Analysis of microRNA expression profiling during paraquat-induced injury of murine lung alveolar epithelial cells

Abstract

Paraquat (PQ) as a non-selective heterocyclic herbicide, has been applied worldwide for over a few decades. But PQ is very harmful to humans and rodents. The lung is the main target organ of PQ poisoning. It is an important event that lung epithelial cells are injured during PQ-induced acute lung injury and pulmonary fibrosis. As a regulator of mRNA expression, microRNA (miRNA) may play an important role in the progress. Our study was to investigate the mechanisms of PQ-induced injury of pulmonary epithelial cells through analyzing the profiling of miRNAs and their target genes. As a result, 11 differentially expressed miRNAs were screened, including 1 upregulated miRNA and 10 downregulated miRNAs in PQ-treated murine lung alveolar epithelial cells (MLE-12 cells). The bioinformatic analyses suggested that the target genes of these miRNAs were involved in mitochondrial apoptosis pathway and DNA methylation, and participated in the regulation of PI3K-Akt, mTOR, RAS, TNF, MAPK and other signal pathways which related to oxidative stress and apoptosis. This indicated that miRNAs were an important regulator of oxidative stress and apoptosis during PQ-induced injury of murine lung alveolar epithelial cells. The findings would deepen our understanding of the mechanisms of PQ-induced pulmonary injury and might provide new treatment targets for this disease.

INTRODUCTION

Paraquat (1,1’-dimethyl-4,4’-bipyridinium dichloride, PQ) is one of the most widely used herbicides with high efficiency. However, PQ is highly toxic to humans and rodents, and even a very small amount may lead to fatal consequences (Serra et al., 2003). Human deaths due to PQ poisoning were firstly reported in the 1960s (Bullivant, 1966). Since then, a large number of PQ poisoning cases have been reported worldwide, especially in developing countries. Oral administration of large amounts of PQ remains a common method of suicide in many countries (Gunnell and Eddleston, 2003). Because of the high mortality rate and the lack of a special antidote, how to treat PQ poisoning patients has become a thorny medical problem.

PQ can cause multiple organ injuries, but lung injury is the most prominent. PQ can be absorbed via multiple ways such as digestive tract, respiratory tract and the skin, and be distributed into different organs. PQ can accumulate intracellularly against a concentration gradient through type II alveolar epithelial cells (AECII) and Clara cells as well as type I alveolar epithelial cells (AECI) (Smith et al., 1990). Thus, alveolar epithelial cells are particularly vulnerable to PQ. The injury of alveolar epithelial cells is a key event in acute lung injury (ALI) and irreversible pulmonary fibrosis which could gradually result in respiratory failure (Fukushima et al., 2002). The injury of alveolar epithelial cells imputes to the activation oxygen species (ROS) system (Suntres, 2002). ROS, as unstable free radicals originated from reduced PQ, can oxidize surrounding lipids and proteins to lead to lipid peroxidation and protein crosslinking deactivation, which may increase or even break the permeability of membrane (Yasaka et al., 1981) and finally result in cell damage (Xu et al., 2014). Besides, apoptosis as a special injury mechanism can also be activated by chromosomal damage and function loss of mitochondria (Rizwan et al., 2020). Therefore, in addition to depleting ROS, regulating oxidative stress and apoptosis induced by ROS is an available countermeasure for PQ poisoning.

MicroRNA (miRNA) is a group of noncoding RNAs with about 20-23 nucleotides encoded by the genome, which is widely distributed in eukaryotes and some viruses (Bartel, 2004). By pairing with the target gene’s mRNA base, it can guide the silencing complex (RISC) to reduce mRNA or hinder its translation (Mohr and Mott, 2015). MiRNA is involved in the regulation of about 50% of gene expression in mammals, and regulates the physiological and pathological activities, such as cell proliferation and differentiation, signal transduction, apoptosis, tumorigenesis and development (Lima et al., 2011). In PQ-poisoned alveolar epithelial cells, miRNA, as a regulatory factor of gene expression, is likely to change and play a role in promoting or inhibiting the damage. Recently, miRNA alterations have proved to play important roles in PQ-induced pulmonary injury and pulmonary fibrosis (Jin et al., 2018; Liu et al., 2019). However, there is no study on the expression profile of miRNA in PQ-induced injury of lung alveolar epithelial cells, and the regulatory mechanism of miRNA remains to be elucidated.

Our study aimed to analyze the profiling of miRNAs in PQ-treated murine lung alveolar epithelial cells (MLE-12 cells) and investigate the mechanisms of PQ-induced injury of lung epithelial cells through bioinformatic analysis. The study might reveal some potential biological and molecular mechanisms contributing to PQ-induced lung injury.

MATERIALS AND METHODS

Chemicals and reagents

MLE-12 cells were purchased from American Type Culture Collection (ATCC). PQ was purchased from Sigma-Aldrich Co. (St. Louis, MO, USA). The culture media and buffer solutions as follows were purchased from Corning, Inc. (Corning, NY, USA): phosphate-buffered saline (PBS), fetal bovine serum (FBS), Dulbecco’s Modified Eagle’s Medium-high glucose (DMEM). The following reagents were purchased from Dojindo Molecular Technologies (Rockville, MD, USA): Annexin V, FITC Apoptosis Detection Kit, Cell Counting Kit-8 (CCK-8). TRIzol™ Reagent was purchased from Invitrogen (Carlsbad, CA, USA). Taq DNA Polymerase, miRNA first strand cDNA synthesis and primers were purchased from Sangon Biotech (Shanghai, China).

Cell cultures

MLE-12 cells were placed in cell culture flask with culture medium (90% DMEM medium+10% FBS) and then cultured at 37°C with 5% CO₂ in a humidified atmosphere. The medium in cell culture flask was replaced every 24 hr. Cell subculture was performed at the confluence of 70%-80%.

Chemical treatment

When the cells were at the confluence of 70%-80%, PQ dissolved in cell medium with different concentrations of was added into the cell culture flask and the cells were cultured for 24 hr. Based on the previous research (Xiong et al., 2019; de Oliveira et al., 2016), PQ with concentration of 100 μM has been used in many models of PQ-induced cell injury. Besides, 100 μM of PQ, equal to 25.72 μg/mL, is widely accepted as a fatal concentration in serum (Proudfoot et al., 1979). Further, according to our previous study, in a concentration-dependent manner, the cell viability was decreased and the apoptotic rate was increased apparently in PQ-treated human alveolar epithelial A549 cells (Sun et al., 2018; Wang et al., 2018b). Thus, to measure the cell viability and apoptotic rate in PQ-treated MLE-12 cells: 50, 100, 200, 400 μM of PQ was used as the cell viability assay concentration, while 50, 100, 150 μM of PQ as the cell apoptosis assay concentration. An equal amount of PBS was dissolved in the medium as a control group. Based on the results above, we chose median inhibition concentration as the appropriate PQ concentration for microarray analysis and the RT-PCR detection. All experiments were repeated three times.

Cell viability assay

The viability of MLE-12 cells was measured by the Cell Counting Kit-8 assay. Cells were seeded onto 96-well plates at a density of 1.0 × 10⁵ cells/well and cultured for 24 hr. When the cells were approximately 70%-80% confluent, according to our previous study, 10 μL PQ solution of different concentrations was added in the wells and incubated for another 24 hr. Then 10 μL CCK-8 solution was added to each well of the culture plate, and the cells were cultured for 4 hr. The absorbance at 450 nm was measured by enzyme scale.

Cell apoptosis assay

The apoptotic rate of PQ-treated MLE-12 cells was measured by Annexin V-FITC apoptosis detection kit. Cells were seeded onto 6-well plates at a density of 1.0 × 10⁶ cells/well and cultured for 24 hr. Then PQ of different concentrations as well as an equal amount of PBS was added to the wells and the cells were cultivated for another 24 hr. The cells were washed with PBS twice before digestion and then suspended in 1 × Annexin V Binding Solution after centrifugation. Then 5 μL Annexin V-FITC and 10 μL PI were added into the 100 μL cell suspension in dark for 15 min at room temperature. For flow cytometric analysis, 400 μL 1 × Annexin V Binding Solution was added to dilution each sample and all the samples were detected with a flow cytometer in 1 hr.

Microarray assays

Total RNAs of MLE-12 cells were extracted using TRIzol™ Reagent according to the manufacturer’s instructions. Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA) was used to detect the amount and quality of RNA. To examine the expression profiles of miRNA, Agilent Mouse miRNA (8*60K) V21.0 (Agilent Technologies, Inc.) was selected. Agilent Scan Control software was used for scanning, while Agilent Feature Extraction software V10.7.1.1 was used for data abstraction. Three parallel replicates were performed for verification.

Real-time PCR

To verify the data of microarray, 6 differential expression miRNAs were chosen to test by RT-PCR. TRIzol™ Reagent was used to extract the total RNAs of MLE-12 cells and miRNA First Strand cDNA Synthesis for syntheses of cDNA. Using Taq DNA Polymerase quantitative fluorescence detection was performed. PCR program was 95°C for 30 sec, 95°C for 5 sec, 60°C for 34 sec for 40 cycles. Primer sequences of the 6 miRNAs for reverse transcription and PCR are listed in Table 1 and Table 2. U6 was chosen for miRNA normalization. All the experiments above followed the manufacturer’s instructions.

Table 1. Stem-loop primer sequence for reverse transcription.

ID	Stem-loop primer sequence (from 5’ to 3’)
mmu-miR-125a-5p	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTCACAG
mmu-miR-19b-3p	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTCAGTT
mmu-miR-23b-3p	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACGGTAAT
mmu-miR-29a-3p	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTAACCG
mmu-miR-106b-5p	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACATCTGC
mmu-miR-130a-3p	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACATGCCC

Table 2. Primer sequence for Real-Time PCR.

ID	Forward Sequence (from 5’ to 3’)	Reverse Sequence (from 5’ to 3’)
U6	CTCGCTTCGGCAGCACA	AACGCTTCACGAATTTGCGT
mmu-miR-125a-5p	GCGTCCCTGAGACCCTTTAAC	AGTGCAGGGTCCGAGGTATT
mmu-miR-19b-3p	CGTGTGCAAATCCATGCAA	AGTGCAGGGTCCGAGGTATT
mmu-miR-23b-3p	GCGATCACATTGCCAGGG	AGTGCAGGGTCCGAGGTATT
mmu-miR-29a-3p	CGCGTAGCACCATCTGAAAT	AGTGCAGGGTCCGAGGTATT
mmu-miR-106b-5p	GCGCGTAAAGTGCTGACAGT	AGTGCAGGGTCCGAGGTATT
mmu-miR-130a-3p	CGCGCAGTGCAATGTTAAAA	AGTGCAGGGTCCGAGGTATT

Enrichment analyses

Downstream target genes of the differently expressed miRNAs were predicted by the following online analysis platforms miRDB (http://mirdb.org/), microRNA.ORG (http://www.microrna.org/microrna/home.do) and TargetScan (http://www.targetscan.org/vert_72/). The intersection of the three platforms was used for the further analysis. DAVID Bioinformatics Resources 6.8 (https://david.ncifcrf.gov/) was the platform which provided gene ontology (GO) analyses and pathway analysis.

The genes were classified according to different functions to achieve the purpose of gene annotation and classification through gene GO analysis which showed the main significant biological processes, including biological process (BP), cellular component (CC), and molecular function (MF) categories for the differential genes. The predicted miRNA target gene is mapped to each term in the GO database, and the gene number of each term was calculated. Furthermore, compared with the whole genomic background, hypergeometric test was used to screen out the GO term which was significantly enriched in the differentially expressed genes.

Besides, pathway analyses, including Kyoto Encyclopedia of Genes and Genomes (KEGG) and Biocarta pathway analysis, were performed to determine the significant pathways for differentially expressed genes. The predicted miRNA target gene is mapped to each term in the KEGG database, and the gene number of each term is calculated. Then, compared with the whole genomic background, hypergeometric test was used to screen out the pathway term which was significantly enriched in the differentially expressed genes.

MiRNA-mRNA network analysis

The downstream target genes of differential expressed miRNAs are predicted using three online analysis platforms: miRDB, microRNA.ORG and TargetScan. The intersection results of the platforms above were regarded to be the target genes. MiRNA-Gene-Network was constructed by Cytoscape 3.61 to describe pathway and GO analysis results intuitively.

Statistical analysis

GraphPad Prism 7.0 and SPSS 20.0 software were used for Statistical analysis. One-way ANOVA and post hoc Tukey’s test were adopted. The repeatability was ensured by three parallel replicates and each treatment group was compared with the control group individually; P-value (< 0.05) and fold change (≥ 2) or (≤ 0.5) were regarded as significant results.

RESULTS

Cell viability

The viability of MLE-12 cells administered PQ for 24 hr was assessed by CCK-8 assay. The survival rates of cells were decreased significantly with the increase of PQ concentration, compared with the control group (Fig. 1). At the concentration of 100 μM, the cell viability decreased to about half of the normal level which could be approximately regarded as median inhibition concentration.

Fig. 1

PQ-induced cytotoxicity of MLE-12 cells. MLE-12 cells were cultured in cell medium with PQ (50, 100, 200, and 400 μM) or with PBS for 24 hr. Cell viability was measured by CCK-8 assay. Data were presented as mean ± S.D. of three independent experiments, *P < 0.05 PQ groups vs. control group at 24 hr, with one-way ANOVA and post hoc Tukey’s test.

Apoptosis rate

To explore the reason of the decreased cell viability we detected the apoptosis rate of MLE-12 cells in different concentrations of PQ. The result is shown in Fig. 2. It indicated that the apoptosis rates increased in a concentration-dependent manner at 24 hr and apoptosis was a vital reason for the decreased cell viability.

Fig. 2

PQ-induced MLE-12 cell apoptosis. MLE-12 cells cultured in cell medium with PQ (50, 100, and 150 μM) or with PBS for 24 hr. (A) The percentage of cellular apoptosis of MLE-12 cells was measured with Annexin V and PI double-stained by flow cytometry. (B) A histogram presents the combination of early apoptosis (Q2) and late apoptosis (Q4) percentages. Data were presented as mean ± S.D. of three independent experiments, ***p < 0.001, PQ groups vs. control group at 24 hr, with one-way ANOVA and post hoc Tukey’s test.

The miRNA profiles analysis

Based on the results above, we chose the median inhibition concentration (100 μM) as optimal PQ concentration to measure the miRNA profiles. Microarray assays showed the holistic changes in miRNA profiles in PQ-treated (100 μM, 24 hr) cells. The miRNAs differentially expressed are summarized in Fig. 3. Thermography reflected the relationship between the sample and gene expression patterns by classifying the mathematical characteristics of different gene data. Hierarchical clustering suggested that miRNA expression profiles were accordant in the same group. There are 11 miRNAs regarded as considerable alternations (fold change /(FC) ≥ 2 or ≤ 0.5, p < 0.05) including 1 up-regulated miRNA (mmu-miR-1892) and 10 down-regulated miRNAs (mmu-miR-125a-5p, mmu-miR-19b-3p, mmu-miR-23b-3p, mmu-miR-26a-5p, mmu-miR-23a-3p, mmu-miR-25-3p, mmu-miR-24-3p, mmu-miR-29a-3p, mmu-miR-106b-5p and mmu-miR-130a-3p).

Fig. 3

The miRNA differential expression profiles. MLE-12 cells cultured in cell medium with 100 μM PQ or with PBS for 24 hr. Total RNAs of cells were extracted and expression profiles of miRNA were tested and confirmed by three parallel replicates. (A) The heat map of differentially expressed miRNAs in MLE-12 cells. Green represents 0, indicating no change in gene expression level; red represents an increase in expression level; black represents a decrease in expression level, and the brightness of the color represents the degree to which the gene expression level increases or decreases. In addition, the expression of similar genes (probes) and samples are clustered together. (B)The volcano plot of miRNAs in MLE-12 cells. One of the coordinates shows the negative log of p-values calculated by t-test, and the other shows the changed value after log2 conversion under the comparison of two conditions. The red dots represent upregulation (fold change ≥ 2, p value < 0.05), and the blue dots represent downregulation (fold change ≤ 0.5, p value < 0.05).

Real-time PCR validation of miRNA expression

Real-time PCR was used to validate the results of microarray. We chose the most obvious differential expression miRNAs (miR-106b-3p, miR-125a-5p, miR-130a-3p, miR-19b-3p, miR-23b-3p and miR-29a-3p) Consistent with the preliminary conclusion of microarray analysis, compared to the normal cells in the control group, the expression of six miRNAs was substantially reduced in PQ treated cells (fold change ≤ 0.5, p < 0.05). The results of Real-time PCR corroborated the effectiveness of microarray to identify the differentially expressed miRNAs (Fig. 4).

Fig. 4

Real-time PCR results. MLE-12 cells cultured in cell medium with 100 μM PQ or PBS for 24 hr. Total RNAs of MLE-12 cells were extracted using TRIzol™ Reagent. The cDNAs were synthesized by miRNA First Strand cDNA Synthesis (Stem-loop Method) following the instructions. Then real-time PCR was performed to determine the miRNA expression. MiRNA expression was standardized with U6. Each value represents the mean ± S.D. from double experiments in triplicate, which is representative of three separate experiments. *p value < 0.05, PQ-treated group vs. CTRL group.

Functional and pathway analysis

The prediction results of online analysis platform indicated that 4183 genes might be the target genes of these differentially expressed miRNAs, and 633 of them could be regulated by multiple miRNAs (Table 3).

Table 3. Target genes of differentially expressed miRNAs.

Systematic Name	Gene Symbol	Gene number
mmu-miR-29a-3p	AGO2 TET3 TET1 etc.	985
mmu-miR-26a-5p	SKP2 KRTAP19-1 NABP1 etc.	383
mmu-miR-25-3p	RPL15 CLDN11 BTG2 etc.	122
mmu-miR-24-3p	ENTPD6 ISM2 BCL2L11 etc.	420
mmu-miR-23b-3p	AUH Apaf1 MRPL18 etc.	93
mmu-miR-23a-3p	SFT2D1 ZNF667 TFPI2 etc.	558
mmu-miR-19b-3p	CHIC1 RAB18 TMEM65 etc.	72
mmu-miR-1892	CDR1as CLIP3 GREM1 etc.	1041
mmu-miR-130a-3p	BCL2 FMR1 TGFBR1 etc.	194
mmu-miR-125a-5p	EIF1AD MSRB3 SCN2B etc.	247
mmu-miR-106b-5p	STAT3 MTERFD2 NR4A3 etc.	69

Besides, GO enrichment analysis was performed to predict the potential biological functions of differentially expressed miRNAs. According to the GO analysis results (Fig. 5), we found that functions such as positive regulation of protein insertion into mitochondrial membrane involved in apoptotic signaling pathway, positive regulation of intrinsic apoptotic signaling pathway in response to osmotic stress as well as the metabolic and catabolic process of 5-methylcytosine were mainly affected by differentially expressed miRNAs.

Fig. 5

Enrichment map of GO analyses. GO terms in descending order of enrich factor value. The first 30 results were taken. Red arrows point at the mentioned functions and processes above.

According to the online analysis platforms, the intersection target genes were used for KEGG pathway analysis to determine the probable changed signaling pathways. The results showed that the target genes of the differentially expressed miRNAs were chiefly associated with processes such as PI3K-Akt signaling pathway (p-value =2.51E-5), mTOR signaling pathway (p-value =2.44E-5), RAS signaling pathway (p-value =1.77E-5), TNF signaling pathway (p-value =3.59E-5) and MAPK signaling pathway (p-value =2.07E-4) (Fig. 6).

Fig. 6

Pathway analysis Enrichment map of KEGG pathway analysis. (A) Pathways in descending order of enrich factor value, the first 30 results were taken. (B) Pathways arranged by p-value, those pathways with p-value < 0.005 were taken. Red arrows point at the mentioned signaling pathways above.

miRNA-mRNA network analysis

For examining the relationship between the differentially expressed miRNAs and the mRNAs that encoded related proteins, we constructed a miRNA-mRNA network using Cytoscape while the predicted downstream target genes were from targetscan, microRNA.org and miRDB. As Fig. 7 shows, it indicated that a miRNA can recognize multiple target mRNAs at the same time, and a gene might also be regulated by multiple miRNAs. Such as miR-29a-3p and miR-23a-3p were miRNAs with obvious expression, and it was predicted that they were highly likely to combine with Sirtuin 1 (SIRT1).

Fig. 7

miRNA-mRNA network. Data were from targetscan, microRNA.org and miRDB. The miRNA-mRNA interaction showed that multiple miRNAs could regulate the same gene, and one miRNA could target multiple different target genes. The red words represent miRNAs. The blue words represent mRNAs. The interactions were marked with yellow background.

DISCUSSION

The mortality rate of PQ poisoning patients remains at a high level and the main reason is that no effective antidotes and effective treatment methods have been found (Yin et al., 2013). At present, the common treatment for PQ poisoning is to control the inflammatory reaction, to regulate the process of oxidative stress and extra-corporeal elimination (Park et al., 2014; El-Aarag et al., 2019; Dinis-Oliveira et al., 2008). Recently, targeting miRNA is considered a potential candidate for therapeutic invention in some diseases (Tiwari et al., 2018). Several miRNA therapies have made progress in phase I and phase II trials (Vicentini et al., 2019). Thus, measuring the miRNA expression profile could help make clear the mechanism in PQ-induced lung epithelial cell injury and provide a new therapy target in the future.

MiRNAs have been found to be involved in PQ poisoning in many studies. In SH-SY5Y cells, miR-34a was found to mediate PQ-induced neurotoxicity and apoptosis (Alural et al., 2015). Down-regulated miR-17-5p promoted PQ-induced degeneration of dopaminergic neurons in Neuro-2a cells (Wang et al., 2018a). Increased miR-210 expression after PQ poisoning aggravates the progression of pulmonary fibrosis through enhancing the stability of hypoxia-inducible factor-1 α, which promotes PQ-induced epithelial-mesenchymal transition (Zhu et al., 2017). These studies indicate that miRNAs participate in different types of PQ-induced cell damage.

In our study, we analyzed miRNA expression profiling in murine lung alveolar epithelial cells administered PQ. A total of 11 miRNAs with different expression were measured and the target genes were predicted. MiR-130a-3p, the most significant one, was predicted to participate in oxidative stress and apoptosis by regulating a variety of target genes, such as BCL2 and TGFBR1. Besides, miR-23b-3p is also involved in the regulation of the expression of many apoptosis-related factors such as Apaf1. In addition, we found that multiple miRNAs could regulate the same target gene. For example, miR-24-3p, miR-125a-5p and miR-29a-3p were involved in the regulation of Sh3glb1 expression, while Sh3glb1 gene can encode endorphin B1 protein, participate in the outer membrane formation of mitochondria and Golgi apparatus, and participate in caspase dependent apoptosis by promoting the activation of Bax/BAK1 (Takahashi et al., 2005).

Many recent studies have confirmed that some differentially expressed miRNAs that we found are closely related to apoptosis. Overexpression of miR-130a-3p could significantly increase cells’ apoptotic rate in triple-negative breast cancer cells and induce apoptosis in hepatic stellate cells (Ouyang et al., 2014; Jiang et al., 2018). MiR-106b-5p could prevent vascular endothelial cells from apoptosis in atherosclerosis diseases (Zhang et al., 2016). Besides, over-expressed miR-19b-3p was reported to suppress renal tubular epithelial cell apoptosis (Chen et al., 2019b). It indicates that our prediction was consistent with these studies.

Besides, we performed bioinformatic analyses to predict the enriched GO terms and regulatory pathways of these miRNAs. GO enrichment analysis indicated that positive regulation of protein insertion into mitochondrial membrane involved in apoptotic signaling and positive regulation of intrinsic apoptotic signaling pathway in response to osmotic stress were significantly enriched. These functions are apparently associated with intrinsic apoptosis. Besides the metabolic and catabolic process of 5-methylcytosine was also enriched. This indicated that methylation activity in cells administrated by PQ might be enhanced, and DNA methylation could further affect gene expression by regulating miRNA synthesis (Hale et al., 2017). Otherwise, according to pathway analysis, PI3K-Akt, mTOR, RAS, TNF and MAPK signaling pathways were significantly enriched. Among them, PI3K-Akt-MAPK signaling pathway, regulated by miR-130a-3p, miR-23a-3p and miR-23b-3p, is involved in cell proliferation and DNA repair. PI3K-Akt-mTOR signaling pathway is related to autophagy (Fan et al., 2010; Ma and Blenis, 2009). RAS and TNF signaling pathways are involved in the process of apoptosis (Wang et al., 2018c). These signaling pathways have been confirmed in the study of PQ lung injury mechanism. For example, the PI3K-Akt-mTOR pathway has been found to be associated with PQ-induced pulmonary fibrosis (Liu et al., 2019). The RAS and MAPK signaling pathways have also been shown to be related to the mitochondrial pathway of apoptosis (Cui et al., 2019; Rezatabar et al., 2019). Therefore, regulating miRNA to affect these signaling pathways and then regulating oxidative stress and apoptosis are worthy subjects to explore.

Moreover, as miRNA-mRNA network analysis showed, SIRT1 was predicted to be regulated by miR-29a-3p and miR-23a-3p. SIRT1, as a member of the sirtuin protein family, is an enzyme responsible for deacetylation of proteins. Previous studies have found that enhancing SIRT1 expression can reduce the regulation of apoptosis and autophagy (Karbasforooshan and Karimi, 2018). Besides, enhanced SIRT1 expression could attenuates PQ-induced lung injury (Li et al., 2016). This indicated that multiple miRNAs had synergistic effects on regulating of gene expression. Besides, apoptosis and autophagy could be vital biological processes in PQ-poisoned cells.

Interestingly, several studies indicate that the same miRNA could play an opposite role in regulating apoptosis in different cell types and environments. Take miR-19b-3p as an example: the up-regulated miR-19b-3p could inhibit apoptosis in renal tubular epithelial cells (Chen et al., 2019b) but inhibit the cell proliferation and contribute to the apoptosis in SH-SY5Y human neuroblastoma cells (Ma et al., 2016). The down-regulated miR-19b-3p in A375 cells and SK-MEL-2 melanoma cells was found to promote apoptosis (Wei et al., 2018). On the contrary, overexpression of miR-19b-3p could cause intracellular ROS accumulation in human aortic endothelial cells, leading to cellular apoptosis (Xue et al., 2015). Similarly, overexpression of miR-23a-3p inhibits the cell apoptosis in renal carcinoma cells (Quan et al., 2019), but promotes apoptosis of oral squamous cell carcinomas by targeting FGF2 (Chen et al., 2019a). Thus, it is a considerable task to verify the role of the above differentially expressed miRNAs in regulating in PQ-induced apoptosis of pulmonary epithelial cells.

There are some limitations in this study. Firstly, although the clinical manifestation of PQ poisoning in mice is very similar to that in human, the alveolar epithelial cells derived from mice differ from those of humans after all so that there may be some differences in response to PQ toxicity between them. Secondly, we only chose the time of 24 hr as the observation time, but the stress response of the body is a continuous process, which may result in missing information about the dynamic changes of the miRNAs during the whole injury process. Some miRNAs we missed may have significantly differential expression at different time points, and that differential expression may be the key factor of modulation. Thirdly, the target genes of the differentially expressed miRNAs need to be confirmed. Consequently, such issues are expected to be investigated in our future studies.

In conclusion, we detected the miRNAs expression profile in PQ-induced injury model of MLE12 cells, and found 11 miRNAs with significant differential expression. Bioinformatics analysis showed that these miRNAs may target BCL2, TGFBR1 and other genes, and these miRNA target genes participate in the regulation of PI3K-Akt and other signal pathways related to oxidative stress and apoptosis. This indicated that miRNAs could be a crucial regulator in the process of PQ-induced pulmonary alveolar epithelial cell injury. Our findings may provide a new view for PQ-induced lung injury research and new probable targets for treatment of lung impairment due to PQ.

ACKNOWLEDGMENT

This research was supported by the National Natural Science Foundation of China (#81471851, #81971821).

Conflict of interest

The authors declare that there is no conflict of interest.

REFERENCES

• Alural, B., Ozerdem, A., Allmer, J., Genc, K. and Genc, S. (2015): Lithium protects against paraquat neurotoxicity by NRF2 activation and miR-34a inhibition in SH-SY5Y cells. Front. Cell. Neurosci., 9, 209.
• Bartel, D.P. (2004): MicroRNAs: genomics, biogenesis, mechanism, and function. Cell, 116, 281-297.
• Bullivant, C.M. (1966): Accidental poisoning by paraquat: report of two cases in man. BMJ, 1, 1272-1273.
• Chen, F., Qi, S., Zhang, X., Wu, J., Yang, X. and Wang, R. (2019a): miR-23a-3p suppresses cell proliferation in oral squamous cell carcinomas by targeting FGF2 and correlates with a better prognosis: miR-23a-3p inhibits OSCC growth by targeting FGF2. Pathol. Res. Pract., 215, 660-667.
• Chen, W., Zhou, Z.Q., Ren, Y.Q., Zhang, L., Sun, L.N., Man, Y.L. and Wang, Z.K. (2019b): Effects of long non-coding RNA LINC00667 on renal tubular epithelial cell proliferation, apoptosis and renal fibrosis via the miR-19b-3p/LINC00667/CTGF signaling pathway in chronic renal failure. Cell. Signal., 54, 102-114.
• Cui, S., Nian, Q., Chen, G., Wang, X., Zhang, J., Qiu, J. and Zhang, Z. (2019): Ghrelin ameliorates A549 cell apoptosis caused by paraquat via p38-MAPK regulated mitochondrial apoptotic pathway. Toxicology, 426, 152267.
• de Oliveira, M.R., Ferreira, G.C. and Schuck, P.F. (2016): Protective effect of carnosic acid against paraquat-induced redox impairment and mitochondrial dysfunction in SH-SY5Y cells: Role for PI3K/Akt/Nrf2 pathway. Toxicol. In Vitro, 32, 41-54.
• Dinis-Oliveira, R.J., Duarte, J.A., Sánchez-Navarro, A., Remião, F., Bastos, M.L. and Carvalho, F. (2008): Paraquat poisonings: mechanisms of lung toxicity, clinical features, and treatment. Crit. Rev. Toxicol., 38, 13-71.
• El-Aarag, B., Magdy, M., AlAjmi, M.F., Khalifa, S.A. and El-Seedi, H.R. (2019): Melittin Exerts Beneficial Effects on Paraquat-Induced Lung Injuries In Mice by Modifying Oxidative Stress and Apoptosis. Molecules, 24, 1498.
• Fan, Q.W., Cheng, C., Hackett, C., Feldman, M., Houseman, B.T., Nicolaides, T., Haas-Kogan, D., James, C.D., Oakes, S.A., Debnath, J., Shokat, K.M. and Weiss, W.A. (2010): Akt and autophagy cooperate to promote survival of drug-resistant glioma. Sci. Signal., 3, ra81.
• Fukushima, T., Tanaka, K., Lim, H. and Moriyama, M. (2002): Mechanism of cytotoxicity of paraquat. Environ. Health Prev. Med., 7, 89-94.
• Gunnell, D. and Eddleston, M. (2003): Suicide by intentional ingestion of pesticides: a continuing tragedy in developing countries. Int. J. Epidemiol., 32, 902-909.
• Hale, V., Hale, G.A., Brown, P.A. and Amankwah, E.K. (2017): A Review of DNA Methylation and microRNA Expression in Recurrent Pediatric Acute Leukemia. Oncology, 92, 61-67.
• Jiang, Y., Wang, W., Liu, Z.Y., Xie, Y., Qian, Y. and Cai, X.N. (2018): Overexpression of miR-130a-3p/301a-3p attenuates high glucose-induced MPC5 podocyte dysfunction through suppression of TNF-α signaling. Exp. Ther. Med., 15, 1021-1028.
• Jin, Y., Liu, W., Liu, X., Ma, T., Yang, C., Cai, Q. and Liu, Z. (2018): Transplantation of endothelial progenitor cells attenuated paraquat-induced acute lung injury via miR-141-3p-Notch-Nrf2 axis. Cell Biosci., 8, 21.
• Karbasforooshan, H. and Karimi, G. (2018): The role of SIRT1 in diabetic retinopathy. Biomed. Pharmacother., 97, 190-194.
• Li, S., Zhao, G., Chen, L., Ding, Y., Lian, J., Hong, G. and Lu, Z. (2016): Resveratrol protects mice from paraquat-induced lung injury: the important role of SIRT1 and NRF2 antioxidant pathways. Mol. Med. Rep., 13, 1833-1838.
• Lima, R.T., Busacca, S., Almeida, G.M., Gaudino, G., Fennell, D.A. and Vasconcelos, M.H. (2011): MicroRNA regulation of core apoptosis pathways in cancer. Eur. J. Cancer, 47, 163-174.
• Liu, M.W., Su, M.X., Tang, D.Y., Hao, L., Xun, X.H. and Huang, Y.Q. (2019): Ligustrazin increases lung cell autophagy and ameliorates paraquat-induced pulmonary fibrosis by inhibiting PI3K/Akt/mTOR and hedgehog signalling via increasing miR-193a expression. BMC Pulm. Med., 19, 35.
• Ma, X.M. and Blenis, J. (2009): Molecular mechanisms of mTOR-mediated translational control. Nat. Rev. Mol. Cell Biol., 10, 307-318.
• Ma, Y., Tian, S., He, S., Chen, Q., Wang, Z., Xiao, X., Fu, L. and Lei, X. (2016): The mechanism of action of FXR1P-related miR-19b-3p in SH-SY5Y. Gene, 588, 62-68.
• Mohr, A.M. and Mott, J.L. (2015): Overview of microRNA biology. Semin. Liver Dis., 35, 3-11.
• Ouyang, M., Li, Y., Ye, S., Ma, J., Lu, L., Lv, W., Chang, G., Li, X., Li, Q., Wang, S. and Wang, W. (2014): MicroRNA profiling implies new markers of chemoresistance of triple-negative breast cancer. PLoS One, 9, e96228.
• Park, J.S., Park, K.H., Kim, H. and Choi, S.Y. (2014): Effects of sivelestat treatment on acute lung injury in paraquat-intoxicated rats. Drug Chem. Toxicol., 37, 114-120.
• Proudfoot, A.T., Stewart, M.S., Levitt, T. and Widdop, B. (1979): Paraquat poisoning: significance of plasma-paraquat concentrations. Lancet, 2, 330-332.
• Quan, J., Pan, X., Li, Y., Hu, Y., Tao, L., Li, Z., Zhao, L., Wang, J., Li, H., Lai, Y., Zhou, L., Lin, C., Gui, Y., Ye, J., Zhang, F. and Lai, Y. (2019): MiR-23a-3p acts as an oncogene and potential prognostic biomarker by targeting PNRC2 in RCC. Biomed. Pharmacother., 110, 656-666.
• Rezatabar, S., Karimian, A., Rameshknia, V., Parsian, H., Majidinia, M., Kopi, T.A., Bishayee, A., Sadeghinia, A., Yousefi, M., Monirialamdari, M. and Yousefi, B. (2019): RAS/MAPK signaling functions in oxidative stress, DNA damage response and cancer progression. J. Cell. Physiol., 234, 14951-14965.
• Rizwan, H., Pal, S., Sabnam, S. and Pal, A. (2020): High glucose augments ROS generation regulates mitochondrial dysfunction and apoptosis via stress signalling cascades in keratinocytes. Life Sci., 241, 117148.
• Serra, A., Domingos, F. and Prata, M.M. (2003): [Paraquat intoxication]. Acta Med. Port., 16, 25-32.
• Smith, L.L., Lewis, C.P., Wyatt, I. and Cohen, G.M. (1990): The importance of epithelial uptake systems in lung toxicity. Environ. Health Perspect., 85, 25-30.
• Sun, D.Z., Song, C.Q., Xu, Y.M., Wang, R., Liu, W., Liu, Z. and Dong, X.S. (2018): Involvement of PINK1/Parkin-mediated mitophagy in paraquat- induced apoptosis in human lung epithelial-like A549 cells. Toxicol. In Vitro, 53, 148-159.
• Suntres, Z.E. (2002): Role of antioxidants in paraquat toxicity. Toxicology, 180, 65-77.
• Takahashi, Y., Karbowski, M., Yamaguchi, H., Kazi, A., Wu, J., Sebti, S.M., Youle, R.J. and Wang, H.G. (2005): Loss of Bif-1 suppresses Bax/Bak conformational change and mitochondrial apoptosis. Mol. Cell. Biol., 25, 9369-9382.
• Tiwari, A., Mukherjee, B. and Dixit, M. (2018): MicroRNA Key to Angiogenesis Regulation: MiRNA Biology and Therapy. Curr. Cancer Drug Targets, 18, 266-277.
• Vicentini, C., Galuppini, F., Corbo, V. and Fassan, M. (2019): Current role of non-coding RNAs in the clinical setting. Noncoding RNA Res., 4, 82-85.
• Wang, Q., Zhan, Y., Ren, N., Wang, Z., Zhang, Q., Wu, S. and Li, H. (2018a): Paraquat and MPTP alter microRNA expression profiles, and downregulated expression of miR-17-5p contributes to PQ-induced dopaminergic neurodegeneration. J. Appl. Toxicol., 38, 665-677.
• Wang, R., Sun, D.Z., Song, C.Q., Xu, Y.M., Liu, W., Liu, Z. and Dong, X.S. (2018b): Eukaryotic translation initiation factor 2 subunit α (eIF2α) inhibitor salubrinal attenuates paraquat-induced human lung epithelial-like A549 cell apoptosis by regulating the PERK-eIF2α signaling pathway. Toxicol. In Vitro, 46, 58-65.
• Wang, X., Li, A.J., Li, W.Z., Lu, W.J. and Cui, W.Z. (2018c): [The effects of Xuebijing injection on apoptosis and expression of regulatory factors TNF-α, NF-κB and Caspase-3 expression in the lung tissues of acute paraquat-induced rats]. Zhonghua Lao Dong Wei Sheng Zhi Ye Bing Za Zhi, 36, 551-555.
• Wei, Y.P., Wang, X.H., Liu, G., Zhang, J.F., Yang, Y.X., Zhang, J., Song, X.L., Li, Z.D. and Zhao, L.D. (2018): Matrine exerts inhibitory effects in melanoma through the regulation of miR-19b-3p/PTEN. Int. J. Oncol., 53, 791-800.
• Xiong, G., Zhao, L., Yan, M., Wang, X., Zhou, Z. and Chang, X. (2019): N-acetylcysteine alleviated paraquat-induced mitochondrial fragmentation and autophagy in primary murine neural progenitor cells. J. Appl. Toxicol., 39, 1557-1567.
• Xu, L., Xu, J. and Wang, Z. (2014): Molecular mechanisms of paraquat-induced acute lung injury: a current review. Drug Chem. Toxicol., 37, 130-134.
• Xue, Y., Wei, Z., Ding, H., Wang, Q., Zhou, Z., Zheng, S., Zhang, Y., Hou, D., Liu, Y., Zen, K., Zhang, C.Y., Li, J., Wang, D. and Jiang, X. (2015): MicroRNA-19b/221/222 induces endothelial cell dysfunction via suppression of PGC-1α in the progression of atherosclerosis. Atherosclerosis, 241, 671-681.
• Yasaka, T., Ohya, I., Matsumoto, J., Shiramizu, T. and Sasaguri, Y. (1981): Acceleration of lipid peroxidation in human paraquat poisoning. Arch. Intern. Med., 141, 1169-1171.
• Yin, Y., Guo, X., Zhang, S.L. and Sun, C.Y. (2013): Analysis of paraquat intoxication epidemic (2002-2011) within China. Biomed. Environ. Sci., 26, 509-512.
• Zhang, J., Li, S.F., Chen, H. and Song, J.X. (2016): MiR-106b-5p Inhibits Tumor Necrosis Factor-α-induced Apoptosis by Targeting Phosphatase and Tensin Homolog Deleted on Chromosome 10 in Vascular Endothelial Cells. Chin. Med. J. (Engl.), 129, 1406-1412.
• Zhu, Y., Wang, J., Meng, X., Xie, H., Tan, J., Guo, X., Han, P. and Wang, R. (2017): A positive feedback loop promotes HIF-1α stability through miR-210-mediated suppression of RUNX3 in paraquat-induced EMT. J. Cell. Mol. Med., 21, 3529-3539.