HDAC1‐mediated deacetylation of LSD1 regulates vascular calcification by promoting autophagy in chronic renal failure

Abstract Chronic renal failure (CRF) is commonly associated with various adverse consequences including pathological vascular calcification (VC), which represents a significant clinical concern. Existing literature has suggested the involvement of histone deacetylases (HDACs) in the progression of CRF‐induced VC. However, the underlying molecular mechanisms associated with HDACs remain largely unknown. Therefore, we established the adenine‐induced CRF rat model and in vitro VC models based on vascular smooth muscle cells (VSMCs) to examine HDAC1/lysine demethylase 1A (LSD1)/SESN2 as a novel molecular pathway in CRF‐induced VC. Our initial results demonstrated that HDAC1 reduced the formation of VC in vivo and in vitro. HDAC1 was found to deacetylate LSD1, which subsequently led to impaired transcriptional activity in CRF‐induced VC. Moreover, our results illustrated that LSD1 diminished the enrichment of H3K4me2 at the SESN2 promoter. Autophagy was identified as a vasculo‐protective element against calcification in VC. Finally, we found that the inhibitory effects of HDAC1 overexpression on VC were partially abolished via over‐expressed LSD1 in adenine‐induced CRF model rats and in high phosphate‐induced VSMCs. Taken together, these results highlight the crucial role of HDAC1 as an antagonistic factor in the progression of VC in CRF, and also revealed a novel regulatory mechanism by which HDAC1 operates. These findings provide significant insight and a fresh perspective into promising novel treatment strategies by up‐regulating HDAC1 in CRF.


| Ethics statement
The current study was performed with the approval of the Ethics

Committee and Experimental Animal Ethics of Yijishan Hospital of
Wannan Medical College (201902042). Extensive efforts were made to minimize animal numbers and their discomfort.

| Adenine-induced CRF rat models
A total of 40 male Sprague-Dawley rats (aged 5-7 weeks old and weighing 200-220 g) purchased from the Experimental Animal Center of the Chinese Academy of Sciences (Shanghai, China) were used in this study. The rats were fed with modified chow containing 0.75% of adenine for first establishment of the CRF model. They then were placed on an additional high-phosphorus diet (1.3% phosphorus, 1.06% calcium, 1000 IU/kg vitamin D3 and 23% protein) to induce VC.
The rats were injected with 4 × 10 8 pfu shRNA lentivirus vector (prepared by Shanghai GenePharma Co. Ltd.) or 8 μL in vivo transfection reagent (Polyplus-transfection Inc) mixed in 600 μL of 10% glucose via tail vein, 16,17 with ongoing access to normal forage. The rats were housed under controlled conditions at 22°C, 50% humidity with a 12hour light/dark cycle and granted free access to food and water. At five weeks after transfection, the rats were anaesthetized with 2% pentobarbital sodium (30 mg/kg) for collection of blood via puncture of the abdominal aorta. The collected blood was heparinized and centrifuged, and the plasma was stored at −80°C for later use. Next, the rats were killed, and a 5 mm segment of the thoracic aorta was resected and fixed by immersion in 4% paraformaldehyde for histological observation. The remaining portion of the aorta was used to determine the intravascular calcium content. In addition, the kidneys were resected and fixed by immersion in 10% formalin for histological analysis. Life Technologies) at 37°C for 30 minutes with gentle shaking. The separated tissues were then added to a culture dish and incubated in DMEM supplemented with 10% foetal bovine serum (FBS) at 37°C with 5% CO 2 for attachment. To induce VC in VSMCs, the cells were then cultured with a calcification medium (pH = 7.4) containing 2 mmol/L inorganic phosphate (Pi) and 3 mmol/L calcium (Ca 2+ ) 18 for a period of 3-6 days, with the medium replaced every two days. The first day of culture in the calcification medium was denoted as day 0. On the sixth day, the cells were rinsed twice with PBS, followed by von Kossa staining to determine the calcium deposition.

| Calcified nodule formation evaluated by alizarin red S staining
Alizarin red S staining was applied to the extracted thoracic aortas. In brief, the thoracic aortas were collected and stored in 70% ethanol. Each tissue sample was then placed in 10 mL of alizarin working solution supplemented with 0.8% alizarin red S and 0.5% KOH and rotated for 24 hours. The unstained tissues were then removed by rotating for an additional 24 hours in 10 mL of 0.05% KOH.
Finally, formation of calcified nodules was analysed under a microscope (Carl Zeiss).

| Determination of calcium content
The cells and tissues were decalcified with 0.6 mol/L HCl at 4°C for 24 hours. The calcium content of the HCl supernatant was determined colorimetrically in accordance with the manufacturer's instructions of the Calcium assay kit (QuantiChrom™ Calcium Assay Kit, BioAssay Systems). In brief, 5 μL of the samples was transferred into a 96-well plate after which absorbance was measured at 570 nm using a microplate ELISA reader (BioTek Instruments). After decalcification, the cells were washed three times with PBS and subsequently solubilized by 0.1 mol/L NaOH and 0.1% SDS. The protein content was determined using a bicinchoninic acid (BCA) Protein Assay kit (Thermo Scientific Pierce). The calcium content of the VSMCs was normalized to the protein content, while that of the tissues was normalized to tissue dry weight.

| Evaluation of kidney function
Rat aortic blood samples collected after 12-14 hours of fasting underwent biochemical analysis. The levels of serum creatinine (SCr), blood urea nitrogen (BUN) and urine protein (U-pro) collected for 24 hours were measured by turbidimetry using an automated analyser Falcor 300 (Menarini Diagnostics Corp.) in conjunction with a protein array instrument (BKMAM Biotechnology Co., Ltd.).

| Von Kossa staining for calcium deposition
Next, to identify calcification of the VSMCs, we conducted von Kossa staining. In brief, the cells were first fixed for 30 minutes with 10% formalin at ambient temperature, rinsed thrice with ddH 2 O and then incubated with 5% silver nitrate solution at ambient temperature for 30 minutes. The cells were then either exposed to ultraviolet light for 2 hours or overnight, until colour development was complete. The silver nitrate solution was discarded, whereupon the cells were washed using ddH 2 O and photographed under a microscope.

| Real-time quantitative polymerase chain reaction (RT-qPCR)
Total RNA content was extracted from the thoracic aorta tissues or VSMCs via a TRIzol kit (#16096020, Thermo Fisher The fold changes were calculated based on the 2 −ΔΔCt relative quantification method. peroxidase-labelled secondary goat anti-rabbit (ab205718, 1:2000, Abcam Inc) for 1 hour. Protein quantification analyses were conducted using ImageJ 1.48u software (National Institutes of Health), with the ratio of the grey value of the target band to β-actin indicating the relative protein expression.

| Cell treatment
The primary VSMCs in the thoracic aorta were collected and treated with 0.25% trypsin and passaged at a ratio of 1:3 for further ex-  Table 2.
In brief, the VSMCs in the exponential phase of growth were crosslinked for 10 minutes using 1% formaldehyde, followed by reaction with 125 mmol/L glycine at room temperature for 5 minutes. The

| Immunofluorescence staining
The VSMCs were cultured in a culture dish with cover glasses placed on top. When cell confluence reached 50%, the cover glass was removed. The cells were then rinsed three times using PBS and fixed in 4% paraformaldehyde for 30 minutes at room temperature. After 15 minutes of permeation using 2% Triton X-100, the cells were sealed for 45 minutes using 2% BSA. The sealing solution was then discarded, whereupon the cells were subjected to overnight incubation at 4°C with LC3 II antibody (ab63817, 1:100, Abcam Inc). After three PBS washes, the cells were re-probed with secondary goat anti-rabbit IgG H&L (ab150080, 1:400, Abcam Inc) at room temperature for 2 hours. Then 4′, 6-diamidino-2-phenylindole DAPI (2 μg/ mL) was added for cell staining followed by mounting on glass slides.

TA B L E 2 Oligonucleotide sequences for transfection
The expression of LC3 II was then detected under a fluorescence microscope, and the ImageJ software was used to quantify the fluorescence intensity.

| Statistical analysis
Statistical analysis was performed using SPSS 21.0 software (IBM Corp.). All measurement data were expressed as mean ± standard deviation (SD). Data following a normal distribution and with homogeneity of variance between two groups were compared using an unpaired t test. Data among multiple groups were compared by one-way analysis of variance (ANOVA) with Tukey's post hoc test.
Any P < .05 was considered to indicate a statistically significant difference.

| Down-regulation of HDAC1 was associated with the formation of VC in vivo and in vitro
To further investigate the expression of HDAC1 in VC in CRF models, the rats were first fed with forage containing 0.75% adenine to establish a CRF model, and then supplemented with a high-phosphorus diet to induce VC. We then examined the rat renal tissues with HE staining. The results depicted in Figure 1A indicate that the number of   detect the expression of HDAC1 in thoracic aorta, which showed low levels of HDAC1 in the CRF + VC group relative to the blank group ( Figure 1G). Next, to verify further the expression of HDAC1 in VC, the primary VSMCs of rats were treated with high Pi (2 mmol/L) in vitro, followed by von Kossa staining to evaluate their induced calcium deposition. The results in Figure 1H demonstrate notable signs of calcium deposition in VSMCs after high Pi induction. Western blot analysis revealed reduced expression of HDAC1 in the high Pi-induced VSMCs ( Figure 1I). Therefore, it was suggested that HDAC1 inhibition correlated with the formation of VC in vivo and in vitro.

| The HDAC1 reduced the formation of VC in vivo and in vitro
Next, to evaluate further the mechanism of HDAC1 in VC of CRF, VSMCs induced by high Pi were transfected with sh-HDAC1 or oe-HDAC1 plasmids, and their calcium deposition was detected by von Kossa staining. As illustrated in Figure 2A,

| Autophagy was vasculoprotective against calcification in vivo and in vitro
Western blot analysis was performed to detect the protein expression of LC3 II and p62 in thoracic aortas, the results of which are depicted in Figure 3A. In the CRF + VC group, the expression of LC3 II was elevated, while that of p62 was reduced. The expression of LC3 II in aortic tissues was detected by immunofluorescence staining, which showed that the expression of LC3 II was up-regulated

| HDAC1 deacetylated LSD1 and thus mediated autophagy in VC in vitro
RT-qPCR analysis showed that LSD1 expression was significantly increased in VSMCs following high Pi induction ( Figure 4A). Next, we used the ChIP to investigate the involvement of HDAC1 regulation on LSD1 acetylation in VSMCs induced by high Pi. Results showed increased HDAC1 enrichment in oe-HDAC1-treated cells ( Figure 4B).
Other ChIP assays showed reduced enrichment of H3K9ac in the LSD1 promoter region in response to oe-HDAC1, while treatment with the HDAC1 inhibitor SAHA (10 μmol/L) led to elevated H3K9ac enrichment in the LSD1 promoter region ( Figure 4C). Furthermore, RT-qPCR and Western blot analysis showed that the mRNA and protein expression of HDAC1 was increased by oe-HDAC1, and the mRNA and protein expression of LSD1 was inhibited by oe-HDAC1, while treatment with SAHA had opposite effects ( Figure 4D,E). The expression of LC3 II and p62 in cells following different treatments was detected by Western blot analysis, which revealed elevated LC3 II expression but reduced expression of p62 in cells with oe-HDAC1 treatment. However, LC3 II expression was reduced, while p62 expression was elevated in SAHA-treated cells, but these effects were abrogated following additional oe-HDAC1 treatment ( Figure 4F).
The formation of autophagosomes was subsequently examined by electron microscopy, which revealed that the cells treated with oe-HDAC1 had increased abundance of autophagosomes, while SAHA treatment led to a decrease in autophagosomes ( Figure 4G). Based on these findings, we conclude that HDAC1 inhibits the expression of LSD1 by modifying the promoter region of LSD1 via H3K9ac, which consequently stimulates cell autophagy.

| LSD1 reduced the expression of SESN2, thus inhibiting autophagy and promoting VC in vitro
Next, to elucidate whether LSD1 affects autophagy through regulating SESN2 in high Pi-induced VSMCs, we overexpressed or knocked down LSD1 in high Pi-induced VSMCs. RT-qPCR and Western blot analysis ( Figure 5A,B) detected elevated SESN2 expression in response to sh-LSD1 treatment, but reduced SESN2 expression following oe-LSD1 treatment, suggesting that LSD1 could indeed regulate the expression of SESN2. ChIP assay to detect the enrichment of LSD1 in the promoter region of SESN2 ( Figure 5C) demonstrated increased enrichment of LSD1 in oe-LSD1-treated cells.

F I G U R E 3 Autophagy confers vasculoprotective against calcification in adenine-induced CRF rats and high phosphate-induced VSMCs.
A, Western blot analysis of LC3 II and p62 proteins in renal tissues. B, Immunofluorescence staining of LC3 II in renal tissues (400×). C, Western blot analysis of LC3 II and p62 proteins in VSMCs. D, Immunofluorescence staining of LC3 II in VSMCs (400×). E, Von Kossa staining of calcium deposition in VSMCs (200×). F, Calcium content in cell supernatant measured using a colorimetric method. G, Western blot analysis of Runx2 and α-SMA proteins in VSMCs. BUN, blood urea nitrogen; CRF, chronic renal failure; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HDAC1, histone deacetylase 1; Pi, inorganic phosphate; Runx2, Runt-related transcription factor 2; SCr, serum creatinine; U-pro, urine protein; VC, vascular calcification; VPA, valproic acid; VSMCs, vascular smooth muscle cells; α-SMA, α-smooth muscle actin. *P < .05 indicates significant difference. Data (mean ± SD) between two groups were analysed using unpaired t test, while data among multiple groups were assessed using one-way ANOVA with Tukey's post hoc test. The experiment was run in triplicate  Figure 5F). Electron microscopy showed that the cells treated with sh-LSD1 had increased autophagosome density, while sh-SESN2 treatment led to reduced autophagosomes ( Figure 5G).
Alizarin red staining showed a reduced calcium deposition in sh-LSD1treated cells, whereas increased calcium deposition was observed in sh-SESN2-treated cells ( Figure 5H). Colorimetric analysis of cell supernatants showed reduced calcium content following sh-LSD1 treatment, whereas sh-SESN2 treatment resulted in increased calcium content ( Figure 5I). These results constituted evidence that LSD1 could inhibit autophagy and may potentially contribute to deteriorated VC by regulating demethylation of SESN2.

| The novel molecular pathway of HDAC1/ LSD1/SESN2 was validated in CRF-induced VC
We have seen that the HDAC1/LSD1/SESN2 pathway could regulate the calcification of VSMCs induced by high Pi by affecting autophagy. To further verify this finding in vivo, rats with CRF were treated with lentiviruses, and effects on HDAC1, LSD1 and SESN2 expression were subsequently detected by RT-qPCR. The results revealed up-regulated HDAC1 and SESN2, but down-regulated LSD1 in aortic tissues of rats treated with oe-HDAC1 ( Figure 6A). In contrast, the oe-LSD1-treated rats exhibited no notable changes in HDAC1 expression, but did show elevated expression of LSD1 and reduced expression of SESN2. The rats treated with oe-HDAC1 + oe-LSD1 showed increased HDAC1 expression but no notable changes in the expression of LSD1 and SESN2. Pathological examination of renal tissue with HE staining ( Figure 6B) showed that rats treated with F I G U R E 4 HDAC1 deacetylates LSD1 and thus mediates autophagy in high phosphate-induced VC models. A, LSD1 expression in VSMCs without treatment or treated with high Pi detected using RT-qPCR. B, The enrichment of HDAC1 in the promoter region of LSD1 measured using the ChIP assay. C, The enrichment of H3K9ac in the promoter region of LSD1 measured using ChIP assay. D, LSD1 expression in cells treated with oe-HDAC1 or SAHA (inhibitor of HDAC1) detected using RT-qPCR. E, Western blot analysis of LSD1 and HDAC1 protein in cells. F, Western blot analysis of LC3 II and p62 proteins in cells. G, The formation of autophagosomes observed under an electron microscope (20 000×). BUN, blood urea nitrogen; CRF, chronic renal failure; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HDAC1, histone deacetylase 1; HE, haematoxylin-eosin; NC, negative control; oe, overexpression; Pi, inorganic phosphate; RT-qPCR, reverse transcription-quantitative polymerase chain reaction; Runx2, Runt-related transcription factor 2; SCr, serum creatinine; U-pro, urine protein; VC, vascular calcification; VPA, valproic acid; VSMCs, vascular smooth muscle cells; α-SMA, α-smooth muscle actin. *P < .05 indicates significant difference. Data (mean ± SD) between two groups were analysed using unpaired t test, while data among multiple groups were assessed using one-way ANOVA with Tukey's post hoc test. The experiment was run in triplicate with oe-HDAC1, but increased calcium deposition with oe-LSD1 treatment ( Figure 6F). Colorimetric analysis revealed a notable reduction in calcium content detected in rats treated with oe-HDAC1, while increased calcium content was detected following oe-LSD1 treatment ( Figure 6G). Finally, Western blot analysis of aortic tissues showed notably diminished expression of Runx2 upon oe-HDAC1 treatment, while α-SMA expression was increased significantly, whereas opposite effects were seen following oe-LSD1 treatment ( Figure 6H). These results suggested that the HDAC1/LSD1/SESN2 pathway likely regulated the VC process in rats with CRF.

| D ISCUSS I ON
Chronic renal failure is a severe clinical disorder, and widely considered to be the final manifestation of a wide variety of chronic kidney diseases. 19  A key initial finding of our study was that renal HDAC1 expression was reduced in rats with CRF in vivo and in vitro. Low HDAC1 expression has been reported in acute kidney injury in vitro, and the unhindered activity of HDAC1 has been regarded as a prerequisite factor for renal protection and regeneration following acute kidney injury. 8 The deletion of the HDAC1 gene from the ureteric bud cell lineage in mice leads to bilateral renal hypodysplasia. 22 Moreover, diminished HDAC1 expression has been frequently reported in studies of kidney ischemia-and reperfusion-induced injury. 23 Our results also demonstrated that HDAC1 could suppress VC in CRF both in vivo and in vitro. Previous literature has indicated VC to be the predominant contributor to cardiovascular dysfunction in CRF. 24 Interestingly, HDAC activity has been identified as a potential therapeutic target in VC due to its active post-translational modification. 25 The deletion of HDAC1 activity via genetic ablation or pharmacological inhibition is capable of augmenting the severity of VC. 10 Thus, enhanced HDAC1 expression could potentially attenuate VC following CRF.
An additional crucial finding of the current study was our observation that promotion of autophagy could inhibit VC in CRF.
Autophagy has been described as a homoeostatic mechanism whereby proteins and organelles are detached and subsequently recycled into a general metabolic pool. 26 Furthermore, autophagy has been regarded as a promising endogenous protective mechanism against Pi-induced VC via its effects on the reduction in matrix vesicle release. 27 Existing literature has suggested that HDAC1 is closely associated with the process of autophagy in mice owing to its ability to regulate skeletal muscle homoeostasis and autophagy flux. 11 Moreover, HDAC1 has been shown to bind to the promoter region of LSD1 and subsequently mediate the deacetylation of LSD1, which consequently affects the expression of LSD1. 13 Notably, the inhibition of LSD1, either by small interference RNA (siRNA) or pharmacological agents, provokes the activation of autophagy in various gynaecologic malignancies. 14 Consistent with the present findings, a previous report concluded that HDAC1 possesses the ability to inhibit the expression of LSD1 by modifying the promoter region of LSD1 via H3K9ac, thereby stimulating cell autophagy.
The current study also revealed that LSD1 could restrain autoph- VSMCs Ac activity and ultimately enhance autophagy. 29 A trend towards increased expression of SESN2 upon LSD1 inhibition was identified during our study, but the expression of SESN2 was markedly reduced in VSMCs following oe-LSD1 treatment. In addition, elevated H3K4me2 expression was observed upon sh-LSD1 treatment. These findings suggest that LSD1 may inhibit SESN2 expression through the demethylation of H3K4me2. Moreover, our in vivo experimental results also demonstrated that the HDAC1/LSD1/SESN2 axis may potentially regulate VC process in rats with CRF.
In conclusion, the in vitro and in vivo results of the current study provide consistent evidence demonstrating that the overexpression of histone deacetylase HDAC1 potentially impedes the onset of VC in CRF via LSD1 inhibition through the SESN2-dependent mTOR signalling pathway (Figure 7). These results provide novel insight into the inhibition of HDAC1-mediated LSD1 as a promising competitive new target in VC after CRF. However, there is a need to confirm present results in animal and cell models by analysis of the HDAC1/LSD1/ SESN2 pathway in tissue specimens from CRF-diagnosed patients.

ACK N OWLED G EM ENTS
We express our sincere gratitude to the reviewers for their valuable suggestions.

CO N FLI C T O F I NTE R E S T
The author declares no competing interest exists.

DATA AVA I L A B I L I T Y S TAT E M E N T
Research data not shared.