Long noncoding RNA UC.98 stabilizes atherosclerotic plaques by promoting the proliferation and adhesive capacity in murine aortic endothelial cells

Abstract

Pathological studies have shown that the vulnerability of plaques affects outcomes in patients with atherosclerosis (AS), a chronic inflammatory disease and common cause of morbidity and mortality worldwide. Although emerging technologies have enabled early diagnosis of AS with high-risk vulnerable plaques, more accurate and noninvasive diagnostic methods are urgently required. To this end, molecules involved in genetic or epigenetic regulation of the vulnerability of atherosclerotic plaques have been extensively studied. Here, we evaluated long noncoding RNA (lncRNA) variability by microarray assay in murine aortic endothelial cells (MAECs) bearing vulnerable plaques and identified the novel functional lncRNA UC.98, whose expression pattern was associated with the vulnerability of atherosclerotic plaques. Consistent with this, clinical statistics comparing the peripheral blood specimens from sets of patients with AS with or without vulnerable plaques confirmed the linear relationship between the expression pattern of UC.98 and plaque instability. Moreover, MTT assays and western blot analysis showed that silencing of intrinsic UC.98 in MAECs not only suppressed cell proliferation but also decreased the expressions of vascular cell adhesion molecule-1 and intercellular adhesion molecule-1, thereby inactivating the nuclear factor-κB pathway. In conclusion, our results highlighted the pivotal role of UC.98 in regulating the vulnerability of plaques during AS progression and suggested that UC.98 may be a biomarker of the early diagnosis and prognosis of AS with vulnerable plaques and a potential therapeutic target for slowing AS progression.

Introduction

Coronary atherosclerosis (AS) is a chronic inflammatory disease that manifests clinically as the deposition of lipids and extracellular materials, reduced vasoactivity, and angiosclerosis induced by calcification within the arterial intima-media [1]. AS has been recognized as the anatomopathological substrate in nearly all cases of acute coronary syndrome (ACS), a life-threatening condition associated with high mortality, unstable angina, and myocardial infarction with or without ST-segment elevation [2,3]. Although early atherosclerotic lesions may develop in childhood or adolescence and are frequently identified in adulthood under angiography, the annual incidence of acute coronary events remains relatively low (~0.2%–1%), suggesting that the mere presence of a coronary atherosclerotic lesion is insufficient to trigger an acute coronary event [4]. Epidemiological studies have shown that the genetic background combined with untreated cardiovascular risk factors (e.g. diabetes mellitus, hypercholesterolemia, hypertension, and/or smoking) determines the outcomes of atherosclerotic lesions [5].

Pathologically, atherosclerotic plaques are regarded as core lesions of AS, and rupture or erosion of these plaques contributes to the occurrence of ACS [6,7]. Atherosclerotic plaques are related to endothelial dysfunction, along with chemically modified lipoprotein-induced adhesion and penetration of leukocytes [8–10]. The plaques are aggravated by a series of local and systemic immune-inflammatory cascades, such as infiltration of macrophages within coronary lesions, recruitment and transition of blood-borne monocytes in the subendothelium, autocrine or paracrine signaling of inflammatory factors and cytokines, and activation and amplification of the immune-inflammatory cycle [11–13]. Although intensive evidence has highlighted the crucial role of the immune-inflammatory cascade in determining plaque vulnerability, the mechanisms through which stable plaques transform into vulnerable plaques remain unclear. Further studies of these mechanisms are necessary for advancing strategies in clinical prevention, diagnosis, and treatment of ACS. Because AS generally progresses without any detectable symptoms and gradually disrupts the function of the cardiovascular system, there is limited time to alleviate symptoms once clinical manifestations are observed or an episode of ACS has developed [14]. Other than traditional angiography, which focuses on detecting only severe angiostenosis, some anatomic or physiological methods that allow early diagnosis of atherosclerotic disease have been developed and applied (e.g. intravascular ultrasound or lipoprotein subclass analysis) [15,16]. Thus, additional noninvasive diagnostic methods are needed for monitoring AS.

Recent studies have demonstrated that specific genetic and/or epigenetic variations are involved in AS development [1]. For example, dysfunctions in some genes involved in lipometabolism (e.g. PLPP3, apoB, and PCSK9) and glycometabolism (e.g. HNF1A and GCK) induce hypercholesterolemia and hyperglycemia, thereby further enhancing inflammation, promoting monocyte/macrophage adhesion via the phosphatidylinositol 3-kinase/Akt/nuclear factor (NF)-κB pathway, and resulting in AS progression [17–20]. Moreover, various noncoding RNAs, including microRNAs and long noncoding RNAs (lncRNAs), play important roles in atherosclerotic disease. For example, overexpression of miR-155-5p and miR-155 accelerates AS through the tumor necrosis factor-α/NF-κB pathway, whereas activation of miR-126 decelerates atherosclerotic plaque formation by inducing endotheliocyte repair and proliferation [21–23]. Additionally, the lncRNA H19 promotes AS through the mitogen-activated protein kinase/NF-κB pathway, whereas the lncRNA SRA attenuates AS by protecting endotheliocytes from dysfunction [24,25].

Currently, few lncRNAs directly or indirectly associated with AS have been identified. Considering the important role of lncRNAs in various biological and pathological processes, we have been aiming to decode more functional lncRNAs responsible for AS progression or prevention. Accordingly, in this study, we isolated murine aortic endothelial cells (MAECs) from GHSR^−/− LDLR^−/− mice, which had been established to mimic atherosclerotic vulnerable plaques in vivo [26]. Then microarray analysis was performed to determine abnormalities in lncRNA profiles in MAECs derived from GHSR^−/− LDLR^−/− mice. Furthermore, RNA interference, western blot analysis and clinical statistics were carried out to confirm the specific bioactivities of candidate lncRNAs in switching plaque vulnerability during AS progression.

Materials and Methods

Animals and cell lines

Both LDLR^−/− and GHSR^−/− LDLR^−/− mice were generated at Shanghai Research Center for Model Organisms (Shanghai, China), as described previously [26]. In this study, LDLR^−/− mice were regarded as donors of stable atherosclerotic plaques, whereas GHSR^−/− LDLR^−/− mice were employed as donors with vulnerable plaques. All animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Guide) and were approved by the Ethics Committee of Shanghai Jiaotong University.

For cytological experiments, MAECs were isolated from LDLR^−/− mice (n = 5) or GHSR^−/− LDLR^−/− mice (n = 5) as previously described [27]. Cells were maintained at 37°C in a humidified atmosphere with 5% CO₂ and routinely cultured in Dulbecco’s modified Eagle’s medium (Hyclone, Logan, USA) containing 1 g/l glucose supplemented with 10% fetal bovine serum (v/v), 2 mM L-glutamine, and 0.1 mM nonessential amino acids.

Microarray analysis

Total RNA was extracted from the indicated MAECs using the TRIzol LS Reagent (cat. no. 10296028; Invitrogen, Carlsbad, USA) and purified using an RNeasy mini kit (cat. no. 74106; Qiagen, Valencia, USA) according to the manufacturer’s instructions with minimal modifications. The integrity of the extracted RNA was assessed using an Agilent Bioanalyzer 2100 (Agilent, Santa Clara, USA) and quantified using a NanoDrop ND-2000 spectrophotometer (Thermo Scientific, Waltham, USA). For microarray profiling, the qualified RNA samples (≥1 μg of each primary cultured cells) were submitted to Kangchen Biotechnology Company (Shanghai, China) and used to generate biotinylated cRNA targets for the Human LncRNA Microarray v3.0 8 × 60 K. After hybridization and washing, processed slides were scanned with an Agilent G2505C microarray scanner. The raw data were extracted using Feature Extraction software 10.7 (Agilent) and normalized using the GeneSpring GX v12.1 software package (Agilent). The threshold for screening of differentially expressed lncRNAs and mRNAs was set at a fold-change greater than 2.0 with a false discovery rate of less than 0.05.

cDNA was reverse transcribed and synthesized using a PrimeScript RT reagent kit (cat. no. RR037A; TaKaRa, Shiga, Japan). Subsequently, real-time quantitative polymerase chain reaction (RT-qPCR) was performed using a TB Green Premix Ex Taq kit (cat. no. RR420A; TaKaRa) on a Light Cycler 480 (Roche, Basel, Switzerland) according to the manufacturer’s recommendations. The relative expression patterns of the candidate lncRNAs were calculated using the 2^-ΔΔCt method, with glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as the housekeeping reference gene. The sequences of the primers used in this study are listed in Table 1.

RNA interference

To confirm the bioactivities of the candidate lncRNAs, RNA interference technology was used to downregulate target lncRNAs in MAECs derived from GHSR^−/− LDLR^−/− mice. Three specific shRNA oligonucleotides against either UC.98 or RP11-80B9.4 were designed and synthesized by GENEWIZ (Suzhou, China), and a scrambled oligonucleotide was designed as a control (Table 1). The annealed double-stranded DNA fragments were inserted into the pLKO.1-TRC lentiviral vector, and the desired plasmids were verified by Sanger sequencing. Lentivirus expressing shlncRNA-UC.98, shlncRNA-RP11-80B9.4, or shCon was packaged and propagated in HEK-293T cells, which were co-transfected with the indicated pLKO.1-shRNA plasmids and psPAX2/pMD2.G packaging plasmids using FuGENE 6 transfection reagent (cat. no. 11814443001; Roche). The lentiviral particles were harvested and purified, and MAECs were infected in 6-well plates at a density of 5 × 10⁴ cells/well with 0.5–1 ml of lentivirus and polybrene (8 μg/ml). Then, the stably silenced cells were screened using puromycin (5 μg/ml), and the efficiency of target silencing was evaluated by RT-qPCR.

MTT assay

MTT assay was carried out to assess the viability of MAECs after silencing of endogenous UC.98 or RP11-80B9.4. Briefly, cells were seeded into 96-well plates at 3 × 10³ cells/well in 200 μl complete growth medium and continuously cultured. At the indicated time points, one parallel set was analyzed using MTT solution (5 mg/ml; cat. no. M2128; Sigma, St Louis, USA). Accumulation of precipitated formazan was evaluated by measuring absorbance at 595 nm using a Multiskan Spectrum spectrophotometer (Thermo Fisher Scientific).

Western blot analysis

Total proteins were extracted from cells using RIPA lysis buffer (cat. no. R0278; Sigma) supplemented with cocktail protease inhibitors (cat. no. 11697498001; Roche) and quantified using a Pierce BCA Protein Assay kit (cat. no. 23250; Thermo Fisher Scientific). Fifty micrograms of each protein sample were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 12% gels and transferred to polyvinylidene difluoride membranes. After blocking nonspecific sites using 5% skim milk, the proteins of interest were probed with the corresponding primary antibodies, including rabbit anti-NF-κB p65 antibodies (1:3000; cat. no. 8242; Cell Signaling Technology, Danvers, USA), rabbit anti-NF-κB p100/p52 antibodies (1:3000; cat. no. 4882; Cell Signaling Technology), mouse anti-intercellular adhesion molecule-1 (ICAM-1) antibodies (1:3000; cat. no. ab33894; Abcam, Cambridge, UK), rabbit anti-vascular cell adhesion molecule (VCAM) antibodies (1:3000; cat. no. ab134047; Abcam), and rabbit anti-GAPDH antibodies (1:5000; cat. no. ab181602; Abcam). Horseradish peroxidase-conjugated goat anti-rabbit IgG (1:6000; cat. no. ab7090; Abcam) or goat anti-mouse IgG (1:6000; cat. no. 97040; Abcam) was used as secondary antibodies. Subsequently, bands were visualized using an ECL Prime Western Blotting system (cat. no. RPN2232; GE Healthcare, Waukesha, USA) and digitized using ImageJ 1.42 software.

Sample collection

Under supervision of the Clinical Research Committees of Shanghai Chest Hospital, Shanghai Jiaotong University, clinical examinations and statistics were carried out. Informed consent was obtained from all participants. Specimens of peripheral blood were collected from 40 cases of ACS with vulnerable plaques and 30 cases with stable plaques diagnosed by virtual histology intravascular ultrasound (VH-IVUS). The expression patterns of circulating lncRNAs or other nucleotides of interest were then detected by RT-qPCR analysis.

Statistical analysis

All results are presented as the mean ± standard deviation (SD), and independent experiments were performed in triplicate. Student’s t-tests and one-way analysis of variance were performed to determine the statistical differences between two groups using SPSS 19.0 software and GraphPad Prism 6 software. Differences with P values of 0.05 or less were considered statistically significant.

Results

Overexpression of UC.98 and RP11-80B9.4 in MAECs is associated with plaque instability

To decode more lncRNAs involved in regulating the vulnerability of atherosclerotic plaques, we evaluated lncRNA and mRNA profiles in MAECs derived from mice with vulnerable plaques (GHSR^−/− LDLR^−/−) or stable plaques (LDLR^−/−). Overall, 13,905 lncRNAs and 13,905 mRNAs were detected in three pairs of MAEC samples from GHSR^−/− LDLR^−/− or LDLR^−/− mice. Compared with those in LDLR^−/− mice, there were 3687 lncRNAs and 2761 mRNAs showing significantly aberrant expression in vulnerable samples (GHSR^−/− LDLR^−/−; Fig. 1A,B). Of the 3687 aberrantly expressed lncRNAs in the vulnerable group, 2232 were upregulated, whereas 1455 were downregulated (fold change ≥ 2.0; Fig. 1C).

Eleven lncRNAs showing elevation of more than 5-fold in the vulnerable group compared with that in the stable group were selected (Table 2). RT-qPCR was used to further validate the expression patterns of candidate lncRNAs in another five pairs of MAEC samples derived from GHSR^−/− or GHSR^−/− LDLR^−/− mice. Surprisingly, only two lncRNAs, UC.98 and RP11-80B9.4, showed dramatic upregulation in the vulnerable samples (Fig. 1D,E). These results suggest that the lncRNAs UC.98 and RP11.80B9.4 may be involved in regulating the physiological status of aortic endothelial cells (AECs), even for determining the vulnerability of atherosclerotic plaques.

Effects of silencing UC.98 or RP11-80B9.4 on the proliferation of MAECs

To verify whether the lncRNAs UC.98 and RP11-80B9.4 could affect the physiology of MAECs, RNA interference technology was introduced to disrupt the intrinsic homeostasis of lncRNAs of interest in MAECs. As expected, all of the established shRNAs decreased the expression level of UC.98 or RP11-80B9.4 in MAECs (GHSR^−/− LDLR^−/−; Fig. 2A). Furthermore, cell viability assays showed that the proliferation of UC98KD MAECs, in which the endogenous lncRNA UC.98 was silenced, was dramatically suppressed more than that in the parental cells (Ctrl). However, there was no marked difference in the proliferation between RP11KD MAECs (MAECs with silencing of RP11-80B9.4) and Ctrl MAECs (Fig. 2B). Thus, these results suggested that the lncRNA UC.98 may act as a positive regulator for AEC proliferation and that the progression of AS may be slowed down by silencing of UC.98 to suppress the aberrant proliferation of AECs.

Effects of silencing UC.98 on the intracellular inflammatory response and adhesive capacity of MAECs

The presence of dysfunctional endotheliocytes is a major characteristic of atherosclerotic plaque formation, and the physiological process through which endotheliocytes develop the mesenchymal phenotype is necessary to promote the proliferation and migration of vascular smooth muscle cells and enhance the adhesion of monocytes/macrophages onto potential lesions [28–30]. Therefore, we assessed the morphological changes in MAECs following silencing of endogenous UC.98 or RP11-80B9.4 by detecting the expressions of ICAM-1 and VCAM-1. A western blot analysis showed that the expression levels of ICAM-1 and VCAM-1 were markedly decreased in UC98KD MAECs (Fig. 3A). Compared with the control, two of three shlncRNA-UC.98 (shUC98–1 and shUC98–2) showed consistent suppression of the cell adhesion in MAECs, whereas shUC98–3 failed to alter this phenotypic transition (Fig. 3B,C). Additionally, a set of shlncRNA-RP11-80B9.4 exerted contradictory effects (Fig. 3A); one of three shlncRNA-RP11-80B9.4 (shRP11–3) suppressed ICAM-1 and VCAM-1 expressions, whereas shRP11–1 increased the expression of ICAM-1 but did not affect VCAM-1 expression, and shRP11–2 did not affect ICAM-1 or VCAM-1 expression (Fig. 3B,C). These results indicated that the lncRNA UC.98 may enable the formation and progression of atherosclerotic plaques by accelerating the cell adhesion process in AECs.

Moreover, to investigate whether the lncRNA UC.98 is responsible for the inflammatory cascades of MAECs during atherosclerotic plaque development, expression of the inflammatory transcription factor NF-κB was monitored in both the parental MAECs (Ctrl) and the UC98KD MAECs. As shown in Fig. 3A,D, the expression of NF-κB (p65) was significantly decreased in all UC98KD MAECs, regardless of which shlncRNA-UC.98 was applied. These observations indicated that the lncRNA UC.98 not only accelerated the cell adhesion of AECs but also promoted the intracellular inflammatory responses of AECs, which resulted in the formation of vulnerable plaques.

Practical application of UC.98 in the clinical diagnosis of vulnerable plaques

To evaluate the diagnostic potential of UC.98 in AS or early stage ACS, we used RT-qPCR to capture and quantify the level of circulating UC.98 in peripheral blood samples from patients with ACS with or without vulnerable plaques and health controls. As shown in Table 3, there were no significant differences between the vulnerable plague group and stable plague group with regard to age, sex, smoking status, blood pressure, presence of diabetes mellitus, total cholesterol, triglycerides, high-density lipoprotein, low-density lipoprotein, troponin I, and C-reactive protein. Importantly, UC.98 was aberrantly overexpressed in specimens from patients who had been diagnosed with ACS with high-risk vulnerable plaques using VH-IVUS (Fig. 4A). Consistent with this, Spearman correlation analysis further confirmed that expression of the lncRNA UC.98 was positively correlated with Gensini score (r = 0.4922711, P = 1.502 × 10⁻⁵; Fig. 4B), suggesting that the expression of circulating UC.98 was associated with the formation of vulnerable plaques in ACS.

Furthermore, to evaluate the practical applications of UC.98 as a marker for clinical diagnosis of AS vulnerable plaques, receiver operative characteristic (ROC) curve was generated. As shown in Fig. 4C, the area under the ROC curve for UC.98 was 0.811 (P < 0.001), and the Yoden index peaked when the optimal threshold of UC.98 was set to 0.132. Accordingly, the sensitivity for the diagnosis of complex lesions was 69.6%, and the specificity reached 94.1%. These clinical statistics confirmed that the lncRNA UC.98 could be considered as a biomarker for the early diagnosis of AS vulnerable plaques and as a molecular target for therapeutic intervention to reduce the risk of ACS.

Discussion

AS, one of the most common causes of morbidity and mortality worldwide, is a chronic inflammatory disease with multiple factors and is recognized as the anatomopathological foundation of ACS [4]. During the initiation and progression of AS, the vulnerability of plaques greatly determines the clinical outcomes of the disease. Other than the conventional cardiovascular risk factors such as hypercholesterolemia and hypertension, genetic and epigenetic abnormalities have been shown to be responsible for the vulnerability of atherosclerotic plaques and play important roles in the development of AS [1]. However, to the best of our knowledge, only a limited number of biomolecules that contribute to the genetic or epigenetic regulation of AS progress have been revealed, particularly for lncRNAs [31,32].

In this study, we evaluated the lncRNA profiles responsible for the vulnerability of atherosclerotic plaques using a high-throughput lncRNA microarray. We identified lncRNAs and mRNAs with aberrant expression in MAECs of vulnerable lesions, and a hierarchical clustering analysis showed distinct expression patterns of some lncRNAs and mRNAs (data not shown). These results may be caused by the multiple biological functions of lncRNAs, which can activate or inhibit their associated coding genes by chromatin modification or transcriptional/post-transcriptional regulation [33,34].

A further analysis identified two candidate lncRNAs (UC.98 and RP11-80B9.4) exhibiting upregulation. UC.98 promoted AEC proliferation, indicating that upregulation of UC.98 may be responsible for the aberrant viability of AECs. Moreover, in UC89KD MAECs, activation of the NF-κB pathway was suppressed by silencing the expression of intrinsic UC.98. As demonstrated by epidemiological and pathological studies, the integrated vascular endothelium is essential for maintaining the physiological functions of cardiovascular system, in which AEC hyperactivation and the subsequent induced inflammatory reaction are critical for the development of AS [35,36]. Additionally, activated NF-κB could induce chromatin remodeling to assemble de novo super-enhancers, which are capable of driving the expression of pro-inflammatory genes, such as CCL2 [31,37]. Thus, the progression of AS could be slowed down by reduction of UC.98 expression, resulting in suppression of AEC proliferation and/or attenuation of intra-inflammatory responses.

We also found that silencing the intrinsic lncRNA UC.98 in MAECs (GHSR^−/− LDLR^−/−) notably decreased the expressions of ICAM-1 and VCAM-1, two adhesion molecules that are commonly used as markers of the mesenchymal phenotype, suggesting a pivotal role of the lncRNA UC.98 in controlling the cellular morphological transition. This result could be related to inadequate activation of NF-κB in UC98KD MAECs, which could decrease the transcriptional activity of twist and Snail. Several studies have demonstrated that phenotypic switching is a major additional feature of AS and that the EndMT process is common in atherosclerotic lesions [28,38]. Related studies suggested that inflammation is closely associated with endothelial expression of cell adhesion molecules, including ICAM-1 and VCAM-1 [39,40]. Additionally, NF-κB, ICAM-1, and VCAM-1, as LPS-induced inflammatory mediators, are also involved in several diseases, including chronic intestinal inflammatory [41] and asthma [42]. Therefore, we assumed that there might be a linear relationship between the expression pattern of UC.98 and atherosclerotic plaque instability, in which UC.98 could be considered as a biomarker for the clinical diagnosis and prognosis of AS/ACS. As expected, clinical statistics from patients with ACS with or without vulnerable plaques confirmed this hypothesis, suggesting that UC.98 may have diagnostic potential for AS with high-risk vulnerable plaques.

In summary, we identified a novel functional lncRNA, UC.98, whose expression pattern was associated with vulnerability of atherosclerotic plaques. Moreover, silencing the intrinsic UC.98 in MAECs suppressed cell proliferation, decreased the expressions of adhesion molecules, and blocked the NF-κB pathway. Therefore, our results suggest that lncRNA UC.98 may be used as a biomarker for the early diagnosis and prognosis of AS/ACS with high-risk vulnerable plaques and be a potential therapeutic target for slowing down the progression of AS. However, more comprehensive studies are needed to evaluate the specific molecular mechanisms through which UC.98 regulates plaque vulnerability.

Funding

This work was supported by the grants from the National Natural Science Foundation of China (No. 80370400) and the Shanghai Municipal Education Commission—Gaofeng Clinical Medicine Grant (No. 20172028).

References