Plasma NORAD levels in the CAD and control groupsIn this study, a total of 75 CAD patients and 76 controls were recruited to determine the plasma lncRNA NORAD levels using qRT-PCR. The baseline characteristics of patients with CAD and controls were presented in Table 1. The characteristic analysis revealed no significant differences in the characteristics of age, gender, smoking, and alcohol consumption. The prevalence of hypertension, diabetes history, as well as plasma total cholesterol (CHOL) and triglyceride (TG) levels were similar between the groups. Concerning the degree of stenosis, the Gensini score for the CAD group was 57.13 ± 42.21, whereas the score for the control group was nearly 0, making it a suitable control group for CAD patients with evident AS plaques.Plasma samples were obtained from both the CAD and control groups, followed by quantification of circulating NORAD levels using qRT-PCR (Fig. 1A). Plasma NORAD levels were significantly elevated in the CAD group compared to the control group which revealed that plasma NORAD exhibited remarkable discriminatory ability in identifying CAD patients from the control group (Fig. 1B). Our findings indicated lower HDL-C levels and higher LDL-C levels in CAD patients, yet the areas under the ROC curves were < 0.7, hence precluding the use of HDL-C and LDL-C as diagnostic indicators (Fig. 1C, D, E, F).Table 1 General demographic characteristics of the research subjects.Fig. 1Circulating NORAD levels and serum lipid levels in the control and CAD groups. (A) Comparison of circulating NORAD between control and CAD group. (B) Receiver operating characteristic (ROC) curve analyses for NORAD. (C) Comparison of HCL-C between the control and CAD groups. (D) ROC curve analyses of HDL-C. (E) Comparison of circulating LDL-C between the control and CAD groups. (F) ROC curve analyses of LDL-C. ***P < 0.001, ****P < 0.0001, ns, not significant.Analysis of the network regulation mechanism of NORADTo investigate the regulation mechanism of NORAD in AS formation, we collected NORAD target miRNAs from starBase and miRcode. At the same time, we screened miRNAs with significantly reduced expression levels from the GSE28858 dataset. A Venn diagram was drawn to identify miRNAs overexpressed in all datasets, resulting in 34 miRNAs (Fig. 2A, Supplementary Table S2). The mRNAs predicted as targets of these miRNAs were obtained from miRBD, miRTarBase, and miRWalk database, and intersected with upregulated genes in GSE100927. Figure 2B shows the network of miRNAs and 138 target genes. We utilize FunRich 3.1.3 to perform GO enrichment analysis on the 34 significantly upregulated miRNAs targeting NORAD in CAD (Fig. 2C). In the Biological Process category, the target miRNAs are predominantly involved in regulating nucleobase, nucleoside, nucleotide, and nucleic acid metabolism, cell communication, and signal transduction. Within the Cellular Component category, the target miRNAs are primarily associated with the nucleus, cytoplasm, and Golgi apparatus. In the Molecular Function category, the target miRNAs are mainly linked to transcription factor activity, protein serine/threonine kinase activity, and ubiquitin-specific protease activity.GO and KEGG pathway enrichment analyses were performed on target genes to explore the biological functions of NORAD. The GO chord diagram of biological processes revealed that NORAD was involved in apoptosis and negative regulation of the MAPK cascade (Fig. 2D). KEGG pathway enrichment analysis revealed that the targets of NORAD were enriched in cancer and the MAPK signaling pathway (Fig. 2E).The 138 targets of NORAD were input into the STRING database to investigate the interrelation of the target proteins. The PPI network included 101 nodes and 448 edges (Fig. 2F). In cytoHubba, the first ten hub genes were identified using the MNC algorithm, including mitogen-activated protein kinase1 (MAPK1), growth factor receptor-bound protein 2 (GRB2), G1/S-specific cyclin-D1 (CCND1), among others (Fig. 2G). MCODE was used to conduct clustering analysis and the functional module was screened out from the PPI network (Fig. 2H). Based on the results of the bioinformatics analysis, we found that NORAD was involved in regulatory processes related to the cell cycle, and CCND1 may be a critical gene in the NORAD regulatory network.Fig. 2Construction of the ceRNA network of NORAD. (A) Venn diagram of miRNA from starBase database, miRcode database and GSE28858. (B) Network of miRNAs – mRNAs. (C) FunRich 3.1.3 to conduct GO enrichment analysis on the 34 significantly upregulated miRNAs targeting NORAD in CAD. (D) GO chord diagram of biological processes based on GO enrichment analysis of target mRNA. The inner circle indicates the mRNAs, and colors represent the log2 (fold change) change in expression. The different colors in the outer circle represent different GO terms. (E) KEGG circle diagram based on KEGG pathway analysis. The target mRNAs are shown on the left side, and distinct colored bands on the right-hand side symbolize different pathways. (F) The PPI network of targets miRNAs using STRING. The node color depth is proportional to the log2 change in expression. (G) The top ten genes subnetwork after CytoHubba filtration with MNC. (H) The significant module from the PPI network with an MCODE score of 4.286.Si-NORAD induces endothelial cell cycle arrest in the G0/G1 phaseTo determine the effect of NORAD on the cell cycle, we measured the expression levels of relevant genes using qPCR. As shown in Fig. 3A, the expression of NORAD was significantly reduced in HUVECs transfected with si-NORAD. Meanwhile CCND1, cell cycle protein A2 (CCNA2), G2/mitotic-specific cyclin-B1 (CCNB1), cyclin-dependent kinase 2 (CDK2), CDK6, and transcription factor 2 (E2F) were significantly decreased after NORAD knockdown in HUVECs. Then we examined endothelial cell proliferation cycles using flow cytometry, and we found that the number of endothelial cells in the G0/G1 phase increased significantly in the NORAD knockdown group. The proportion of cells in the S phase was significantly reduced (Fig. 3B). These results suggest that NORAD knockdown induces endothelial cell cycle arrest in the G0/G1 phase via downregulated cyclin D1 and cyclin-dependent kinases, then inhibits endothelial cell proliferation.Fig. 3NORAD regulated the cell cycle. (A) The expression level of cell cycle-related genes in HUVECs. (B) Cell cycle determination using flow cytometry. N = 3–7. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant.Si-NORAD inhibited HUVECs migrationTo investigate the effect of NORAD on the migration of HUVECs, we performed the wound-healing assay. As shown in Fig. 4A and B, the wound-healing areas at 6 and 12 h were estimated by a cell scratch test, and both significantly decreased in HUVECs with si-NORAD. Cell proliferations were measured at 0 h, 24 h, 48 h and 72 h using the CCK8 assay. The results suggest that si-NORAD significantly inhibited the proliferation of HUVECs compared with the control group (Fig. 4C). These findings suggest that si-NORAD inhibits HUVECs proliferation and migration. As indicated by the previous analysis, CCND1 may be the target of NORAD. We then detected the CCND1 protein levels using western blot. The result showed that CCND1 protein expression decreased in HUVECs transfected with si-NORAD (Fig. 4D).Fig. 4Si-NORAD inhibited HUVECs migration. (A and B) Wound healing was estimated using a scratch assay in HUVECs at 0, 6, and 12 h. (C) Cell proliferation was determined using CCK8 analysis. (D) CCND1 protein levels were measured using western blot assay. N = 3 or 4. *P < 0.05, **P < 0.01.NORAD regulated the migration of HUVECs by targeting miR-106aBased on miRNA targets prediction tool starBase database, miRcode database and combined with the data of GSE28858, we predicted NORAD could be targeted by miR-106a. And we suspected that miR-106a might be a sponge of NORAD involved in regulating CCND1 protein levels. To determine the targeting relationship between NORAD and miR-106a, we examined the expression of miR-106a in HUEVCs, and found elevated levels of miR-106 in HUVEC with si-NORAD (Fig. 5A). We predicted the potential binding site of the NORAD 3´UTR terminal with miR-106a (Fig. 5B). We verified the targeting relationship between them by a dual luciferase assay. The luciferase activity of the WT reporter was lower in NORAD-WT treated with miR-106a mimic compared to the NORAD-WT treated with miR-NC (Fig. 5B), but the repression was abrogated after mutating the putative binding sites of NORAD gene. The results supported the notion that miR-106a was directly targeted by NORAD. We also measured the effect of miR-106a on endothelial cells. After miR-106a was overexpressed in HUVECs, cell proliferation, and migration were detected using a cell scratch assay at 6 and 12 h. As shown in Fig. 5C, miR-106a mimic inhibited HUVEC proliferation and migration, while the miR-106a inhibitor promoted the proliferation and migration. These results suggest that NORAD may interact with miR-106a, potentially modulating its availability or activity, and thereby regulating the proliferation and migration of HUVECs.Fig. 5NORAD acts as a sponge for miR-106a. (A) MiR-106a levels in HUVECs measured by qPCR. (B) A dual-luciferase reporter gene was used to verify the targeted relationship between NORAD and miR-106a. (C) Wound-healing was tested using a scratch assay in HUVECs treated with miR-106a mimic or inhibitor. N = 3–5. *P < 0.05, **P < 0.01.MiR-106a directly targets CCND1In addition, miR-106a may target CCND1 directly, and the binding site was predicted using the starBase database. To confirm that miR-106a specifically binds to the 3’UTR of CCND1 mRNA to regulate the expression of CCND1, we performed a dual-luciferase reporter assay. The results showed that the luciferase activity in the CCND1-WT + miR-106a mimic co-transfected group was lower than that of the CCND1-WT + miR-NC co-transfected group, but there was no significant difference when miR-106a mimic or NC was co-transfected with CCND1-MUT (Fig. 6A). The result supported that miR-106a directly targeted the CCND1 3′UTR. To verify whether miR-106a affects the expression of CCND1 in HUVECs, we performed qPCR and western blot to measure the CCND1 expression. CCND1 mRNA and protein expression levels decreased (Fig. 6B and C) in HUVECs treated with miR-106a mimic. These results suggest that miR-106a directly targets CCND1, inhibiting the proliferation of endothelial cells.Fig. 6MiR-106a directly targets CCND1. (A) The putative binding sites and corresponding mutant region for miR-106a within CCND1 and dual-luciferase reporter assay to verify the targeted relationship. (B) The effect of miR-106a on the mRNA level of CCND1 was determined using qPCR. (C) The effect of miR-106a on the protein expression of CCND1 was determined using western blot. N = 3–6. *P < 0.05, **P < 0.01, ***P < 0.001, ns, not significant.Si-NORAD regulated cell apoptosisThe apoptosis of HUVEC induced by NORAD was detected by PI and Annexin V-FITC staining. As shown in Fig. 7A, compared with HUVEC transfected with NC, HUVEC transfected with si-NORAD exhibited high proportions (7.4% vs. 4.5%, > 20%) of Annexin V + PI− (early apoptotic), but there was no significant difference in the proportions of Annexin V + PI+ (late apoptotic) and Annexin V- PI+ (necrotic) cells between those transfected with NC or si-NORAD. These results collectively suggest that si-NORAD induced HUVEC early apoptotic.Additionally, the expression levels of apoptotic key molecules were examined. The enzyme activity assay revealed no statistically significant difference in Caspase 3 activity between the two groups (Fig. 7B). Furthermore, Western blot analysis demonstrated that following transfection with si-NORAD, there was no remarkable alteration in the protein expression levels of Caspase 3 and Bax in HUVECs, as compared to the NC group (Fig. 7C). Based on the above results, it may sugguest that si-NORAD dose not activate the Caspase 3-mediated apoptotic pathway, thereby not proceeding to the terminal stage of apoptosis.Protein electrophoresis results showed that compared with the control group, the protein levels of PPARγ increased, but FOXO3a protein levels were down-regulated in HUVEC with NORAD knockdown. Besides, the protein level of p53 did not increase, but the phosphorylated AMPK decreased, as induced by si-NORAD in HUVECs (Fig. 7D).Fig. 7Si-NORAD induced early apoptosis. (A) Cell apoptosis analysis of HUVEC transfected with si-NORAD by flow cytometry. (B) Intracellular Caspase 3 activity assay. (C) and (D) The protein levels were measured using western blot. GAPDH or β-Tubulin was used as the protein loading control. N = 3–5. *P < 0.05, **P < 0.01, ns, not significant.Modulation of endothelial redox state by si-NORADThe impact of si-NORAD transfection on endothelial ROS levels and redox state was investigated using H2DCFD staining and high-content analysis. Transfection of HUVECs with si-NORAD resulted in a significant increase in intracellular ROS levels. However, the addition of liproxstatin-1 attenuated this increase (Fig. 8A and B). Additionally, si-NORAD transfection led to a decrease in intracellular glutathione (GSH) levels, which were restored upon the addition of liproxstatin-1, indicating a restorative effect of liproxstatin-1 on the cellular redox balance (Fig. 8C). Further analysis of mitochondrial membrane potential using the JC-1 probe revealed that si-NORAD transfection induced a decrease in mitochondrial membrane potential. Although liproxstatin-1 was administered, it did not significantly affect the mitochondrial membrane potential, suggesting that the protective effects of liproxstatin-1 may not be mediated through this particular mitochondrial pathway (Fig. 8D and E). To assess lipid peroxidation, the liprofluo fluorescent probe was employed, which revealed that si-NORAD transfection significantly increased the level of oxidized lipids in HUVECs. However, this increase was reversible with the application of liproxstatin-1, indicating that liproxstatin-1 can effectively reduce the oxidative stress induced by si-NORAD (Fig. 8F). Collectively, these results suggest that si-NORAD induces oxidative stress in HUVECs.Fig. 8Si-NORAD induced HUVEC oxidation. (A) and (B) The images of HUVEC stained with H2DCFD for ROS and Hoechst (blue fluorescent) for nuclei by the high-content live cell imaging system. (C) The GSH level in HUVEC transfected with si-NORAD. (D) HUVEC stained with JC-1 for mitochondria and Hoechst (blue fluorescent) for nuclei. JC-1 forms aggregates (red fluorescent) under high mitochondrial potential condition and becomes monomers (green fluorescent) under low mitochondrial potential condition. Bar = 100 μm. (E) The ratios of red fluorescent intensity to green fluorescent intensity were shown. (F) Representative flow cytometric profiles are shown to demonstrate lipid peroxides with Liperfluo signals. N = 3–8. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant.Regulatory roles of NORAD in ferroptosisThrough the application of Gene Set Variation Analysis (GSVA), our investigation into the biological processes and potential mechanisms of CCND1 expression in atherosclerosis (AS) revealed distinct pathway enrichments. High expression of CCND1 was associated with the significant enrichment of metabolic pathways such as linoleic acid metabolism, α-linolenic acid metabolism, and ether lipid metabolism. Conversely, low expression of CCND1 correlated with the enrichment of pathways including butanoate metabolism, limonene and pinene degradation, and taurine and hypotaurine metabolism (Fig. 9A).Besides, our research demonstrated that the silencing of NORAD via si-NORAD led to a marked increase in cellular oxidative stress levels. This oxidative stress response was found to be reversible by the administration of liproxstatin-1, a ferroptosis inhibitor, suggesting a potential role for NORAD in the regulation of ferroptosis, a form of cell death linked to iron metabolism. Utilizing the String database, we conducted an analysis to elucidate the potential interactions between CCND1 and ferroptosis-related factors (Fig. 9B). Analysis of the GSE100927 dataset revealed a positive correlation between the expression levels of CCND1 and ferroptosis inhibitors such as GPX4 and FTH1 in AS plaques (Fig. 9C and D). This analysis provided a framework for understanding the molecular connections that may underlie the observed correlations between CCND1 expression and ferroptosis in AS.To further delineate the effects of NORAD silencing on ferroptosis-related proteins, we transfected HUVECs with si-NORAD and assessed the expression levels of key proteins. Our results indicated that NORAD silencing led to a decrease in the expression of GPX4, FTH1, KEAP1, NCOA4, and Nrf2, while Xct levels were increased (Fig. 9E). These alterations in protein expression were found to be reversible by the treatment with liproxstatin-1, confirming the involvement of ferroptosis in the cellular response to NORAD silencing.Fig. 9Si-NORAD regulated cell ferroptosis-related genes. (A) GSVA analysis on CCND1. (B) The PPI network of CCND1 and ferroptosis-related proteins. (C) Relative expression of CCND1 and ferroptosis-related genes in GSE100927. (D) Corrplot of CCND1 and ferroptosis-related genes in GSE100927. (E) Protein levels were measured using western blot. N = 3–10. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant.
