Introduction
Diabetic nephropathy (DN), which is regarded as the common and primary microvascular complication of diabetes mellitus, has been proven to induce terminal-stage renal diseases [
1,
2]. Glycated haemoglobin (HbA1c), as an indicator of long-term glycaemic control, has been proposed as a screening target for diabetic nephropathy and is also a glycaemic marker in patients with gestational diabetes mellitus and advanced chronic kidney disease [
3‐
6]. In addition to HbA1c, the clinical manifestations of diabetic nephropathy include a gradual increase in urinary albumin and a decrease in the glomerular filtration rate [
7,
8]. Currently, if DN develops into renal failure, the patient’s treatment costs and mortality will increase exponentially, and few effective treatment approaches for DN are available [
9,
10]. Therefore, the search for the underlying mechanism of DN has practical value.
A recent report found that endothelial-to-mesenchymal transition (EndMT) occurs in glomerular endothelial cells, which is important for the progression of DN [
11,
12]. EndMT occurs in the glomerular endothelium of patients with DN, as shown by a decrease in CD31 but an increase in α-SMA expression [
13]. In EndMT, endothelial cells develop mesenchymal characteristics, replacing the endothelial phenotype [
14], which is a specific type of epithelial-mesenchymal transition (EMT). Its representative genes include Snail, α-SMA, vimentin and CD31. Among these genes, Snail is a critical transcriptional regulator of EndMT [
15]. Thus, EMT modulators may be involved in the regulation of EndMT by changing the expression of Snail [
16].
Previous reports revealed that Bric-a-brac/Tramtrack/Broad (BTB) and cap'n'collar (CNC) homology 1 (bach1) impairs angiogenesis and mediates oxidative stress in vascular endothelial cells [
17,
18]. In addition, bach1 participates in EMT in cancer cells [
19]. However, the mechanism by which bach1 participates in hyperglycaemia-mediated EndMT has not yet been studied.
Moreover, some studies have demonstrated that ETS domain-containing protein (ELK1), a member of the E26 transformation-specific (ETS) oncogene family, is involved in the modulation of cell proliferation, apoptosis, differentiation and tumorigenesis [
20‐
25]. Moreover, ELK1 was reported to participate in oxidized low-density lipoprotein-induced endothelial cell apoptosis [
24]. ELK1 plays an important role in transforming growth factor-beta-induced EndMT [
26,
27]. However, the role of ELK1 in hyperglycaemia-induced EndMT is unclear.
In addition to the above, SETD8, as the sole nucleosome-specific methyltransferase, can regulate the monomethylation of histone H4 lysine 20 (H4K20me1) [
28]. Our previous studies demonstrated that suppression of SETD8 aggravates high glucose-induced vascular endothelial injury [
29‐
31]. In addition, SETD8 was reported to mediate EMT [
32]. At present, no research has shown that SETD8 participates in high glucose-induced EndMT.
The present study showed that SETD8 interacts with bach1 to regulate the transcription of Snail, leading to the occurrence of EndMT, which is involved in the progression of DN. In addition, SETD8 cooperates with ELK1 to regulate the transcription of bach1 by affecting histone methylation at the promoter region of bach1. Thus, SETD8 plays a key role in the progression of DN, and can be employed as the new target in the treatment of DN.
Methods
Subjects
This experiment was approved by the Ethics Committee of Huzhou Central Hospital (licence number: 20191209-01) and followed the Declaration of Helsinki. The present study recruited thirty patients with DN and type 2 diabetes. Additionally, thirty diagnosed renal cancer patients with normal renal function served as controls. All the participants signed informed consent forms. Exclusive criteria include advanced liver disease, renal failure, stroke, and other cardiovascular diseases.
Rat model of DN
Under the provisions of the Guide for the Care and Use of Laboratory Animals of Fudan University Shanghai Cancer Center and the Guide for the Care and Use of Laboratory Animals published by the US NIH (2011), male SD rats weighing 300–400 g were employed. The animals were kept in a temperature-controlled environment (22 °C to 25 °C) and maintained in a 12-h light/dark cycle. All rats underwent unilateral nephrectomy (Unx) under isoflurane anaesthesia (3–4% induction, 1.5–2.5% maintenance, 100% oxygen) and were sent back to the care facility for 9 weeks. Three weeks after Unx, the rats that received a single intraperitoneal injection of citrate buffer (0.1 M, pH 4.5) were defined as the control group (con). Rats treated with a high-sugar and high-fat diet for 9 weeks after Unx and intraperitoneal injection of streptozotocin (STZ, 50 mg/kg) 3 weeks after Unx were defined as the DN group (n = 10) [
33]. To clarify the protective effect of SETD8 overexpression in DN, we injected the rats with DN with AVV-SETD8 or control vectors into the contralateral kidney at the time of Unx. The rats were defined as the DN-AVV group (n = 10) and DN-AVV-SETD8 group (n = 10) accordingly [
34,
35].
Immunohistochemistry (IHC)
Tissue slides were deparaffinized and then stored in methanol containing 3% hydrogen peroxide. After the background was blocked, the slides were incubated with anti-SETD8 (dilution 1:200, ProteinTech, Wuhan, China), anti-ELK1 (dilution 1:200, ProteinTech), anti-bach1 (dilution 1:200, ProteinTech, anti-Snail (dilution 1:200, ProteinTech,), anti-vimentin (dilution 1:200, ProteinTech), anti-α-SMA (dilution 1:200, ProteinTech), and anti-CD31 (dilution 1:200, Abcam, Cambridge, UK) antibodies at 4 °C overnight. The next day, the slides were cultured with secondary antibodies at 37 °C. Finally, a DAB Detection Kit (GeneTech, Shanghai, China) was applied to stain the slides, and haematoxylin was used for counterstaining.
Cell culture and intervention
HGECs were procured from Procell (Wuhan, China) and incubated with 5 mM glucose and 10% foetal bovine serum in an incubator at 37 °C in a humidified 5% carbon dioxide atmosphere. Cells were cultivated in high glucose (25 mM) DMEM for 3 days for the high glucose treatment. The apoptosis control used glucose (5 mM) mixed with mannitol (20 mM).
Western blot
The samples were extracted from different cell groups and boiled with loading buffer for 10 min. The proteins were separated by 8–10% SDS-PAGE. The PVDF membranes were cultured with primary antibodies at 4 °C overnight. The primary antibodies were antibodies against β-actin (Dilution 1:2000, ProteinTech), SETD8, ELK1, bach1, Snail, vimentin, α-SMA, CD31 and H4K20me1 (Dilution 1:1000, Abcam). The next day, the membranes were cultured with the secondary antibodies. After that, the membranes were detected by the ECL system.
Quantitative real-time PCR (qPCR)
Hieff UNICON® qPCR TaqMan Probe Master Mix (Yeasen, Shanghai, China) was used to perform quantitative qPCR to determine the target genes. The primer sequences are shown in Additional file
1: Table S1. The relative gene expression was calculated by the 2
−△△CT method. Data are shown as the fold change relative to the control group. In addition, the ratio of the control group was set as 1.
Coimmunoprecipitation (Co-IP)
Cell protein lysates were isolated with lysis buffer, which was mixed with primary antibodies against bach1, SETD8, ELK1 and IgG at 4 °C overnight for endogenous IP. The next day, the lysates were incubated with 50 μL of protein Dynabeads (Thermo, MA, USA) for 6 h at 4 °C. Furthermore, the beads were washed with RIPA 3 times to remove impurities. Finally, 20 µL of IP lysates was added to 2× loading buffer and boiled together. The results were analysed by Western blots.
Immunofluorescence assay
After HGECs were seeded onto glass slides, the cells were fixed with 4% paraformaldehyde for 10 min. The cells were incubated with anti-SETD8 and anti-ELK1 antibodies at 4 °C overnight. Then, they were sequentially incubated with fluorescent secondary antibodies for 1 h at 37 °C. Next, DAPI (Yeasen) was used to stain nuclei. Finally, a confocal fluorescence microscope (Leica) was employed to capture images.
GST pulldown assay
We purchased His-SETD8 (ProteinTech) and GST-ELK1 (ProteinTech) fusion proteins for the experiment. The fusion proteins were mixed for 12 h at 4 °C in GST binding buffer. Anti-His or anti-GST beads were added and incubated with the fusion protein for an additional 4 h. The beads were washed three times, and the proteins were detected by western blotting.
siRNA treatments
In the experiment, HGECs were transfected with siRNA against ELK1 and bach1 using Lipofectamine 3000. ELK1 siRNA (Biotend) sequences were as follows: siRNA-a, 5′-GGUACUACUAUGACAAGAAdTdT-3′ and siRNA-b, 5′-GCAGCUGCUGAGAGAGCAAdTdT-3′. The bach1 siRNA sequences were as follows: siRNA-a, 5′-CAGACAUAUGAGUCCAUGUdTdT-3′ and siRNA-b, 5′-CAGCAAUUUAACAGCUUGAdTdT-3′.
Short hairpin RNA (shRNA) and mutant SETD8
SETD8 shRNAs and mutant SETD8
R295G plasmid [
27] were transfected into HGECs. The sequences of shRNA were as follows: shRNA-a, 5′-CAACAGAATCGCAAACTTA-3′ and shRNA-b, 5′-CAACAGAATCGCAAACTTA-3′.
Chromatin immunoprecipitation (ChIP) assay
The ChIP assay kit (Millipore, MA, USA) was used in the study. Briefly, cells (1 × 107) were settled with 1% formaldehyde. Then, glycine (2.5 M) was added to stop the crosslinking reaction. Chromatin was subjected to ten ultrasounds. After centrifugation, the supernatant was incubated with anti-bach1, anti-ELK or anti-H4K20me1 antibodies and IgG at 4 °C. Agarose beads were applied to connect with immunoprecipitants. Furthermore, DNA–protein crosslinking was reversed by using a water bath at 65 °C for 6 h. The enriched sequences of the purified DNA were analysed by PCR. The bach1 oligonucleotide primer sequences were forward 5′-ACTGGCTCAAGGTGGAAGGA-3′, and reverse 5′-CAGGCTGCCTCAGTTCATGG-3′. The Snail oligonucleotide primer sequences were forward 5′-TAAATTGACACGGGACGGGG-3′, and reverse 5′-CTGGTTCTAGCTGGAGAGCG-3′. Furthermore, a re-ChIP assay was performed to verify whether SET8 and ELK1 occupied the same binding site on the bach1 promoter region. The chromatin from the beads was eluted by 10 mM DTT after the standard ChIP procedure. The eluent was then diluted with sonication buffer before undergoing the ChIP process again.
Dual-luciferase assay
The Dual-luciferase Assay Kit was employed to measure the impact of ELK1 on the bach1 promoter. The bach1 promoter was amplified from genomic DNA of HGECs and ligated into the pGL3-Basic vector. Moreover, the deleted promoter site was constructed for comparison. Then, pGL3-DAPK3 and pGL3-DAPK3Del were transfected into HGECs. The relative luciferase activity was used to evaluate the influence of ELK1 on bach1 promoter activity.
Statistical analysis
In this study, the sample size of the in vivo experiment was 10, the sample size of the in vitro experiment was 5, and statistical significance was obtained.
The data from separate experiments were analysed by GraphPad Prism 8 Project software. The comparison of means of two groups was performed by two-tailed unpaired t tests. We used a one-way ANOVA test to compare the means of more than 2 groups. p < 0.05 was considered statistically significant, and the data were plotted using GraphPad Prism 8.
Discussion
The novel findings of this research demonstrated that hyperglycaemia, by augmenting bach1 expression, participated in EndMT of glomerular endothelial cells. Moreover, SETD8 interacted with bach1 to regulate the transcriptional activity of Snail, which regulated the occurrence of EndMT. In addition, a high glucose environment enhanced ELK1 levels and restrained SETD8 expression. Meanwhile, SETD8 cooperated with ELK1 and occupied the bach1 promoter region at the same site, thereby regulating the transcription of bach1, thus mediating EndMT in hyperglycaemia-cultured HGECs.
EMT is a sophisticated cell phenotype reprogramming process that participates in organ injury [
36]. Previous reports demonstrated that EMT plays a vital role in renal interstitial myofibroblasts, thus underpinning the progression of renal fibrosis [
37]. Moreover, studies have found that EndMT is involved in the progression of DN [
11,
12]. Similarly, our representative images of HE and Masson staining exhibited collagen deposition and fibrosis. Moreover, IHC staining illustrated that CD31 expression was suppressed, while Snail, vimentin and α-SMA levels were increased in the glomerular endothelial cells of DN. These data were in agreement with EndMT in kidney-aggravated DN [
13]. Importantly, among these genes, Snail is the key regulator of EndMT in endothelial cells [
38]. To determine whether EndMT in DN was due to a high glucose environment, we employed high glucose-cultured HGECs. The changes in the levels of Snail, CD31, vimentin and α-SMA confirmed that hyperglycaemia induced EndMT, and the result coincided with our previous study [
39]. In addition, bach1 was verified to participate in EMT in cancer cells [
19]. It was deduced that EMT and EndMT share cooperative modulators [
16]. We next investigated whether bach1 was involved in regulating EndMT in hyperglycaemic HGECs. Bach1 levels were confirmed to be higher in DN. High glucose treatment increased bach1 expression and EndMT in HGECs, while inhibition of bach1 expression reversed these trends. ChIP assays suggested that bach1 can accumulate in the Snail promoter region. These data showed that upregulated bach1 expression is required for high glucose-mediated EndMT by regulating Snail in HGECs. Our previous studies indicated that SETD8 participates in hyperglycaemia-mediated increases in endothelial adhesion molecule expression [
40], proinflammatory enzymes, proinflammatory cytokine production [
29], and antioxidant imbalance [
30], thus mediating vascular endothelial injury [
29‐
31,
40]. In this study, our data indicated that SETD8 could regulate the expression of Snail. Co-IP indicated that SETD8 interacted with bach1 in HGECs. The ChIP assay confirmed that not only bach1 but also SETD8 occupied the Snail promoter. This finding indicated that SETD8 cooperated with bach1 to regulate Snail expression, thus participating in EndMT in DN.
ELK1 was reported to modulate cell proliferation, apoptosis, differentiation, and tumorigenesis [
20‐
23,
25]. In addition, ELK1 mediates oxidized low-density lipoprotein-induced endothelial cell apoptosis [
24]. Moreover, upregulated ELK1 expression has been confirmed to mediate cell death [
20,
41]. Furthermore, ELK1 is activated by hypoxia in vascular endothelial cells [
21,
42] and is involved in transforming growth factor-beta-induced EndMT [
26,
27]. In this study, we found that high glucose augmented ELK1 expression in HGECs. Then, we investigated whether ELK1 regulated EndMT via modulation of bach1 in HGECs. We demonstrated that si-ELK1 reversed the upregulation of bach1 expression and EndMT under hyperglycaemic conditions. Moreover, overexpression of ELK1 aggravated bach1 levels and EndMT, which was neutralized by si-bach1. All these results indicated that ELK1 orchestrated EndMT in HG-cultured HGECs by augmenting bach1 expression.
To date, few reports have described the regulation of EndMT by SETD8. Indeed, it was reported that SETD8 participates in EMT by modulating TWIST [
43] and cooperating with zinc finger E-box-binding homeobox 1 [
32]. As a specific form of EMT, EndMT and EMT may possess shared modulators [
16]. The present study confirmed that SETD8 overexpression neutralized high glucose-mediated upregulation of bach1 expression and EndMT in HGECs. Moreover, sh-SETD8 upregulated bach1 expression and mediated EndMT in HGECs, resulting in the same effect as high glucose treatment. Furthermore, si-bach1 neutralized sh-SETD8-mediated EndMT. Downstream of SETD8, H4K20me1 accumulates at the bach1 promoter region. All these findings indicated that high glucose-induced downregulation of SETD8 expression caused EndMT in HGECs by upregulating bach1 expression.
Previous reports have shown that epigenetic modifications can modulate the transcriptional activity of ELK1 [
44,
45]. In this study, ELK1 was found to associate with SETD8, which is the sole nucleosome-specific methyltransferase. The results of the luciferase assay illustrated that upregulated ELK1 expression augmented bach1 promoter activity. Moreover, our data showed that SETD8 and ELK1 were located at the same promoter region of bach1 by the re-ChIP assay. Moreover, the occupancy of ELK1 on the bach1 promoter region was augmented in cells in which SETD8 was silenced. These data indicated that the transcriptional activity of ELK1 in DN was modulated by SETD8. Furthermore, SETD8 overexpression inhibited HG-induced bach1 expression, and the SETD8 mutant did not function. Our results indicated that SETD8-modulated H4K20me1 participates in the modulation of bach1 levels.
To further confirm the crucial and protective role of SETD8 in DN, we performed SETD8 overexpression experiments in vivo. Our data illustrated that the increase in ELK1 and bach1 expression and EndMT, as well as the decline in renal function, were reversed by SETD8 overexpression in in vivo experiments. Thus, the SETD8/ELK1/bach1 axis may be a potential therapeutic target for blocking the occurrence and development of EndMT in DN.
There are some limitations in our study. First, HGECs were employed in our experiment; our results need to be verified in other basic endothelial cells. Second, how SETD8 and bach1 regulate Snail in high glucose-treated HGECs should be further investigated. Third, safe and effective SETD8 agonists that can be used in humans need to be further explored.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.