MYCT1 represses apoptosis of laryngeal cancerous cells through the MAX/miR‐181a/NPM1 pathway

MYCT1 is an important gene known to regulate cell viability and apoptosis of laryngeal cancer cells. However, the underlying molecular mechanism remains unclear. Here, we show that MAX enhances the expression of miR‐181a by directly binding to its promoter, whereas miR‐181a targets NPM1 and suppresses its expression in laryngeal cancer cells. MYCT1 and miR‐181a decrease cell viability and colony formation through enhanced apoptosis, whereas NPM1 displays opposite effects in laryngeal cancer cells. Their opposing functions are further supported by the findings (a) that miR‐181a is down‐regulated, while NPM1 is up‐regulated in laryngeal cancer, and (b) that either inhibition of miR‐181a or overexpression of NPM1 can revert the pro‐apoptotic effects of MYCT1 on laryngeal cancer cells through extracellular and intracellular apoptotic pathways. Our data suggest that MYCT1 may synergistically interact with MAX as a co‐transcription factor or a component of MAX transcriptional complex, to transcriptionally regulate the expression of miR‐181a, which, in turn, decreases NPM1 expression at post‐transcriptional levels, leading to enhanced apoptosis in laryngeal cancer cells. These factors may serve as potential targets for early diagnosis and treatment of laryngeal cancer.


Introduction
Laryngeal carcinoma is a malignant tumor originated from the epithelial cells of laryngeal mucosa; it is also the second most common malignant head and neck tumor in the world [1,2]. Both environmental and genetic factors contribute to laryngeal carcinogenesis and the incidence of laryngeal cancer has been increasing steadily [3,4]. Despite various advanced treatment interventions, the 5-year survival rate of laryngeal cancer patients has not been significantly improved over the past several decades, with a 5-year survival rate of 50-60%, and the molecular mechanism underlying laryngeal cancer occurrence and development remains unclear [5,6]. Inhibition of apoptosis is a common event in carcinogenesis [7]. Laryngeal carcinoma is an aggressive and lethal malignant tumor resistant to various apoptosis stimulating factors and its poor prognosis is also related to inhibition of apoptosis [8,9]. Studies have revealed that expression of BCL2 (B-cell lymphoma 2) and Survivin in laryngeal squamous cell carcinoma is associated with pathological grade and recurrence rate of laryngeal cancer, indicating that these are potential biomarkers for predicting the survival of laryngeal cancer patients [10][11][12]. Livin expression is significantly higher in laryngeal squamous cell carcinoma than in vocal cord polyp, and its overexpression has been correlated with laryngeal pathological stages [13]. Mutant P53 shows differential expression in different stages of laryngeal carcinoma and precancerous lesions, and is also associated with laryngeal cancer cell viability and prognosis [14,15]. In addition, the expression of Caspases-3, -8 and -9 has also been related to the clinical prognosis of laryngeal cancer patients and thus, can be potential biomarkers for laryngeal cancer treatment and prognosis [16][17][18]. Available data all imply that apoptosis plays an important role in the occurrence, development, and prognosis of laryngeal carcinoma.
We previously identified and cloned human MYCT1, which we named as MTLC standing for myc-target from laryngeal carcinoma [19]. Our data have demonstrated that MYCT1 inhibits laryngeal cancer cell viability and promotes apoptosis as a tumor suppressor [20,21]. However, the molecular mechanism through which MYCT1 modulates laryngeal cancer cell viability and apoptosis still remains unknown. NPM1 gene encodes a shuttle protein that travels between the nucleus and cytoplasm [22]. In addition to constituting the nucleolus, NPM1 is also involved in the regulation of cancer cell viability and apoptosis via both extracellular death receptor pathway and intracellular mitochondrial pathway [23]. Studies have shown that NPM1 inhibits apoptosis by blocking the translocation of P53 from nucleus to mitochondria [24,25]. In acute promyelocytic leukemia, NPM1 represses apoptosis by preventing TNF (tumor necrosis factor)-induced extracellular apoptosis pathways [26]. NPM1 was also reported to maintain the resistance of NIH-3T3 cells to UV-induced apoptosis by binding IRF1 (interferon regulatory factor 1) [27]. Although NPM1 has been studied in the other cancers, whether and how NPM1 is involved in the pathogenesis of laryngeal cancer is yet to be investigated.
In our subsequent study using next-generation RNA deep sequencing (RNA-seq) analyses, we identified NPM1 as the most dysregulated apoptosis-related genes in Hep2 cells with stable overexpression of MYCT. This prompted us to explore whether MYCT1 directly controls NPM1 expression, which, in turn, affects laryngeal cancer apoptosis. Here, we show that MYCT1 may interact synergistically with MAX as a co-transcription factor or a component of MAX transcript complex, which, in turn, directly enhances the transcriptional activity of miR-181a. Moreover, NPM1 is a direct target of miR-181a. Thus, MYCT1 appears to participate in the regulation of laryngeal cancer cell viability and apoptosis through the MAX/miR-181a/ NPM1 pathway.

MYCT1 is a negative regulator of NPM1 in laryngeal cancer cells
To identify the downstream effectors of MYCT1, we performed RNA-seq to compare the gene expression profiles between laryngeal cancer cells stably transfected with MYCT1 expression plasmids and control cells with the empty vector. Raw data were archived in the Gene Expression Omnibus under accession number GSE75544. Our transcriptomic analyses identified 326 down-regulated genes and 1 up-regulated gene (≥2-fold) (Fig. 1A). Gene ontology (GO) analyses identified that these genes were involved in a wide variety of biological processes and cellular functions (Fig. 1B). Among these dysregulated genes, NPM1 was down-regulated by~3-fold (Data S1). To validate the RNA-seq results, we conducted both quantitative real-time PCR (qRT-PCR) and western blot assays using stable MYCT1-expressing Hep2 cells. MYCT1 mRNA was indeed significantly overexpressed in Hep2 cells transfected with MYCT1 expression plasmids, as compared to the control cells transfected with empty vectors (Fig. 1C). In contrast, both mRNA and protein levels of NPM1 significantly decreased in Hep2 cells with stable MYCT1 overexpression (Fig. 1D,E). These results suggest that MYCT1 down-regulates NPM1 in laryngeal cancer cells.

NPM1 plays an oncogenic role in laryngeal cancer cells
We then examined NPM1 levels in 45 cases of laryngeal cancer tissues (T) and paired nontumorous tissues (R). NPM1 mRNA levels were up-regulated in 37 out of 45 cases (82%) of laryngeal cancer tissues ( Fig. 2A) and the average levels were significantly greater in laryngeal cancer tissues than in paired nontumorous tissues (P < 0.001, Fig. 2B). Moreover, NPM1 mRNA levels were significantly higher in Hep2 cells than in HEK293T cells (P < 0.05, Fig. 2C). NPM1 protein was overexpressed in 14 out of 16 laryngeal cancer cases (87.5%) and the average levels were significantly higher in laryngeal cancer tissues than in controls (Fig. 2D). We then assayed the effects of NPM1 on laryngeal cancer cell viability and apoptosis. Three NPM1 sh-RNAs were used to knock down the NPM1 gene. Both NPM1 mRNA and protein levels significantly decreased in Hep2 cells transfected with the three sh-RNAs compared to those in Hep2 cells transfected with a scrambled shRNA (Fig. 2E,F). Since sh-NPM1-2# displayed the highest knockdown efficiency, it was used in the subsequent experiments. Hep2 cells were transfected with the expression plasmids and sh-RNA 2# of NPM1, and the effect on cell viability was measured 48 h after transfection. Cell Counting Kit-8 (CCK8) and colony-formation assays revealed that NPM1 knockdown significantly inhibited laryngeal cancer cell survival and colony formation, as compared to empty vector-transfected and untransfected cells. In contrast, both survival and colony formation were enhanced significantly after the laryngeal cancer cells were transfected with NPM1 expression plasmids (Fig. 2G,H). We also examined the effect of NPM1 overexpression on Hep2 cell apoptosis. For the apoptosis assay, Hep2 cells were harvested 48 h after transfection, stained with Annexin V-PE and 7-AAD and subsequently analyzed with flow cytometry. As a result, early apoptosis significantly increased in the NPM1 knockdown Hep2 cells compared to the empty vector-transfected cells and untransfected cells. In contrast, there was no significant effect of NPM1 overexpression on early apoptotic event in laryngeal cancer cells (Fig. 2I). Together, these data suggest that NPM1 has an oncogenic role in laryngeal cancer cells.

MAX may interact synergistically with MYCT1 in Hep2 cells
Based on online-software STRING prediction, we found four potential MYCT1-interacting proteins including MAX (Fig. 3A). Co-immunoprecipitation (Co-IP) assays revealed that both anti-GFP (for MYCT1) and anti-MAX antibodies could precipitate MAX and MYCT1 in MYCT1-overexpressing Hep2 cells and wild-type Hep2 cells, respectively (Fig. 3B). Immunofluorescence analyses revealed enhanced co-localization of MYCT1 (green) and MAX (red) proteins in the nuclei of Hep2 cells overexpressing MYCT1, as compared to the controls (Fig. 3C), suggesting that nuclear translocation of MAX was promoted by overexpression of MYCT1, leading to the enhanced co-localization of MYCT1 and MAX in Hep2 cell nuclei. These results suggest that there was a protein-protein interaction between MYCT1 and MAX which may be direct by combining with each other or indirect by depending on a third protein.
MAX positively regulates miR-181a transcription in laryngeal cancer cells qRT-PCR analyses indicated that endogenous miR-181a levels were down-regulated in the majority (85%, 40 of 47) of the laryngeal cancer cases (Fig. 4A), and its average levels were significantly lower in laryngeal cancer tissues than that in paired noncancerous tissues (P < 0.001, Fig. 4B). qRT-PCR analyses also showed that endogenous miR-181a levels decreased significantly in Hep2 cells compared to HEK293T cells (P < 0.001, Fig. 4C). These findings suggest that miR-181a might act as a cancer suppressor gene in laryngeal cancer. Compared to the control cells, miR-181a was significantly up-regulated in Hep2 cells stably expressing MYCT1 (P < 0.01, Fig. 4D), indicating that miR-181a is also a potential downstream target of MYCT1.
Based on bioinformatics prediction, we found two potential-binding sites of MAX in the miR-181a gene promoter region (Fig. 4E). Three specific small interfering RNAs against MAX mRNA, si-MAX 1#, si-MAX 2#, and si-MAX 3#, significantly repressed MAX expression at both mRNA and protein levels in Hep2 cells as compared to negative control-treated and untransfected Hep2 cells (Fig. 4F,G), suggesting that MAX was successfully knocked down. Meanwhile, we found that miR-181a was significantly down-regulated in Hep2 cells when MAX was suppressed compared to negative control-treated and untransfected Hep2 cells (P < 0.01, Fig. 4H).
To explore whether miR-181a was regulated by MAX in Hep2 cells, we examined the miR-181a core promoter region for two putative MAX-binding sites at the regions -292 to -283 bp (R1) and -611 to -602 bp (R2) in the miR-181a promoter (Fig. 4E). PCR products spanning two putative MAX-binding sites of the miR-181a promoter region were detected from chromatin fragment precipitated by anti-MAX antibody in our chromatin immunoprecipitation assays (ChIP) (Fig. 4I). The result showed that endogenous MAX indeed bound to miR-181a promoters. Furthermore, MAX knockdown significantly reduced levels of MAX on the miR-181a promoter compared to the negative controls (P < 0.01, Fig. 4J). These results imply that MAX binds directly to the miR-181a promoter region in vivo and promotes its expression as the transcriptional factor. To corroborate this notion, we then explored the binding of MAX to miR-181a in Hep2 cells by luciferase reporter assays. The fragments of miR-181a promoter regions containing one (-292 bp) or two (-611 bp) MAX-binding sites were cloned into GV148 vector and transiently transfected into Hep2 cells along with pRL-TK. Luciferase results revealed~7fold and~5-fold increases in transcriptional activities  of GV148-miR-181a promoters, as compared to the GV148-basic vector in Hep2 cells, indicating that the miR-181a promoter regions that we used were indeed active (P < 0.01, Fig. 4K). Meanwhile, deletion of the MAX-binding site R2 impaired the effect of MAX on miR-181a transcription activation, suggesting that MAX binds to their special binding regions to regulate miR-181a transcription. Furthermore, MAX and MYCT1 knockdown significantly inhibited luciferase activities in Hep2 cells transfected with constructs containing either both two binding sites (R1 + R2) or one binding site (R1), as compared to their own activities (P < 0.05, Fig. 4K), suggesting that interactions between MYCT1 and MAX occur on the miR-181 promoter in vitro. Taking together, these data display that MAX directly promotes miR-181a transcription.
Here, qRT-PCR and western blot analyses showed that both NPM1 mRNA (Fig. 5C) and protein (Fig. 5D) levels were significantly down-regulated in Hep2 cells treated with miR-181a mimics, but up-regulated in Hep2 cells treated with the inhibitor, as compared to negative control-treated cells and untransfected Hep2 cells. A negative correlation was identified between miR-181a and NPM1 mRNA levels (r = -0.4185, P < 0.01, Fig. 5E). We also performed luciferase reporter assays with the wild-type and mutant 3 0 UTRs of NPM1. Our results demonstrated that miR-181a significantly decreased the relative luciferase activity of the wild-type NPM1 3 0 UTR compared to the controls (P < 0.01, Fig. 5F). However, miR-181a had no significant effect on mutant 3 0 UTR of NPM1, suggesting that miR-181a indeed targets NPM1.
CCK8 and colony formation assays showed that miR-181a mimics and inhibitor significantly suppressed and promoted Hep2 cell survival and colony formation compared to negative control-treated cells and untransfected Hep2 cells, respectively (Fig. 5G,H). For apoptosis assays, Hep2 cells were harvested 48 h after incubation, stained with Annexin V-PE and 7-AAD for flow cytometry. As a result, miR-181a overexpression and knockdown significantly enhanced and inhibited laryngeal cancer cell apoptosis, respectively (Fig. 5I). Moreover, the effects of miR-181a on laryngeal cancer cell viability and early apoptosis were rescued by NPM1 expression (Fig. 5G-I). Collectively, these results suggest that miR-181a suppresses cell viability and promotes early apoptosis directly via NPM1 in laryngeal cancer cells.

MYCT1 promotes cell viability and inhibits apoptosis through miR-181a and NPM1
MYCT1 significantly inhibited laryngeal cancer cell survival and colony formation compared to controls (empty vector-transfected cells and untransfected Hep2 cells) (Fig. 6A,B). Both NPM1 overexpression and   both NPM1 overexpression and miR-181a knockdown significantly rescued the effects of MYCT1 on laryngeal cancer cell early apoptosis compared to controls (Fig. 6C). These results suggest that MYCT1 regulates laryngeal cancer cell viability and apoptosis via miR-181a/NPM1. As NPM1 acts as a regulator in mitochondrial-dependent and -independent apoptotic pathways, we explored the effects of MYCT1 on protein levels and activities of Caspases-3, -8, and -9 in laryngeal cancer cells. MYCT1 significantly increased the activities of Caspases-3, -8, and -9 compared to untransfected Hep2 cells and both NPM1 overexpression and miR-181a knockdown rescued the effects of MYCT1 on Caspases-3, -8, and -9 activities in Hep2 cells compared to controls. (Fig. 6D). Similarly, MYCT1 significantly promoted Caspases-3, -8, and -9 cleavages compared to untransfected Hep2 cells and both NPM1 overexpression and miR-181a knockdown significantly rescued the effects of MYCT1 on Caspases-3, -8, and -9 activities in Hep2 cells compared to controls (Fig. 6E). Our data suggest that MYCT1 regulates laryngeal cancer cell apoptosis mediated by NPM1 in mitochondrial-dependent and -independent manners.

Discussion
The imbalance between cell viability and apoptosis represents an important cause of carcinogenesis [28,29]. Apoptosis is a unique form of cell death essential for many physiological processes [30]. Apoptosis is controlled by both the extracellular death receptor pathway and the intracellular mitochondrial pathway [31,32]. Apoptotic activators, Caspases-8 and -9 are necessary for the extracellular death receptor pathway and intracellular mitochondrial pathway, respectively, both of which have the same effector Caspase-3 [33].
MAX is a ubiquitously expressed and highly conserved transcription factor that regulates various aspects of cell behaviors including cell viability, differentiation, and apoptosis via the MYC/MAX/MAD network of basic helix-loop-helix leucine zipper (bHLHZip) transcription factors. Max was originally discovered based on its ability to associate with c-Myc and is required for Myc to bind DNA and activate transcription [34]. MAX has been viewed as a central component of the transcriptional network, forming homodimers as well as heterodimers with other members of the Myc and Mad families [35]. MYCT1 is mainly associated with the cytoplasmic membrane, endoplasmic reticulum, and Golgi body. MYCT1 has three forms of protein's secondary structure: alpha-helix, extended strand, and random coil [36]. Consistent with the notion that MYCT1 is not a transcriptional activator, we did not find any potential MYCT1-binding site in miR-181a promoter region (data not shown). Instead, we discovered that MYCT1 interacts with MAX, which binds the miR-181a gene promoter and promotes its transcription. Thus, it appears that MYCT1 acts as a co-transcription factor or a component of MAX transcript complex in the regulation of miR-181a transcription. We will conduct more studies in the future to identify the specific mechanism and relationship between MYCT1 and MAX.
Nucleophosmin (NPM/B23) family contains three subtypes, including NPM1, NPM2, and NPM3 [37]. NPM1 is a multifunctional protein that modulates ribosome biosynthesis, centrosome replication, cell viability, and apoptosis through a variety of signaling pathways [38]. Studies have shown that NPM1 is overexpressed in multiple solid tumors, and accelerates malignant transformation and viability and inhibits apoptosis [39]. Furthermore, NPM1 overexpression has been positively correlated with cancer progression and thus, a valuable biomarker for many types of cancer, including gastric cancer, colon cancer, liver cancer, and ovarian cancer [40][41][42][43]. Similar to the above findings, we also found overexpression of NPM1 in laryngeal carcinoma tissues and cells, suggesting that NPM1 is involved in laryngeal tumorigenesis. We reveal for the first time the key roles of NPM1 in laryngeal cancer. Our data here suggest that NPM1 is one of the downstream targets of MYCT1, and MYCT1 regulates laryngeal cancer cell viability and apoptosis by NPM1 through the extracellular death receptor pathway and the intracellular mitochondrial pathway to at least maintain the balance between viability and apoptosis.
Here, miR-181a is an important member of the miR-181 family and has been shown to take part in cancer cell viability, apoptosis, invasion, and migration by targeting different genes [44]; miR-181a is also abnormally expressed in lots of malignancies such as gastric cancer, myeloma and breast cancers, suggesting that it is an important gene in carcinogenesis [45][46][47]. Interestingly, miR-181a is considered tumor-specific depending on different targets. For instance, miR-181a enhances cell viability and diminishes apoptosis in clear cell renal cell carcinoma by targeting Kr€ uppellike factor 6 (KLF6) [48]. In contrast, miR-181a reduces cell viability and inhibits apoptosis of cutaneous squamous cell carcinoma by down-regulating Kirsten rat sarcoma 2 viral oncogene homolog (KRAS) [49]. Although these various observations for the roles of miR-181a may be highly dependent on the cellular and disease context, we provide strong evidence that miR-181a-targeting NPM1 inhibits laryngeal cancer cell viability and promotes apoptosis in vitro for the first time. The restoration of NPM1 rescued the effects of miR-181a on laryngeal cancer cell viability and apoptosis. Our finding that NPM1 is a direct target of miR-181a and miR-181a is the target of MYCT1-MAX complex revealed a complex network operating in laryngeal cancer cells (Fig. 7). Taken together, our data, for the first time, revealed the roles of NPM1 and miR-181a in laryngeal cancer and the association with MYCT1. We conclude that MYCT1 decreases NPM1 expression via MAX-regulated miR-181a expression and the net effects lie in the inhibition of cell viability and promotion of apoptosis through extracellular and intracellular apoptotic pathways. The factors involved in this pathway may serve as biomarkers for early diagnosis and treatment of laryngeal cancer. Further investigations to gain insights into the functional and clinical implications of the pathway in vivo will be pivotal in the prevention and treatment of laryngeal cancer.

Specimens and cell culture
Cancerous and paired adjacent tissues were obtained from patients with laryngeal squamous cell cancer (LSCC), who underwent surgery at the NO. 463 Hospital of PLA (Shenyang, China). The study was approved by the ethics committee of China Medical University and written informed consents were obtained from each patient. No patient had received radiotherapy or chemotherapy prior to surgery. Pathological verification of these tissues was conducted by a pathologist and the samples were stored at À80°C immediately after been removed from the patients.
Lentiviral GV358-MYCT1 and control, AgeI, were constructed by GENECHEM, to overexpress MYCT1 and to establish stable cell lines. The lentiviruses were transduced

RNA-sequencing and analysis
RNA samples were sequenced by the Beijing Genomics Institute (BGI, Shenzhen) with the HiSeq TM 4000 SE50 platform (for small RNA) and Hiseq TM 4000 PE101 platform (for large RNA). After the total RNA extraction and DNase I treatment, magnetic beads with Oligo (dT) are used to isolate mRNA (for eukaryotes). Mixed with the fragmentation buffer, the mRNA is fragmented into short fragments. Then cDNA is synthesized using the mRNA fragments as templates. Short fragments are purified and resolved with EB buffer for end reparation and single nucleotide A (adenine) addition. Thereafter, the short fragments are connected with adapters. After agarose gel electrophoresis, the suitable fragments are selected for the PCR amplification as templates. During the QC steps, Agilent 2100 Bioanaylzer and ABI StepOnePlus Real-Time PCR System are used in quantification and qualification of the sample library. At last, the library could be sequenced using Illumina HiSeq TM 4000.
We used cuffdiff for quantitative and differential analysis. Based on the quantitative results, NOIseq was used to analyze the differences between groups, and the subsequent model expression cluster analysis. Primary sequencing data that was produced by Illumina HiSeq TM 4000, called as raw reads, was subjected to quality control (QC) that determines if a resequencing step is needed. After QC, raw reads were filtered into clean reads which were aligned to the reference sequences. QC of alignment was performed to determine if resequencing is needed. The alignment data was utilized to calculate distribution of reads on reference genes and mapping ratio. When the alignment result passed QC, we proceed with downstream analysis, including gene and isoform expression, deep analysis based on gene expression (PCA/correlation/screening differentially expressed genes and so on), exon expression, gene structure refinement, alternative splicing, novel transcript prediction and annotation, SNP detection, Indel detection, gene fusion. Further, we also performed deep analysis based on DEGs, including Gene Ontology (GO) enrichment analysis, Pathway enrichment analysis, cluster analysis, protein-protein interaction network analysis, and finding transcription Factor (Data S2).
We made a multi-hypothesis test correction for the Pvalues of the difference test, and determined the domain values of the P-value by controlling the FDR. We used 'FDR ≤ 0.001 and the absolute value of Log2 Ratio ≥ 1' as the threshold to judge the significance of gene expression difference.

Co-immunoprecipitation
Hep2 cells were lysed in IP lysis buffer (Beyotime) containing 1% PMSF for 30 min on ice and centrifuged at 12 000 g for 20 min at 4°C. Extracts (containing 1 mg of total protein) were pre-cleared with 100 lL of protein G Plus/protein A agarose beads at 4°C for 10 min and centrifuged at 12 000 g for 15 min to remove the beads for supernatant. Since no commercial IP-grade antibody for MYCT1 is available at present, here anti-GFP antibody for MYCT1 was used. And considering the insufficient of GFP antibody, we selected the IP-grade antibody for MAX to perform co-IP experiment. A quantity of 2lg anti-GFP antibody or 2lg anti-MAX antibody together with 50 lL protein G Plus/protein A agarose beads were added to the immunoprecipitate and incubated by rotation at 4°C overnight. Then the mixtures were centrifuged at 12 000 g for 1 min to remove the precipitates and the beads containing immune complexes were washed for three times with cold PBS buffer following centrifugation at 300 9 g for 2 min. Normal mouse IgG was used as a negative control. Then western blotting assay was performed with another related primary antibody.

Immunofluorescence analysis
Hep2 cells were grown on glass coverslips in 12-well culture plates for 4 h, respectively, and Hep2 cells were transiently transfected with MYCT1 overexpression plasmid. After 24 hours post-transfection, cells were washed with PBS, fixed for 15 min at room temperature with 4% paraformaldehyde. After being washed in PBS with 0.5% Triton X-100 at room temperature for three times, cells were blocked in PBS with 1% bovine serum albumin for 1 h. Then, cells were incubated with the desired primary antibody (1 : 100) overnight at 4°C, followed by incubation with a specific fluorescently labeled IgG for 1 h in the dark. Subsequently, cells were stained with DAPI (blue), mounted for 5 min and examined using fluorescence microscopy.

Cell viability assay
Hep2 cells (5 9 10 3 cells per well) were seeded into 96-well plates. Cell viability was measured at specified time points using the Cell Counting Kit-8 (KeyGEN, Nanjing, China) according to manufacturer's instruction. After incubating for 2 h, viable cells were counted and growth curve was constructed using OD450 nm as the ordinate axis.

Colony formation assay
Hep2 cells (5 9 10 3 cells per well) were seeded into six-well plates and allowed to grow until visible colonies formed. After 4-6 days, colonies were fixed with methanol and stained with hematoxylin. Cell colonies were photographed under a microscope and determined by counting the number of colonies.

Cell apoptosis assay
Hep2 cells were grown in six-well plates to reach about 70% confluence. After 48 hrs transfection, cells were digested and collected. Thereafter cells were harvested and stained with Annexin V-PE/7-AAD Apoptosis Detection Kit (KeyGEN, Nanjing, China) according to the manufacturer's protocol, and finally analyzed by Flow Cytometer (FACS Calibur, Becton Dickinson, USA).

Caspase activity assay
Caspase-3, caspase-8, and caspase-9 activities were determined using the Colorimetric Assay Kits (KeyGEN, Nanjing, China) according to the manufacturer's instruction. Briefly, Hep2 cells were harvested after treatment with Cells were incubated on ice in 100 lL lysis buffer containing 1 lL DTT for 60 min and proteins were then isolated by centrifugation at 11 200 g for 1 min. Protein concentration was measured by the BCA Protein Assay kit (Byotime) and 200 lg protein was diluted in 50 lL reaction buffer containing 0.5 lL DTT for each assay. A quantity of 5 lL of Caspase-3, -8 or -9 substrates was added, respectively. The reaction mixture was incubated at 37°C for 4 h in the dark. Levels of caspase activities were measured at 405 nm by a microplate reader, and were calculated by ODexperiment/ODcontrol.

Statistical analysis
All data were subjected to statistical analysis using SPSS program (IBM, SPSS, New York, NY, USA) and shown as mean AE standard error of the mean (SEM). Each experiment was executed at least three times. NPM1 and miR-181a expression levels were analyzed by nonparametric ttest. Correlation analysis between of miR-181a and NPM1 mRNA expression was performed by Pearson correlation analysis. The student's t-test and one-way ANOVA were used in estimating the significance of differences in mean values. Symbols *, **, *** and ns represent P < 0.05, P < 0.01, P < 0.001 and no significance, respectively.