The ameliorative effect of terpinen-4-ol on ER stress-induced vascular calci cation depends on sirt1-mediated regulation of PERK acetylation

Endoplasmic reticulum (ER) stress-mediated phenotypic switching of vascular smooth muscle cells (VSMCs) is key to vascular calcification (VC) in patients with chronic kidney disease (CKD). Studies have shown that activation/upregulation of SIRT1 has a protective effect on CKD-VC. Meanwhile, although terpinen-4-ol has been shown to exert a protective effect against cardiovascular disease, its role and underlying mechanism in VC remain unclear. Herein, we explored whether terpinen-4-ol alleviates ER stress-mediated VC through sirtuin 1 (SIRT1) and elucidated its mechanism to provide evidence for its application in the clinical prevention and treatment of VC. To this end, a CKD-related VC animal model and β-glycerophosphate (β-GP)-induced VSMC calcification model were established to investigate the role of terpinen-4-ol in ER stress-induced VC, in vitro and in vivo. Additionally, to evaluate the involvement of SIRT1, mouse and VSMC Sirt1-knockdown models were established. Results show that terpinen-4-ol inhibits calcium deposition, phenotypic switching, and ER stress in VSMCs in vitro and in vivo. Furthermore, pre-incubation of VSMCs with terpinen-4-ol or a SIRT1 agonist, decreased β-GP-induced calcium salt deposition, increased SIRT1 protein level, and inhibited PERK-eIF2α-ATF4 pathway activation, thus, alleviating VC. Similar results were observed in VSMCs induced to overexpress SIRT1 via lentivirus transcription. Meanwhile, the opposite results were obtained in SIRT1-knockdown models. Further, results suggest that SIRT1 physically interacts with, and deacetylates PERK. Specifically, mass spectrometry analysis identified lysine K889 as the acetylation site of SIRT1, which regulates PERK. Finally, inhibition of SIRT1 reduced the effect of terpinen-4-ol on the deacetylation of PERK in vitro and in vivo and weakened the inhibitory effect of terpinen-4-ol against ER stress-mediated VC. Cumulatively, terpinen-4-ol was found to inhibit post-translational modification of PERK at the K889 acetylation site by upregulating SIRT1 expression, thereby ameliorating VC by regulating ER stress. This study provides insights into the underlying molecular mechanism of terpinen-4-ol, supporting its development as a promising therapeutic agent for CKD-VC.


Introduction
Medial vascular calci cation (VC) is frequently observed in patients with chronic kidney disease (CKD), and it increases the incidence of cardiovascular events and mortality. [1] VC promotes stiffness of the vascular wall, resulting in increased pulse pressure, left ventricular hypertrophy, and heart failure. [2] Studies have identi ed the phenotypic switching of vascular smooth muscle cells (VSMCs) as a key event in VC. Osteoblastic differentiation is characterized by the reduction of VSMC markers, such as αsmooth muscle actin (α-SMA) and smooth muscle 22α (SM22α), and the upregulation of the expression of osteogenic markers, such as osteopontin (OPN), runt-related transcription factor 2 (Runx2), and bone morphogenetic protein 2 (BMP2). [3][4][5] However, the osteogenic/chondral transdifferentiation of VSMCs involves the regulation of multiple complex intracellular signal networks, which is not yet fully understood.
The endoplasmic reticulum (ER), as "the largest factory", is an important organelle in eukaryotic cells, is mainly responsible for lipid biosynthesis, Ca 2+ homeostasis, and the processing, folding, and secretion of almost all proteins. However, excessive activation of ER stress causes abnormal ER structure and function, thereby promoting the dissociation of the chaperone protein glucose regulatory protein 78 (GRP78) and causing the unfolded protein reaction (UPR) activation. [6,7] The classic UPR is mainly composed of three pathways: protein kinase RNA-like ER kinase (PERK), inositol-requiring enzyme 1 (IRE1), and activating transcription factor 6 (ATF6). [8,9] Among them, the PERK-eIF2α-ATF4 signaling is a key signal that induces VC. Phosphorylation of PERK subsequently leads to eukaryotic translationinitiation factor 2α (eIF2α) phosphorylation. Phosphorylated eIF2α can activate downstream activating transcription factor 4 (ATF4) in the nucleus to promote the phenotypic transformation of VSMCs, thereby promoting the development of VC. [10][11][12] However, the exact molecular mechanism of PERK-eIF2α-ATF4 signaling in CKD-dependent VC is still unclear.
Sirtuin-1 (sirt1) is an NAD + -dependent lysine deacetylase that can deacetylate various proteins and produce nicotinamide as a by-product, which then acts as a negative regulator of sirt1 activities, including regulation of histone and non-histone acetylation in the aging process, apoptosis, and energy metabolism in the anti-stress ability of cells. [13] A recent study showed that sirt1 downregulation promoted calci cation of VSMCs under osteogenic conditions [14,15] and that sirt1-knockdown mice showed accelerated calci cation induced by phosphate.
[16] Moreover, mechanistically, the downregulation of sirt1 expression promotes the acetylation of the Runx2 promoter region, which increases VSMC calci cation. [17] ER stress is an important mechanism in this phenomenon. Increasing evidence shows that sirt1 plays an active role in various ER stress-induced diseases. A study showed that decreased expression of sirt1 can promote the acetylation and phosphorylation of eIF2α, in which sirt1 regulates UPR by regulating eIF2α acetylation on lysine residues K141 and K143 as well as eIF2α phosphorylation on serine Ser51/Ser52. The PERK/eIF2α pathway protects cardiomyocytes from ER stress-induced apoptosis.
[18] More recently, Shufang Wu and colleagues [19] showed that sirt1 inhibition promotes both the hyperacetylation and phosphorylation of PERK, and then triggers the PERK-ATF4 signaling of ER stress. However, to date, there are no studies on whether and how sirt1 improves ER stress and inhibits VC at a molecular level by regulating the PERK-eIF2α-ATF4 axis of ER stress.
Terpinen-4-ol, a monomer component, is widely found in most plant essential oils and possesses effective anti-in ammatory, antitumor, and antibacterial effects. [20][21][22][23][24] At present, the protective effect of terpinen-4-ol on VC has not been reported. In this study, we aimed to investigate whether terpinen-4-ol ameliorates VC by regulating the PERK-eIF2α-ATF4 axis of ER stress via sirt1. Our ndings would provide a novel mechanism to support the potential of terpinen-4-ol as a promising candidate therapeutic agent for CKD.

Materials And Methods
Relevant materials and reagent information have been described in the supporting materials.

Animal treatment
The animal model was established according to the experimental method of Li et al. and Huang et al. [25,26] Seven-week-old male C57BL/6J mice were purchased from the Experimental Animal Center of Guizhou Medical University (Guizhou, China). After 1 week of acclimatization period, the mice were divided randomly into four groups and received the following treatment: control group, model group, and two terpinen-4-ol treatment groups. The control group was fed a normal diet, whereas the other groups were fed a high phosphorus diet supplemented with adenine at a dose of 0.2% (w/w). Two terpinen-4-ol treatment groups were administered terpinen-4-ol at a dose of 10 and 20 mg/kg/day respectively. After 6 weeks, all mice were euthanized with an overdose of Pentobarbital sodium (240 mg/kg, IP injection), and the thoracic arteries were collected and stored under speci c conditions for further experiments. All animal experiments followed the national guidelines and were approved by the Animal Ethics Committee of the Guizhou Medical University of Technology.
Cell culture and treatment VSMCs were cultured in a culture medium with DMEM supplemented with 10% FBS, 100 ng/mL FGF cytokine and 1% streptomycin and penicillin at 37 ℃ and 5% CO 2 .

Immunohistochemistry
According to standard procedures, artery sections were depara nized and rehydrated. Next, the sections were immersed in 0.05 M sodium citrate buffer (pH 8.0) for heat-mediated antigen retrieval and then in 3% hydrogen peroxide for 10 min to remove endogenous peroxidase. After that, the slides were blocked with 10% goat serum at 37 ℃ for 30 min and then incubated with the primary antibody overnight in a humid chamber at 4 ℃. The slides were incubated with an appropriate secondary antibody at 37 ℃ for 30 min and then reacted with 3,3′-diaminobenzidine solution. The tissue sections were observed and then photographed.

Immuno uorescence
VSMCs were seeded in a six-well plate containing coverslips. After treatment, the adherent cells were gently washed with cold PBS three times, xed with 4% paraformaldehyde for 20 min, and permeabilized with 0.2% Triton-X100 for 15 min. After blocking with goat serum for 40 min, the coverslips were incubated with primary antibodies overnight at 4 ℃. After washing with PBS, the coverslips were incubated with FITC-conjugated secondary antibodies for 1 h at room temperature, followed by incubation with DAPI for 30 min to stain the nucleus. Finally, immuno uorescence images were captured using a Leica DMi8 microscope and the Leica X software (Wetzlar, Germany) at ×200 magni cation.
Assessment of ALP activity and calcium content ALP activity was detected using a speci c kit according to the manufacturer's instructions. The results were normalized to the total protein level determined with the BCA protein assay kit to correct the ALP activity in the cells.
Cytosolic Ca 2+ levels were measured via ow cytometric estimation using Fluo-4 AM. [29,30] The cells were collected and incubated with 5 μmol/L Fluo-4 AM for 30 minutes at 37℃ in the dark, and then resuspended in 500 μL of phosphate buffered saline (PBS). The uorescence intensity was recorded at Ex/Em = 488/525 nm using a ow cytometer and analyzed using the NovoExpress software (NovoCyte, ACEA Biosciences, San Diego, CA, USA).

Alizarin red staining
Alizarin Red staining was performed to detect calcium nodules, as described previously. [31,32] Thoracic aortic rings and VSMCs were xed with 4% paraformaldehyde for 30 minutes at room temperature. After rinsing with PBS, incubation of 1% alizarin red S (pH 4.2) for 30 min was used to stain followed by rinsing ve times with PBS, and then photos were captured with a Leica DMi8 microscope (Wetzlar, Germany).

Sirius Red Staining
The mice thoracic aortic were were xed with 4% paraformaldehyde and embedded in para n. After conventional dewaxing treatment, samples were stained in picro-sirius red solution (0.1% with 1.2% picric acid). Finally, specimens were dehydrated with ethanol. Then, uorescent images were captured using P250 Pannoramic Scanner (3D Histech, Hungary) and observed by Caseviewer 2.3.

Lentivirus transfection and RNA interference of sirt1
LV overexpressing sirt1 was designed and synthesized by Shanghai Gene Chemistry Co., Ltd. (Shanghai, China). LV expressing GFP was used as a carrier negative control (NC-GFP). According to the manufacturer's instructions, VSMC was transfected with LV-sirt1 or LV-NC. Twenty-four hours after transfection, the medium was removed, and then the cells were incubated in a medium containing β-GP with or without addition of terpinen-4-ol.
Quantitative reverse-transcription PCR (qRT-PCR) Total RNA was extracted from cells, and the expression of related mRNA in VSMCs was determined by quantitative real-time polymerase chain reaction (qRT-PCR) in accordance with the manufacturer's instructions and detected in a CFX96 real-time PCR detection system (Bio-Rad, Hercules, CA, USA). The housekeeping gene GAPDH was used as an endogenous control to normalize the amount of RNA in each sample. qRT-PCR was performed using primers purchased from Sangon Biotech (Shanghai, China). The primers are listed in After treatment, total extraction of VSMCs or mice thoracic aorta was lysed in lysis buffer containing 99% e cient RIPA tissue/cell fast lysis solution (R0010) and 1% PMSF (R0100) from Solarbio. Protein concentration in the supernatant was detected using a BCA Protein Assay Kit with a microplate spectrophotometer (Thermo Fisher Scienti c, Inc.). The total proteins (20-40 μg) were separated by 10% or 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to a polyvinylidene uoride membrane (Cat#T8060; Solarbio). The membranes were blocked with 5% bovine serum albumin for 2 h at room temperature and then incubated with the appropriate primary antibodies overnight at 4 ℃. The membranes were washed with Tris-buffered saline containing Tween 20 and then incubated with secondary antibodies for 2 h at room temperature. The protein blot intensities were quanti ed using the Image Lab Software (Bio-Rad) and normalized to the housekeeping protein (GAPDH) levels.

Co-immunoprecipitation assay
The total protein was extracted from lysis buffer and incubated with a speci c antibody against PERK, sirt1, or acetyl-lysine under frequent mixing overnight at 4 ℃ and added into Protein A/G-agarose beads (Millipore, USA) to generate protein complexes, followed by incubation overnight at 4 ℃ for immunoprecipitation. Immunoprecipitated protein complexes were washed with wash buffer at least six times, boiled in SDS sample buffer for 10 min, and subjected to immunoblotting as described above using an acetyl-lysine antibody (1:200), sirt1 antibody (1:200), or PERK antibody (1:100).

Mass spectrometry analysis
To identify the acetylated sites of PERK, cell lysates were collected and incubated with anti-PERK antibody overnight at 4 ℃ for immunoprecipitation. The immunoprecipitated PERK protein was separated by SDS-PAGE (Figure in the online-only supplemental data). The gel fragments were subjected to in-gel trypsin digestion, and then extracted with 50% acetonitrile/5% formic acid and 100% acetonitrile. The peptide was dried to completion and resuspended in 2% acetonitrile/0.1% formic acid. The peptide was dissolved in 0.1% formic acid and loaded onto a self-made reverse phase analytical column (length 15 cm, inner diameter 75 μm). The peptides were processed from NSI sources and then tandem mass spectrometry (MS/MS) was performed in Q ExactiveTM Plus (Thermo, USA) connected online to UPLC.
Site modi cations were performed by PTM Aims (Shanghai, China), as shown in Figure 6e.
Data and statistical analysis All data are representative of more than three independent experiments, the graph shows the mean + standard deviation. Statistical analysis was performed in GraphPad Prism 7.0 (Inc, La Jolla, CA). multiple sets of data were analyzed through one-way analysis of variance, followed by the Bonferroni post-hoc test. Differences between two groups were assessed by Student's t-test. P-values < 0.05 were considered signi cantly different ( * P < 0.05; ** P < 0.01; *** P < 0.001).

Results
Terpinen-4-ol improves VC in CKD mice induced by adenine After 4 weeks of intragastric adenine administration, mice thoracic aorta was calci ed, as indicated by increased expression of the VSMC osteogenic phenotype-related markers ALP, BMP2, and Runx2 and decreased expression of the contractile phenotype markers α-SMA; both of these changes were timedependent (Fig. 1b). Western blotting results showed that the expression of α-SMA in the blood vessels of CKD mice was signi cantly downregulated, whereas that of BMP2, Runx2, and ALP was upregulated. Nevertheless, terpinen-4-ol upregulated the expression of α-SMA in calci ed blood vessels and inhibited the expression of BMP2, Runx2, and ALP (Fig. 1c). These results are consistent with the results of immuno uorescence assay (Fig. 1d). We con rmed that in the thoracic aorta of CKD-VC mice, the expression of Runx2 was upregulated, whereas that of α-SMA was downregulated, and that these effects were reversed by terpinen-4-ol treatment. Alizarin red staining were conducted to evaluate the calci cation and of the thoracic aorta of mice in each group. The results showed that the vascular media of CKD mice showed the typical orange-red staining, suggesting calci cation of vascular media, this pathological change was reduced in the terpinen-4-ol treatment groups compared with model group. Sirius red staining were conducted to evaluate the expression of collagen bers in the thoracic aorta of mice. The results showed that compared with the normal group, the accumulation of vascular collagen bers in the vascular media of CKD mice increased, and the rupture of vascular media elastic bers, structural disorders, partial irregular arrangement of blood vessels, and loss of normal wavy contractions were observed. After terpinen-4-ol treatment, the vascular collagen content decreases, and the vascular elastic bers tend to be regular and continuous (Fig. 1e). The above results suggest that terpinen-4-ol reduces vascular calcium salt deposition in CKD mice and inhibits phenotypic switching of VSMC.
Terpinen-4-ol inhibits β-GP-induced phenotypic switching and calcium deposition in VSMCs It has been shown that 10 mM β-GP enhanced calcium deposition and phenotypic switching in VSMCs.
[32] First, we determined changes in the levels of classic osteoblastic differentiation-related markers, and our nding showed signi cant upregulation of Runx2, ALP, and BMP2 protein levels as well as downregulation of α-SMA protein levels (Fig. 2a) in VSMCs stimulated with β-GP for 3 to 7 days. In addition, after treatment with β-GP, VSMCs showed obvious calci cation with many calci ed nodules, as revealed by alizarin red S staining (Fig. 2b). Then, we determined the changes in the levels of classic osteoblast differentiation markers (including Runx2, ALP and BMP2) and the VSMC marker α-SMA. Our results showed that terpinen-4-ol enhanced α-SMA expression and reduced Runx2, ALP, and BMP2 expression at both the protein and mRNA levels (Fig. 2c, d). These results showed that co-treatment with terpinen-4-ol reversed the β-GP-induced changes in osteoblastic differentiation-related markers.
The effect of terpinen-4-ol was further con rmed by immuno uorescence analysis. In β-GP-induced VSMCs, the expression and nuclear translocation of Runx2 increased and the expression of α-SMA decreased compared with those in the control; however, these changes were restored by terpinen-4-ol treatment (Fig. 2e). In addition, terpinen-4-ol inhibited the β-GP-induced increase in BMP2 uorescence intensity (Fig. 2f). VSMCs were stimulated with β-GP and different concentrations of terpinen-4-ol (5 and 20 μM) for 7 days, and the results showed that terpinen-4-ol reduced the calcium deposition of VSMCs ( Fig. 2g) and ALP activity (Fig. 2h). However, the increase in intracellular Ca 2+ content was inhibited by terpinen-4-ol, as con rmed via ow cytometry using the Ca 2+ -sensitive uorescence indicator Fluo-4 AM (Fig. 2i). These data suggest that terpinen-4-ol can suppress phenotypic switching and calcium deposition in VSMCs.
Terpinen-4-ol alleviates VC by inhibiting PERK-eIF2α-ATF4 pathway in vivo and in vitro Studies have indicated that the CKD environment could induce the activation of the PERK-eIF2α-ATF4 signaling pathway, which leads to ER stress and the development of VC. [33,34] Taking into account our current ndings, we aim to reveal the mechanism of potential endorphins protection. terpinen-4-ol against VC. Therefore, we evaluated the protein expression of the PERK-eIF2α-ATF4 pathway. Western blotting results revealed that the expression of p-PERK, p-eIF2α, and ATF4 increased in CKD mice, but this increase was restored by terpinen-4-ol (Fig. 3a). In vitro experiments in VSMC further proved above results (Fig. 3b).
The expression of proteins of the PERK-eIF2α-ATF4 pathway was decreased by terpinen-4-ol treatment for 7 days. To examine whether the PERK-eIF2α-ATF4 signaling pathway is related to the effect of terpinen-4ol on VC, we further used ER stress agonist tunicamycin (TM, 0.1μmol/L) and inhibitor 4-PBA (5 mmol/L) to clarify the effect of terpinen-4-ol on VC. [35,36] Flow cytometry con rmed that 4-PBA inhibited the β-GP-induced increase in Ca 2+ in VSMCs (Fig. 3c). The results showed that β-GP and TM activated ER stress, upregulating the expression of osteogenic marker proteins and ER stress-related proteins; however, terpinen-4-ol reversed these effects. Notably, we found that 4-PBA decreased the expression levels of GRP78, p-PERK, p-eIF2α, ATF4, Runx2, ALP, and BMP2, showing no signi cant difference from those in the group co-treated with terpinen-4-ol and 4-PBA (Fig. 3d, e). Overall, these results indicate that terpinen-4-ol alleviates ER stress in vivo and in vitro, which may help improve CKD-VC.

Terpinen-4-ol inhibits VC by upregulating sirt1
It has been reported that the upregulation of sirt1 can improve the VSMC phenotype, whereas the downregulation of sirt1 contributes to CKD-related VC. [37] In the present study, we found that terpinen-4ol upregulated the expression of sirt1 (Fig. 4a, b). In vivo results showed that sirt1 was down-regulated in CKD mouse aortas but was up-regulated in mice treated with terpinen-4-ol (Fig. 4c, d). To determine the regulatory effect of sirt1 on the phenotype transformation of VSMCs, we performed an in vitro loss-offunction assay in VSMCs (Fig. 4e). Overexpression of sirt1 was achieved by transfecting VSMCs with a stable LV (Fig. 4g, h). VSMCs transfected with sirt1 siRNA or negative control were cultured with β-GP to induce osteoblast differentiation. The results showed that siRNA-mediated silencing of sirt1 in VSMCs further increased the β-GP-induced increase in calcium salt deposition (Fig. 4f). Silencing of sirt1 further increased the β-GP-induced expression of ALP, BMP2, and Runx2, and further decreased the expression of α-SMA protein. In contrast, LV-mediated sirt1 overexpression decreased the protein levels of osteogenic markers and increased the expression of α-SMA (Fig. 4i). These results indicate that sirt1 plays a negative role in the osteoblast differentiation of VSMCs. We nd that no signi cant changes in the expression of BMP2 and Runx2 were observed after terpinen-4-ol treatment (Fig. 4j). These results once again proved that knockdown sirt1 blocked the effect of terpinen-4-ol on the phenotype transformation of VSMCs.

Terpinen-4-ol alleviates β-GP-induced ER stress in a sirt1-dependent manner
To examine the protective of sirt1 in ER stress-mediated VC, we conducted an experiment using the sirt1 activator resveratrol (50 μmol/L).
[38] We noted that terpinen-4-ol and resveratrol reduced the expression of ER stress markers (Fig. 5a). We nd that higher expression of p-PERK, p-eIF2α, and ATF4 in the transfected sirt1 siRNA in β-GP-induced VSMCs. It is worth noting that in VSMCs with silenced sirt1, no signi cant changes were observed after terpinen-4-ol treatment (Fig. 5b). In contrast, sirt1 overexpression further decreased the protein levels of ER stress markers in VSMCs (Fig. 5c).These results indicate that terpinen-4-ol could regulate PERK-eIF2α-ATF4 pathway to inhibit VC by upregulating the expression of sirt1.
We further examined the effects of sirt1 gene silencing on VC in adenine-induced CKD mice. Knockdown of the sirt1 gene in vivo was con rmed by both uorescence microscopy and western blotting (Fig. 4d). After 6 weeks of adenine administration, western blotting results showed no change in the expression of ER stress and osteogenic differentiation marker proteins following NC-GFP injection compared with that in the normal group. Compared with NC-GFP, LV-sirt1 RNAi decreased the expression of α-SMA and increased the expression of BMP2, Runx2, p-PERK, p-eIF2α, and ATF4, suggesting that sirt1 knockdown promoted ER stress and VC, consistent with the in vitro results. Moreover, LV-sirt1 RNAi injection signi cantly reduced the expression of sirt1 and the terpinen-4-ol-induced decrease in the expression of ER stress and VC-related markers in the aorta of adenine-fed mice, once again proving that sirt1 is a key signal involved in ER stress and may be an important molecular target of terpinen-4-ol in inhibiting VC.
We also found, through immuno uorescence staining, that PERK and sirt1 were co-localized in the thoracic aorta of CKD-VC mice (Fig. 5f). These results suggest that terpinen-4-ol can improve β-GP-induced ER stress by downregulating the expression of sirt1 and that sirt1 knockdown diminishes the ameliorative effect of terpinen-4-ol on ER stress.

Terpinen-4-ol improves ER stress-induced VC through sirt1-mediated PERK deacetylation
To con rm the role of PERK in ER stress-induced VC, we performed an experiment with terpinen-4-ol and the PERK inhibitor GSK2606414 (5 μmol/L). [39] The results showed that GSK2606414 suppressed the expression of ER stress markers and VC-related proteins.. In order to study the mechanism of sirt1 attenuating PERK signal pathway, the acetylated protein on the lysine residues were removed from the VSMC lysate. PERK is present in the anti-acetyl-lysine immunoprecipitate. Mutual immunoprecipitation further con rmed the acetylation of PERK (Fig. 6b). Moreover, knockdown of sirt1 greatly increased the acetylation of PERK (Fig. 6c). We also con rmed that sirt1 and PERK existed in a protein binding mode (Fig. 6d). The immunoprecipitation results showed that β-GP treatment promoted PERK acetylation, and terpinen-4-ol can reverse this effect by inhibiting the acetylation modi cation of PERK (Fig. 6e). Consistent with the in vitro ndings, terpinen-4-ol inhibited PERK acetylation in CKD mice (Fig. 6f). In VSMCs transfected with an empty vector, β-GP exposure increased PERK acetylation, which was signi cantly reversed by terpinen-4-ol treatment. In sirt1 siRNA-transfected VSMCs treated with or without terpinen-4-ol, PERK acetylation was increased (Fig. 6g). The in vitro results also showed that the inhibitory effect of terpinen-4-ol on PERK acetylation was signi cantly blocked by LV-sirt1 RNAi injection (Fig. 6h).
To identify the sites of PERK acetylation, PERK was subjected to immunoprecipitation, followed by proteolytic digestion and nano-LC-MS/MS analysis (Fig. 6i). The analysis revealed different trypsincleaved peptides near K889 (peptides ENLKDWMNR). In addition, an increase of 42 011 Da was observed in the mass of lysine K889, which is equivalent to an increase of an acetyl group. These results revealed not only that the PERK protein was acetylated but also that its prominent acetylation site was lysine K889. The PERK protein sequence is highly conserved among species(e.g. R. norvegicus, H. sapiens, D. Rerio and X. Tropicalis et.al) (Fig. 6j). The structure of PERK was predicted using I-TASSER, and the pymol algorithm was constructed for visualization. The crystal structure of PERK showed that K889 (red) was located in the protein kinase-like domain (yellow) (Fig. 6k). The above results suggest that terpinen-4-ol improves ER-induced CKD-VC by mediating PERK deacetylation through sirt1.

Discussion
CKD-VC threatens human health owing to its high morbidity and mortality, and this disease is di cult to treat. Mechanistically, the osteogenic phenotype transformation of VSMCs promotes the development of VC. [40][41][42] Therefore, preventing the phenotype switching of VSMCs may be a new therapeutic option for CKD-VC. Terpinen-4-ol, a monoterpene from aromatic plants, has previously been investigated for therapeutic potentials against VC by biological approaches. Our previous research showed that terpinen-4-ol inhibits oxidative stress damage in VSMCs induced by high glucose. [43] Thus, we purposed to elucidate the effect of terpinen-4-ol on CKD-VC. Disturbance of mineral homeostasis and abnormal deposition of calcium and phosphorus in blood vessel walls are core processes of VC. [44][45][46] Our present results showed that terpinen-4-ol reduced calcium deposition and ALP activity in CKD mouse arteries. Furthermore, our results con rmed that terpinen-4-ol inhibits the β-GP-induced increases in calcium deposition, calcium concentration, and ALP activity in rat VSMCs. However, evidence has shown that VC is not simply caused by calcium salt deposition and that phenotype switching of VSMCs. We showed that terpinen-4-ol inhibited the phenotype switching of VSMCs.
Accumulating evidence has indicated that ER stress contributes to the progression of CKD through increased VSMC differentiation. [47] It is well established that the PERK-eIF2a-ATF4-CHOP pathway is upregulated in animal models of VC. [11,48] In vitro models suggest that ER stress also increases the expression of ATF4, which binds to the Runx2 promoter, affecting VSMCs calci cation. [49][50][51] In order to clarify the relationship between ER stress and VC, we used 4-PBA, a classic ER stress inhibitor. 4-PBA signi cantly alleviated VC-related protein levels. This indicates that ER stress participates in the development of VC. Moreover, our results also indicated that terpinen-4-ol decreased the levels of ER stress markers and VC-related marker proteins, which were increased by the ER stress inducer TM. Taken together, this study showed for the rst time that the protective effects of terpinen-4-ol was related to the inhibition of the PERK signal pathway. However, it is not clear whether IRE1 and ATF6 signaling pathways of ER stress are related to the effects of terpinen-4-ol on VC, thus further research is needed.
Numerous studies have shown that sirt1 could attenuate VC by reversing the osteoblastic differentiation of VSMCs. [52] More importantly, the downregulation of sirt1 contributes to CKD-associated VC. [15,53] In addition, sirt1 activators, such as resveratrol, [54][55][56] have been proposed as therapeutic strategies for treating and preventing VC, as they can alleviate the calci cation of VSMCs by increasing the expression of the calci cation inhibitors such as OPG and OPN. More importantly, sirt1 directly regulates VC through the deacetylation of the Runx2 promoter. [14] Consequently, sirt1 possesses anti-calci cation activity. Consistent with these ndings, our study con rmed that sirt1 expression was signi cantly lower in β-GPinduced VSMCs and in the thoracic aorta of CKD mice, but sirt1 expression was upregulated by terpinen-4-ol treatment. Moreover, we used LV-sirt1 to induce sirt1 overexpression, and our results indicated that sirt1 inhibits the osteoblast differentiation of VSMCs, as evidenced by the reduced expression of osteoblast differentiation markers. In contrast, siRNA-mediated silencing of sirt1 promoted osteoblast differentiation of VSMCs. Altogether, these data suggest that sirt1 functions as a negative regulator of osteoblast differentiation in CKD-VC. Upregulation of sirt1 leads to inhibition of IRE1α, PERK, and ATF6 signaling pathways that activate ER stress. [57][58][59][60] We observed that pre-incubation of VSMCs with terpinen-4-ol or the sirt1 agonist resveratrol and transfection of VSMCs with LV overexpressing sirt1 (LV-sirt1) decreased β-GP-induced calcium salt deposition, inhibited ER stress, and improved the phenotypic transformation of VSMCs. These ndings following in vitro silencing of sirt1 and in vivo injection of LV-sirt1 RNAi are contrary to overexpressing sirt1 results. Therefore, our results provide that terpinen-4-ol alleviates β-GP-induced ER stress and osteoblast phenotypic switching of VSMCs in a sirt1-dependent manner The ability of cells to respond to ER stress is essential for cell survival, but the unrecoverable level of ER stress could lead to VC.[61, 62] The knockdown of the PERK-eIF2α-ATF4-CHOP pathway blocks osteoblastic differentiation in VSMCs. PERK siRNA reduced the protein levels of PERK and its downstream target, p-eIF2α, which notably diminished calci cation and ALP activity. [34,49] We found that the PERK inhibitor GSK2606414 suppressed ER stress and reversed the phenotypic switching of VSMCs. PERK is a strong positive regulator of ER stress and VC. Acetylation and deacetylation of proteins modulates a wide variety of cellular biological processes, such as cell proliferation, gene

Limitations
The present study has several limitations. We used LV to knockdown sirt1 in CKD mouse aortas. Although the results of uorescence microscopy and western blotting assay proved that sirt1 was effectively knocked-down in the mouse thoracic aorta, there are currently many studies using LV overexpression or interference of genes in the thoracic aorta, but it is better than when experiments were conducted on sirt1-/-knockout mice. In addition, we proved that sirt1 can regulate the acetylation/deacetylation of PERK, and PERK is acetylated at position K889; therefore, the function of acetylated PERK protein needs to be further clari ed by a point mutation in the next step.

Conclusion
In summary, our study showed that terpinen-4-ol can inhibit the ER-mediated phenotypic transformation of VSMCs by regulating PERK acetylation modi cation by sirt1. As shown in Fig. 7, this indicates that sirt1 is essential for terpinen-4-ol alleviated VC, which provides a novel mechanism supporting terpinen-4ol or activation of sirt1 signaling as a promising clinical therapeutic agent or strategy, respectively, for the treatment of CKD-VC.  Expression of α-SMA in mouse aortas was determined by immuno uorescent staining of aortic root cross-sections (scale bar: 50 μm). (e) Alizarin red staining and sirius red staining of mouse aortas (scale bar: 50 μm). All data represent the mean ± SEM of more than three experiments. *P < 0.05, **P < 0.01, ***P < 0.001.   Terpinen-4-ol improves ER stress-induced VC through sirt1-mediated PERK deacetylation. (a) VSMCs treated with or without β-GP (10 mmol/L) were treated with terpinen-4-ol (20 μmol/L) or the PERK inhibitor GSK2606414 (5 μmol/L) for 7 days. Western blotting analysis of p-PERK, PERK, p-eIF2a, eIF2a, ATF4, BMP2, and Runx2 was performed in VSMCs, and GAPDH was used as a loading control. (b) PERK was immunoprecipitated from VSMC lysates, and its acetylation was analyzed by immunoblotting with an anti-acetyl-lysine antibody. Immunoprecipitation of acetylated proteins from VSMC lysates was followed by immunoblotting with the indicated antibodies; Input: supernatant before immunoprecipitation; IP: immunoprecipitate; IgG: negative control. (c) VSMCs were transfected with sirt1 siRNA or negative control, PERK was immunoprecipitated, and its level of acetylation was determined by immunoblotting with an anti-acetyl-lysine antibody. Ratios of acetylated versus total PERK are presented.
(d) Physical interaction between endogenous sirt1 and PERK was shown by co-immunoprecipitation. sirt1 was precipitated from VSMC lysates with anti-sirt1 antibody and blotted with anti-PERK antibody, and vice versa. (e) Immunoprecipitation of acetylated proteins from VSMC lysates was followed by immunoblotting with the indicated antibodies. Input: supernatant before immunoprecipitation, IgG: negative control, rabbit IgG; IP: immunoprecipitated. (f) Acetylated lysine and total PERK expression in the aortas of C57BL/6J mice as described in Figure 1 by immunoprecipitation and western blotting analysis. The 3D crystal structure of PERK shows that K889 (red) is located in the protein kinase-like domain (yellow). Figure 7