LncRNA HCG11/miR‐26b‐5p/QKI5 feedback loop reversed high glucose‐induced proliferation and angiogenesis inhibition of HUVECs

Abstract Acute coronary syndrome caused by the rupture of atherosclerotic plaques is one of the primary causes of cerebrovascular and cardiovascular events. Neovascularization within the plaque is closely associated with its stability. Long non‐coding RNA (lncRNA) serves a crucial role in regulating vascular endothelial cells (VECs) proliferation and angiogenesis. In this study, we identified lncRNA HCG11, which is highly expressed in patients with vulnerable plaque compared with stable plaque. Then, functional experiments showed that HCG11 reversed high glucose‐induced vascular endothelial injury through increased cell proliferation and tube formation. Meanwhile, vascular‐related RNA‐binding protein QKI5 was greatly activated. Luciferase reporter assays and RNA‐binding protein immunoprecipitation (RIP) assays verified interaction between them. Interestingly, HCG11 can also positively regulated by QKI5. Bioinformatics analysis and luciferase reporter assays showed HCG11 can worked as a competing endogenous RNA by sponging miR‐26b‐5p, and QKI5 was speculated as the target of miR‐26b‐5p. Taken together, our findings revered that the feedback loop of lncRNA HCG11/miR‐26b‐5p/QKI‐5 played a vital role in the physiological function of HUVECs, and this also provide a potential target for therapeutic strategies of As.


| INTRODUC TI ON
Vascular endothelial cells (VECs) injury and the blood vessel functional imbalance induce atherosclerosis plaque. The instability and rupture of vulnerable atherosclerotic plaque are the main reasons for cardiovascular and cerebral vascular events. The local neovascularization is closely associated with plaque stability, 1,2 because excessive angiogenesis is associated with intraplaque haemorrhage, which may contribute to plaque progression and rupture. 3 Hyperglycaemia is known to impair endothelial cells (ECs) angiogenesis, such as in diabetic retinopathy (DR) or myocardial infarction (MI). [4][5][6] Regarding the mechanism of angiogenesis, many angiogenic factors and signalling pathways have been shown to regulate blood vessel growth and morphogenesis. 7 Recent studies have revealed important functions of lncRNAs in angiogenesis and cardio-cerebrovascular disease. 8 Long non-coding RNAs (lncRNAs) are characterized by a length of more than 200 nucleotides and a lack of coding protein function. It is becoming widely accepted that several types of lncRNAs, which are involved in regulating the ECs proliferation, migration and angiogenic capacity, are considered to be the new markers for the diagnosis and prognosis of cardiovascular and cerebrovascular diseases. [9][10][11] In atherosclerosis, as a molecular sponge for miR-195, overexpression of lncRNA activated by tumour growth factor-β (lncRNA-ATB) increases ECs viability, migration and angiogenesis, along with up-regulation of matrix metalloproteinase-2 (MMP-2), MMP-9, and vascular endothelial growth factor (VEGF) that associated with stability of plaque. 12 lncRNA TCONS_00024652 acts as a competitive endogenous RNA (ceRNA) that affects ECs proliferation and angiogenesis after TNF-α stimulation by modulating miR-21 expression. 13 According to previous reports, ln-cRNA HCG11 was up-regulated in atherosclerotic plaques, but the role of HCG11 in the progression of atherosclerotic plaque is unclear. 14 Subsequently, Zhang Y et al showed that down-regulation of HCG11 in prostate cancer (PCa) tissues was associated with poor survival of PCa patients. 15 Xu Y et al found HCG11 increased cell viability, proliferation and migration ability in HepG2 cells via interaction with IGF2BP1, leading to activation of MAPK signalling, which eventually promoted the progression of HCC. 16 As it is well known that the stability of advanced atherosclerotic plaque is closely related to intraplaque neovascularization, and in our present study, we found HCG11 was high expressed in vulnerable plaque and implemented preliminary research on the effect of angiopoiesis of HUVECs.
Mechanistically, in general, lncRNA exerts its biological effects by combining with competing endogenous RNAs to regulate target gene expression. 17,18 Bioinformatic evidence reveals the binding sites between HCG11 and miRNA-26b-5p. MiR-26b-5p is a member of the miR-26 family. It has been confirmed as an important regulator in breast cancer and other pathological processes. 19 MiR-26b-5p suppressed vascular mimicry (VM) and angiogenesis by down-regulating the expression of VE-cadherin, Snail and MMP2 and could inhibit the apoptosis of HCC cells 20 ; MiR-26b-5p inhibited liver fibrogenesis and angiogenesis through directly targeted PDGF receptor-β. However, the role of miR-26b-5p in the progression of atherosclerosis is still unclear.
Using bioinformatics analysis approaches, we identified the RNA-binding protein (RBP) quaking (QKI-5) as a miR-26b-5p target mRNA in addition, catRAPID analysis predicted the binding potential of HCG11 and QKI-5. QKI-5 is a member of the RNA Signaling and Activation (STAR) family of RNA-binding proteins. It plays an important regulatory role in accumulating ECs and to induce angiogenesis.
Azam SH et al found that endothelial QKI-5 expression remarkably correlated with angiogenic indices. 21 Cochrane A et al found that the mouse ECs overexpressing QKI-5 significantly improved angiogenesis and neovascularization and blood flow recovery in experimental hind limb ischaemia. The results highlight a clear functional benefit of QKI-5 in neovascularization, blood flow recovery and angiogenesis. 22 The present study was designed to explore the effect of HCG11 in the stability of atherosclerotic plaque. We measured the expression of HCG11 in human plasma samples. The previous study mentioned that high concentrations of glucose impaired angiogenesis. 5 Our experiments found HG-induced HUVECs reduced the expression of HCG11. HCG11 levels paralleled with QKI-5, a key molecule involved in triggering angiogenesis. In addition to being able to combine directly with QKI-5, we also found that HCG11 acts as sponge to regulate miR-26b-5p and further affects the expression of the target gene, QKI-5. All results showed that HCG11 reversed HGinduced HUVECs proliferation and angiogenesis inhibition via the miR-26b-5p/QKI-5 feed back loop signalling pathway.

| Ethics statement
The clinical study was approved by the ethics committees of Inner Mongolia Medical University of China (No. YKD2015061). All procedures were in accordance with the Helsinki declaration. All patients gave their written informed consent to participate in the study. The data do not contain any information that could identify patients.

| Clinical sample collection
The study population consisted of 290 patients with atherosclerosis vulnerable plaque, 178 stable plaque and 547 normal con-  Table 1.

| Definition of CT plaque characteristics
There are four points for diagnosing vulnerable atherosclerotic plaques: (1) Low attenuation plaque (LAP): average density ≤ 30 HU from 3 random region-of-interest measurements, with approximately 0.5 to 1.0 mm2 in non-calcified low CT attenuation portion of the plaque; (2) Positive remodelling (PR): remodelling index ≥ 1.1; (3) Spotted calcification (SC): average density> 130 HU, diameter < 3 mm in any direction, length of the calcium < 1.5 × the vessel diameter, and width of the calcification less than two-thirds of the vessel diameter and; (4) "Napkin ring" sign: ring-like attenuation pattern with peripheral high attenuation tissue surrounding a central lower attenuation portion. According to CT characteristics, patients were three groups: vulnerable plaque (VP) group, stable plaque (SP) group and normal group (NG).

| Fluorescence in situ hybridization (FISH)
The HCG11 FISH probes were designed and synthesized by RiboBio

| EdU proliferation assay
For the cell proliferation assay, HUVECs were seeded in 96-well plate. After transfection of the cells as described above, the cells According to the manufacturer's instructions, firstly, the cells were incubated with EdU for 4 h, following fixation, permeabilization and Edu staining, and last, the cells were stained with DAPI for 5 min, the percentage of the cells incorporated EdU was assayed using a fluorescence microscope.

| Matrigel-based tube formation assay
The capability of HUVECs to form capillary tube-like structures was assessed by the Matrigel-based tube formation assay. 50 μl of Matrigel was added to a pre-cooling 96-well plate and solidified at 37°C for 1 hour. After that, the pre-transfected HUVECs were harvested and cultured on the Matrigel-coated plate for another 8 hours, with or without glucose treatment (33mM). The tube formation effects were observed, and representative images were captured using a light microscope (magnification, ×400).

| Real-time fluorescent quantitative PCR (qRT-PCR)
After a series of above-mentioned processing, the total RNA was extracted from HUVECs using RNA isolation kit (Invitrogen,

| Western blot and antibodies
HUVECs were cultured and transfected as described in Cell Culture and Transfection. After transfection, cells were treated with glucose (33 mg/ml) for another 24 hours. After the treatment, the cells were collected and washed twice with phosphate-buffered saline.
The total proteins were extracted by adding 50 μl radio immuno-

| Bioinformatics analysis and luciferase assay
To explore the regulatory mechanism of lncRNA HCG11 in AS,

| Statistical analysis
All data were statistically analysed using the Statistical Package for the Social Sciences (SPSS) version 20.0 software (SPSS Inc, Chicago, IL, USA). All experiments were repeated 3 times independently, and the data were expressed as mean ± SD. Comparison of paired design between the two groups with positive distribution and homogeneity of variance was analysed by paired t test, otherwise non-paired t test was applied. Statistical analysis among more than two groups was conducted by one-way analysis of variance (ANOVA) or two-way ANOVA followed by a Tukey post hoc test. Correlations between HCG11 and miR-26b-5p, HCG11 and LDL or FBG were analysed using the Spearman's correlation test. Statistical significance is shown as described in the figure legends. P < .05 indicates the difference was statistically significant.   Figure 1D). Using Spearman's correlation test, we analysed the negative correlation between the lncRNA HCG11 expression and the FBG expression level in patients of vulnerable plaque ( Figure 1E), while there is no correlation between lncRNA HCG11

| Association of plaque characteristics with MACCE and clinical analysis
and LDL ( Figure 1F). Together, these findings indicated the potential involvement of lncRNA HCG11 in stability of plaque.

| High glucose inhibited proliferation and tube formation of HUVECs, simultaneously reduced lncRNA HCG11 expression
To investigate the role of LncRNA HCG11 in the progression of HUVECs, firstly, based on the RNA FISH and subcellular fractionation assay, LncRNA HCG11 was enriched in the cytoplasm of HUVECs (Figure 2A-B). The above data indicated that LncRNA These finding indicated that HG-induced inhibition of proliferation and tube formation may corrected with LncRNA HCG11.

| Overexpression of lncRNA HCG11 reversed HG-induced inhibition of proliferation and tube formation of HUVECs
To explore the function of lncRNA HCG11 in the proliferation and tube formation in HG treatment, we constructed lncRNA HCG11 overexpression vectors in HUVECs, lncRNA HCG11 is overexpressed after transfected with pcDNA/HCG11, the pcDNA3.1 was regarded as a negative control, qRT-PCR and RNA FISH assay were used to examine the overexpression efficiency ( Figure 3A-B). We performed proliferation assay and tube formation assays to gain insight into the role of lncRNA HCG11 in HG-induced cell proliferation and tube formation inhibition. As shown in (Figure 3E), EdU proliferation assay found that (F) the luciferase reporter assay was used to examine the relative luciferase activity in HUVECs cotransfected with lncRNA HCG11-WT or lncRNA HCG11-Mut reporter plasmid together with miR-26b-5p mimics, ** P < .01 vs miR-NC group. (G) The expression of lncRNA HCG11 in the present with miR-26b-5p mimics or inhibitors together with NG or HG condition, ** P < .01, && P < .01 vs miR-NC

| QKI-5 is a target gene of HCG11 and is involved in lncRNA HCG11 mediated HUVECs proliferation and angiogenesis
Previous studies have demonstrated that QKI-5 is a key molecule involved in triggering angiogenesis 23 and is dysregulated in many cancers. 24 Quantitative RT-PCR showed that the level of QKI-5 mRNA was significantly higher in vulnerable plaque group, compared with that in stable plaque group (p＜0.0001) ( Figure 4A). In addition, correlation analyses showed that QKI-5 expression was positively associated with LncRNA HCG11 in plasma of patients with vulnerable plaques (p＜0.05) ( Figure 4B).
In vitro experiment also showed that QKI-5 was down-regulated in HGtreated HUVECs ( Figure 4C). To verify whether the angiogenesis effects of lncRNA HCG11 are mediated by QKI-5, we further detected the expression of QKI-5 with lncRNA HCG11 overexpression in HG condition.
Western Blot Analysis indicated that QKI-5 protein levels were increased after overexpression lncRNA HCG11( Figure 4D). Moreover, qRT-PCR was used to examine the transfection efficiency of QKI-5 ( Figure 4E), EdU and tube formation assay showed that knockdown of QKI-5 suppressed lncRNA HCG11-induced HUVECs proliferation and tube formation ( Figure 4F-G). These results indicate that lncRNA HCG11 might play a role in proliferation and angiogenesis through QKI-5.

| LncRNA HCG11 directly interacted with QKI-5 in HUVECs
As a RNA-binding protein, QKI-5 favoured interacted with RNAs to perform its function. To verified potential interaction between lncRNA HCG11 and QKI-5, catRAPID was used to rapidly evaluate the interaction tendency of lncRNA HCG11 and QKI-5 based on the secondary structure, hydrogen bonding and molecular interatomic forces. The prediction revealed that there existed an interaction between lncRNA HCG11 and QKI-5 with a IP value of 150 and DP value of 100% ( Figure 5A-B). Then, to further verified this prediction, RIP experiment was applied to detected whether they can interact with each other, the relative enrichment levels of lncRNA HCG11 was increased in the anti-QKI-5 group than the anti-IgG group ( Figure 5C).
Our previous data showed QKI-5 is a down regulator of lncRNA HCG11, weather QKI-5 played a role in the regulation of lncRNA HCG11, this deserves further research. Knockdown of QKI-5 rapidly reduced the enrichment of lncRNA HCG11 with anti-QKI-5 ( Figure 5D-F), and the total expression level of lncRNA HCG11 in HUVECs also reduced ( Figure 5G).

| LncRNA HCG11 functions as miR-26b-5p sponge in HG-induced HUVECs
In order to research the mechanism of lncRNA HCG11 regulated QKI-5, Starbase prediction software showed miR-26b-5p is a competitive miRNA between lnRNA HCG11 and QKI-5. So we assessed whether lncRNA HCG11 acts as a competitive endogenous RNA to miR-26b-5p and inhibits the expression of miR-26b-5p. Quantitative RT-PCR (qRT-PCR) shows that miR-26b-5p level was significantly lower in vulnerable plaque group compared with stable plaque group (p＜0.01). And it was no significantly difference between stable plaque group and normal controls ( Figure 6A).
In vitro data showed treatment with glucose (33mM) for 24h up-regulated the expression of miR-26b-5p ( Figure 6B). Correlation analyses showed that lncRNA HCG11 expression was negatively associated with miR-26b-5p level in plasma of patients with atherosclerotic vulnerable plaques ( Figure 6C). Overexpression of lncRNA HCG11 significantly down-regulated the level of miR-26b-5p ( Figure 6D). To verify whether the angiogenesis effects of lncRNA HCG11 are completely mediated by miR-26b-5p, we next conducted dual-luciferase reporter assays. The results demonstrated that cotransfection with wild-type lncRNA HCG11 (HCG11-WT) plasmids and miR-26b-5p mimics remarkably reduced the luciferase activities in HUVECs, whereas the luciferase activities were not changed when the cells were cotransfected with mutant lncRNA HCG11 (LUCAT1-MUT) plasmids as well as miR-26b-5p mimics ( Figure 6E-F). Moreover, the relative expression of lncRNA HCG11 was significantly decreased in HG-induced HUVECs when they were transfected with miR-26b-5p mimics, whereas silence of miR-26b-5p mimics resulted in notably decreased expression of lncRNA HCG11 in HG-induced HUVECs ( Figure 6G). Increased expression of miR-26b-5 reduced the endothelial cells proliferation and tube formation ( Figure 7A-C). Further research found miR-26b-5p mimics abrogated the effects of lncRNA HCG11 on high glucose-induced angiogenesis inhibition ( Figure 7D-E). All results suggested that miR-26b-5p was a target of lncRNA HCG11 and interacts with HCG11 to promote angiogenesis.

| LncRNA HCG11/MiR-26b-5p/QKI-5 axis formed a feedback loop to regulate proliferation and tube formation of HUVECs
StarBase analysis indicated that QKI-5 was predicted to be a tar- well as miR-26b-5p mimics ( Figure 8A), suggesting the interaction between miR-26b-5p and QKI-5. Furthermore, we cotransfected with pcDNA-HCG11 in both groups, the former effect of the luciferase activities was partly reversed and the latter did not ( Figure 8B).
Deservedly, both the mRNA and protein expression level of QKI-5 were decreased after up-regulation of miR-26b-5p ( Figure 8C-D).
While this inhibition effect was impaired in response to overexpression of lncRNA HCG11 ( Figure 8E-F). For function assay, the proliferation and angiogenesis promoting of lncRNA HCG11 overexpression were reduced by miR-26b-5p, but partially rescued by the up-regulation of QKI-5 ( Figure 8G-H), Collectively, lncRNA HCG11 promoted HUVECs proliferation and angiogenesis through regulating miR-26b-5p/QKI-5 feedback loop axis.  miRNAs usually repressed gene expression by combined with 3'UTR of specific mRNA, in our study, bioinformatics predicted miR-26b-5p might bind to the position of 017-022 in QKI-5 3'-UTR.

| D ISCUSS I ON
Further reporter vectors construction and dual-luciferase reporter were implemented to verify the binding sites between miR-26b-5p and QKI-5, respectively. In HG-treated HUVECs, overexpression of lncRNA HCG11 significantly increased the expression of QKI-5.
While gain or out of miR-26b-5p obviously reduced or increased QKI-5 expression. Functional experiments showed that, overexpression of QKI-5 increased the proliferation and tube formation of HUVECs, and lncRNA HCG11 induced cell proliferation could be inhibited by overexpression of miR-26b-5p or knockdown of QKI-5.
In this study, we found that the combined sequence (AAUGAACU) between miR-26b-5p and lncRNA HCG11 is much of the same as the bind to the QKI-5 3'UTR. Above-mentioned results indicated that lncRNA HCG11 worked as a ceRNA binding with miR-26b-5p, and then attenuated the inhibitory effect of miR-26b-5p on QKI-5 3'UTR, finally affected the progression of HUVECs.
Interestingly, the combination between QKI-5 and lncRNA HCG11 is predicted by catRAPID and confirmed by RIP assay.
Overexpression of QKI-5 increased the expression of lncRNA HCG11, simultaneously, lncRNA HCG11 sponged miR-26b-5p and reduced the suppression of miR-26b-5p on QKI-5 3'UTR. Taking together, these results suggested that lncRNA HCG11/miR-26b-5p/ QKI-5 formed a positive-feedback loop to regulate the physiological function of HUVECs (Supplementary Figure 2). These findings provide new insights for the diagnose of the vulnerable plaque, as well as a new therapeutic target for the treatment of AS in the future.

| CON CLUS ION
Our study revered that the feedback loop of lncRNA HCG11/ miR-26b-5p/QKI-5 played a vital role in the physiological function of HUVECs and that could provide a potential target for therapeutic strategies of As.

CO N FLI C T O F I NTE R E S T
We declare that the authors have no competing interests that might be perceived to influence the results and/or discussion reported in this paper.