miR‐1322 regulates ChREBP expression via binding a 3′‐UTR variant (rs1051943)

Abstract The carbohydrate response element‐binding protein (ChREBP), also referred to as MLXIPL, plays a crucial role in the regulation of glucose and lipid metabolism. Existing studies have shown an association between genetic variations of the ChREBP gene and lipid levels, such as triglycerides and high‐density lipoprotein cholesterol. However, mechanistic studies of this association are limited. In this study, bioinformatic analysis revealed that the polymorphism rs1051943A occurs in the complementary binding sequence of miR‐1322 in the ChREBP 3′‐untranslated region (UTR). Studies of potential mechanisms showed that the A allele could facilitate miR‐1322 binding, and luciferase activity significantly decreased when co‐transfected with a ChREBP 3′‐UTR luciferase reporter vector and miR‐1322 mimics in HepG2 cells. Furthermore, miR‐1322 significantly regulated the expression of ChREBP downstream genes and reduced the synthesis of lipids. The expression of miR‐1322 was up‐regulated by glucose and palmitic acid stimulation. Population studies showed that rs1051943‐A allele was only found in the Han Chinese and Uighur ethnic groups, different from European populations (G allele frequency = 0.07). In summary, we provide evidence that the rs1051943 A allele creates a functional miR‐1322 binding site in ChREBP 3′‐UTR and post‐transcriptionally down‐regulates its expression, possibly associated with levels of plasma lipids and glucose.

(SNPs) in or near the ChREBP gene have been found to be associated with plasma triglyceride levels by genome-wide association studies (GWASs). [8][9][10] During the last decade, more than 20 studies confirmed the association between ChREBP gene variations and plasma triglyceride levels, as well as the association with coronary artery disease (CAD). [11][12][13][14][15][16] MicroRNAs (miRNAs) are small non-coding RNAs (19-22 nucleo-tides) known to be regulators of gene expression at post-transcriptional levels. 17 The previous studies have shown that miRNAs could alter gene expression and individual susceptibility to disease traits, and variants located in these miRNA binding sites may alter clinical characteristics and disease susceptibility. [18][19][20] To our knowledge, whether polymorphism in the 3′-UTR of ChREBP could modify gene expression and further influence triglycerides and glucose levels has not been clearly demonstrated. Therefore, we performed a bioinformatics analysis to identify the potential functional variant in the 3′-UTR of ChREBP. As a result, the known polymorphism, rs1051943, was found in the seed binding site of miR-1322, one of the 69 miR-NAs previously identified to be associated with abnormal lipid levels. 21 Based on these results, we hypothesized that the human rs1051943 polymorphism could regulate ChREBP expression by influencing miR-1322 binding and thus affect plasma levels of lipids and glucose.

| Recruitment for hypertriglyceride (HTG) patients and control participants
The sequencing cohort included 169 HTG individuals and 313 con-

| Genetic variation screening
In order to identify existing variants in the ChREBP gene, Sanger sequencing was performed. Details on sample sequencing were ZHANG ET AL. | 5323 described in our previous report. 20 Briefly, PCR fragments covering the 3′-UTR of ChREBP (ChREBP consensus sequence, NC_000007.13, GRCh37.p13) were generated using primers (forward primer: 5′-AGCTGGGCACATCTACCAGTAT-3′ and reverse primer: 5′-CACTGCCAACAGGCTCTCTCT-3′). Applied Biosystems 3130xl capillary sequencer (Applied Biosystems, Foster City, CA) was used to analyse fluorescent dye-terminator cycle products from the PCR fragments. Putative polymorphisms were identified through the Chromas program (Technelysium Pty. Ltd., Helensvale, Queensland, Australia). Results were then confirmed by two independent observers. All identified variants were confirmed by repeat sequencing. The supernatant was collected after centrifuging at 12,000 g at 4°C for 20 minutes. Protein concentration was determined by BCA protein assay kit (Boster Biological Technology, Wuhan, China). Cell lysates were separated by 8% SDS-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride (PVDF) membranes.

| Western blots
After 2 hours blocking with 5% non-fat milk at room temperature, the ChREBP antibody (Abcam, Boston, USA) was utilized to perform immunoblots at 4°C, followed by incubation with a peroxidase-con-  After transfection for 24 hours, cell lysates were extracted as described previously. 20 Triglyceride and total cholesterol levels were determined using a commercial assay kit (Nanjing Jiancheng Bioengineering Institute, China) normalized to protein concentration following the manufacturer's instructions. The assay sensitivity was 0.01 mmol/mL, and average intra-and inter-assay coefficients of variation were 3% and 5%, respectively. The absorbance was measured using a BioStack Microplate Stacker (BioTek, WI, USA) according to the instruction manual.

| Quantitative real-time PCR
Total RNA was extracted from human normal tissues including liver, adipose tissue, large intestine, small intestine, heart, lung and cell  Table S3.

| Bioinformatics analysis of ChREBP 3′-UTR
Bioinformatics analysis demonstrated that 30 miRNAs might contain binding sites in 3′-UTR of ChREBP (Table S4) in which 13 miRNAs are confirmed by TargetScan 7.1 as shown in Table 2. Among these miR-NAs, only miR-1322 was found to be associated with abnormal lipid levels. 21 The variant rs1051943 is an A to G change (mRNA sequence as reference), and computer alignment demonstrated that this SNP was located in the miR-1322 binding site in ChREBP 3′-UTR ( Figure 1A).

| Effects of polymorphisms on activity of ChREBP 3′-UTR in vitro
In HepG2 cells, the luciferase reporter activity of the pMIR-G vector (containing G allele of rs1051943) is significantly increased (58.1 ± 7.3%, P < 0.0001) when compared with the pMIR-A construct (containing A allele of rs1051943) as shown in Figure 1B. No significant difference was found between pMIR-G and pMIR-A in HEK293T cells ( Figure 1C). These results indicate that endogenous hepatic regulator factor could target ChREBP 3′-UTR and decrease ChREBP 3′-UTR luciferase expression.

| Rs1051943 of ChREBP occurring in miR-1322 binding site
Given that rs1051943 was located in the conserved region of

| The effects of miR-1322 on insulin resistance
To further explore the association between miR-1322 and the metabolism of lipids and glucose, cell models for insulin resistance were induced with palmitate or high glucose cell culture media. miR-1322 levels are significantly increased after high glucose ( Figure 4A) or palmitate stimulation ( Figure 4B). Although the cell viability was obviously decreased after PA (1.0 mmol/L) treatment ( Figure S4), the expression of miR-1322 was positively correlated with the concentration of PA stimulation. In addition, we found that miR-1322 was more highly expressed in HepG2 cells ( Figure 4D) compared with other cell lines, and an abundant amount of miR-1322 was also detected in the intestine ( Figure 4C), indicating that miR-1322 could affect insulin resistance.

| Re-sequencing results of 3′-UTR in ChREBP
A total of 482 Han Chinese participants were enrolled, including 169 HTG patients and 313 control participants (Table 1). Cumulatively, we identified two DNA variants in this population. Of the two variants, one was a common polymorphism (minor allele frequency >1%); the other was a rare polymorphism. No significant differences of genotype frequencies were discovered between the two groups (Table 3). We did not find significant associations between rs1051921 polymorphism and TG, TC, LDL-C or HDL-C (Table S5).
According to the luciferase assay, significant differences of rs1051921-C allele and T allele luciferase activity were discovered ( Figure S5). However, no transcription factor or miRNA was predicted in this SNP. Only rs1051943-A allele was found in both control and HTG patients. To eliminate racial differences, 96 Chinese Uighur ethnic participants were also included in the sequencing cohort, of which similar results were found (Table 3).

| Conservation of 3′-UTR in ChREBP in different species
Sequence alignment showed that 3′-UTR in ChREBP flanking the miR-1322 binding site was conserved in mice and rat ( Figure 5A).
However, the homology allele of the rs1051943 was a T allele in both species, which could lead to the destruction of the miR-1322 binding site. According to the existing database, the hsa-miR-1322 homologue was not found in both mice and rat. Therefore, we performed bioinformatics analysis with UCSC (http://genome.ucsc.edu/).

| DISCUSSION
In this study, we identified a functional variant, the A allele of  Figure 6).
It is clear that ChREBP plays important roles in modulating the process of glycolysis and lipogenesis. 23 Iizuka et al reported that the ChREBP knockout mice showed reduced glycolysis and lipogenesis and exhibited insulin resistance, 24 whereas Benhamed et al showed that the ob/ob mice with ChREBP deletion exhibited reduced lipogenesis and improved insulin sensitivity. These controversial results in different mice models were confirmed by multiple laboratories. 25,26 Recently, an American obese adolescents study revealed that the expression of ChREBP was significantly increased in the liver with high insulin resistance, which was in contrast to that in adipose tissue. 27 Regarding the aforementioned evidence, ChREBP mechanisms in glucose and lipid metabolism warrant further exploration.
Given the important roles of ChREBP in glucose and lipid meta-   No statistical significant difference of genotype frequencies between control and HTG patient groups was discovered using chi-squared test.

R E F E R E N C E S
1. Alberti KGMM, Eckel RH, Grundy SM, et al. Harmonizing the metabolic syndrome: a joint interim statement of the international diabetes federation task force on epidemiology and prevention; national heart, lung, and blood institute; american heart association; world heart federation; international atherosclerosis society; and