Introduction

Aflatoxins were highly toxic, mutagenic and carcinogenic compounds, which produced by fungi Aspergillus flavus and Aspergillus parasiticus1. The aflatoxins were discovered as a contaminant of human and animal food in the late 1950s and early 1960s2. Aflatoxins are potent liver toxins, lethal when consumed in large doses. Studies showed that dietary exposure to aflatoxins is one of the major risk factors for HCC, the most frequent liver cancer in the world3. More than 600,000 people die each year from HCC, mostly (48%) in developing countries4. AFB1, the major aflatoxin product, is a genotoxic hepatocarcinogenic compound which may also cause tumors in other organs, such as colon and kidney5. It is bio-activated in liver by cytochrome P450 and its epoxide metabolite attacks DNA, forming AFB1-DNA adducts6 that might evolve to secondary injuries such as apurinic sites (AP) or imidazole AFB1 formamidopyrimidine opened rings (AFB1-FAPY)7. Therefore, the removal or inactivation of this toxic fungal metabolite is a major concern to salvage-contaminated foodstuffs and feedstuffs8. The human hepatocellular carcinoma cell line HepG2 has been widely used as a model system to evaluate toxic effects of various substances, especially for the study of the most potent hepatocarcinogen, AFB19,10. MiRNAs are endogenous, non-coding small RNAs that usually are 21–30 nt11. MiRNAs can regulate gene expression at post-transcriptional level by degrading target mRNA or inhibiting translation as a result of complementary matching between miRNAs and specific sites in target genes. A large fraction of protein-coding genes can be miRNA targets, while a single miRNA can target hundreds to a thousand or more mRNAs as well. In the recent years, several studies revealed that miRNAs aberrantly expressed in human HCC in comparison with matched non-neoplastic tissue12. Furthermore, results from some reports suggested that changes in the expression of miRNAs may occur early in a variety of essential cellular processes, including cell growth, differentiation, metabolism and apoptosis13 and these changes may be related to specific etiological factors, such as AFB1. Some studies showed that genotoxical environmental agents, including AFB1, caused a variety of non-genotoxic alterations. These still preliminary evidences raised the possibility of using miRNAs as early markers for aflatoxins exposure14.

Recently, miR-34a has been demonstrated to be a key direct transcriptional target of p53 in most HCC15,16. Many of the biological targets of miR-34a have been recently identified. Studies showed that the ectopic expression of miR-34a induces cell-cycle arrest, senescence and apoptosis by regulation of critical cell cycle motors or apoptosis inhibitors including CDK4/6, cyclin E2 (CCNE2), cyclin D1(CCND1), E2F3 and Bcl-217,18. Interestingly, miR-34a also exerts tumor suppressor activity by regulating Wnt/β-catenin signaling which is a master regulator of cell proliferation, differentiation and movement19. Aberrant regulation of the Wnt/β-catenin signaling pathway by the mutation of one of the critical members of this pathway appears to play an important role in the development of hepatocellular cancers20. However, mechanisms of the regulation of miRNA in hepatocellular cancers development remain to be clarified.

Considering the effects of AFB1 as one of the most important reasons in HCC, we hypothesized that AFB1 might also trigger the differential expression of miRNAs which contribute to hepatocellular cancer development. Moreover, there are a lack of knowledge on the relationship between miRNAs and AFB1 in vitro, which is important to explain the role of miRNA in carcinogenesis. Based on these, by using Illumina deep sequencing, the changes of miRNAs profiling induced by AFB1 can be well studied. What’s more, the role of miR-34a will be explored in the hepatotoxicity induced by AFB1.

Methods

Cell culture and treatment

The human HCC cell lines HepG2 were cultured in monolayer in Dulbecco’s Modified Eagle’s Medium (DMEM, Neuronbc, Beijing) supplemented with 10% of fetal bovine serum (FCS, Hyclone, USA) and 1% of antibiotics (100 U/mL Penicillin Streptomycin Amphotericin B, Maichen). Cells were grown at 37 °C and 5% CO2 in a humidified atmosphere. For cell counting and subculture, the cells were dispersed with trypsin. HepG2 cells were treated with AFB1 at different concentrations of 0 and 10 μg/mL for 24 h. We labeled the 10 μg/mL treatment as group N (N1 and N2 for duplication), while the control as group CK (CK1 and CK2 for duplication). AFB1 were dissolved in DMSO and added to the culture media. The final concentration of DMSO in the media was less than 0.1%. Every group was designed two repeats, while the R2 were 0.971 and 0.964 of CK and the AFB1 treatment group, respectively.

RNA extraction

About 5.0 × 106 cells per sample were used for RNA isolation using miRcute miRNA Isolation Kit (Tiangen, Beijing) according to the manufacturer’s protocol. RNA degradation and contamination were monitored on 1% agarose gels. RNA purity was checked using the Nano Photometer® spectrophotometer (IMPLEN, CA, USA). RNA concentration was measured using Qubit® RNA Assay Kit in Qubit® 2.0 Flurometer (Life Technologies, CA, USA), while the RNA integrity was assessed using the RNA Nano 6000 Assay Kit of the Agilent Bioanalyzer 2100 system (Agilent Technologies, CA, USA) with the parameters: RIN ≥ 7.5, concentration ≥ 200 ng/μL.

Library preparations for Small RNA sequencing

RNA samples were stored at −80 °C and sequenced with the Illumina HiSeqTM2000/MiSeq platform. An amount of 3 μg total RNA per sample was used as input material for the small RNA library. Sequencing libraries were generated using NEB Next Multiplex Small RNA Library Prep Set for Illumina (NEB, USA.) following manufacturer’s recommendations and index codes were added to attribute sequences to each sample. Briefly, NEB 3′ SR Adaptor was direct and specifically ligated to 3′ end of miRNA, siRNA and piRNA. After the 3′ ligation reaction, the SR RT Primer hybridized to the excess of 3′ SR Adaptor (that remained free after the 3′ ligation reaction) and transformed the single-stranded DNA adaptor into a double-stranded DNA molecule. This step is important to prevent adaptor-dimer formation, besides, dsDNAs are not substrates for ligation mediated by T4 RNA Ligase 1 and therefore do not ligate to the 5′SR Adaptor in the subsequent ligation step. 5′ends adapter was ligated to 5′ends of miRNAs, siRNA and piRNA. Then first strand cDNA was synthesized using M-MuLV Reverse Transcriptase (RNase H−). PCR amplification was performed using Long Amp Taq 2×Master Mix, SR Primer for illumina and index (X) primer. PCR products were purified on a 8% polyacrylamide gel (100 V, 80 min). DNA fragments corresponding to 140 ~ 160 bp (the length of small noncoding RNA plus the 3′ and 5′ adaptors) were recovered and dissolved in 8 μL elution buffer. At last, library quality was assessed on the Agilent Bioanalyzer 2100 system using DNA High Sensitivity Chips.

The clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq SR Cluster Kit v3-cBot-HS (Illumia) according to the manufacturer’s instructions. After cluster generation, the library preparations were sequenced on an Illumina Hiseq 2500/2000 platform and 50 bp single-end reads were generated.

Target gene predictions

Target predictions, determined with avariety of algorithms including DIANA-microT, miRanda, PicTar and TargetScanS, were collected for the validated miRNA with the miRGendatabase (version 4.0). Multihit miRNA target analysis was also performed with miRanda alone (http://www.microRNA.org)to enrich for target genes with three or more target sites for the validated schizophrenia-associated miRNA. The frequency of miRNA targeting for any given target gene was determined for the pooled gene list of validated miRNA with PASW Statistics 18, (SPSS, Chicago, Illinois; IBM, Armonk, New York). In the multihit analysis, target genes with more putative miRNA binding sites were assumed to have more potential for post-transcriptional regulation and greater intensity than those with only one or two.

GO and KEGG Enrichment Analysis

Pathway analysis of these lists was achieved with the functional annotation tools on the Database for Annotation, Visualization and Integrated Discovery (DAVID; http://david.abcc.ncifcrf.gov/)21,22. Gene Ontology (GO) enrichment analysis was used on the target gene candidates of differentially expressed miRNAs. GOseq based Wallenius non-central hyper-geometric distribution23, which could adjust for gene length bias, was implemented for GO enrichment analysis.

Quantification of miRNAs and mRNAs

Semiquantitative real-time PCR (qRT-PCR) was performed on 200 ng of total RNA extracts that had been polyadenylated and reverse transcribed into cDNA using an anchored oligo (dT) primer (Tiangen, Beijing, China). The miRNA was transcribed into first-strand cDNA using miRcute miRNA first-strand cDNA synthesis kit (Tiangen, Beijing, China). The miRNA primers were listed in the supplementary Table 1A.

mRNA was transcribed into first-strand cDNA using Quantscript RT Kit (Tiangen, Beijing, China). qRT-PCR was run using the RealMasterMix (SYBR green I) (Tiangen, Beijing, China). The genes include Drosha, DRCG8, Dicer1 and the target genes of some miRNAs. RT-PCRs were run on the CFX96 Real-time PCR machine (BIO-RAD, Richmond, CA). The primers were marked in the supplementary Table 1B. Each RT reaction contained 11.25 uL 2.5 × RealMasterMix, 0.3 uL stem loop RT specific primer, 200 ng sample in 25 uL according to the Kit Introduction. Target sequences were amplified by being incubated at 94 °C for 2 min, followed by 40 cycles of 94 °C for 20s and 58 °C for 34s. β-actin and 5S RNA were used as endogenous normalization control for genes and miRNA separately. All assays were analyzed using the delta–delta–Ct method.

Transfection of miR-34a inhibitor

MiR-34a inhibitor oligonucleotide (ACAACCAGCUAAGACACUGCCA) and negative control were chemically synthesized bygenepharma (Shanghai, China). HepG2 cells were transfected with 150 nM oligonucleotides using HTF of Nuolanxin (Beijing, China), according to the manufacturer’s protocol. Six hours after transfection, the cells were treated with 5, 10 and 15 μg/mL AFB1 untreated for an additional 24 h. The cells were harvested for subsequent experiments.

Cell-cycle analysis

The cells treated by AFB1 after the miRNA inhibitor and NC transfections were harvested and washed once in phosphate-buffered saline and fixed in 70% ethanol over night. The DNA content was then examined using the Cell Cycle Analysis Kit of Beyotime (Beijing, China) and analyzed using BD FACS Calibur FlowCytometer. The data were analyzed by ModiFit analysis software.

Statistics

The data were expressed as the means ± standard deviation. The experiments were repeated at least twice and each experiment included at least triplicate treatments. The data from different treatments were subjected to an analysis of variance (ANOVA) and the comparisons of the means were performed using Duncan’s multiple range test. All of the statistical analyses were performed using the SPSS 16.0 software program. Differences with values of p < 0.05 were considered to be significant.

Results

Summary of small RNA sequencing data

Total RNA was extracted from each group and thus two datasets were obtained from CK and treatment (CK1, 6889737 reads; CK2, 6528991 reads; N1, 5652409 reads; N2, 9962058 reads), respectively. Clean reads (about 90% of total reads) were obtained by removing reads containing ploy-N, with 5′adapter contaminants, without 3′adapter or the insert tag, containing ploy A or T or G or C and low quality reads from raw data (Table 1). Then, we chose a certain range of length (18–35 nt) from clean reads to do all the downstream analyses. Analyses of small RNA length distribution peaked at the size of 22 nt as shown in Fig. 1. The small RNA tags were mapped to reference sequence by Bowtie24 without mismatch to analyze their expression and distribution on the reference. Mapped small RNA tags were used to looking for known miRNA. miRBase20.0 was used as reference, modified software mirdeep225 and srna-tools-cli were used to obtain the potential miRNA and draw the secondary structures. Custom scripts were used to obtain the miRNA counts as well as base bias on the first position of identified miRNA with certain length and on each position of all identified miRNA respectively.

Table 1 Parameters of small RNA sequences.
Full size table
Figure 1
figure 1

Small RNA reads percentage of different length distribution in CK group and N group.

HepG 2 cells were treated with AFB1 (0 and 10 μg/ml), denoted as CK and N group, respectively for 24 hours.

Full size image

The matched reads were used to identify mature miRNAs and the numbers of their reads were accounted. The reads that did not yield a match were used to predict novel miRNAs using MIREAP. The numbers of miRNA reads were normalized by Tags per million (TPM) values26 (TPM = (readCount *1,000,000)/libsize) to express miRNAs in CK and N comaprable in one table.

Novel miRNA analyses

A number of criterions were used for evaluating whether a small RNA was a genuine miRNA, such as formation of a stable hairpin structure, lower minimal free energies for hairpin structure of its precursors and detection of miRNAs27. Given these analyses, novel miRNAs were identified and examined by real-time PCR. The negative peak and the positive peak were different. Based on this, the novel miRNA was existing (Fig. 2, supplementary Fig. 1).

Figure 2
figure 2

PCR analyses of the novel miRNA.

(A) the novle miRNA of CK1, N1 and NC was detected by PCR. (B) The melt peak of the novle miRNA of qRT-PCR product in CK1, N1 and NC group was detected.

Full size image

Down regulation of miRNA biogenesis after AFB1 treatment

The microprocessor component genes DiGeorge syndrome critical region 8 (DGCR8), Drosha and Dicer are the miRNA processing enzymes that are required for the maturation of miRNAs. They are reported to be down-regulated in the human cancer mostly, so we examined the expression of these three miRNA genes to estimate the influence of AFB1. Dicer and DRCG8 displayed a robust 2.0-fold decrease in the HepG2 cell lines after AFB1 administration (p = 0.002 and p = 0.02 respectively). The Drosha displayed a 1.5-fold decrease in expression (p = 0.005) (Fig. 3). These results indicated that AFB1 affected the mechanism of miRNAs synthesis processing.

Figure 3
figure 3

The mRNA expressions of Drosha, DRCG8 and Dicer1 of the HepG2 in CK and N group.

HepG 2 cells were treated with AFB1 (0 and 10 μg/mL) for 24 hours, denoted as CK and N group. Expression levels were normalized using β-actin. Data are presented as the means ± SD. Three independent experiments are conducted in HepG2 cells. *p < 0.05, **p < 0.01.

Full size image

Differently expressed miRNAs

We further analyzed the differentially expressed miRNA between two conditions/groups using the DESeq2R package28. The read count data of the miRNA expression level was analyzed based on negative binomial distribution. Volcano Fig. 4 was used to show the global distribution of the differentially expressed miRNA, corrected P-value of 0.05 was set as the threshold for significantly differential expression by default (padj < 0.05) (Fig. 4).We used hierarchical cluster to analyze differentially expressed sRNA of every sample (Fig. 5). First we got a set of difference of miRNAs from each combination, then compared the TPM of every sample finite union from all the sets of differential expression miRNAs to make hierarchical cluster. The P-values was adjusted using the Benjamini & Hochberg method (padj < 0.05). The results showed that hsa-miR-19b, hsa-miR-19a, hsa-miR-34a, hsa-miR-99a, hsa-miR-190a and hsa-miR-16 were up-regulated as a result of AFB1 treatment compared to CK, while hsa-miR-1307, hsa-miR-99b, hsa-miR-100-5p showed the opposite trend (Table 2).

Table 2 Deferential expressed miRNAs between N and CK group.
Full size table
Figure 4
figure 4

Different miRNA expression in volcano diagram.

The x axis stands for the fold change of different miRNAs. The Y axis stands for significant difference of miRNA expression changes. Every miRNA are represented with the dots. The blue dots indicate no significant difference miRNAs; The red dots indicate up-regulated miRNAs; The green dots mean down-regulated miRNAs.

Full size image
Figure 5
figure 5

Hierarchical clustering of miRNA expression.

miRNA profiles from four groups of HepG2 were clustered. Treatment groups are in columns, miRNAs in rows. Cluster analysis based on log10 (TPM + 1). The red means up-regulated miRNAs and the blue means down-regulated miRNAs.

Full size image

To validate the results, we analyzed the expression of these miRNA using qPCR, which revealed that hsa-miR-19a displayed a 2.15-fold increase in expression in the treatment group (p = 0.006), while hsa-miR-1307 displayed a 0.54-fold decrease (p = 0.001) (Fig. 6). The detailed data were listed in the supplementary Table 2.

Figure 6
figure 6

qRT-PCR analyses of 9 miRNAs expression levels.

Expression levels were normalized by 5S RNA. Data are presented as the means ± SD. Three independent experiments are conducted in HepG2 cells. *p < 0.05, **p < 0.01.

Full size image

The miRNA targets prediction and qRT-PCR validation

The targets were predicted according to the method above by different miRNA target prediction algorithms, which can be helpful in minimizing the number of putative or false positive targets. The mRNA expression of AGO/IGF1/mTOR (miR-99a), E2F3/Cyclin D1(miR-16), PTEN (miR-19a/b), Cyclin E/β-catenin/Bcl2 (miR-34a) were strongly correlated with its corresponding miRNAs shown in the parentheses. The mRNAs were decreased significantly after AFB1 treatment. All the primers used in the RT-PCR analyses were listed in supplementary Table 1B. The expression tendency of these mRNA targets was opposite to the expression of their corresponding miRNAs, as shown in Fig. 7.

Figure 7
figure 7

qRT-PCR analyses of mRNA expression levels.

Expression levels were normalized using β-actin. Data are presented as the means ± SD. Three independent experiments are conducted in HepG2 cells. *p < 0.05, **p < 0.01.

Full size image

KEGG pathway and GO Analysis

The predicted target genes were subjected to KEGG pathway enrichment analysis (Supplementary Table 3). Six pathways including “Axon guidance” “Pathways in cancer” and “Wnt signaling pathway” were analyzed in both miRNA expression groups (all up-regulated miRNA were divided into one group while down-regulated ones the other). The Wnt signaling pathway has been studied to have a closely relationship with AFB1-induced hepatocellular cancers.

The top 30 enriched GO terms based on FDR for gene targets of the differently expressed miRNAs were shown in Table 3. Detailed classifications about the biological process, cellular component and molecular function of the miRNA targets were shown in Fig. 8. Twenty one of the top 30 GO terms belonged to biological process, including “microtubule-basedprocess” “single-organism process” “biological regulation”. The other 9 GO terms were cellular component including “intracellular” “organelle”.

Table 3 The top 30 enriched GO terms.
Full size table
Figure 8
figure 8

Function classification of the target genes of 9 miRNAs.

GO terms were applied to enriched the target genes.

Full size image

For further study on the regulatory mechanism of miRNA after AFB1 treatment, we focus on the Wnt signaling pathway, based on the results that several miRNA targets, such as Cyclin D1, all included in the Wnt signaling pathway. The other genes in this pathway, including MACF1, CDK4 and c-Myc, were all validated to have a decrease trend after AFB1 treatment (Fig. 9).

Figure 9
figure 9

qRT-PCR analyses of the mRNA expression levels of MACF1, c-Myc and CDK4.

Expression levels were normalized using β-actin. Data are presented as the means ± SD. Three independent experiments are conducted in HepG2 cells. *p < 0.05, **p < 0.01.

Full size image

Anti-miR-34a oligo transfection negatively correlates to levels of β-catenin and Wnt signaling pathway

We previously showed that AFB1 down-regulated β-catenin expression and Wnt signaling pathway. Thus, to further test whether abnormal expression of miR-34a could induce its target genes to regulate cellular processes, we treated HepG2 cells with 150 nM anti-miR-34a oligo transfection before AFB1 adminstration and performed qRT-PCR to detect. The results indicated that 150nM can effectively inhibited the miR-34a expression and the levels of miR-34a were up-regulated in after AFB1 treatment both in NC group (marked NC) and anti-miR-34a group (marked A) (Fig. 10A). The miR-34a target gene, β-catenin, cyclin E and Bcl2 (Fig. 10B) and CDK4, c-Myc of the Wnt signaling pathway all showed the opposite trend of miR-34a change (Fig. 10C). Cyclin D1 showed no significant change.

Figure 10
figure 10

qRT-PCR analyses of miRNAs and genes expression.

The detection was conducted in HepG2 cells transfected with an NC-inhibitor or hsa-miR-34a before treated with AFB1 (0, 5, 10 and 15 μg/mL). (A) The expression of miR-34a was detected. Expression levels were normalized using 5S.#p < 0.05 (B) The mRNA expression levels of β-catenin, cyclin E and Bcl-2. Expression levels were normalized using β-actin. (C) The expression levels of CDK4, c-myc of the Wnt signaling pathway. Expression levels were normalized using β-actin. The significant difference were labelled with a, b, c and d. Data are presented as the means ± SD. Three independent experiments are conducted.

Full size image

Deregulated miRNAs involved in the cell cycle

The results analyzed by Modifit show that AFB1 induced S-phase arrest in HepG2 cell lines, especially the arrest were significant when 10 and 15 μg/mL AFB1 treatment (p = 0.053 and .003, respectively). However, the 150 nM hsa-miR-34a inhibitor seemed to reduce the S-phase arrest, which seemed there was no signification in the hsa-miR-34a inhibitor group. It can be seen in the Fig. 11, between the NC and hsa-miR-34a inhibitor group, the S-phase arrest were reduced significantly in 10 and 15 μg/mL AFB1 as shown. The G0/G1 phases were reversed with the S phase, while the G2/M phases were unchanged (Data were not shown).

Figure 11
figure 11

Cell-cycle analyses in NC and anti-miR-34a inhibitor group.

The S phase of HepG2 cells transfected with an NC-inhibitor or hsa-miR-34a before treated with AFB1(0, 5, 10 and 15 μg/mL) were detected. Data are presented as the means ± SD. Three independent experiments are conducted. #p < 0.05.

Full size image

Discussion

Previous studies have revealed that AFB1 could induce cytotoxicity in hepatoma cells. Several studies also revealed that miRNAs aberrantly expressed in human HCC29. However, if miRNAs are involved in cell regulation of AFB1-induced HCC has received little attention. Previously, Yang et al. provided miRNA level changes in AFB1-induced hepatic injury which may lead to HCC through high-throughput profiling of miRNA in rat liver tissue before and after treatment with AFB130. To further study the functional complexity of miRNAs in AFB1-induced HCC, we applied Illumina sequencing to investigate high-throughput profiling of miRNA in HepG2 treated with AFB1 in vitro. Different from the AFB1 miRNA profiling in vivo, we studied further about an important miRNA which seemed regulate several signaling pathways. To our knowledge, this is the first study to investigate the miRNA profiling under AFB1 stress and screen one single miRNA to further study its regulating pathways. These studies pave a new way to gain a better understanding of the mechanisms of AFB1 induced hepatocellular toxicology and carcinoma, in human cells.

The microprocessor component genes DGCR8, Drosha and the type III ribonuclease are responsible for cleavage of the pre-miRNA hairpin structure to form the mature miRNA. Both Drosha and DGCR8 are abundant and ubiquitous, but the expression level of these proteins depends on cell types. Dicer has been revealed to degrade their target mRNAs which is required for the maturation of short interfering RNAs (siRNAs). Decreased Dicer expression in cancer conferred increased proliferative ability and an invasive phenotype31,32,33. Considering the important role of the integrity of miRNA processing mechanism, we studied DGCR8, Drosha and Dicer, which all displayed a decrease after 24 h AFB1 administration. Our expression analysis of the miRNA biosynthesis pathway revealed that three key molecules were simultaneously down-regulated in HepG2, indicating miRNA biosynthesis was damaged.

We generated a differential miRNA expression profile of HepG2 under different treatments (normal and AFB1 treated) and RT-PCR was performed to validate the profile. After background subtraction, normalization and correction for multiple testing with SAM, we identified 9 miRNAs that were differentially expressed after AFB1 treatment. Significantly, 6 of these miRNAs were up-regulated, including several cancer-related miRNAs, hsa-miR-34a, hsa-miR-19a/b and hsa-miR-99a. However, these cancer-related miRNAs showed contradictory changes in expression relative to their known functions. For example, miR-34a, which is considered to be a tumour suppressive miRNA, was up-regulated in our study. In the past experiment, the over-expression of miR-34a was found in Fischer 344 male rats after AFB1 treatment for 3 days30, the parallel trend led us to the view that there may be different expression of miRNAs between the cancer development process and cancer formation.

We also predicted a set of potential target genes of the differentially expressed miRNAs. KEGG pathway and GO enrichment analysis based on the predicted targets revealed that the effect of abnormally expressed miRNAs upon HepG2 on a metabolomic scale. Based on the KEGG and GO analysis, we predicted the abnormally expressed miRNAs triggered by AFB1 might contribute to tumorigenesis in liver cancer, disorder of cell cycle and apoptosis. Furtherly, potential target genes of the differential expressed miRNAs, such as AGO/IGF1/mTOR (miR-99a), E2F3/Cyclin D1 (miR-16), PTEN (miR19a/b), Cyclin E/β-catenin/Bcl2 (miR-34a), were verified by q-PCR. IGF-1R and mTOR were characterized as direct targets of miR-99a, which exerted function of miR-99a as a cell cycle progression inhibitor. AGO is a key enzyme involved in the regulation of the processing step from pre-miRNA to mature miRNA, which is involve in HCC34. PTEN is a tumor suppressor gene and essential for regulating PI3K/AKT signaling pathway. There was a correlation of the down-regulation of PTEN mRNA with tumor TNM stage and metastasis in HCC35,36. We focus on the miR-34a and its targets in AFB1 administration. The previous studies showed that Cyclin D1, cyclin E and β-catenin are included in cell cycle induced by the Wnt/β-catenin signaling pathway and β-catenin has been proved to be a critical component of this signaling pathway37. So we checked the expression of other key genes including CDK4, MACF1 and c-Myc, and the results showed that these genes were all down-regulated by AFB1. The past experiments have studied that microtubule-actin cross-linking factor 1 (MACF1) appeared to be inactivate GSK3β by phosphorylation. Then, β-catenin was released and entered the nucleus. Our results showed the AFB1 exposure down-regulated MACF1 and then the expression of β-catenin with its downstream genes (C-myc and cyclin D1) was reduced as well38. Cyclin D1, with the help of CDK4, is a crucial mediator of the G1 to S progression39. The RT-PCR showed that CDK4 and other member of the cyclin family, Cyclin E, were down-regulated upon exposure to AFB1. C-Myc, considered as the proto-oncogene, is always up-regulated in the tissue of cancer. However, it was down-regulated in AFB1-treated HepG2 cell. This indicated the different mechanisms of AFB1 in vivo and in vitro.

To further study, the miR-34a inhibitor was used before AFB1 treatment. RT-PCR showed that miR-34a inhibitor relieved the down expression of miR-34a targets genes and CDK4, MACF1 and c-myc. Considering that these genes generally mediates cell cycle, we inspected the distribution of different phases of cell cycle. Cell cycle dysregulation was an essential step in the initiation and development of human malignancies, including HCC40. The results showed that AFB1 can induce S-phase arrest and the miR-34a inhibitor can relieved. We predicted that miR-34a induced Cyclin D1 and Cyclin E down-regulated. there may be some other miRNAs or mechanisms induced S-phase arrest more than G1-arrest.

In summary, we preliminarily established the relationship between miRNAs, especially miR-34a and AFB1 in HepG2. We applied Illumina sequencing and bioinformatics to compare the expression profile of AFB1-treated HepG2 and control HepG2 cell lines. We identified some abnormally expressed cancer-related miRNAs. This is the first study to investigate the miRNA expression and sequence profile in HepG2 under AFB1 administration. Among these abnormally expressed miRNA, miR-34a was proved to up-regulated by AFB1, which can participated in the regulation of Wnt/β-catenin signaling pathway and cell cycle. Based on multifaceted toxicology studies on AFB1, we hypothesized that AFB1 might trigger abnormal expression of miRNAs which are involve in the mechanism of liver tumorigenesis.

Additional Information

How to cite this article: Zhu, L. et al. miR-34a screened by miRNA profiling negatively regulates Wnt/β-catenin signaling pathway in Aflatoxin B1 induced hepatotoxicity. Sci. Rep. 5, 16732; doi: 10.1038/srep16732 (2015).