Knockdown of lncRNA AK139328 alleviates myocardial ischaemia/reperfusion injury in diabetic mice via modulating miR‐204‐3p and inhibiting autophagy

Abstract This study was aimed at investigating the effects of lncRNA AK139328 on myocardial ischaemia/reperfusion injury (MIRI) in diabetic mice. Ischaemia/reperfusion (I/R) model was constructed in normal mice (NM) and diabetic mice (DM). Microarray analysis was utilized to identify lncRNA AK139328 overexpressed in DM after myocardial ischaemia/reperfusion (MI/R). RT‐qPCR assay was utilized to investigate the expressions of lncRNA AK139328 and miR‐204‐3p in cardiomyocyte and tissues. Left ventricular end diastolic diameter (LVEDD), left ventricular end systolic diameter (LVESD), left ventricular ejection fraction (LVEF) and fractioning shortening (FS) were obtained by transthoracic echocardiography. Haematoxylin‐eosin (HE) staining and Masson staining were utilized to detect the damage of myocardial tissues degradation of myocardial fibres and integrity of myocardial collagen fibres. Evans Blue/TTC staining was used to determine the myocardial infarct size. TUNEL staining was utilized to investigate cardiomyocyte apoptosis. The targeted relationship between lncRNA AK139328 and miR‐204‐3p was confirmed by dual‐luciferase reporter gene assay. MTT assay was used for analysis of cardiomyocyte proliferation. Western blot was utilized to investigate the expression of alpha smooth muscle actin (α‐SMA), Atg7, Atg5, LC3‐II/LC3‐I and p62 marking autophagy. Knockdown of lncRNA AK139328 relieved myocardial ischaemia/reperfusion injury in DM and inhibited cardiomyocyte autophagy as well as apoptosis of DM. LncRNA AK139328 modulated miR‐204‐3p directly. MiR‐204‐3p and knockdown of lncRNA AK139328 relieved hypoxia/reoxygenation injury via inhibiting cardiomyocyte autophagy. Silencing lncRNA AK139328 significantly increased miR‐204‐3p expression and inhibited cardiomyocyte autophagy, thereby attenuating MIRI in DM.

Diabetic patients are at higher risk of ischaemic heart disease as well as myocardial ischaemia/reperfusion (I/R) injury (MIRI). 4 Moreover, researchers have revealed that diabetes mellitus aggravates MIRI and influences the functioning of ischaemic preconditioning and some cardioprotective pharmacologic agents. 5 Therefore, the investigation of pathological mechanism of diabetes in interaction with MIRI novel targets would be contributing to diabetic myocardium protection.
Non-coding RNAs including microRNAs (about 20-22 nt) and long non-coding RNAs (lncRNAs, >200 nt) have already been heated research topics. 6 Previous researches have suggested that some lncRNAs were closely associated with myocardial ischaemiareperfusion. Li et al provide evidence for lncRNA KCNQ1OT1 knockdown being a preventive solution to MIRI. 7 Zhao et al demonstrated that lncRNA MALAT1 was involved in myocardial ischaemia-reperfusion and down-regulation of MALAT1 contributed to the reduction of cardiomyocyte apoptosis. 8 LncRNA AK139328, which has been indicated by Chen et al to be involved in liver I/R injury, is a vital member of lncRNAs. 9 However, the mechanism and role of lncRNA AK139328 in MIRI have not been completely investigated.
Apoptosis or type I programmed cell death, a main type of cell death that occurs when DNA damage is irreversible, is characterized by distinct morphological and biochemical changes. 10,11 I/R induced myocardium apoptosis, which becomes a key factor in MIRI pathogenesis. 12 Type II programmed cell death, namely autophagy, is by mechanism distinct from apoptosis. 13 Reduced superoxide dismutase (SOD) and increased lactate dehydrogenase (LDH) and malondialdehyde (MDA) concentration appeared in line with autophagy. 14 Recent studies emphasized the dynamic process of autophagy, that is autophagic flux in understanding the impairment caused by MIRI. 15 In addition, while basal autophagy maintains cell recycling and thereby inhibits cell apoptosis in infarct area, excessive acceleration of autophagy would probably lead to an aggravation in MIRI. 11,15 A deep investigation into the apoptotic and autophagic mechanism underlying MIRI is of pivotal importance in the advancement of myocardial ischaemia treatment.
Recent studies have also suggested that various miRNAs could regulate cell apoptosis, such as miR-1, miR-133, miR-208, miR-21 and miR-204. MiR-204 has already been identified to exert antiautophagy effects and may also regulate metabolic recovery in MIRI. 16 Nonetheless, the mechanisms of how such miRNAs regulate myocardial cell apoptosis and autophagy in MIRI as modulated by lncRNAs remain unknown. Understanding the influence of their interaction on autophagy in MIRI will contribute to develop new diagnostic and therapeutic strategies.
In this study, we constructed myocardial I/R models in normal mice (NM) and diabetic mice (DM). The expression level of AK139328 and miR-204-3p in cardiomyocyte and tissues was detected, along with their targeting relationship. After AK139328 silencing, the apoptosis ratio and the autophagy level of cardiomyocyte were evaluated. Tissue damage and infarct size were determined by staining. Our study may supply novel therapeutic strategies for the treatment of MIRI.

| Experimental animals and treatments
Adult healthy male C57BL/KsJ db/+ mice (6-8 weeks), weight 20-25 g, and adult male C57BL/KsJ db/db mice were purchased from Cavens Lab Animal Co., Ltd. (Changzhou, China) and kept in a temperature-controlled room with a constant temperature of 26 ± 2°C, a humidity of 60%-80% and a programmed 12-hours light/12-hours dark cycle for circadian control. Fasting for one day was requested before the experiments. The mice were kept at specific pathogenfree (SPF) level to prevent the infection of specific disease. C57BL/ KsJ db/+ mice were used as normal mice, and C57BL/KsJ db/db mice were used as diabetic mice. The biological characteristics of C57BL/ KsJ db/+ and C57BL/KsJ db/db mice are listed in Table S1. C57BL/ KsJ db/+ were randomly divided into 2 groups: the sham (normal mice, NM) group and the I/R (NM) group. C57BL/KsJ db/db mice were randomly assigned to 4 groups: the sham (DM) group, the I/R (diabetic mice, DM) group, the I/R+Lv-shRNA1-AK139328 group and the I/R+Lv-shRNA2-AK139328 group. Each of these 6 groups contained 40 mice. Animal protection measures and all experimental procedures were strictly implemented following the agreement of the animal experiment regulations approved by the Second Xiangya Hospital, Central South University.

| Myocardial ischaemic reperfusion injury (MIRI) surgery
According to the explanation on myocardial I/R model construction by Gao et al,17 all mice were anesthetized with inhaled ether (3 minutes) and fixed for endotracheal intubation by the use of a small animal respirator. A longitudinal incision was made from the third to fourth ribs, exposing the heart. Then, a 5-0 Prolene suture was placed around the 2 cm of the root of left anterior descending coronary artery (LAD). The suture was loosened after occlusion for 30 minutes, which was followed by 120-minutes reperfusion of LAD.

| Microarray analysis
Mouse LncRNA Expression Array V3.0 chip (Aksomics, Shanghai, China) was utilized to screen differentially expressed lncRNAs in NM and DM after MIRI. Total lncRNAs were extracted for chip hybridization. The expression levels of lncRNAs were represented by fluorescence intensity values. The differential expression criteria were set as fold change >2, P < .05.

| RT-qPCR assay
TRIzol ™ reagent (Invitrogen, Carlsbad, CA, USA) was used to obtain total RNA. Two microgram of total RNA was used for reverse transcription by PrimeScript® 1st Strand Synthesis Kit (TaKaRa,Tokyo, Japan). The real-time RT-qPCR was performed with QuantiTect YU ET AL.

| Myocardial enzyme determination
Mice were anesthetized with inhaled ether 24 hours after reperfusion. Blood sampled from the inner canthus vein was centrifuged for 15 minutes. The serum was kept at a temperature of −20°C. Semiautomatic biochemical analyser was used to assess the expression of phosphocreatine kinase (CK), creatine kinase MB (CK-MB) and lactate dehydrogenase (LDH) in the serum.

| Echocardiographic assessment
At the end of reperfusion, mice were re-anesthetized with isoflurane,

| Histopathology examination
The histopathology examination of myocardial tissues was performed with haematoxylin-eosin (HE) staining. The left ventricle of heart samples was put in 10% formaldehyde solution, dehydrated in ethanol gradient, embedded in paraffin and cut down into slices of 4 μm.
After deparaffinage, the samples were stained with haematoxylin and eosin. Then, the slices were mounted and observed under a light microscope (Leica Microsystems, Wetzlar, Germany).

| Masson staining
Mice cardiac tissues were fixed with 10% formaldehyde for 24 hours at room temperature, then decalcified, dehydrated, permeabilized with xylene, embedded in wax and finally sliced into

| Infarct size determination
To prove the effects of I/R, the infarct size of the hearts was tested in this study. After the haemodynamic testing, the LAD was sutured. Evans Blue dye (1 mL of a 2.0% solution) was injected through a carotid artery catheter into the coronary circulation to delineate the in vivo area at risk. The heart was rapidly excised and frozen at a temperature of −20°C for 30 minutes and then serially cross-sectioned in 1-mm-thick sections which were then incubated in pH7.4 1.0% 2,3,5-triphenyl tetrazolium chloride (TTC) formulated in phosphate-buffered solution (PBS) for 15 minutes at 37°C. Sizes of the normal region (area not at risk, ANAR), the ischaemic region (area at risk, AAR) and the infarct area (INF) were assessed by Image-Pro Plus software, results of which were taken as INF/AAR × 100%.

| TUNEL staining
The heart was rapidly excised after reperfusion and sectioned into

| Caspase-3 activity assay
Caspase-3 activities were assessed by caspase-3 activity assay kit (BestBio, Shanghai, China). The cardiomyocyte were lysed with 90 μL lysis buffer. Subsequently, 10 μL Ac-DEVD-ρNA was added to cell lysates, and the mixtures were incubated for 2 hours at room temperature. The absorbance at 405 nm of apoptosis cells and control cells was determined by enzyme immunoassay detector. Caspase-3 activity was taken as the ratio of these 2 absorbance values.

| Cell isolation and culture
Cardiomyocytes were isolated from adult C57BL/KsJ db/+ mice heart.
In brief, mice were narcotized with 2% isoflurane, and hearts were taken off and perfused at 37°C for 3 minutes with a Ca 2+ -free bicarbonate- After the myocytes were pelleted by gravity for 10 minutes, the supernatant was aspirated, and the myocytes were resuspended in bicarbonate-based buffer containing 250 mmol/L Ca 2+ . One hour after plating, cells were washed with PBS and non-adhering cells were removed from the culturing system. Cells were randomized to receive into 2 groups: (i) control group (culture medium containing 5 mmol/L glucose) and (ii) high-glucose group (culture medium containing 33 mmol/L glucose).

| Hypoxia/reoxygenation (H/R) injury model
By the model of hypoxia/reoxygenation (H/R) injury, we analysed hypoxia-induced and reoxygenation-induced cell death. The cardiac myocytes were cultured under a hypoxic gas mixture supplemented with 95% N 2 and 5% CO 2 , and the medium was placed in a hypoxic incubator (95% N 2 and 5% CO 2 ) for 3 hours which was followed by reoxygenation incubator (95% O 2 and 5% CO 2 ) for 3 hours. Cells of control group were cultured with 5% CO 2 at 37°C for 6 hours.

| Construction of shRNA plasmid vector
Interference sequences of lncRNA-AK139328 were designed by BLOCK-iT TM RNAi designer (Invitrogen). Relevant sequences are shown in Table S3. BamH Ι and Hind III restriction sites were led into 3′-UTR and 5′-UTR of shRNA to synthetize oligonucleotide single strand. shRNA was linked with pGFP-V-RS plasmid vector. The plasmids were extracted after amplification.
After transfection, all cells were collected for further experiments.

| LDH, SOD and MDA determination
After cardiomyocyte H/R model establishment, the expression level of LDH was detected with LDH cytotoxicity assay kit (Beyotime, Shanghai, China). The expression level of SOD was detected with total superoxide dismutase assay kit (Solarbio, Beijing, China), and the expression level of MDA was detected with malondialdehyde assay kit (Solarbio). The level of MDA was taken as the disparity of the absorbance at 532 nm and 600 nm.

| MTT assay
Myocardial cells were seeded onto 96-well plates at a density of 1000-10 000 cells/well and cultured in 200 μL cell culture medium YU ET AL.
| 4889 F I G U R E 1 Knockdown of lncRNA AK139328 suppressed CK, CK-MB and LDH expressions. A, The results of chip analysis showed the differentially expressed lncRNAs in normal mice (NM) and diabetic mice (DM) after myocardial ischaemia/reperfusion (MI/R). B, The expression of lncRNA AK139328 was indicated to be most up-regulated by RT-qPCR. *P < .05, **P < .01, compared with sham (NM) group. ## P < .01, ### P < .001, compared with sham (DM) group. C, The expression of lncRNA AK139328 in shRNA-AK139328 group was inhibited. **P < .01, compared with NC group. D, The expression of CK after inhibiting AK139328 in DM was reduced significantly. **P < .01, compared with sham (NM) group, && P < .01, compared with sham (DM) group, and ## P < .01, compared with I/R (DM) group. E, The expression of CK-MB after inhibiting AK139328 in DM was reduced significantly. **P < .01, compared with sham (NM) group, && P < .01, compared with sham (DM) group, and ## P < .01, compared with I/R (DM) group. F, The expression of LDH after inhibiting AK139328 in NM and DM was reduced significantly. **P < .01, compared with sham (NM) group, && P < .01, compared with sham (DM) group, and ## P < .01, compared with I/R (DM) group at 37°C for 3-5 days. Ten microlitre MTT (5 mg/mL, pH = 7.4, prepared with PBS) was added to culture the cells for 4 hours. After the medium was turned away, the precipitate was made soluble in 100 μL DMSO. An enzyme-linked immunosorbent plate reader was utilized to determine the absorbance of each well.

| Statistical analysis
The cell experiments were performed at least 3 times. The animal experiments were performed at least 6 times. The data were exhibited as mean ± SD. Student's t test was utilized to evaluate data between 2 groups, while the comparison among multiple groups was conducted by one-way ANOVA. Statistical analysis was performed with Graph-Pad Prism 6.0 software. Significance level was set as P < .05.

| Knockdown of lncRNA AK139328 suppressed the expression of CK, CK-MB and LDH
As shown in Table S1, the blood glucose and serum insulin levels of C57BL/KsJ db/db mice were significantly greater than those of C57BL/KsJ db/+ mice. So, C57BL/KsJ db/+ mice were used as normal mice, and C57BL/KsJ db/db mice were used as diabetic mice. Then, differentially expressed lncRNAs in normal mice (NM) and diabetic mice (DM) after MIRI were selected with LncRNA Microarray v3.0 chip analysis ( Figure 1A). The results of RT-qPCR assay showed that lncRNA AK139328 was most up-regulated in DM compared with other lncRNAs ( Figure 1B). shRNA1 and shRNA2 were transfected into heart tissues of DM to inhibit the expression of lncRNA AK139328. The results of RT-qPCR indicated that AK139328 was significantly down-regulated compared with NC group, indicating successful transfection ( Figure 1C). The expression of CK after I/R was higher in DM than that in NM, whereas knockdown of lncRNA AK139328 significantly suppressed the expression of CK ( Figure 1D). The expression of CK-MB after I/R was higher in DM than that in NM. Knockdown of lncRNA AK139328 repressed the expression of CK-MB significantly in comparison with the I/R group ( Figure 1E). Likewise, the expression of LDH after I/R was higher in DM in comparison with NM, while knockdown of lncRNA AK139328 marked a significant decrease in concentration of LDH ( Figure 1F). The above results indicated that ischaemia in myocardium was alleviated by downregulated AK139328.

| Knockdown of lncRNA AK139328 relieved myocardial ischaemia/reperfusion injury in DM
Haematoxylin-eosin (HE) staining results revealed the degree of myocardial tissue impairment in mice models. In Figure 3A inhibitor group was tested with RT-qPCR assay. Relative AK139328 expression was promoted in H/R (control) group and H/R (high-glucose) group, compared with NC groups ( Figure 5C). On the contrary, relative miR-204-3p expression was inhibited in H/R (control) group and H/R (high-glucose) group, compared with NC groups ( Figure 5D). Therefore, miR-204-3p was the direct target of lncRNA AK139328.

| AK139328 knockdown and miR-204-3p overexpression relieved H/R injury via inhibiting autophagy
Relative AK139328 expression was reduced in shRNA1-AK139328 group ( Figure 6A). Relative miR-204-3p expression was increased in miR-204-3p mimic group, while reduced in miR-204-3p inhibitor group ( Figure 6B). The above results all indicated successful F I G U R E 3 Knockdown of lncRNA AK139328 relieved myocardial ischaemia/reperfusion injury in diabetic mice (DM). A, The results of HE staining showed that knockdown of lncRNA AK139328 relieved the dreadful effects of I/R on myocardial tissue. B, Masson staining results displayed I/R that caused myocardial fibres generated a disorder arrangement and large amounts of collagen deposition. The myocardial collagen fibres were stained blue, and the myocardial fibres were stained red. C, Evans Blue/TTC staining results showed that I/R could lead myocardial infarction to normal mice (NM) and DM. The myocardial infarct size was reduced in DM on account of low expression of lncRNA AK139328. D, The expression of α-SMA in MI/R groups was inhibited compared with sham groups. The knockdown of lncRNA AK139328 significantly improved the level of α-SMA. **P < .01, compared with sham (NM) group, && P < .01, compared with sham (DM) group, and # P < .05, ## P < .01, compared with I/R (DM) group YU ET AL. Atg5 and LC3-II/LC3-I, the level of p62, which was negatively associated with autophagy, acted likewise but towards an opposite direction ( Figure 7A,B). Therefore, lncRNA AK139328 relieved H/R injury through inhibiting cardiomyocyte autophagy via regulating miR-204-3p.

| DISCUSSION
Autophagy has been reported by the recent literature to be protective for myocardial I/R, and there are increasing studies on F I G U R E 5 LncRNA AK139328 targeted and modulated miR-204-3p. A, Bioinformatics prediction and dual-luciferase reporter assay. MiR-204-3p-binding site in lncRNA AK139328 wild-type form (AK139328-wt) and the mutated form (AK139328-mut) were shown in the upper panel. HEK293T cells were transfected with miR-204-3p mimics or mimic-NC and then transfected with the luciferase constructs of AK139328-wt or AK139328-mut. The luciferase activity was analysed. B, The expression of miR-204-3p was inhibited in I/R (NM) group and I/ R (DM) group and the expression of miR-204-3p rose in I/R + shRNA groups. **P < .01, compared with sham (NM) group, && P < .01, compared with sham (DM) group, and ## P < 0.01, compared with I/R (DM) group. C, Relative AK139328 expression was promoted in hypoxia/ reoxygenation (H/R) (control) group and H/R (high-glucose) group, compared with NC groups. **P < .01, compared with NC (control) group, and && P < .01, compared with NC (high-glucose) group. D, Relative miR-204-3p expression was inhibited in H/R (control) group and H/R (highglucose) group, compared with NC groups. **P < .01, compared with NC (control) group, and && P < .01, compared with NC (high-glucose) group autophagy in interaction with apoptosis in ischaemic myocardium. 18 For instance, Zheng et al investigated berbamine pre-treatment on protecting I/R injury in hearts on the basis of autophagy mechanism. 19 Ma et al emphasized ALDH2 as therapeutic target for MIRI and it prompting autophagy during ischaemia and remained elevated autophagy at reperfusion. 20 Yao et al argued the MIRI preventive role of vitamin D receptor in that its overexpression repressed abnormal apoptosis and autophagy. 21 Our work figured out that knockdown of lncRNA AK139328 relieved the damage of I/R to myocardium tissue and reduced infarction size. We identified miR-204-3p as the direct target of lncRNA AK139328 and negatively regulated autophagy and apoptosis both in animal normal or in F I G U R E 6 LncRNA AK139328 knockdown and miR-204-3p overexpression relieved hypoxia/reoxygenation (H/R) injury in high-glucose environment. A, Relative AK139328 expression was reduced in shRNA1-AK139328 group. **P < .01, compared with NC group. B, Relative miR-204-3p expression was promoted in miR-204-3p mimic group and reduced in miR-204-3p inhibitor group. **P < .01, compared with NC group. C, The expression of LDH increased after H/R injury. shRNA1, miR-204 mimics and 3-MA reduced the expression of LDH, which relieved H/R injury, while miR-204 inhibitor promoted it. D, H/R injury significantly reduced the number of viable cells, which was lower in H/ R + miR-204 inhibitor group and higher in H/R + shRNA1 group, H/R + miR-204 mimic group and H/R + 3-MA group. E, SOD concentration was inhibited after H/R injury both in control group and in high-glucose group. SOD concentration was recovered in H/R + shRNA1 group, H/ R + miR-204 mimic group and H/R + 3-MA group but inhibited in H/R + miR-204 inhibitor group. F, MDA concentration was promoted after H/R injury both in control group and in high-glucose group. MDA concentration was promoted in H/R + miR-204 inhibitor group and recovered in H/R + shRNA1 group, H/R + miR-204 mimic group and H/R + 3-MA group. **P < .01, compared with NC (control) group, && P < .01, compared with NC (high-glucose) group, and # P < .05, ## P < .01, compared with H/R (high-glucose) group diabetic models and in vitro experiments of normal or high-glucose condition. We concluded that by targeting miR-204-3p, silencing MiR-204-3p has been widely accepted as a tumour suppressor.
Chen et al stated that miR-204-3p demoted glioma cell proliferation by enhancing glioma cell apoptosis. 26  Limitations still exist in this study. Firstly, as all our assays were performed in mice, the experimental results might not be directly extrapolated to humans. For future experiments, we may introduce tests on other animal models such as rates. We also required further identification and confirmation of the precise mechanisms underlying AK139328 and miR-204-3p. Above all, we made a successful attempt in identifying lncRNA AK139328 as MIRI-influencing factor.
Ongoing researches showed that lncRNAs and miRNAs played a powerful role in almost all cellular events including myocardial I/R injury. Higher levels of apoptosis and autophagy have been detected after MIRI. AK139328 and miR-204-3p have been indicated to modulate such cellular processes. Knockdown of AK139328 attenuated myocardial ischaemia/reperfusion injury and inhibited cardiomyocyte autophagy and apoptosis in DM. It was confirmed that AK139328 directly modulated miR-204-3p, thereby modulating apoptosis and autophagy.

CONFLI CT OF INTEREST STATEMENT
The authors confirm that there are no conflict of interests.