MAGI2-AS3 modulates proliferation, apoptosis and glycolysis in acute lymphoblastic leukemia via miR-452-5p/FOXN3 axis

Background Long non-coding RNA MAGI2 antisense RNA 3 (MAGI2-AS3) has been identified as a tumor suppressor in various cancers. Acute lymphoblastic leukemia (ALL) is a prevalent kind of leukemia among children. In this study, we aimed at evaluate the role of MAGI2-AS3 in ALL and its underlying mechanisms. Methods qPCR was adopted to determine MAGI2-AS3, miR-452-5p, and FOXN3 expression. The malignant properties of ALL cells were assessed by CCK8 assay and flow cytometry analysis. The glucose uptake, lactate production, and ATP level were measured to evaluate glycolysis. Western blotting was performed to detect PCNA, Bcl-2, Bax, and HK2 protein levels. The interaction between MAGI2-AS3/FOXN3 and miR-452-5p was validated by luciferase reporter assay. The in vivo growth of ALL cells was determined in xenograft model. Results MAGI2-AS3 was strikingly down-regulated in ALL samples and cells. Overexpression of MAGI2-AS3 restrained growth, glycolysis and triggered apoptosis of ALL cells. Mechanistically, MAGI2-AS3 could sponge miR-452-5p to up-regulate FOXN3. Silencing of FOXN3 abolished the anti-tumor effect of MAGI2-AS3. Finally, MAGI2-AS3 suppressed the in vivo growth of ALL cells via modulating miR-452-5p/FOXN3 axis. Conclusion Our findings demonstrate that MAGI2-AS3 delays the progression of ALL by regulating miR-452-5p/FOXN3 signaling pathway.


Introduction
Acute lymphoblastic leukemia (ALL) is one of prevalent hematological malignant tumors among children, which accounts for about 75% of children's acute leukemia [1].
According to immunophenotyping, ALL patients are generally divided into B-ALL or T-ALL. Although conventional chemotherapy has greatly improved the survival rates of ALL patients, the recurrent cases are hard to get remission [2]. In addition, adverse effects induced by conventional chemotherapy may increase the risk of developing a secondary tumor [3]. Hence, it is urgent to uncover the pathogenesis of ALL for providing novel therapeutic options.
Long non-coding RNAs (lncRNAs) are a kind of non-protein coding transcripts more than 200 nucleotides in length [4]. Growing evidence has revealed that lncRNAs are implicated in tumorigenesis and development of various malignancies [5,6]. Previous publications have demonstrated that lncRNAs, served as tumor suppressor or promotor, play crucial roles in tumor occurrence, metastasis or recurrence [7,8].
Besides, lncRNAs have also been confirmed to affect the tumorigenesis and progression of ALL [9]. MAGI2 antisense RNA 3 (MAGI2-AS3) has been documented to be down-regulated and play oncogenic roles in different types of malignancies, such as breast cancer, lung cancer, hepatocellular carcinoma, bladder cancer and so on [10]. Notably, a significant downregulation of MAGI2-AS3 has been found in ALL patient samples [11]. A recent study by Chen et al reported that MAGI2-AS3 exhibited a poor expression in acute myeloid leukaemia, which was responsible for self-renewal of leukaemia stem cells [12]. Yet, the biological functions of MAGI2-AS3 in ALL remains obscure.
MiR-452-5p has been verified as tumor promoter in multiple tumors. For instance, increased expression of miR-452-5p facilitated colorectal cancer development via modulating PKN2/ERK/MAPK feedback loop [15]. In lung squamous cell carcinoma, miR-452-5p up-regulation remarkably correlated with tumor-node metastasis and exerted vital role in tumor progression via targeting CDKN1B [16]. However, whether miR-452-5p is involved in ALL development has not been revealed.
Forkhead box N3 (FOXN3), a member of forkhead box family, is involved various pathophysiological processes, including tumorigenesis [17]. As reported by previous researches, a decreased expression of FOXN3 is confirmed in a wide variety of tumors, including ALL [18,19]. More importantly, bioinformatics analysis indicates that both MAGI2-AS3 and FOXN3 possess binding sites in miR-452-5p. Therefore, MAGI2-AS3 might modulate the progression of ALL through miR-452-5p/FOXN3 axis.
To validate the above hypothesis, this study probed into the regulatory function and molecular mechanisms of MAGI2-AS3 in ALL, which provided new insights into the pathogenesis of ALL.
(Applied Biosystems, USA). Evaluation of relative RNA levels was carried out with 2 −ΔΔCt method. Table 1 presents the primers of target genes.

Cell counting kit-8 (CCK-8)
ALL cells seeded in 96-well plates were reacted with 10 μ l of CCK-8 reagent (Dojindo, Japan) at 37°C at various time points. Four hours later, the absorbance intensity was detected at a wavelength of 450 nm on a microplate reader (Tecan, Switzerland).

Flow cytometry
The apoptosis of ALL cells was assessed using a FITC Annexin V/Dead Cell Apoptosis Kit (Thermo Fisher). In short, the collected ALL cells were resuspended in binding buffer, followed by staining with FITC Annexin V (5 μ L) and PI (1 μ L) solution for 30 min away from light. Then the stained cells were detected on a flow cytometer (Millipore, USA).

Detection of glucose uptake, lactate production and ATP level
To evaluate glycolysis of ALL cells, glucose uptake, lactate production and ATP level were determined using Glucose Uptake Assay Kit (Abcam, UK), Lactate Assay Kit II (Sigma-Aldrich, USA), and ATP Detection Kit (Solarbio, Beijing, China) following the protocols provided by the manufacturers.

Animal experiment
BALB/c nude mice (18-22 g) were obtained from SLAC Laboratory Animal Co., Ltd (Shanghai, China). 1 × 10 7 Jurkat cells stably transfected with OE-MAGI2-AS3 or vector were intraperitoneally implanted the nude mice (n=6 per group). Tumor length and width were measured every three days. Tumor volume was calculated as follow: length × width 2 /2 (mm 3 ). Tumors were collected from mice 30 days later, weighed and subjected to subsequent experiments. All experimental protocols were approved by the Animal Ethics Committee of The First Affiliated Hospital of Zhengzhou University.

TUNEL
To examine apoptosis in tumor tissues, the Colorimetric TUNEL Apoptosis Assay Kit (Beyotime) was adopted. Briefly, the tumor tissues were taken to paraffin imbedding and cut into 5-μm sections. After deparaffination and hydration, antigen retrieval was carried out using microwave. The sections were treated with Proteinase K (20 mg/mL) and then blocked in 3% H 2 O 2 for 20 min. Thereafter, the reaction solution was added to the sections, followed by incubation at 37° C for 1 h. After reaction with streptavidin-HRP and DAB solution, the sections were observed under an optical microscope (Micro-shot Technology Co., Ltd, Guangzhou, China).

Statistical analysis
Data are presented as mean± standard deviation (SD). GraphPad Prism 6.0 was adopted for statistical analysis using Student's t test for two group or One-Way ANOVA for multiple group comparison. A p-value less than 0.05 was considered as statistically significant.

Up-regulation of MAGI2-AS3 in ALL tissues and cells
First, we determined the differential expression of MAGI2-AS3 in bone marrow specimens from ALL patients and healthy volunteers. qPCR results revealed that MAGI2-AS3 was down-regulated in ALL bone marrow samples relative to normal controls (Fig. 1A). Consistently, a decreased expression of MAGI2-AS3 was validated in a series of ALL cells in comparison with PBMCs (Fig. 1B). Thus, MAGI2-AS3 might exert crucial regulatory roles in ALL progression.

Effect of MAGI2-AS3 on proliferation, apoptosis and glycolysis of ALL cells
To assess the regulation of MAGI2-AS3 in ALL, ALL cells were transfected with OE-MAGI2-AS3 plasmid to overexpress MAGI2-AS3. As verified by qPCR, OE-MAGI2-AS3 plasmid remarkably raised the expression of MAGI2-AS3 in Jurkat and Reh cells ( Fig. 2A). Cell proliferation was determined by CCK8 assay, and we found that the proliferative ability was significantly reduced in MAGI2-AS3-overexpressed group (Fig. 2B). Flow cytometry was used to test the apoptosis of ALL cells. As illustrated in Fig. 2C, ectopic expression of MAGI2-AS3 evidently triggered apoptosis of Jurkat and Reh cells. Warburg effect in cancer cells has been well recognized, which is featured by an increase in glucose uptake and lactate production [20]. We demonstrated that overexpression of MAGI2-AS3 led to a striking decrease in glucose uptake, lactate production and ATP level in ALL cells ( Fig. 2D-F). In addition, the protein levels of PCNA, Bcl-2, and HK2 were declined, but Bax level was elevated after overexpression of MAGI2-AS3 ( Fig. 2G-I). These data implied that MAGI2-AS3 overexpression restrained proliferation, glycolysis and induced apoptosis of ALL cells.

Silencing of FOXN3 reversed the anti-cancer effect of MAGI2-AS3
To determine the involvement of FOXN3 in MAGI2-AS3-mediated anti-cancer effects, the ALL cells were transfected with OE-MAGI2-AS3 together with or without shFOXN3. As detected by CCK8 assay, MAGI2-AS3 overexpression-induced proliferation inhibition of ALL cells could by restored by FOXN3 knockdown (Fig.   4A&B). In addition, the enhanced apoptosis of MAGI2-AS3-overexpressed ALL cells could be reversed by silencing of FOXN3 ( Fig. 4C-E). Depletion of FOXN3 also recovered the decreased glucose uptake, lactate production, and ATP level induced by MAGI2-AS3 overexpression ( Fig. 4F-K). Besides, transfection with OE-MAGI2-AS3 led to reduction in PCNA, Bcl-2, HK2 and elevation in Bax expression, which was partly counteracted by co-transfection with shFOXN3 ( Fig. 4L-N). These findings suggested that MAGI2-AS3 delayed the progression of ALL via modulating FOXN3.

Effect of MAGI2-AS3 on ALL tumor growth in vivo
To confirm the function of MAGI2-AS3 in tumorigenesis in vivo, the nude mice were inoculated with Jurkat cells stably transfected with OE-MAGI2-AS3 or vector. As presented in Fig. 5A-C, the tumor volume and weight were distinctly reduced by overexpression of MAGI2-AS3. Moreover, TUNEL assay revealed that apoptosis in tumor tissues was enhanced in response to MAGI2-AS3 overexpressing (Fig. 5D).

Discussion
Over the past several decades, great improvement has been made in understanding the pathogenesis of ALL. However, the detailed mechanisms underlying the malignant progression and recurrence of ALL cells remain not entirely clear. This study focused on a novel modulatory pathway MAGI2-AS3/miR-452-5p/FOXN3 during ALL development. We demonstrated that MAGI2-AS3 and FOXN3 were up-regulated, and miR-345-5p was down-regulated in ALL specimens and cells. MAGI2-AS3 sequestered miR-345-5p to facilitate FOXN3 expression, which contributed to ALL cell growth, apoptosis inhibition and glycolysis. Recently, the carcinogenic role of MAGI2-AS3 in acute myeloid leukemia has been identified [12]. According to our data, MAGI2-AS3 was strikingly lowly expressed in ALL clinical samples and cells. Furthermore, enforced expression of MAGI2-AS3 suppressed proliferation and apoptosis inhibition of ALL cells. Thus, our study was the first to reveal that MAGI2-AS3 functioned as a tumor suppressor lncRNA during ALL development.
Glucose metabolic reprogramming has been considered as a striking hallmark of tumors [25]. Being different from healthy cells, tumor cells are prone to metabolize glucose by glycolysis, leading to enhanced glucose uptake and lactate production [26].
Glycolysis confers the malignant phenotypes as well as drug resistance of ALL cells [27]. HK2 has been recognized as a specific enzyme catalyzing glycolysis process [28]. In this study, we demonstrated that MAGI2-AS3 overexpression restrained glucose uptake, lactate production, ATP level, as well as reduced HK2 expression in ALL cells. Therefore, repression of glycolysis was involved in the anti-cancer effect of MAGI2-AS3 in ALL.
To further probe the regulatory mechanisms of MAGI2-AS3 in ALL, we focused on the ceRNA hypothesis that cytoplasmic lncRNA adsorbs miRNA to release the expression of miRNA target gene. miR-452-5p is a crucial member of miRNA, and its involvement in cancers has been reported. The interaction between miR-452-5p and . Our data showed that miR-452-5p was aberrantly highly expressed in ALL, which was negatively correlated with MAGI2-AS3 level. Moreover, the direct binding relationship between MAGI2-AS3 and miR-452-5p was validated. These data indicated that MAGI2-AS3 down-regulation resulted in abnormal higher miR-452-5p expression, which contributed to ALL progression. To sum up, we recognized MAGI2-AS3 as a tumor suppressor in ALL. MAGI2-AS3 overexpression repressed growth, glycolysis and induced apoptosis of ALL cells via sequestering miR-452-5p to elevate FOXN3 level. This work helps to further uncover the pathological mechanisms of ALL and identifies MAGI2-AS3 as a potential diagnostic and therapeutic target.

Funding
None.

Declaration of Interests
The authors have no interests to disclose.    The interaction between FOXN3 and miR-452-5p was determined by dual-luciferase reporter assay. FOXN3 expression in ALL cells with multiple transfections was assessed by qPCR (L-M) and Western blotting (N-P). Data are expressed as mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001. Flow cytometry was adopted to assess apoptosis of ALL cells. (F-K) The glucose uptake, lactate production, and ATP levels were measured by commercial kits. (L-N)