LincRNA‐p21 alleviates atherosclerosis progression through regulating the miR‐221/SIRT1/Pcsk9 axis

Abstract Atherosclerosis (AS) is the main aetiology of coronary heart disease, cerebral infarction and peripheral vascular disease in humans. Long‐noncoding RNA (LincRNA)‐p21 has been reported to participate in the development of AS. Therefore, this study was designed to investigate the mechanism of LincRNA‐p21 on suppressing the development of AS. We fed ApoE−/− mice with a high‐fat diet to induce an AS mouse model where the lesion area of AS and the extent of lipid deposition were measured. The binding of LincRNA‐p21 and miR‐221 or miR‐221 and SIRT1 was measured using a dual luciferase reporter gene assay and RIP. Following loss‐ and gain‐ function assays, CCK8, EdU, Transwell assay and scratch test were performed to determine the biological processes of human aortic endothelial cells (HAECs). miR‐221 was highly expressed while SIRT1 was poorly expressed in AS. LincRNA‐p21 acted as a sponge for miR‐221. miR‐221 targeted and negatively regulated the expression of SIRT1. LincRNA‐p21 promoted the deacetylation of Pcsk9 by SIRT1 by competitively binding to miR‐221, whereby promoting HAEC proliferation, migration and tube formation. In conclusion, LincRNA‐p21 acted as a molecular sponge for miR‐221 to promote deacetylation of the promoter region of Pcsk9 by SIRT1, therefore preventing the development of AS.


| INTRODUC TI ON
Atherosclerosis (AS) is an arterial disease process that is characterized by focal intravascular aggregation of lipoproteins containing apolipoprotein B, immune and vascular wall cells and extracellular matrix. 1 AS is the basis of many high mortality rates of cardiovascular diseases worldwide, 2 which is induced by accumulation of lipids in the walls of the blood vessels. As a natural component of blood vessels, human aortic endothelial cells (HAECs) play a key role in maintaining the integrity of the cardiovascular system by serving as the barrier between the blood and the vascular wall, and HAEC damage is a key factor in the onset of AS. 3 Oxidized low-density lipoprotein (LDL) could activate HAECs, which exert pivotal function on the onset and development of AS. 4 Understanding the molecular mechanisms driving the phenotypic regulation of HAECs is therefore critical for the better treatment of AS.
A previous study has shown that miR-221 is involved in the development of AS. 5 miR-221 is widely distributed in eukaryotes and deeply involved in posttranscriptional regulation of gene expression. 6 Moreover, long intergenic noncoding RNA (LincRNA) acts as an endogenous RNA sponge for miRNA, regulates the expression of miRNA and interacts with miRNA to mediate the progression of AS. 7 LincRNA-p21 is located on chromosome 17, approximately 15 KB upstream of the Cdkn1a (p21) gene. 8 The relation between LincRNA-p21 and AS was discussed in previous studies; for example, LincRNA-p21 is especially poorly expressed in AS plaques of apolipoprotein E-deficient (ApoE (−/−) ) mice and LincRNA-p21 inhibits cell proliferation and induces apoptosis of vascular smooth muscle cells and mouse mononuclear macrophages in vitro. 9 LincRNA-p21 acted as a sponge for miR-221. Next, Sirtuin 1 (SIRT1) was predicted to be a downstream target of miR-221 through bioinformatics website. The epigenetic regulation of SIRT1 is controlled by miRNA. 10 SIRT1 and NAD (+)-dependent histone deacetylase exert significant effects on a variety of biological processes, including lifespan, stress response and cell survival. 11 SIRT1 is downregulated and involved in the development of AS. 12,13 Besides, lack of SIRT1 has been linked to an increased expression of the proprotein convertase subtilisin/kexin type 9 (Pcsk9) gene and increased LDL-cholesterol levels. 14 Pcsk9 is a secreted protein that promotes low density lipoprotein receptor degradation, hence regulating the plasma level of LDL cholesterol. 15 Furthermore, Pcsk9 has been revealed to be a gene that regulates and promotes AS. 16 Therefore, the objective of this study is to investigate the mechanism of LincRNA-p21 preventing the development of AS through miR-221/SIRT1/Pcsk9 axis using ox-LDL-treated HAECs and AS mouse model.

| Ethics statement
The study was conducted with the approval of Huaihe Hospital of

| Collection of clinical samples
Peripheral blood samples from 25 patients with AS aged 50.80 ± 6.72 years (AS-H group, among which 14 males and 11 females were included) and 18 healthy volunteers aged 54.72 ± 6.11 years (Normal-H group, among which 10 males and 8 females were included) were collected from Huaihe Hospital of Henan University from February 2018 to January 2019.

| AS mouse model induction
Sixty specific pathogen-free male ApoE −/− mice and 15 normal mice (C57BL/6; aged eight-week and weighting 16-21 g) were obtained from Beijing University of Medicine Laboratory (Beijing, China). All mice were housed at 18-23°C with a humidity of 40-60%. Normal mice (Normal-M group) were fed with basal diet and ApoE −/− mice (AS-M group) fed with a high-fat diet (21% fat and 0.25% cholesterol) for 12 weeks. After that, 15 AS modelled mice set aside as the control, and the remaining 45 modelled mice were injected with lentiviral vectors expressing oe-LincRNA-p21, oe-Pcsk9, or oe-NC (n = 15). 200 μL of corresponding adenovirus (Fubio Biological Technology Co., Ltd.) was injected into the tail veins of mice once a week (1 × 10 10 pfu/mL) for 12 weeks, followed with intraperitoneal injection with 3% pentobarbital sodium (P3761; Sigma-Aldrich) in anaesthetized mice; carotid artery tissues were collected at the last step.
All shRNA, mimic and inhibitors were purchased from Sangon Biotech Co., Ltd. Plasmid construction was completed by the company according to the protocols. HAECs in the exponential growth phase were treated with 0.25% trypsin-EDTA, passaged and counted (1 × 10 6 cells/mL). HAECs were then plated into a six-well plate, and shRNA, mimic or inhibitor were mixed with Lipofectamine 2000 transfection reagent (11668019; Thermo Fisher Scientific), and transfected into the cells when cell confluence reached 60%.
After 24 hours of transfection, the cells were collected for further analysis.

| Haematoxylin and eosin and Oil red O staining
Haematoxylin and eosin (HE) and Oil red O staining were performed according to the methods mentioned in a study. 17 The atherosclerotic lesions in mice aortic sinus were observed under the microscope (BX63, Olympus Optical Co., Ltd.).

| Reverse transcription quantitative polymerase chain reaction (RT-qPCR)
Total RNA content of tissues and cells was extracted with TRIzol (16096020; Thermo Fisher Scientific). RNA of miRNA was reversely transcribed using reverse transcription kit TaqMan ™ MicroRNA Reverse Transcription Kit (4366596; Thermo Fisher Scientific) and the RNA of mRNA was reversely transcribed using reverse transcription kit High-Capacity cDNA Reverse Transcription Kit (4368813; Thermo Fisher Scientific). The primer sequences of lincRNA-p21, miR-221, SIRT1, Pcsk9, GAPDH and U6 were designed and provided by Sangon Biotech. The primer sequences are shown in Table S1, with GAPDH as internal reference. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) kit (11732020; Thermo Fisher Scientific) was used to carry out the RT-qPCR experiment according to the instructions using 2 −ΔΔCt method.

| Western blot analysis
The total protein content of tissues and cells was extracted by following the instructions provided by the protein extraction kit (78501; Thermo Fisher Scientific). The total protein concentration of each sample was determined using the BCA kit (23229; Thermo Fisher Scientific). The protein sample was then added to each lane with a micro sampler for electrophoretic separation and transferred onto the PVDF membrane (1620177; Bio-Rad, Hercules, CA, USA). The membranes were blocked in 5% BSA for 1 hour, added with primary and secondary antibodies (Abcam; Table 1), followed by incubation for 1 hour. The membrane was immersed in ECL reaction solution (1705062; Bio-Rad) for 1 minute and exposed by an ImageQuant LAS 4000 mini imaging system (GE Healthcare). U6 served as the internal reference for miR-221, while glyceraldehyde-3phosphate dehydrogenase (GAPDH) as the internal reference, the relative expression level of each protein was measured by the ratio of grey value between the target band and the internal reference band.

| Dual luciferase reporter gene assay
The online analysis website (http://starb ase.sysu.edu.cn/) was used to predict targeted binding sites of miR-221 to SIRT1. Synthesis of miR-221 and LincRNA-p21, miR-221 and SIRT1 gene fragments containing targeted binding (LincRNA-p21 WT and SIRT1 WT) and mutant fragments (LincRNA-p21 Mut and SIRT1 Mut) were introduced into the PmirGLO vector (LM1439; LMAI Bio) using T4 DNA ligase (M0204S; New England Biolabs). Renilla fluorescent plasmids and constructed luciferase reporter plasmids were co-transfected into HEK293T cells with mimic-NC and miR-221 mimic. Cells were collected and lysed after 48 hours of routine transfection. Dual luciferase ® Reporter Assay System kit (E1910; Promega Corporation) was applied for luciferase activity measurement by using a GloMax ® 20/20 Luminometer detector (E5311; Promega Corporation) and relative luciferase activity was determined as previously conducted.
All vectors were constructed by Sangon Biotech.

| RNA immunoprecipitation
An RNA immunoprecipitation (RIP) kit (17-701; Millipore) was used to detect the possible binding sites between LincRNA-p21 and miR-221 and Ago2 protein in HAECs by following the protocols of the manufacturer. Cells at 80%-90% confluency were lysed in an ice bath with RIPA lysate (P0013B; Beyotime Institute of Biotechnology) for 5 minutes, and the supernatant was taken after centrifugation at 4°C for 10 minutes at 10 000 g. Part of the cell extract was taken as the input, and the remainder was incubated with antibodies for co-precipitation.
The magnetic beads-antibody complex was re-suspended in 900 μL RIP Wash Buffer, and 100 μL cell extracts were added and incubated overnight at 4°C. The sample was placed on a magnetic base to collect the magnetic bead-protein complex. The samples and input were digested by protease K, and RNA was extracted for subsequent RT-qPCR. The antibody used for RIP was Ago2 (ab32381; 1:100, Abcam) and was allowed to mix for 30 minutes, and anti-human IgG (ab182931; 1:100, Abcam) was used as an NC.

| Co-immunoprecipitation
Co-immunoprecipitation (Co-IP) assay was performed by following the methods indicated in a study. 19 The primary antibody SIRT1

| Tube formation test
Capillary-like network formation was performed to detect the angiogenic ability of HAECs. In short, HAECs were inoculated at a density of 2 × 10 4 cells/well on a 96-well plate coated with 60 μL Matrigel (BD Bioscience). After 48 hours of culture, the average number of capillary-like branches was counted in five random microscope fields with a computer-aided microscope.

| Terminal-deoxynucleotidyl transferase mediated nick end labelling staining
The experiment was conducted according to the operating instruc-

| Cell-counting kit 8 assay
The transfected cells were cultured in a 96-well plate (1 × 10 4 cells/well). Next, 10 μL Cell-counting kit 8 (CCK-8) solution was added to each well and incubated at 37℃ for 4 hours. The absorbance value at 450 nm was then measured with microplate reader Synergy 4 (BioTek) to indirectly identify the number of living cells.

| Statistical analysis
All experimental data were analysed using the SPSS 21.0 statistical software (IBM Corp, Armonk, NY, USA). The measurement data were expressed by the mean ± standard deviation. The two groups of data conforming to the normal distribution were compared using an independent sample t test. The data comparisons between multiple groups were performed using one-way ANOVA, followed by Tukey's post hoc test. Cell activity at multiple time points was analysed using two-way ANOVA. Pearson correlation analysis was used for analysing miR-221 and LincRNA-p21.
P < .05 indicated that the difference was statistically significant.

| miR-221 expression is upregulated in AS
The bioinformatics website predicted that miR-221 was highly expressed in AS. RT-qPCR displayed that miR-221 expression was markedly elevated in patients with AS ( Figure 1A). Next, AS mouse model was constructed (AS-M) and HE staining showed that AS mice model exhibited larger lesion area than that in normal mice ( Figure 1B). Oil red O staining indicated that the lipid deposition area in AS mice was strikingly increased when compared to that in normal mice ( Figure 1C), confirming the successful construction of the AS model. Meanwhile, carotid tissue samples were collected from the mouse model and RT-qPCR and FISH revealed that miR-221 expression was notably increased in AS mice model when compared to that in normal mice ( Figure 1D, E).

| miR-221 inhibits the proliferation, migration and tube formation of HAECs in vitro
To

| LincRNA-p21 acts as a sponge of miR-221
Because miR-221 played key roles in regulating the HAEC phenotype, we investigated the potential regulatory mechanisms of miR-221 in depth. LncRNAs can act as molecular sponges for endogenous RNAs, interacting with miRNAs and affecting the expression of target genes and interacting with AS. 5 A study has also shown that LincRNA-p21 is reduced in patients with coronary heart disease and is potentially involved in the inhibition of AS. 7 To further explore whether LincRNA-p21 affected the development of AS through miR-221, we first detected LincRNA-p21 in peripheral blood samples of patients with AS, which showed a striking decrease in LincRNA-p21 expression in the peripheral blood of patients with AS compared with that of normal individuals ( Figure 3A). Furthermore, Pearson correlation analysis identified a negative correlation between LincRNA-p21 and the expression of miR-221 ( Figure 3B). We also predicted the binding site of LincRNA-p21 and miR-221 through the bioinformatics website (http://starb ase.sysu.edu.cn/) ( Figure 3C). The luciferase activity of miR-221 against LincRNA-p21 was verified by dual luciferase reporter gene assay. The results showed that overexpressing miR-221 could significantly inhibit the luciferase activity of the wild-type binding site of LincRNA-p21, but there was no effect on luciferase activity at the mutant binding site ( Figure 3D). RIP indicated that LincRNA-p21 and miR-221 in HAECs were more abundantly expressed in Ago2 precipitates than that in IgG precipitates ( Figure 3E). Subsequently, FISH analysis showed that LincRNA-p21 and miR-221 co-localized at the cytoplasm ( Figure 3F). In addition, we also found that ox-LDL treatment inhibited the expression of LincRNA-p21 in HAECs ( Figure 3G). We later found that miR-221  Figure 3H). Furthermore, overexpressing miR-221 contributed to a downregulated LincRNA-p21 expression, whereas silencing miR-221 resulted in a significantly increased LincRNA-p21 expression ( Figure 3I).

| LincRNA-p21 promotes the proliferation, migration and tube formation of HAECs by inhibiting miR-221 expression
The results showed that overexpressing LincRNA-p21 had signifi-

| miR-221 inhibits SIRT1 expression to repress proliferation, migration and tube formation of HAECs
To further study the downstream regulatory mechanism of miR-221,  miR-221 may have the potential to bind to SIRT1 ( Figure 5E). RT-qPCR and western blot analysis showed that overexpressing miR-221 contributed to decreased SIRT1 expression in HAECs, whereas inhibiting miR-221 resulted in increased SIRT1 expression ( Figure 5F, G).
We then silenced SIRT1 in the presence of miR-221 inhibitor, and performed EdU and CCK-8 assays, which revealed that inhibiting miR-221 promoted HAEC proliferation following ox-LDL treatment, which was reversed by the further treatment of si-SIRT1 ( Figure 5H, I). Transwell assay, scratch test and tube formation test indicated that inhibiting miR-221 had promoted the migrative and tube formation abilities of HAECs exposed to ox-LDL, while silencing SIRT1 blocked the inhibition of miR-221 on cell migrative and tube formation abilities in ox-LDL-treated HAECs ( Figure 5J-L).

| miR-221 inhibits the deacetylation of Pcsk9 by targeting SIRT1 and inhibits HAEC proliferation, migration and tube formation
RT-qPCR showed that Pcsk9 expression was increased in HAECs exposed to ox-LDL ( Figure 6A). Afterwards, HAECs were transfected with SIRT1-mimic and SIRT1-inhibitor, and the expressions of SIRT1 and Pcsk9 were detected using RT-qPCR and western blot analysis, which suggested that Pcsk9 expression was notably decreased in SIRT1-mimic-transfected HAECs, whereas Pcsk9 expression was significantly increased in SIRT1-inhibitor-transfected HAECs ( Figure 6B, C). The results of Co-IP showed that in HAECs, SIRT1 was enriched in Pcsk9 gene ( Figure 6D), and after overexpressing SIRT1, the correlation between SIRT1 and Pcsk9 was decreased and the level of acetylation was reduced ( Figure 6E).
However, the co-treatment with miR-221 inhibitor +oe-Pcsk9 only resulted in an increased expression of Pcsk9 ( Figure 6G). Similar results were obtained using western blot analysis ( Figure 6H).
Apoptosis of carotid endothelial cells in mice was detected using TUNEL staining, which showed that overexpressing LincRNA-p21 noticeably reduced the number of TUNEL positive cells, but the outcome was reversed by further treatment of oe-Pcsk9 ( Figure 7C).
Oil red O staining showed that overexpressing LincRNA-p21 resulted in declined lipid deposition area but the co-treatment of oe-LincRNA-p21 + oe-Pcsk9 contributed to increased lipid deposition area ( Figure 7F).

| DISCUSS ION
AS is a complex chronic disease characterized by the deposition of lipids on the walls of blood vessels, which can lead to the proliferation of smooth muscle, endothelial cells and immune cells. 20 A previous study has revealed that LincRNAs can act as endogenous RNA sponges for miRNAs that can further alter the expression of miRNAs

F I G U R E 4
Overexpression of LincRNA-p21 promotes HAEC proliferation, migration and tube formation through suppressing the expression of miR-221. HAECs were co-transfected with oe-LincRNA-p21-1 + mimic NC or oe-LincRNA-p21 + miR-221 mimic exposed to ox-LDL. A, The expressions of LincRNA-p21 and miR-221 in HAECs were detected using RT-qPCR. B, HAEC proliferation ability was assessed using EdU. C, HAEC proliferation ability was examined using CCK-8. D, HAEC migrative ability was evaluated using Transwell assay. E, Wound healing ability in HAECs was monitored by the means of scratch test. F, Angiogenic ability in HAECs was tested using tube formation test. The experiment was repeated three times Besides, miR-221-3p expression was also validated to be strikingly upregulated in atherosclerotic vessels. 22 We also found that miR- The binding site between SIRT1 and miR-221 was predicted by bioformatics website. B, SIRT1 expression in peripheral blood from patients with AS (n = 25) and normal patients (n = 18) was measured using RT-qPCR. C, The mRNA expression of SIRT1 in AS tissues was detected using RT-qPCR. D, The mRNA expression of SIRT1 in ox-LDL-treated HAECs was measured using RT-qPCR. E, The relationship between SIRT1 and miR-221 was verified using dual luciferase reporter gene assay. F, The mRNA expression of SIRT1 in HAECs was measured using RT-qPCR. G, The mRNA expression of SIRT1 in HAECs was measured using western blot analysis. H, HAEC proliferation ability was assessed using EdU in response to miR-221 inhibitor +siRNA-SIRT1. I, HAEC proliferation ability was examined using CCK-8. J, HAEC migrative ability was evaluated using Transwell assay. K, Wound healing ability in HAECs was monitored by the means of scratch test. L, Angiogenic ability in HAECs was tested using tube formation test. The experiment was repeated three times proliferation and induced apoptosis of vascular smooth muscle cells and mouse mononuclear macrophages in vitro. 9 Furthermore, miR-221 was highly expressed in diabetic mice, whereas LincRNA p21 was poorly expressed. Overexpressing LincRNA p21 suppressed the apoptosis of hippocampal neuron in diabetic mice by sponging miR-221. 24 miR-221 was demonstrated to target and inhibit the expression of SIRT1. Similarly, SIRT1 was confirmed as a target gene of miR-221 and miR-221 inhibited the proliferation and fibrosis-related proteins by targeting SIRT1. 25 Indeed, miR-221 was negatively correlated with SIRT1 mRNA 26 and we also found that the expression of SIRT1 was decreased in AS. Related research showed that deficiency of SIRT1 promoted the development of AS. 27 The expression of SIRT1 in ox-LDLtreated human umbilical vein vessel endothelial cells was decreased. 28 The experiments showed that miR-221 inhibited the deacetylation of Pcsk9 by targeting SIRT1 and inhibiting HAEC proliferation, migration and tube formation. Gain-of-function of SIRT1 activation reduced plasma LDL-C levels through inhibiting Pcsk9 secretion. 29 SIRT1 activator SRT3025 provided AS protection in Apoe -/mice by reducing hepatic Pcsk9 secretion. 30 Moreover, the expression of Pcsk9 in AS was significantly higher and Pcsk9 knockdown inhibited ox-LDL-induced F I G U R E 6 miR-221 inhibits the deacetylation of Pcsk9 to suppress HAEC proliferation, migration, and tube formation through targeting and inhibiting the expression of SIRT1. HAECs were transfected with SIRT1-mimic and SIRT1-inhibitor, or co-transfected with miR-221 inhibitor +oe-NC or miR-221 inhibitor +oe-Pcsk9. A, The mRNA expression of Pcsk9 in ox-LDL-treated HAECs was detected using RT-qPCR. B, The mRNA expressions of Pcsk9 and SIRT1 in ox-LDL-treated HAECs were detected using RT-qPCR. C, The protein expressions of Pcsk9 and SIRT1 were detected using western blot analysis. D, The interaction between SIRT1 and Pcsk9 in HAEC was detected by Co-IP. E, After overexpressing the expression of SIRT1, the interaction between SIRT1 and Pcsk9 and the acetylation level were detected by Co-IP. F, Pcsk9 expression was measured using RT-qPCR. G, The expressions of miR-221, SIRT1 and Pcsk9 were evaluated using RT-qPCR. H, The protein expressions of SIRT1 and Pcsk9 were detected using western blot analysis. I, The proliferation of HAECs was detected using EdU. J, HAEC proliferation was examined using CCK-8. K, HAEC migration was measured using Transwell assay. L, Wound healing ability was detected using scratch test. M, HAEC angiogenesis was detected using tube formation test. The experiment was repeated for 3 times   16 Lastly, the in vivo experiments further validated that LincRNA-p21 significantly reduced the plaque area in mice, inhibited Cleaved-Caspase-3 and Bax expression but upregulated Bcl-2 expression, which inhibited the development of aortic AS.
Conversely, plaque stability was characterized by an increased plaque collagen content and thickened fibrous cap, highlighting the involvement of SIRT in the inflammatory pathways of diabetic atherosclerotic lesions modulated by incretin. 31 Overexpressing LincRNA-p21 contributed to decreased expressions of Cleaved caspase-3 and Bax but upregulated the expression of Bcl-2. 24

| CON CLUS IONS
In conclusion, elevated LincRNA-p21 expression enhanced the promoter deacetylation of SIRT1 on Pcsk9 to protect against F I G U R E 7 Overexpression of LincRNA-p21 inhibits endothelial cell apoptosis and the development of ApoE −/− mice via mediating miR-221/SIRT1/Pcsk9 axis. A, LincRNA-p21, SIRT1, miR-221 and Pcsk9 expressions in plaque area were detected using RT-qPCR. B, The protein expressions of SIRT1 and Pcsk9 were detected using RT-qPCR. C, Apoptosis of carotid endothelial cells in mice was detected using TUNEL staining. D, The protein expressions of apoptosis-related factors Cleaved-Caspase-3, Caspase-3, Bax and Bcl-2 were examined using western blot analysis. E, Carotid artery injury in mice was examined using HE staining. F, The plaque area of carotid artery in mice was detected by oil red O staining. n = 15. The experiment was repeated three times   -221 (Figure 8). These results draw attention to a potential molecular mechanism of LincRNA-p21 in the treatment of AS. As stated by a prior study, increased inflammation is involved with the adipose tissue hypo-expression of SIRT1, which exerts local and systemic effects and affect the cardiac performance in overweight patients with pre-diabetes. 32 However, this effect was not evaluated in the study population given the limited time and funding, thus warranting further investigations in the future.

ACK N OWLED G EM ENTS
The authors would like to acknowledge the helpful comments on this paper received from its respected reviewers.

CO N FLI C T S O F I NTE R E S T
The authors declare no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets generated/analysed during the current study are available.