LncRNA FENDRR Inhibits ox-LDL Induced Mitochondrial Energy Metabolism Disorder in Aortic Endothelial Cells via miR-18a-5p/PGC-1α Signaling Pathway

Atherosclerosis (AS) is the main cause of morbidity and mortality in the world. Mitochondrial dysfunction is closely related to AS. At present, several signaling pathways related to mitochondrial dysfunction have been found, one of which is around PGC-1α. PGC-1α is a transcription activator, which is related to mitochondrial biogenesis and antioxidant defense. In this study, we explored the effect of miR-18a-5p/PGC-1α signaling pathway on mitochondrial energy metabolism in HAECs with ox-LDL treatment. The results showed that the mitochondrial energy metabolism disorder in HAECs treated by ox-LDL was related to the downregulation of LncRNA FENDRR and PGC-1α. FENDRR could reverse ox-LDL induced mitochondrial energy metabolism disorder and upregulate the PGC-1α expression. FENDRR could be used as ceRNA to inhibit the miR-18a-5p expression and reduce the negative regulation of miR-18a-5p on PGC-1α. Downregulation of miR-18a-5p expression or upregulation of PGC-1α in ox-LDL treated HAECs could reverse mitochondrial energy metabolism disorder. In conclusion, these findings suggested that FENDRR/miR-18a-5p/PGC-1α signaling pathway regulated mitochondrial energy metabolism in HAECs; ox-LDL downregulated the expression of PGC-1α and cause mitochondrial energy metabolism disorder by inhibiting this signal pathway.


INTRODUCTION
Atherosclerosis (AS), which leads to ischemia-reperfusion (I/R) injury of the heart, brain, and other tissues, is a chronic inflammatory disease and is an important risk factor for premature death (1). AS plays a key role in the pathogenesis of cardiovascular diseases. The course of AS is complex and is mainly manifested by the highly expressed adhesion molecules of arterial endothelial cells; they promote the infiltration of monocytes and differentiate into highly active lipid foam cells, accompanied by the migration of vascular smooth muscle cells (VSMCs) into the intima (1,2). AS is caused by the long-term and excessive reaction of vascular wall to injury; it starts from oxidative damage and inflammatory changes of endothelial cells (3). The apoptosis of vascular endothelial cells caused by mitochondrial ROS disorder is the initial stage of AS (4). In vascular disorder process, mitochondrial ROS increased in endothelial cells induced by a variety of factors, such as hypertension, hyperglycemia, and hyperlipidemia (5). Oxidized low-density lipoprotein (ox-LDL) regulates the migration, apoptosis, proliferation, and differentiation of vascular endothelial cells (VECs) (6). ox-LDL induced oxidative stress of VEC by promoting the production of ROS (7). There are many sources of ROS production in cells, such as mitochondria, NADPH oxidase or xanthine oxidase (XO), but the mitochondria have been considered as the major source of ROS in human endothelial cells. ROS produced by the mitochondria is the main promoter of cell signal under stress stimulation or risk factor exposure (8).
Peroxisome Proliferator Activated Receptor (PPAR) g-Coactivator 1-a (PGC-1a), a transcription factor, is involved in the upstream regulation of lipid catabolism, mitochondrial function. It is also a key molecule that regulates mitochondrial energy metabolism (9). Valle reported that an increased PGC-1a expression in human umbilical vein endothelial cells reduces ROS level, increases mitochondrial membrane potential, and reduces apoptosis under oxidative stress (10). PGC-1a could also resist excessive oxidative stress by increasing the transcription of antioxidant genes, containing manganese superoxide dismutase (MnSOD) (11). PGC-1a could restore the physiological membrane potential by increasing the activity of ATP/ADP transposase and inhibiting the production of excessive ROS (12). These results suggested that PGC-1a played an important role in regulating mitochondrial energy metabolism of VECs.
Non-coding RNAs include miRNA, piRNA, tRNA, and lncRNA (13)(14)(15). They do not have a coding function. However, they have a transcriptional regulation function and play a variety of roles. miRNA is a small RNA with a length of about 22 nt. It can bind with 3′UTR of mRNA through base complementary pairing, inducing mRNA splicing or translation block and resulting in gene silencing so as to regulate gene expression (16). Some lncRNAs regulate gene expression by recruitment of transcriptional regulatory complexes to the proximal (CIS) or distal (trans) genomic binding sites; others can be used as scaffold or decoys to assist protein or miRNA binding (17). Many studies have shown that a variety of miRNAs and lncRNAs were involved in the regulation of VEC function (18)(19)(20).
In recent years, some lncRNAs have been found to be differentially expressed in human cardiovascular diseases, which can be used as predictors of cardiovascular disease (21). FENDRR is an important lncRNA related to cardiac development and pathology; it is the regulator of two important histone modified complexes PRC2 and TrxG/MLL, which induces histone H3 lysine 27 trimethylation (H3K27Me3) by binding to PRC2 complex and inhibits many gene expressions at the transcriptional level (22). The expression level of FENDRR decreased in atherosclerotic plaques, which may promote the VSMC proliferation through the decrease of H3K27Me3 (23). FENDRR is a tumor suppressor gene in the tumor, which is related to the proliferation, invasion, and metastasis of tumor cells. In gastric cancer, the FENDRR expression decreased and caused the increased expression of fibronectin, which increased the invasion and metastatic ability of gastric cancer cells (24). FENDRR inhibited the progression of NSCLC by regulating TIMP2 expression through binding with miR-761 (25). FENDRR may play a role of tumor suppressor gene by inhibiting SOX4 in colon cancer (26). In addition, the expression level of FENDRR was decreased in primary liver cancer, chronic myeloid leukemia, osteosarcoma, prostate cancer, renal cell carcinoma, and cholangiocarcinoma, leading to the upregulation of tumor cell proliferation, invasion, and metastasis (27)(28)(29)(30)(31)(32). To our knowledge, the role of FENDRR in aortic endothelial cells has not been reported.
In this study, we speculated that ox-LDL may cause dysfunction of endothelial cells by affecting the expression of PGC-1a. We analyzed miRNA and lncRNA that may regulate PGC-1a expression by ENCORI (33) and explored the effect of FENDRR/miR-18a-5p/PGC-1a regulatory axis on mitochondrial energy metabolism in HAECs with ox-LDL treatment.

Cells
HAECs were purchased from ATCC (Manassas, VA, USA). They were cultured with DMEM/F12 medium containing FBS (10%), 1% penicillin and streptomycin at 37°C, 5% CO2. ox-LDL (final concentration 50 mg/ml) was added into the ox-LDL treated group. The cells were subcultured by not more than five times in our laboratory. This study was approved by the biosafety committee of the First Hospital of Shanxi Medical University and complied with biosafety procedures rules.
form plasmid/Lipofectamine 3000 complex and then added to the cells cultured in six-well plate for 24 h. Similarly, 75 pmol miR-18a-5p mimics or miR-18a-5p inhibitor was mixed with 125 ml Opti-MEM medium; 7.5 ml Lipofectamine 3000 was mixed with 125 ml Opti-MEM medium and incubated at room temperature for 5 min. They were mixed and incubated at room temperature for 5 min to form plasmid/Lipofectamine 3000 complex and then added to the cells cultured in six-well plate for 24 h.

Cell Apoptosis Detection
Apoptosis detection was performed using AnnexinV-PI Analysis Kit (Beyotime Biotechnology, Shanghai, China) following the manufacturer's instructions. The cells were inoculated in a sixwell plate. ox-LDL (12.5 ml and final concentration was 50 mg/ ml) was added to the ox-LDL treatment group when the cells grew to 70% confluent. Then the same amount of solvent was added to the control group and cultured for 48 h. Cells were collected, digested, and washed three times with ice-cold PBS. 1 × 10 6 cells/ml were collected and resuspended in 300 ml 1× binding buffer. 195 ml Annexin V-FITC, 5 ml Annexin V-FITC, and 10 ml PI were added according to the kit's protocol. Pre-cooled 1× binding buffer (200 ml) was added after 10 min of incubation at 4°C refrigerator in the dark. The apoptosis rate was detected using flow cytometry. The results were analyzed by CELLQUEST software (BD Biosciences). All tests were repeated three times.

Mitochondrial ROS Determination
The cells were inoculated in a six-well plate. ox-LDL (12.5 ml and final concentration was 50 mg/ml) was added to the ox-LDL treatment group when the cells grew to 70% confluent. Then the same amount of solvent was added to the control group and culture for 48 h. The cells were cultured at 37°C with fresh media containing 5 mM MitoSox Red mitochondrial superoxide indicator (Invitrogen) for 10 min. The cells were washed in PBS, and then mean fluorescence intensity (MFI) was detected by flow cytometry (FACS Aria, Becton Dickinson, San Jose, CA). All tests were repeated three times.

Mitochondrial Complexes Activity
The cells were inoculated in a six-well plate. ox-LDL (12.5 ml and final concentration was 50 mg/ml) was added to the ox-LDL treatment group when the cells grew to 70% confluent. Then the same amount of solvent was added to the control group and culture for 48 h. Mitochondria were first isolated from cultured cells using Mitochondria Isolation Kit for cultured cells following the manufacturer's instructions. Activities of respiratory chain complexes in the mitochondria were measured in microplates using the specific kits complex I-IV from Abcam (Cambridge, UK) normalized to protein concentration as measured by DC Protein Assay (BioRad). All tests were repeated three times.

Mitochondrial DNA Estimation
The cells were inoculated in a six-well plate. ox-LDL (12.5 ml and final concentration was 50 mg/ml) was added to the ox-LDL treatment group when the cells grew to 70% confluent. Then the same amount of solvent was added to the control group and culture for 48 h. Mitochondrial DNA copy number was estimated as previously described (34). Briefly, the total DNA was isolated with DNeasy Tissue Kit (Qiagen) according to instructions. Quantitative PCR was performed by using an equal amount of DNA from each sample, Power SYBR Green Master Mix (LifeTec Garland Island, NY), and 5 mM of each primer, i.e. mitochondrial ND1, mitochondrial Cyt B, and nuclear H19 DNA. To compare mitochondrial copy number, ratio of ND1:H19 and/or Cyt B:H19 was calculated. Results are presented as percent change in the mitochondrial DNA in the experimental versus wild-type/control samples. All tests were repeated three times.

Mitochondrial ATP Determination
The cells were inoculated in a six-well plate. ox-LDL (12.5 ml and final concentration was 50 mg/ml) was added to the ox-LDL treatment group when the cells grew to 70% confluent. Then the same amount of solvent was added to the control group and culture for 48 h. ATP levels were determined using the Cell Titre-Glo Luminescent Cell Viability Assay Kit (Promega), as previously described (35). All tests were repeated three times.

Seahorse Metabolic Analyzer Assays
Cell energy metabolism analysis was performed on cell energy metabolism analyzer (Seahorse, USA). The cells in different groups were inoculated (5 × 10 4 cells/well) to 24 well XF24 plate (Seahorse Biosciences). Cartridge plates for metabolic stress injections were hydrated with calibrant solution (Seahorse Biosciences) for at least 8 h at 37°C prior to the assay. One hour before detection, the medium of each well was aspirated, and the cells were washed twice with 1 ml of XF detection solution (Seahorse Biosciences). The detection solution was added to each well until a final volume of 500 ml/well was achieved. The cell plate was placed in an incubator (CO 2 free) at 37°C for 1 h. Calibrant solution hydrated for 8 h was added to each well. Assay conditions and set-up were performed according to instructions. Oxygen consumption rate (OCR) was reported in the unit of picomoles per minute. The results were normalized to cell number using CyQUANT Direct Cell Proliferation Assay kit (Molecular Probes). The area under curve (AUC) was calculated as previously described (36). All tests were repeated three times.

RT-PCR
Total RNA was extracted by TRIzol reagent according to the instructions. The purity and content of RNA were determined by NanoDrop ™ 1000 spectrophotometer. MMLV Reverse Transcriptase kit (Takara Bio Inc) was used for reverse transcription according to the manual. miR-18a-5p mRNA expression was detected by GenePharma Hairpin-it TM microRNA RT-PCR Quantitation Kit according to the instructions. The thermocycler conditions were 95°C for 10 min, 40 cycles of 95°C for 12 s and 62°C for 40 s. lncRNA FENDRR and PGC-1a mRNA expression was detected by SYBR Premix Ex Taq ™ Kit (Takara Bio Inc). GAPDH and U6 were used as internal control. Quantifications was performed by the2 −△△ Ct method. Primers' sequences were listed in Table 1.

Western Blot
The cells were lysed by Cell Lysis Solution. They were centrifuged at 4°C (1,000 rpm) for 5 min; the supernatant was collected. Proteins were extracted. Their concentration was measured with BCA. Proteins (50 mg/lane) were separated with 10% SDS− PAGE, then they were transferred to a membrane. The PVDF membrane was washed and blocked by buffer. PGC-1a antibody (Novus Biologicals, LLC, Centennial, CO, USA, NBP1-04676, 1:1,000) and b-actin antibody (Abcam, Cambridge, UK, ab8226, 1:2000) were added, and they were incubated at 4°C overnight. The membrane was rinsed, and the secondary antibody was added; they were incubated at RT for 1 h. They were determined by an enhanced chemiluminescence kit (Perkin-Elmer Inc., Waltham, MA, USA). They were quantified with Imagequant LAS4000 (GE Healthcare, Japan).

Double Luciferase Reporter Gene Assay
Wild type wt-pGL3-FENDRR and mutant type mut-pGL3-FENDRR plasmids were constructed based on the prediction. FENDRR luciferase reporter gene, miR-18-5p, and Renilla luciferase were co-transfected HEK293 cells. They were split by Dual Luciferase Assay following the protocol after they were cultured for 48 h. The results were analyzed by Panomics Luminometer. They were homogenized by sea renin fluorescence.

RNA Binding Protein Immunoprecipitation Assay
RIP experiments were conducted by EZMagna RIP kit (Catalog No. 17-701, Millipore, Billerica, MA, USA) according to the instruction. The cells were lysed when they grew to 80-90% full. Then cell lysate (100 ml) was incubated with anti-Ago2 antibody bound magnetic beads, normal mouse IgG (negative control) or anti-SNRNP70; they were incubated at 55°C for 30 min to remove protein using proteinase K. RNA was purified with RNeasy Micro Kit (Qiagen, Dusseldorf, Germany). The content and purity of RNA were detected by NanoDrop 1000 spectrophotometer. They were detected by RT-PCR.

Statistical Analyses
The statistical tests were conducted by SPSS 20.0 software. The differences among groups were tested by one-way ANOVA or Student's t-test. The correlation was examined by Spearman correlation test. P < 0.05 was considered statistically significant. In order to observe the effect of ox-LDL on mitochondrial function of HAECs, we analyzed the changes of mitochondrial function of HAECs treated with ox-LDL. ox-LDL also reduced the copy number of mitochondrial DNA in HAECs ( Figure 1B) and inhibited the activities of complexes I and III ( Figures 1D, E), but it had no effect on the activities of complexes II and IV ( Figures  1F, G). ox-LDL decreased ATP levels in HAECs ( Figure 1H). The basic and maximal oxygen consumption rates (OCRs) in HAECs treated by ox-LDL were significantly decreased ( Figures 1I-K).
The expression levels of FENDRR and PGC-1a in HAECs decreased significantly after ox-LDL treatment ( Figures 1L-N).
These results suggested that ox-LDL caused mitochondrial energy metabolism disorder, upregulated mitochondrial ROS level and apoptosis in HAECs ( Figures 1A, C), which may involve the regulation of FENDRR and PGC-1a expression.

LncRNA FENDRR Reversed ox-LDL Induced Mitochondrial Energy Metabolism Disorder and Upregulated PGC-1a Expression in HAECs
Downregulation of FENDRR expression in HAECs promoted apoptosis (Figure 2A), decreased mtDNA copy number ( Figure 2B), and resulted in mitochondrial ROS accumulation ( Figure 2C). Downregulation of FENDRR expression also inhibited complex I ( Figure 2D) and III activity ( Figure 2E) and downregulated ATP level ( Figure 2F), which were the same as that of ox-LDL treatment. The basic and maximal OCRs in HAECs were significantly decreased after FENDRR was downregulated ( Figures 2G-I). FENDRR overexpression reversed the regulation of ox-LDL on mitochondrial energy metabolism in ox-LDL treated HAECs. PGC-1a expression was downregulated in HAECs after FENDRR was downregulated and vice versa (Figures 2J-L). These results suggested that LncRNA FENDRR reversed ox-LDL induced mitochondrial energy metabolism disorder and upregulated PGC-1a expression in HAECs.
LncRNA FENDRR Could Be Used as ceRNA to Adsorb miR-18a-5p and Upregulate the PGC-1a Expression The above results suggested that there may be a direct regulatory relationship between FENDRR and PGC-1a. Six miRNAs were found to bind to FENDRR and PGC-1a 3′-UTR simultaneously ( Figure 3A). RT-PCR was used to detect the expression of these six miRNAs in HAECs after ox-LDL treatment. The results showed that the miR-18a-5p expression level increased most significantly after ox-LDL treatment ( Figure 3B). RIP assay showed that miR-18a-5p, FENDRR, and PGC-1a mRNA were enriched in the precipitate ( Figure 3C), which indicated that there was a direct combination of the three. Further analysis of gene sequences showed that there were target regulatory sites between FENDRR and miR-18a-5p and between miR-18a-5p and PGC-1a 3′-UTR ( Figure 3D). The results of luciferase assay also showed that FENDRR and miR-18a-5p ( Figure 3E) and miR-18a-5p and PGC-1a 3′-UTR ( Figure 3F) had targeted regulatory effects. These results suggested that FENDRR, as a ceRNA, inhibited miR-18a-5p level, further promoted the PGC-1a expression, and maintained the stability of mitochondrial energy metabolism in HAECs.

Downregulation of miR-18a-5p Expression in ox-LDL Treated HAECs Reversed Mitochondrial Energy Metabolism Disorder and Upregulated PGC-1a Expression
In miR-18a-5p mimics transfected HAECs, apoptosis, and mitochondrial ROS levels were significantly increased; mitochondrial DNA copy number, complex I and III activities, and ATP levels were significantly decreased, and basic and maximal OCRs were also significantly decreased. After miR-18a-5p inhibitor transfection in ox-LDL treated HAECs, apoptosis and mitochondrial ROS levels were significantly decreased, mitochondrial DNA copy number, complex I and III activities, and ATP levels were significantly increased, and basic and maximal OCRs were also significantly increased ( Figures 4A-I). PGC-1a expression was downregulated in HAECs after transfection of miR-18a-5p mimics in HAECs and vice versa (Figures 4J-L). These results suggested that downregulation of miR-18a-5p expression in ox-LDL treated HAECs reversed mitochondrial energy metabolism disorder and upregulated PGC-1a expression.

Upregulation of PGC-1a Expression in ox-LDL Treated HAECs Reversed Mitochondrial Energy Metabolism Disorder
PGC-1a shRNA expression plasmid was transfected into HAECs, and PGC-1a overexpression plasmid was transfected into ox-LDL treated HAECs to observe the changes of mitochondrial energy metabolism. The results showed that in HAECs transfected with PGC-1a shRNA, cell apoptosis and mitochondrial ROS levels were significantly increased, mitochondrial DNA copy number, complex I and III activity and ATP levels were significantly decreased, and basic and maximal OCRs were also decreased. In ox-LDL treated HAECs transfected with overexpressed PGC-1a, cell apoptosis and mitochondrial ROS levels significantly decreased, mitochondrial DNA copy number, complex I and III activity, ATP levels and basic and maximal OCRs significantly increased ( Figures 5A-I). Transfection of PGC-1a shRNA into HAECs decreased the PGC-1a expression and vice versa ( Figures  5J, K). These results suggested that upregulation of PGC-1a expression in ox-LDL treated HAECs reversed mitochondrial energy metabolism disorder.

Dose Effect of PGC-1a
We transfected 7.5 ml (0.3125 mg/ml, 0.625 mg/ml, 1.25 mg/ml, and 2.5 mg/ml) PGC-1a eukaryotic expression vectors into HAECs treated with ox LDL, respectively. The apoptosis rates, mito SOX, mitochondrial copy numbers, Complex I activity, Complex III activity, and relative ATP levels were determined. The results showed that the apoptosis rates and mito Sox decreased with the increase of PGC-1a transfection, and the mitochondrial copy numbers, Complex I activity, Complex III activity and relative ATP levels increased with the increase of PGC-1a transfection ( Figures 6A-I).

DISCUSSION
In this study, we found that ox-LDL inhibited the expression of LncRNA FENDRR in HAECs and downregulated the expression of PGC-1a through FENDRR/miR-18a-5p/PGC-1a signal axis, resulting in the increase of ROS level and the disorder of mitochondrial energy metabolism in HAECs. VECs maintain intravascular homeostasis mainly by regulating vascular relaxation, leukocyte adhesion, platelet activation, and smooth muscle cell proliferation and migration (37,38). The first pathophysiological change is endothelial dysfunction in the early stage of AS, which characterized by the decrease of NO synthesis and secretion (39). Endothelial nitric oxide (NO) inhibits the endothelial cell adhesion molecules and chemokine expression and inhibited platelet activation and aggregation. NO is synthesized by converting L-arginine to L-citrulline by endothelial nitric oxide synthase (eNOS) in endothelial cells (40). The main reason for the reduction of NO production is ROS mediated oxidative stress-induced eNOS degradation rather than the eNOS gene inhibition (41). eNOS dysfunction could produce more ROS, further damaged endothelial function, and promoted the AS development (40). The mitochondria are involved in ROS generation, apoptosis and intracellular signal transduction of VECs, and play a key role in the VEC function regulation (42). Excessive ROS production leads to endothelial cell senescence, apoptosis and AS. VEC dysfunction with mitochondrial damage and dysfunction preceded cardiovascular disease (43). Mitochondrial respiratory chain is the main pathway of energy production. Under pathological conditions, the effective utilization of respiratory chain decreased, ATP synthesis decreased, and more electrons were released to produce ROS (42). mtDNA is highly susceptible to ROS attack (44). mtDNA mitochondrial damage is an early event of AS in ApoE−/− mice (45). In this study, we used ox-LDL to treat HAECs and found that ATP synthesis decreased, mitochondrial OCRs, mtDNA copy  The study confirmed that non-coding RNA was involved in a variety of cardiovascular diseases, including AS (46). In this study, ENCORI analysis showed that FENDRR/miR-18a-5p/PGC-1a signal axis may be involved in the damage of ox-LDL on mitochondrial energy metabolism of HAECs. RIP results showed that miR-18a-5p, FENDRR, and PGC-1a mRNA could bind directly. The results of luciferase reporter gene also indicated that there were targeted regulatory effects between FENDRR and miR-18a-5p, miR-18a-5p and PGC-1a 3′-UTR. In patients with hypertrophic cardiomyopathy, the circulating miR-18a-5p expression was significantly increased, which was positively correlated with the disease progression and could also be used as a biomarker of diffuse myocardial fibrosis (47). In this study, the miR-18a-5p expression was upregulated, and FENDRR expression was downregulated in ox-LDL treated HAECs. Transfection of FENDRR or miR-18a-5p inhibitor into ox-LDL treated HAECs could reverse the effect of ox-LDL and reduce the level of mitochondrial ROS and apoptosis. Transfection of miR-18a-5p mimics or sh-FENDRR in normal HAECs could cause mitochondrial energy metabolism disorder, upregulate mitochondrial ROS level and apoptosis level. These results suggested that FENDRR/miR-18a-5p/PGC-1a signal axis existed in HAECs, and ox-LDL caused mitochondrial energy metabolism disorder in HAECs by FENDRR/miR-18a-5p/PGC-1a signal axis. PGC-1a is a main regulator of mitochondrial biosynthesis. PGC-1a activates mitochondrial transcription factors A and B, which regulate the mtDNA encoded genes' expression, including the basic subunit of electron transport chain (48). PGC-1a significantly inhibits endothelium-dependent vasodilation induced by fatty acids and recovers the function of NO (49). PGC-1a repair fatty acid oxidation of endothelial cells, so as to improve the activity of ATP/ADP transporter, increase ATP synthesis, reduce ROS production, so that endothelial cells can adapt well with high lipid load (12). PGC-1a overexpression increased the UCP-2 and mitochondrial antioxidant enzymes' expression, including catalase, manganese SOD, and thioredoxin 2, but silencing PGC-1a effect was the opposite (49). PGC-1a can also prevent the VECs' apoptosis, limit the activation of inflammation, prevent the activation of c-Jun N-terminal kinase, and increase the bioavailability of NO (50). We found that the PGC-1a expression was decreased in ox-LDL treated HAECs. Overexpression of PGC-1a in ox-LDL treated HAECs could reverse the effect of ox-LDL and reduce the level of mitochondrial ROS and apoptosis. Downregulation of PGC-1a expression in normal HAECs could cause mitochondrial energy metabolism disorder, increase mitochondrial ROS level and apoptosis level. These results further confirmed that FENDRR/ miR-18a-5p/PGC-1a signal axis is involved in the regulation of mitochondrial energy metabolism in HAECs.

CONCLUSIONS
In conclusion, this study showed that FENDRR acted as a molecular sponge to adsorb miR-18a-5p in normal HAECs and inhibited the negative regulation of miR-18a-5p on PGC-1a, thus maintaining normal mitochondrial energy metabolism in HAECs. ox-LDL could downregulate the FENDRR expression in HAECs, decrease the expression of PGC-1a, cause the mitochondrial energy metabolism disorder, increase the level of ROS and apoptosis. This study presented the transcriptional regulation spectrum of key energy metabolism signal transduction pathways in vascular tissue and suggested potential therapeutic targets for vascular endothelial dysfunction related diseases. The data is too limited to ensure that the observed effects are due to direct regulatory events; further research is needed.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.