu, China). In the experiments, all siRNAs and shRNAs were transfected using Lipofectamine 3000 (Invitrogen, CA, United States) according to the manufacturer’s instructions. The overexpression plasmids containing either β-arrestin-2 or ATF6 were purchased from BioSune (China). GENCs were also transfected with the pCDNA-β-arrestin-2 or pCDNA-ATF6 plasmids with Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions.

AAV-mediated gene delivery into the kidney

Recombinant AAV containing shRNA-β-arrestin-2 (5’-GGAACUCUGUGCGGCUUAUTT-3’) and a negative control (AAV-Null: 5’-UUCUCCGAACGUGUCACGUTT-3’) were purchased from BioSune. Details of intrarenal AAV delivery are provided in the Supplementary Materials. The mice were sacrificed, and blood, urine, and kidney samples were collected for subsequent-experimentation (eight mice of each group; Supplementary Table 3)[17-19].

Real-time PCR

Total RNA was extracted from the kidney cortex and GENCs using TRIzol reagent (Invitrogen) following the manufacturer’s instructions. The levels of β-arrestin-1/2 were detected by real-time (RT)-PCR and the primers used in this study are listed in Supplementary Table 4.

Western blot analysis

Proteins were extracted from the kidney cortex and GENCs. Then, Western blot (WB) analyses were performed as previously described[20-22], with detailed procedures provided in the Supplementary Materials. A nuclear protein extraction kit (P0028; Beyotime, China) was used to extract proteins from the nucleus. The antibodies used in this study are summarized in Supplementary Table 5.

Morphological, immunohistochemical, and immunofluorescence staining

The fixed renal tissues were embedded in paraffin and cut into 4 μm sections. The sections were stained with periodic acid-Schiff (PAS), and immunohistochemical (IHC) staining and terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) kits according to the manufacturer’s protocols (The details of the methods are provided in the Supplementary Materials)[23].

Statistics analysis

All data are expressed as the mean ± SD unless otherwise specified. Statistical analyses was performed with GraphPad Prism (version 8.0, GraphPad Software, San Diego, CA, United States). Normally distributed data were compared using the two-tailed t-test to analyze the differences between groups. One-way ANOVA followed by post hoc Tukey’s test was used to analyze differences among multiple groups. P values < 0.05 were considered statistically significant.

RESULTS

Upregulation of β-arrestin-2 in patients with DN and GENCs from DN mice

IHC staining revealed upregulation of β-arrestin-2 in renal biopsies from patients with DN compared to that of the normal subjects (Figure 1A). As detailed in Supplementary Table 1, STZ-induced DN mice exhibited hyperglycemia and a lower body weight than control mice. Additionally, the urinary albumin-to-creatinine ratio (UACR) and relative kidney weight were significantly higher in DN mice, whereas there was no significant difference in heart rate between the two groups. Supplementary Figure 1 further shows renal injury in model mice. RT-PCR and WB analyses indicated that the expression of β-arrestin-2 was increased in the renal cortex of DN mice (Figure 1B and C). The same results were observed in the paraffin sections of renal tissue using IHC (Figure 1D). Immunofluorescence double staining and quantifications further confirmed that the expression of β-arrestin-2 (green) was significantly increased in the GENCs (red) of DN mice (Figure 1E and F), thereby indicating the upregulation of β-arrestin-2. Moreover, the upregulation of β-arrestin-1 in the renal biopsies of patients with DN (Supplementary Figure 2A) and in the DN model mice (Supplementary Figure 2B–D) was also detected. However, the expression of β-arrestin-1 increasing in GENCs (Supplementary Figure 2E) was not determined.

Figure 1 β-Arrestin-2 is increased significantly in renal biopsies from patients with diabetic nephropathy and glomerular endothelial cells from mice with diabetic nephropathy. A: Images of immunohistochemical (IHC) staining to detect the expression of β-arrestin-2 in paraffin sections of human renal tissue from normal controls and patients with diabetic nephropathy (DN) (black bars = 10 μm, n = 5); B: Relative mRNA levels of β-arrestin-2 in the renal cortex of DN mice (mean ± SD, ^cP < 0.001 vs control, n = 8); C: The expression of β-arrestin-2 in the renal cortex of DN mice was analyzed by immunoblotting (^cP < 0.001 vs control, n = 8); D: Representative images of IHC staining to detect the expression of β-arrestin-2 in paraffin sections of the kidneys from control and DN mice (red bars = 20 μm, n = 8); E: Detection of β-arrestin-2 expression in glomerular endothelial cells (GENCs) in streptozotocin (STZ)-induced DN mouse model by immunofluorescence double labeling: Endothelin (red, mark protein in GENCs) and β-arrestin-2 (green) (white bars = 50 μm, n = 8); F: Quantifications showing expression of β-arrestin-2 in the kidneys from different groups of mice (^bP < 0.01 vs control, n = 8).

Expression of β-arrestin-2 is increased in GENCs under various stimuli

The expression of β-arrestin-2 was increased in GENCs treated with glucose at a concentration of 20 or 40 mmol/L, but the expression of β-arrestin-2 remained the same when the cells were treated with mannitol at a concentration of 40 mmol/L (Figure 2A). Furthermore, the expression of β-arrestin-2 was increased in GENCs under common detrimental factors in DN such as advanced glycation end-product (AGE) (Figure 2B), tumor necrosis factor-α (TNF-α) (Figure 2C), angiotensin II (Ang II) (Figure 2D), and transforming growth factor-β1 (TGF-β1) (Figure 2E). The increased β-arrestin-2 expression in GENCs occurred in a concentration-dependent manner.

Figure 2 β-Arrestin-2 is upregulated in glomerular endothelial cells under high glucose levels and other stimuli in vitro. A: Representative images and summarized data showing increased expression of β-arrestin-2 in glomerular endothelial cells (GENCs) with high glucose (HG) (concentrations of 20 and 40 mmol/L; 40 mmol/L mannitol was used as osmolarity control) stimulation for 24 h by Western blotting (^aP < 0.05 vs control, ^bP < 0.01 vs control, n = 6); B: Representative images and summarized data showing upregulation of β-arrestin-2 in GENCs treated with advanced glycation end products (0, 50, 100, and 200 μg/mL) for 24 h by immunoblotting (^bP < 0.01 vs control, ^cP < 0.001 vs control, n = 6); C: Representative images and summarized data showing increased expression of β-arrestin-2 in GENCs treated with tumor necrosis factor-α (0, 20, 40, and 80 ng/mL) for 24 h by Western blotting (^aP < 0.05 vs control, ^bP < 0.01 vs control, n = 6); D: Representative images and summarized data showing increased β-arrestin-2 expression in angiotensin II (Ang II: 0, 10^-7, 10^-6, and 10^-5 mol/L) treated GENCs for 24 h by immunoblotting (^aP < 0.05 vs control, ^bP < 0.01 vs control, n = 6); E: Images and summarized data reflecting upregulation of β-arrestin-2 in GENCs treated with transforming growth factor-β1 (0, 2, 4, and 8 ng/mL) for 24 h by Western blotting (^aP < 0.05 vs control, ^bP < 0.01 vs control, n = 6). TGF-β1: Transforming growth factor-β1; HG: High glucose; AGE: Advanced glycation end product; TNF-α: Tumor necrosis factor-α.

Silencing of β-arrestin-2 ameliorates HG induced GENC injury

To investigate the effect of β-arrestin-2 in GENCs under HG stimulation, knockdown of β-arrestin-2 by siRNA was done (Figure 3A). The membrane proteins zonula occludens-1 (ZO-1) and occludin are related to the permeability of endothelial tight junctions. Decreased ZO-1 and occludin levels are closely associated with DN progression through the disruption of GENC function[24]. Inhibition of eNOS reduces NO production, leading to endothelial injury; eNOS-knockout mice develop nodular diabetic glomerulosclerosis[25,26]. WB analyses showed that the expression of ZO-1, occludin, and eNOS decreased in GENCs under HG stimulation. Silencing of β-arrestin-2 recovered occludin, ZO-1, and eNOS expression in GENCs under HG treatment (Figure 3B). Changes in apoptosis-related proteins, such as Bcl-2, Bax, and cleaved caspase 3 were detected. HG induced upregulation of apoptosis related proteins was decreased by knockdown of β-arrestin-2 (Figures 3C and D). Flow cytometry further confirmed these results (Figure 3E and F).

Figure 3 Knockdown of β-arrestin-2 by siRNA ameliorates high glucose induced glomerular endothelial cell injury and apoptosis. A: Images showing efficiency of silencing β-arrestin-2 with small interfering RNA (siRNA) by Western blotting (^aP < 0.05 vs control, n = 6); B: Representative images showing effect of β-arrestin-2 knockdown on expression of occludin, ZO-1, and eNOS in high glucose (HG)-treated glomerular endothelial cell (GENCs) by Western blotting (^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, n = 6); C: Immunoblotting images showing effect of β-arrestin-2 knockdown on expression of apoptosis related proteins Bcl-2 and Bax in HG-treated GENCs (^aP < 0.05 vs scramble/control, ^cP < 0.001 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, n = 6); D: Immunoblotting images showing effect of β-arrestin-2 knockdown on expression of cleaved caspase 3 and caspase 3 in GENCs with HG treatment (^cP < 0.001 vs scramble/control, ^eP < 0.01 vs scramble/HG treatment, n = 6); E: Flow cytometric images showing decreased apoptosis of GENCs with HG treatment by knockdown of β-arrestin-2; F: Summarized flow cytometric data showing reduced apoptosis of GENCs with HG treatment by silencing β-arrestin-2 (^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, n = 6). HG: High glucose; siRNA: Small interfering RNA.

Knockdown of β-arrestin-2 suppresses endoplasmic reticulum stress induced apoptosis in GENCs through ATF6

ER stress plays a vital role in the progression of DN, and BiP and CHOP are marker proteins associated with endoplasmic reticulum (ER) stress. The results indicated that HG induced the upregulation of BiP and CHOP, which could be decreased by knockdown of β-arrestin-2 in GENCs (Figure 4A), indicating that silencing of β-arrestin-2 suppressed ER stress which was activated by HG in the GENCs. The role of ER stress in GENCs was not clear; therefore, using ER stress activators, such as tunicamycin (TM) and thapsigargin, and the ER stress inhibitor 4-phenyl butyric acid (4-PBA), the apoptosis of GENCs was determined by Annexin V/propidium iodide staining. Flow cytometry results confirmed that activated ER stress was induced by GENC apoptosis, which was alleviated by the ER stress inhibitor 4-PBA (Figure 4B). Furthermore, silencing of β-arrestin-2 could partially reduce apoptosis of GENCs induced by the ER stress activator TM (Figure 4C).

Figure 4 Knockdown of β-arrestin-2 reduces endoplasmic reticulum stress and apoptosis of glomerular endothelial cells in high glucose treatments by decreasing expression of activating transcription factor 6. A: Representative images and summarized data showing that knockdown of β-arrestin-2 inactivated endoplasmic reticulum (ER) stress by downregulating expression of BiP and CHOP in HG-treated glomerular endothelial cells (GENCs) by Western blotting (^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, n = 6); B: Flow cytometric analysis showing that the ER stress activators tunicamycin (TM) and thapsigargin (TG) induced GENC apoptosis and the inhibitor 4-phenyl butyric acid (4-PBA) reduced GENC apoptosis (^bP < 0.01 vs scramble/control, ^fP < 0.05 vs TM/TG treatment, n = 5); C: Summarized flow cytometric data showing apoptosis of GENCs under different stimuli (^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, ^fP < 0.05 vs TM/TG treatment, n = 4); D: Western blotting images showing effects of β-arrestin-2 knockdown on expression of ATF6 in GENCs under HG treatment (^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, n = 6). HG: High glucose; siRNA: Small interfering RNA; TM: Tunicamycin; TG: Thapsigargin; 4-PBA: 4-Phenyl butyric acid.

Protein kinase R-like ER kinase (PERK), inositol requiring 1α (IRE1α), and ATF6 associated pathways are three signal pathways responsible for ER stress. To determine the exact pathway via which β-arrestin-2 induces ER stress, the IRE1α and PERK signaling pathways were first detected. Immunoblot of the related proteins in the two pathways showed that HG dramatically enhanced expression of p-IRE1α, XBP1, p-eIF1α, p-PERK, and ATF4, but knockdown of β-arrestin-2 did not change the upregulation of these proteins under HG treatment (Supplementary Figure 3). The protein levels of ATF6, a transcription factor involved in ER stress, were then measured. WB results showed that HG increased the expression of ATF6 in GENCs, whereas knockdown of β-arrestin-2 significantly reduced the upregulation of ATF6 in GENCs under HG treatment (Figure 4D).

β-Arrestin-2 activated ER stress induced GENC injury is blocked by knockdown of ATF6

The overexpression plasmid of β-arrestin-2 (Supplementary Figure 4A) and shRNA of ATF6 (Supplementary Figure 4B) were synthesized. WB results showed that overexpression of β-arrestin-2 potentially activated ER stress in GENCs through increase of BiP and CHOP, which could be blocked by knockdown of ATF6 (Figure 5A). Moreover, silencing of ATF6 could upregulate the expression of ZO-1 and occludin that was decreased by overexpression of β-arrestin-2 in GENCs (Figure 5B). WB and flow cytometry both confirmed that the apoptosis of GENCs induced by overexpression of β-arrestin-2 could be attenuated by knockdown of ATF6 (Figure 5C and D). Additionally, overexpression of β-arrestin-2 could upregulate ATF6. However, knockdown of ATF6 did not decrease the expression of β-arrestin-2 in GENCs (Figure 5E).

Figure 5 Overexpression of β-Arrestin-2 aggravates glomerular endothelial cell injury and apoptosis by activating endoplasmic reticulum stress through upregulation of activating transcription factor 6. A: Representative images showing that overexpression of β-arrestin-2 enhanced expression of BiP and CHOP in glomerular endothelial cells (GENCs), which could be blocked by knockdown of activating transcription factor 6 (ATF6) (^aP < 0.05 vs scramble/control, ^gP < 0.05 vs pCDNA-β-arrestin-2 treatment, n = 6); B: Representative images showing that overexpression of β-arrestin-2 reduced expression of ZO-1 and Occludin in GENCs, which could be recovered by knockdown of ATF6 in GENCs (^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^hP < 0.01 vs pCDNA-β-arrestin-2 treatment, n = 6); C: Representative images showing that overexpression of β-arrestin-2 promoted apoptosis related protein expression in GENCs, which was blocked by knockdown of ATF6 (^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^gP < 0.05 vs pCDNA-β-arrestin-2 treatment, n = 6); D: Summarized flow cytometric data showing that overexpression of β-arrestin-2 aggravated apoptosis of GENCs, which was inhibited by knockdown of ATF6 (^aP < 0.05 vs scramble/control, ^gP < 0.05 vs pCDNA-β-arrestin-2 treatment, n = 6); E: Representative Western blot images showing that overexpression of β-arrestin-2 upregulated expression of ATF6, but knockdown of ATF6 did not affect expression of β-arrestin-2 (^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^gP < 0.05 vs pCDNA-β-arrestin-2 treatment, n = 6). ATF6: Activating transcription factor 6.

β-Arrestin-2 increases expression of ATF6 and promotes ATF6 tranlocation into the nucleus

In HG treated GENCs transfected with the overexpression plasmid of ATF6 (Supplementary Figure 4C), the protein levels of BiP and CHOP were downregulated by knockdown of β-arrestin-2 (Figure 6A). Knockdown of β-arrestin-2 attenuated HG induced GENC injury, which was reversed by overexpression of ATF6 through the increased expression of ZO-1 and occludin (Figure 6B). Apoptosis of GENCs under HG treatment was reduced by knockdown of β-arrestin-2, but overexpression of ATF6 aggravated this apoptosis (Figure 6C and D). These results were confirmed based on the expression of cleaved caspase-3 (Supplementary Figure 4D). Silencing of β-arrestin-2 reduced expression of ATF6, but overexpression of ATF6 did not change the expression of β-arrestin-2 (Figure 6D). In the ATF6 signaling pathway, ATF6 is cleaved and translocates to the nucleus to promote the transcription of downstream target genes. Immunoblotting and immunofluorescence were performed to detect the expression of activated ATF6 in the nucleus. The results indicated that HG induced ATF6 translocation into the nucleus was blocked by silencing of β-arrestin-2. Overexpression of β-arrestin-2 promoted ATF6 being translocated into the nucleus (Figure 6E) which was inhibited by knockdown of ATF6. Immunofluorescence and IHC results showed that ATF6 translocated into the nucleus in DN, which was inhibited by silencing of β-arrestin-2 (Supplementary Figure 5). Similar results in the mRNA levels of the ATF6 target genes GRP78 and GRP94 (Figure 6F and Supplementary Figure 4E) were observed.

Figure 6 β-Arrestin-2 enhances expression of activating transcription factor 6 and promotes activating transcription factor 6 translocation into the nucleus of glomerular endothelial cells treated with high glucose. A: Representative images and summarized data showing that knockdown of β-arrestin-2 reduced protein levels of BiP and CHOP, which was reversed by overexpression of activating transcription factor 6 (ATF6) [^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/high glucose (HG) treatment, ⁱP < 0.05 vs Si-β-arrestin-2/HG treatment]; B: Representative images and summarized data showing that knockdown of β-arrestin-2 recovered protein levels of ZO-1 and Occludin, which was reversed by overexpression of ATF6 (^aP < 0.05 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, ⁱP < 0.05 vs Si-β-arrestin-2/HG treatment); C: Flow cytometric analysis showing that knockdown of β-arrestin-2 reduced apoptosis of glomerular endothelial cells treated with HG was reversed by overexpression of ATF6; D: Representative images and summarized data showing that knockdown of β-arrestin-2 reduced apoptosis related protein expression, which was reversed by overexpression of ATF6. Silencing of β-arrestin-2 decreased expression of ATF6, but overexpression of ATF6 did not change expression of β-arrestin-2 (^aP < 0.05 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, ⁱP < 0.05 vs Si-β-arrestin-2/HG treatment); E: Representative Western blot images and summarized data showing expression of cleaved ATF6 in the nucleus of cells under different stimuli (^bP < 0.01 vs scramble/control, ^cP < 0.001 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, ^gP < 0.05 vs pCDNA-β-arrestin-2 treatment, ⁱP < 0.05 vs Si-β-arrestin-2/HG treatment); F: Relative mRNA levels of GRP78 (^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/HG treatment, ^fP < 0.05 vs TM/TG treatment, ^gP < 0.05 vs pCDNA-β-arrestin-2 treatment, ⁱP < 0.05 vs Si-β-arrestin-2/HG treatment). ATF6: Activating transcription factor 6; HG: High glucose; TM: Tunicamycin; 4-PBA: 4-Phenyl butyric acid.

Silencing of β-arrestin-2 ameliorates kidney injury in vivo

To examine the therapeutic efficiency of β-arrestin-2 targeting in mice with DN, AAV containing shRNA-β-arrestin-2 was delivered into the experimental mice by tail vein injection. The results confirmed that AAV-shRNA-β-arrestin-2 significantly reduced the mRNA and protein levels of β-arrestin-2 in the renal cortex of the mice (Figure 7A and B and Supplementary Figure 6A). Silencing of β-arrestin-2 reduced the UACR (Figure 7C). Using PAS staining, a widened mesangial area and an increased mesangial matrix were observed in DN mice, whereas the mesangial area and matrix decreased in AAV-shRNA-β-arrestin-2 transduced DN mice (Figure 7D). TUNEL analysis showed that there were fewer apoptotic cells in AAV-shRNA-β-arrestin-2 transduced DN mice compared to control DN mice (Figure 7E and Supplementary Figure 6B). WB and IHC results indicated that in AAV-shRNA-β-arrestin-2 transduced DN mice, the levels of ZO-1, eNOS, and occludin were significantly increased (Figure 7F and Supplementary Figure 6C and D). The ER stress was inhibited as evidenced by the decreased protein level of BiP, CHOP, and ATF6 in DN mice transduced with AAV-shRNA-β-arrestin-2 (Figure 7G).

Figure 7 Adeno-associated virus injection induced gene silencing of β-arrestin-2 ameliorates renal injury in streptozotocin-induced diabetic nephropathy mice. A: Relative mRNA levels of β-arrestin-2 in the renal cortex of mice after adeno-associated virus (AAV)-null (scramble) and AAV-shRNA-β-arrestin-2 (shRNA-β-arrestin-2) tail injection (mean ± SD, ^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/streptozotocin (STZ) treatment, n = 8); B: Western blot images showing protein level of β-arrestin-2 in the renal cortex of mice after AAV-null (scramble) and AAV-shRNA-β-arrestin-2 (shRNA-β-arrestin-2) tail injection (mean ± SD, ^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/STZ treatment, n = 8); C: Urinary albumin-to-creatinine ratio (UACR) in different groups of mice (mean ± SD, ^aP < 0.05 vs scramble/control, ^bP < 0.01 vs scramble/control, ^dP < 0.05 vs scramble/STZ treatment, n = 8); D: Periodic-acid Schiff staining showing changes of glomerular structure in different groups of mice (black bars = 20 μm, n = 8); E: TUNEL staining showing apoptosis of cells in the kidneys in different groups of mice (white bars = 50 μm, n = 8); F: Representative images showing protein levels of eNOS, ZO-1, and Occludin in the renal cortex of different groups of mice (mean ± SD, ^aP < 0.05 vs scramble/control, ^dP < 0.05 vs scramble/STZ treatment, n = 8); G: Western blot analysis showing expression of endoplasmic reticulum stress related proteins BiP, CHOP, and activating transcription factor 6 in the renal cortex of different groups of mice (mean ± SD, ^aP < 0.05 vs scramble/control, ^dP < 0.05 vs scramble/STZ treatment, n = 8). ATF6: Activating transcription factor 6; STZ: Streptozotocin.

DISCUSSION

As a key component of the GFB, GENCs have been intensively studied in recent decades[27]. GENCs are the first layer of the GFB; therefore, GENC dysfunction occurs early and plays a key role in the initiation and development of DN. Increasing evidence suggests that GENC injury plays an important role in the pathogenesis of DN[28-30]. Previous studies have demonstrated that the mRNA levels of β-arrestin-1/2 were negatively correlated with estimated glomerular filtration rate in all available subjects individually, but the functional role of β-arrestins in GENCs is unclear. In this study, the expression of β-arrestin-2, but not β-arrestin-1, was increased significantly in the GENCs of DN mice. The renin-angiotensin-aldosterone system (Ang II), HG, and AGE formation are important pathways in the development and progression of DN. Each pathway causes damage via multiple mediators or by interacting with other pathways[31]. Although DN has not traditionally been considered an inflammatory disease, immune and inflammatory responses play important roles in its pathogenesis. Inflammatory factors, such as IL-6, TNF-α, TGF-β1, and IL-18, are elevated in blood and have been shown to be involved in the development and progression of DN[32]. Moreover, the upregulation of β-arrestin-2 in GENCs treated with detrimental factors (such as HG and AGE) in vitro was also observed. Then the role of β-arrestin-2 in GENCs was investigated and it was determined that HG reduced the expression of eNOS, occludin, and ZO-1 in vitro; however, knockdown of β-arrestin-2 could recover the expression of these proteins. Furthermore, it was also determined that knockdown of β-arrestin-2 could decrease HG induced GENC apoptosis. These data show that knockdown of β-arrestin-2 alleviates HG induced GENC injury and apoptosis.

ER stress is a major cellular mechanism involved in kidney injury in DN[33]. Numerous studies have demonstrated that ER stress dysfunction is associated with the onset and progression of DN, and that ER stress inhibitors decrease ER stress and halt DN progression[34,35]. However, some studies have demonstrated that ER stress plays a protective role in DN, reflecting the bidirectional control of ER stress in DN[36]. In this study, it was determined that activated ER stress induced apoptosis of GENCs and that ER stress was activated in GENCs treated with HG. Additionally, silencing of β-arrestin-2 not only decreased the upregulation of ER stress related proteins BiP and CHOP, but also reduced ER stress induced apoptosis in GENCs with HG treatment.

There are three ER stress sensor pathways, namely, IRE1/sXBP1, PERK/EIf2α, and ATF6[37]. It has been reported that β-arrestin-2 interacted with eIF2α in intestinal epithelial cells, which contributed to promoting ERS/PUMA, thereby inducing mucosal apoptosis in colitis through the mitochondrial apoptotic pathway[38]. However, the present study observed different outcomes for GENCs under different HG conditions. Although p-PERK, p-eIF2α, and ATF4 were increased in HG-treated GENCs, knockdown of β-arrestin-2 had no effect on the expression of these proteins. The mixed results showed different contributions of β-arrestin-2 to ER stress in different diseases or different cell types. Therefore, this study aimed to investigate how β-arrestin-2 regulates ER stress in GENCs under different HG conditions. However, this study did not observe the interaction between the β-arrestin-2 and the IRE1α pathway in GENCs; therefore, the focus was placed on the ATF6 sensor pathway. As a type II transmembrane protein, ATF6 is an important molecule in the ER stress pathway and participates in the regulation of cell apoptosis. In this study, it was observed that upregulation of ATF6 could be reduced by knockdown of β-arrestin-2 in GENCs under different HG treatments. Overexpression of β-arrestin-2 increased the expression of ATF6 directly in GENCs, but overexpression or knockdown of ATF6 did not change the expression of β-arrestin-2. This study investigated whether injury and apoptosis of GENCs induced by overexpression of β-arrestin-2 could be alleviated by knockdown of ATF6, and if silencing of β-arrestin-2 could reduce injury and apoptosis of GENCs, they would be reversed by overexpression of ATF6. The results confirmed that β-arrestin-2 regulated the expression of ATF6 in GENCs. When unfolded proteins accumulate, ATF6 is transported to the Golgi complex, where it is proteolytically cleaved by S1P and S2P to release its NH2 terminal-domain[39]. Cleaved ATF6 translocates to the nucleus, where it activates the transcription of target genes, such as GRP78 and GRP94[40,41]. ATF6 was detected in the nucleus, and the results showed that β-arrestin-2 promoted ATF6 translocation to the nucleus in GENCs treated with HG. The mRNA levels of the ATF6 target genes GPR78 and GPR94 also showed the same results. But the manner in which β-arrestin-2 promotes ATF6 expression and translocates to the nucleus is unclear.

Gene therapy has recently been used to treat several diseases. AVV with low toxicity and antigenicity is a promising vehicle for gene therapy, and commercial AAV gene therapy products have been approved by regulatory agencies and are used clinically to treat spinal muscular atrophy, Leber congenital amaurosis, and hemophilia A[42-44]. In the present study, AAV containing shRNA-β-arrestin-2 was delivered into the mice via tail vein injection. This study demonstrated that AAV could stably transduce renal cells for 3 mo and silencing of β-arrestin-2 could attenuate kidney damage in DN mice. To enhance cell-specific impacts, we employed an endothelial-specific ICAM2 promoter to induce β-arrestin-2 knockdown in endothelial cells. Although β-arrestin-2 expression was inhibited in GENCs, its expression in other organs remained uncertain. The role of β-arrestin-2 in endothelial cells of other organs is unclear. We plan to investigate the effects of β-arrestin-2 silencing in the endothelial cells of these organs in future studies. As shown in Supplementary Table 2, the sample size for both normal controls and DN patients was limited to five in each group. Whether silencing of β-arrestin-2 is also effective in a clinical setting warrants further investigation.

CONCLUSION

In conclusion, this study determined that the level of β-arrestin-2, but not β-arrestin-1, was increased in the GENCs of DN mice. Silencing of β-arrestin-2 alleviated GENC injury and apoptosis under HG conditions in vitro. In vivo, adeno-associated virus (AAV) containing shRNA-β-arrestin-2 was injected through the tail vein of DN mice, and it was determined that the damage to the kidneys was significantly reduced. Mechanistically, β-arrestin-2 activated ER stress through the ATF6 signal pathway by increasing its expression and promoting its entry into the nucleus. Therefore, the findings of this study suggest β-arrestin-2 as a potential therapeutic target for the treatment of DN in a clinical setting. This discovery offers a new perspective on the critical role of GENCs in DN.