Long Non-coding RNA GAS5 Maintains Insulin Secretion by Regulating Multiple miRNAs in INS-1 832/13 Cells

Type-2 diabetes mellitus (T2DM) is a complex disease characterized by reduced pancreatic islets β-cell mass and impaired insulin release from these cells. Non-coding RNAs, including microRNAs (miRNA) and long non-coding RNAs (lncRNAs), play a role in the progression of T2DM. Decreased serum lncRNA GAS5 levels were reported to be associated with T2DM. However, the role of GAS5 in regulating islet cell functions remain unknown. In this study, we found that the serum levels of GAS5 were significantly lower in patients with T2DM compared with healthy control subjects, and the low serum GAS5 levels were associated with high levels of HbAlc and fasting glucose in patients with T2DM. In addition, we found that serum GAS5 levels were negatively correlated with the serum levels of miR-29a-3p, miR-96-3p, and miR-208a-3p in patients with T2DM. Consequently, using INS-1 832/13 rat β-cell line, we found that overexpression of GAS5 by lentivirus infection increased glucose-stimulated insulin secretion and insulin content compared with negative control, whereas knockdown of GAS5 expression reduced both them. Moreover, GAS5 interacted with miR-29a-3p, miR-96-3p, and miR-208a-3p in INS-1 832/13 cells, as judged by pull-down assay and dual luciferase reporter assay. GAS5 overexpression caused the decrease in expression of miR-29a-3p, miR-96-3p, and miR-208a-3p in INS-1 832/13 cells and conversely caused the increase in expression of insulin receptor, insulin receptor substrate, and phosphoinositide-3-kinase regulatory subunit 1. Thus, these results reveal a novel mechanism whereby GAS5 is involved in maintaining insulin secretion and may represent a novel therapeutic target for T2DM.


INTRODUCTION
Type-2 diabetes mellitus (T2DM) is a complex disease characterized by diminishing pancreatic islet β-cell mass and impaired insulin release from these cells (Wessel et al., 2015). T2DM has become one of the leading causes of morbidity and mortality and costs astronomical sums of money in annual healthcare in worldwide. Non-coding RNAs, including microRNAs (miRNAs) and long non-coding RNAs (lncRNAs) have been reported to play a role in the progression of T2DM (Chang, 2017). These non-coding RNAs are crucial for controlling the expression of genes related to insulin release, which points to unknown regulatory network governing islet cell functions (Akerman et al., 2017). LncRNAs function as biomarkers in the personalized treatment of T2DM (Lin et al., 2019;Liu et al., 2019). Researchers have discovered thousands of lncRNAs that highly enriched or specifically expressed in pancreatic islet β-cells (Qi et al., 2017). Carter et al. (2015) demonstrated that decreased serum LncRNA GAS5 (growth arrest-specific transcript 5) levels were associated with higher risk for having T2DM (Carter et al., 2015). GAS5 is extensively expressed in various tissues and organs and plays a key role in cell apoptosis and growth. Jin et al. found that the GAS5 level was significantly decreased in db/db mice. Knockdown of GAS5 expression led to cell cycle G1 arrest and impaired insulin synthesis and secretion in Min6 cells (Jin et al., 2017). A recent study demonstrates that modulation of GAS5 in the human beta cell alleviated the glucocorticoid-induced insulin secretion defect, suggesting the potential of GAS5 as a new regulator, maintaining β cell identity and function by affecting insulin synthesis and secretion (Esguerra et al., 2020).
However, the underlying mechanisms how GAS5 regulates islet cell functions are not fully clear. LncRNA, as a competitive endogenous RNA (ceRNA), binds to miRNA and acts as a miRNA sponge, thereby reducing the activity of miRNA and indirectly up-regulating the expression of miRNA-related target genes (Goustin et al., 2019). It has been reported that GAS5 targets to miR-452-5p and miR-221-3p in renal tubular cells and mesangial cells, which are involved in diabetic nephropathy (Ge et al., 2019;Xie et al., 2019).
This study aimed to identify the potential mechanism of GAS5 as a miRNA sponge in maintaining islet cell function. We demonstrated that serum GAS5 level was downregulated in patients with T2DM compared with healthy controls and was associated with high levels of HbAlc and fasting glucose. Overexpression of GAS5 by lentivirus infection increased glucose-stimulated insulin secretion and insulin content. By the bioinformatics analysis, pull-down and luciferase assay, we revealed that GAS5 interacted with miR-29a-3p, miR-96-3p, and miR-208a-3p, thereby regulating the expression of insulin receptor (INSR), insulin receptor substrate 1 (IRS-1), and phosphoinositide-3-kinase regulatory subunit 1 (PIK3R1) in INS-1 832/13 cells. Therefore, GAS5 might serve as a regulator in pancreatic β cell function and T2DM development through acting as a miRNA sponge.

Tissue Samples
One hundred and twenty-two serum samples from health people and 88 serum samples from type-2 diabetes were collected at Xiangya Hospital, Central South University (Changsha, China). The statistical number required for this study was estimated by G * Power software (Faul et al., 2007). According to the WHO diagnostic criteria in 1999 and oral glucose tolerance test (OGTT), the patients with fasting blood glucose (FPG) ≥7.0 mmol/L and 2 h postprandial blood glucose (2 h PG) ≥ 11.1 mmol/L were included in this study. Exclusion criteria: the patients with hypertension, malignant tumors, lung diseases, polycystic ovary syndrome, autoimmune diseases; patients with other diseases affecting glucose metabolism, such as hyperthyroidism, acromegaly, and Cushing's syndrome, acute metabolic abnormalities syndrome or other stress conditions and smokers were excluded. The subjects with fasting blood glucose (FPG) <6.1 mmol/L and 2 h postprandial blood glucose (2 h PG) <7.8 mmol/L were included as healthy controls in this study. The basic clinical information of 122 healthy subjects and 88 patients with T2DM was shown in Table 1. The samples were snap-frozen in liquid nitrogen, and then stored at −80 • C for further use. This project was approved by the Ethic Committee of Xiangya Hospital, Central South University. All subjects have signed the informed consent.

Cell Culture
Rat β-cell line INS-1 832/13 and pancreatic β cell line Min6 were purchased from National infrastructure of Cell Line Resource (Beijing, China) and grown routinely in RPMI-1640 medium with 11.1 mM D-glucose (cat no. 984313, Invitrogen, CA, United States) supplemented with 10% fetal bovine serum (cat no. 16140071, Gibco, CA, United States) and 50 µmol/L 2-mercaptoethanol (cat no. 21985023, Invitrogen, CA, United States). Culture conditions were maintained at 37 • C in a humidified atmosphere of 5% CO 2 . The cells were used for performing experiment after 4 times of passages.

Over-Expression and Knock-Down of GAS5
To knock down GAS5, the shRNA sequence of GAS5 (shRNA 1# 5 -GCCTAACTCAAGCCATTGG-3 or shRNA 2# 5 -GGTATGGAGAGTCGGCTTG-3 ) was inserted into psi-LVRU6GP vector (GeneCopoeia, Guangzhou, China), which was packaged into lentivirus by GeneCopoeia Co., Ltd. (Guangzhou, China). The cells transfected with lentivirus containing scramble sequence (scramble sequence: 5 -GCGACACGCGCATCTATAT-3 ) were used as a control for shRNA transfection. To overexpress GAS5, the sequence of rat GAS5 (based on the rat GAS5 sequence, NR_002704.1, 429 bp) was directly synthesized and inserted into pReceiver-Lv218 vector (GeneCopoeia, Guangzhou, China), which was packaged into lentivirus by GeneCopoeia Co., Ltd. (Guangzhou, China). The cells transfected with lentivirus carrying empty pReceiver-Lv218 vector were used as a control for overexpression transfection. Briefly, cells were plated in 6-well clusters or 96-well plates for 12 h and then infected with lentivirus at multiplicity of infection (MOI) = 50 for 48 h. Transfected cells were used in further assays or RNA/protein extraction.

MTT Assay
Each group of INS-1 832/13 (5 × 10 3 each wells) was seeded into a 96-well plate fulfilled with 200 µl medium and continually cultured for 0, 24, 48, or 72 h. At the indicated end points, 20 µl of MTT solution (5 mg/ml, cat no. M6494, Invitrogen, CA, United States) was added to each well, and incubated for 4 h. And then 150 µl of dimethyl sulfoxide was added to each well to dissolve the crystals completely. The absorbance at 570 nm of each well was measured by a microplate reader.

Pull-Down Assay With Biotinylated GAS5 Probe
The biotinylated GAS5 probe was synthesized by GenePharma Co. Ltd. (Shanghai, China). The probes were incubated with streptavidin-coated magnetic beads (Sigma) at room temperature for 4 h to generate probe-coated magnetic beads. INS-1 832/13 cells were lysed by RIPA buffer. The lysates were incubated with probe-coated beads at 4 • C overnight. The next day, the beads were washed with wash buffer for three times. The RNA complexes bound to the beads were eluted and extracted by for quantitative PCR analysis.

Statistical Analysis
All data from three independent experiments were expressed as mean ± S.D. and processed using SPSS17.0 statistical software. The difference among the groups was estimated by Student's t-test or one-way ANOVA depending on the conditions. The correlation of GAS5 and miRNAs was analyzed by Spearman correlation analysis. The association of GAS5 and miRNAs with demographic and biochemical parameters of patients with type 2 diabetes mellitus was analyzed by chi-square test. A P value of <0.05 was statistically significant.

The Association of GAS5 Levels With Biochemical Parameters in Patients With T2DM
We also analyzed the associations between GAS5 or miRNAs and various clinical parameters of the 88 patients with T2DM (Tables 2, 3). We found that the low serum levels of GAS5 were associated with high levels of HbAlc, fasting glucose, and LDLc in patients with T2DM ( Table 2). In addition, the percentage of patients with high serum GAS5 was higher in patients treated with oral hypoglycemic agents or insulin than in patients without the treatments (Table 3). By contrast, the high levels of miR-29a-3p, miR-96-3p, and miR-208a-3p were associated with high levels of HbAlc and fasting glucose in patients with T2DM ( Table 2). The percentage of patients with low serum miR-96-3p was higher in patients treated with oral hypoglycemic agents than in patients without the treatment; while the percentage of patients with low serum miR-29a-3p and miR-208a-3p was higher in patients treated with insulin than in patients without insulin treatment ( Table 3).

The Effects of GAS5 on Insulin Secretion in INS-1 832/13 Cells
To investigate whether GAS5 affected insulin secretion in pancreatic β cells, we overexpressed or knocked down (KD) GAS5 in INS-1 832/13 cells by lentivirus transfection. GAS5 expression was significantly increased in the cells infected lentivirus carrying GAS5 DNA sequence compared with the controls (Figure 3A). GAS5 expression was significantly reduced in the cells infected lentivirus carrying GAS5 shRNA sequence compared with the scramble controls ( Figure 3B). In addition, the GAS5 shRNA 2# showed a higher efficiency in decreasing GAS5 level compared with the shRNA 1# ( Figure 3B). Therefore, shRNA 2# was used for the all next experiments if without specifically referring. We performed MTT assay to determine the cell viability after overexpression and knockdown of GAS5 in INS-1 832/13 cells. We found that GAS5 overexpression significantly increased the cell viability compared with the control, while GAS5 knockdown significantly reduced the cell viability compared with the scramble control (Supplementary Figure S2). Importantly, overexpression of GAS5 increased about 20% of insulin secretion and up to 30% of insulin content compared to negative control (Figures 3C,D). In contrast, knockdown of GAS5 led to significant reduction of glucose-stimulated insulin secretion (30%) and insulin content (40%) compared to negative control (Figures 3E,F). These results suggest that GAS5 has a direct regulatory role on insulin secretion in the β cells.
GAS5 Interacts With miR-29a-3p, miR-96-3p, and miR-208a-3p and Regulates Their Expression Bioinformatics analysis was performed to search the potential targeted microRNAs of GAS5 as previous described (Jeggari et al., 2012), which was based on the comprehensive GENCODE gene annotation, including >10000 long non-coding RNA genes and site conservation is evaluated based on 46 vertebrates species, including human and rat. By the bioinformatics analysis, there was conserved binding site for GAS5 on miR-29a-3p, miR-96-3p, and miR-208a-3p in rat and human (Figure 4).

DISCUSSION
In this study, we found that the serum levels of GAS5 were significantly decreased in patients with T2DM compared with healthy control subjects. These results were consistent with previous report that decreased serum GAS5 levels were associated with T2DM (Carter et al., 2015). We further demonstrated that FIGURE 4 | GAS5 negatively regulates the expression of miR-29a-3p, miR-96-3p, and miR-208a-3p. (A) QPCR was performed to determine the expression of miR-29a-3p, miR-96-3p, miR-208a-3p, miR-452-5p, and miR-221-3p in INS-1 832/13 cells after transfection with lentivirus carrying GAS5 DNA sequence or empty pReceiver-Lv218 vector control. (B) QPCR was performed to determine the expression of miR-29a-3p, miR-96-3p, miR-208a-3p, miR-452-5p, and miR-221-3p in INS-1 832/13 cells after transfection with lentivirus carrying GAS5 shRNA sequence or scramble control. (C) Pull-down assay with biotinylated GAS5 probe (Bio-GAS5) was performed to determine the interaction of GAS5 with miR-29a-3p, miR-96-3p, and miR-208a-3p in INS-1 832/13 cells.  were significantly enriched in the complex pulled down by Bio-GAS5 compared with Bio-NC. (D-F) The wild type binding sites of GAS5 and miRNAs (indicated by green), and the mutant type (indicated by red) were shown (upper). Dual luciferase reporter gene assay was performed to validate the interaction between . Data were presented as mean ± SD. *p < 0.05 vs. control. KD, knockdown; ns, no significant. overexpression of GAS5 increased glucose-stimulated insulin secretion and insulin content compared to negative control. In contrast, knockdown of GAS5 led to significant reduction of glucose-stimulated insulin secretion and insulin content.
GAS5 plays a widespread role in different diseases, including cancers, inflammation-related diseases, obesity and autoimmune diseases (Ma et al., 2016;Zhou et al., 2019). GAS5 is downregulated in multiple cancers and acts a tumor suppressor in breast cancer, prostate cancer, lung adenocarcinoma, and pancreatic cancer (Lu et al., 2013;Liu et al., 2018). The levels of GAS5 are significantly decreased in pancreatic cancer tissues compared with normal control. Overexpression of GAS5 in pancreatic cancer cells inhibits cell proliferation by negatively regulating cyclin-dependent kinase 6 (Song et al., 2014). In addition, GAS5 participates in the inflammation progress. For example, GAS5 contributes to the pathogenesis of osteoarthritis by acting as a negative regulator of miR-21 and thereby regulating chondrocytes survival (Mayama et al., 2016). The abnormal levels of GAS5 also alter glucocorticoid response in part through modulation of the glucocorticoid receptor transcriptional activity via its decoy RNA "glucocorticoid response element" (Madhyastha et al., 2012;Lucafo et al., 2015). Qi et al. (2017) analyzed the differentially expressed lncRNAs in lymphatic endothelial cells from non-diabetic patients and T2DM patients and found that GAS5 was dysregulated in lymphatic endothelial cells of T2DM patients (Qi et al., 2017), suggesting that GAS5 is involved in the pathogenesis of diabetes-related complications. In addition, knockdown of GAS5 arrests cell cycle G1 and impair insulin synthesis and secretion in Min6 cells, and decreased the expression of insulin gene and transcription factors, Pancreatic And Duodenal Homeobox 1 (PDX1) and MAF BZIP Transcription Factor A (MAFA) in primary isolated islets (Jin et al., 2017). Recently, Esguerra et al. (2020) also find that GAS5 knockdown or dexamethasone treatment resulted in reduced GAS5 levels, leading to concomitant reductions in glucocorticoid receptor (GR), PDX1, NK6 Homeobox 1 (NKX6-1), and synaptotagmin 13 (SYT13) proteins. In line with these results, we found that knockdown of GAS5 led to significant reduction of glucose-stimulated insulin secretion and insulin content in rat pancreatic β cells. These results indicate that GAS5 might perform as a regulator, maintaining β cell identity and function by affecting insulin synthesis and secretion.
The functions of lncRNAs are achieved by interacting with endogenous miRNAs. We demonstrated that decreased serum GAS5 levels were negatively correlated with high serum levels of miR-29a-3p, miR-96-3p, and miR-208a-3p in patients with T2DM. MiR-96-3p is dysregulated during healing of diabetic wounds (Locke et al., 2014) and is important for β-cell function (Chakraborty et al., 2013;Yang et al., 2016). MiR-96-3p is induced strongly by saturated fatty acids in the development of hepatic insulin resistance. Overexpression of miR-96-3p can cause an impairment of insulin signaling and glycogen synthesis in hepatocytes by repressed the expression of INSR and IRS-1 at the post-transcriptional level (Yao et al., 2019).
In addition, miR-29 family has important roles in insulin secretion and insulin resistance. Overexpression of miR-29a/b/c in 3T3-L1 adipocytes could largely repress insulin-stimulated glucose uptake through inhibiting Akt activation and cause insulin resistance (He et al., 2007). MiR-29 expression in human hepatoma cells is controlled by forkhead box A2 (FOXA2), a key gene in hepatic energy homeostasis. In turn, miR-29 fine-tunes FOXA2-mediated activation of key lipid metabolism genes (Kurtz et al., 2014). Global or hepatic insufficiency of miR-29 potently prevents the onset of diet-induced insulin resistance by negatively regulating insulin signaling via PIK3R1 regulation (Dooley et al., 2016), suggesting that miR-29 is an important regulatory factor in metabolism homeostasis (Massart et al., 2017).
MiR-208a-3p was a heart-specific microRNA and was required for cardiomyocyte hypertrophy and fibrosis in response to stress and hypothyroidism (van Rooij et al., 2007). In this study, by the bioinformatic tools (see text footnote 1), we also found that miR-96-3p conversely targeted 3 UTR of INSR, PIK3R1, and IRS1, miR-29a-3p conversely targeted PIK3R1 and IRS1, and miR-208a-3p conversely targeted PIK3R1 among mammals, including human and rat (Data not show). In addition, we found that GAS5 negatively regulated the expression of miR-29a-3p, miR-96-3p, and miR-208a-3p in rat pancreatic β cells. Moreover, GAS5 positively regulated the expression of INSR, PIK3R1 and IRS1 rat pancreatic β cells. Therefore, we supposed that GAS regulated the expression of INSR, PIK3R1 and IRS1 through competitively interacting miR-29a-3p, miR-96-3p, and miR-208a-3p, but we cannot exclude other mechanisms play a role in this regulation, such as epigenetic mechanism (Sun et al., 2017). MiR-452-5p and miR-221-3p are two valid targets of GAS5 in renal tubular cells and mesangial cells, respectively, and both of them are involved in diabetic nephropathy (Ge et al., 2019;Xie et al., 2019). However, the levels of miR-452-5p and miR-221-3p were comparable between high GAS5 group and low GAS5 group. In addition, knockdown or overexpression of GAS5 did not alter the expression of miR-452-5p and miR-221-3p. We supposed that the targets of GAS5 were cell-or tissue-specific.

CONCLUSION
We demonstrated that decreased serum GAS5 levels were negatively associated with HbAlc and fasting glucose in patients with T2DM. Overexpression of GAS5 increased glucosestimulated insulin secretion and insulin content by inhibiting miR-29a-3p, miR-96-3p, and miR-208a-3p. Thus, our findings reveal that GAS5 is a biomarker and novel therapeutic target for T2DM treatment.

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

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Ethics Committee of Xiangya Hospital, Central South University. The patients/participants provided their written informed consent to participate in this study.