Fusaric acid induces hepatic global m6A RNA methylation and differential expression of m6A regulatory genes in vivo - a pilot study

ABSTRACT

N6-methyladenosine (m6A) is an abundant epitranscriptomic mark that regulates gene expression to execute cellular developmental programmes and environmental adaptation. Fusaric acid (FA) is a mycotoxin that contaminates agricultural foods and exerts toxicity in humans and animals; however, its epitranscriptomic effects are unclear. We investigated the effect of FA on global m6A RNA methylation and mRNA expression levels of key m6A regulatory genes in C57BL/6 mouse livers. C57BL/6 mice (n = 6/group) were orally administered 0.1 M phosphate-buffered saline (PBS) or 50 mg/kg FA. Mice were euthanized 24 h after oral administration, livers were harvested, and RNA was isolated. RNA samples were assayed for global m6A levels using an m6A RNA Methylation Quantification Kit. The mRNA expression of m6A regulators i.e. writers, erasers, and readers were measured by qRT-PCR. FA increased global m6A RNA methylation (p < 0.0001) in mouse livers. FA increased the expression of METTL3 (p = 0.0143) and METTL14 (p = 0.0281), and decreased the expression of FTO (p = 0.0036) and ALKBH5 (p = 0.0035). The expression of YTHDF2 (p = 0.0007), YTHDF3 (p = 0.0061), and YTHDC2 (p = 0.0258) were increased by FA in mouse livers. This study shows that the liver m6A epitranscriptome can be modified by FA exposure in an in vivo model and can be useful for identifying the molecular mechanisms whereby m6A RNA modifications influence the toxicological outcomes of FA exposure.

KEYWORDS:

• mycotoxin
• fusaric acid
• epitranscriptomics
• m6A RNA methylation
• m6A writers
• m6A erasers
• m6A readers

Introduction

Chemical modifications of RNA transcripts are involved in regulating RNA-protein and RNA-RNA interactions. N6-methyladenosine (m6A) is an epitranscriptomic mark and the most abundant post-transcriptional modification of RNA in eukaryotes. It is enriched in three prime untranslated regions and near stop codons at the RRACH (R = G or A; H = A, C, or U) consensus motif, where it regulates gene expression [1], development [2], and disease [3].

As a dynamic and reversible modification, m6A RNA methylation is regulated by writers and erasers [4]. The m6A writers comprise the methyltransferase complex, which includes methyltransferase-like 3 (METTL3), methyltransferase-like 14 (METTL14), and Wilm’s tumour 1-associated protein (WTAP), whose main function is to catalyse the formation of m6A [5]. METTL3 is the catalytic subunit of the methyltransferase complex that transfers a methyl group from S-adenosylmethionine to the N6 position of RNA adenine bases; METTL14 stabilizes METTL3 and enables RNA substrate binding; and WTAP facilitates nuclear localization of the METTL3-METTL14 complex [4,5]. The demethylases, fat mass and obesity-associated protein (FTO) and alkylation repair homolog 5 (ALKBH5) are considered m6A erasers that remove the N-methyl groups of m6A [6,7].

The biological functions of m6A RNA methylation are mediated by m6A RNA binding proteins, also known as readers, which decode the m6A information [8–10]. The m6A readers primarily consist of the YT521-B homology (YTH) domain family proteins 1–3 (YTHDF1–3) and the YTH domain containing proteins 1–2 (YTHDC1–2) [11]. YTHDF1, YTHDF3, and YTHDC2 promote translation of messenger RNA (mRNA) by interacting with ribosomal proteins and other translational machinery [1,8]; YTHDF2 destabilizes mRNA by promoting the deadenylation and decay of m6A modified mRNAs [9,12]; and YTHDC1 regulates cellular localization and alternative splicing by mediating nuclear export and recruiting pre-mRNA splicing factors [10,13].

The contamination of agricultural foods with pathogenic fungi and mycotoxins is a global problem that threatens food security and safety. In the body, mycotoxins are biotransformed into toxic metabolites that interact with various biomolecules such as DNA and RNA, thus affecting their normal functions through epigenetic modifications [14–17]. The role of epitranscriptomics in mycotoxicology is an emerging field of epigenetics that is crucial to understanding the adverse health effects and diseases caused by food-borne mycotoxins. Only a few studies have elucidated the effects of mycotoxins on m6A RNA methylation, and found alterations in global and transcript-specific m6A methylation levels as well as differential expression of m6A regulatory genes [18–23].

Fusaric acid (FA) is a mycotoxin produced by the Fusarium species that frequently contaminate agricultural foods and cause toxicity in humans and animals [24–27]. FA alters brain neurochemistry [28], blood pressure [24], platelet function [27], bone ossification [26], and notochord formation [25]. It is also associated with lethargy, feed refusal, vomiting, and decreased growth in animals [28,29], and akathisia in humans [30]. Studies have also reported synergistic effects of FA with other food-borne mycotoxins [29,31,32] as well as the role of FA in plant pathogenesis [33].

Despite the numerous studies on FA, little is known about its epitranscriptomic effects. In this study, we investigated the effect of FA on global m6A RNA methylation and mRNA expression levels of key m6A regulatory genes in C57BL/6 mouse livers. The liver was chosen for study as it is susceptible to damage by chemical substances and mycotoxins found in the foods ingested; this is due to its detoxification, biotransformation, and metabolizing functions as well as its close relationship with the gastrointestinal tract. In addition, this research builds on our previous in vitro study of FA and m6A RNA methylation in the human liver (HepG2) cell line [18]. Dysregulation in the liver m6A RNA methylation pattern and expression of m6A regulatory genes may influence liver function, lipid metabolism, and promote liver carcinogenesis [34–36]. Our findings provide the first evidence that the liver m6A epitranscriptome can be modified by FA exposure in an in vivo model and lay the foundation for future studies to determine the molecular mechanisms by which m6A RNA modifications influence FA toxicity.

Materials and methods

Animal treatment

All animal procedures were performed in full compliance with the Animal Research: Reporting In Vivo Experiments (ARRIVE) guidelines and were approved by the Institutional Animal Research Ethics Committee at the University of KwaZulu-Natal (AREC/079/016). Eight-week-old C57BL/6 male mice were housed at the Africa Health Research Institute under standard laboratory conditions (temperature = 25°C, humidity = 40–60%, 12 h light/dark cycle) with ad libitum access to a commercial mice feed and normal drinking tap water. After one week of habituation, twelve mice with a body weight of 17–23 g were randomly divided into two groups, control and FA, with six mice per group. FA (F6513) was purchased from Sigma-Aldrich and prepared in 0.1 M phosphate-buffered saline (PBS; pH 7.4). Mice were orally administered by gavage with 0.1 M PBS (control group) or 50 mg/kg FA (FA group) [26]. At the end of the experimental period (24 h after gavage), the mice were euthanized using gas inhalation anaesthesia, Isofor, and the livers were harvested. The livers were rinsed three times in 0.1 M PBS and stored in 500 µl Qiazol reagent (Qiagen, 79306) for RNA isolation. All experiments including animal treatments were performed two-independent times.

RNA extraction

Total RNA was extracted from control and FA-treated mouse livers stored in Qiazol reagent. Briefly, mouse livers were homogenized separately in Qiazol reagent and chloroform (100 µl) was added. The samples were centrifuged (12,000 xg, 4°C, 15 min) and isopropanol (500 µl) was added to the aqueous phase followed by overnight incubation at −80°C. The samples were centrifuged (12,000 xg, 4°C, 20 min), supernatants were removed, and RNA pellets were washed in 75% ethanol (500 µl). Samples were centrifuged (7,400 xg, 4°C, 15 min), RNA pellets were air dried (30 min, room temperature (RT)), and resuspended in nuclease-free water (15 µl). RNA concentration and purity were assessed with the Nanodrop2000 spectrophotometer (Thermo-Fisher Scientific). Samples with A260/A280 ratios of 1.9–2.1 were considered pure and used for all subsequent assays. RNA concentration was adjusted as required for the respective assays.

Quantification of global m6A RNA methylation

The m6A content in total RNA was measured using an m6A RNA Methylation Quantification Kit (Abcam, ab185912). Briefly, RNA samples standardized to 200 ng, positive control standards (0.01–0.50 ng), and negative control from the kit were added to a plate containing strip wells with 80 µl RNA binding solution. The plate was sealed with Parafilm M and gently shaken to allow the RNA to mix in solution before incubation at 37°C for 90 min. Thereafter, the binding solution was removed and each well was washed three times with 1X wash buffer (150 µl) before the addition of the m6A capture antibody (50 µl (1: 1,000), 1 h, RT). The wells were washed three times with 1X wash buffer (150 µl) and the detection antibody (50 µl (1: 2,000), 30 min, RT) was added. Subsequently, the wells were washed four times with 1X wash buffer (150 µl), followed by the addition of the enhancer solution (50 µl (1: 5,000), 30 min, RT). After five washes with 1X wash buffer (150 µl), the developer solution (100 µl, 10 min, RT) was added. The developer solution turns blue in the presence of m6A. Stop solution (100 µl) was added to inhibit the enzymatic reaction and the absorbance was measured using the SPECTROstar Nano spectrophotometer (BMG Labtech) at a wavelength of 450 nm. The mean absorbances of the positive control standards were used to construct a standard curve from which the percentage m6A in total RNA was calculated using the following equation:

$\mathrm{T}\mathrm{O}\mathrm{T}\mathrm{O}\mathrm{X}=2\hspace{0.17em}\mathrm{P}\mathrm{V}\hspace{0.17em}+p\mathit{-}AV$

Quantitative real-time polymerase chain reaction (qRT-PCR)

qRT-PCR was used to determine the mRNA expression of METTL3, METTL14, WTAP, FTO, ALKBH5, YTHDF1, YTHDF2, YTHDF3, YTHDC1, and YTHDC2. Total RNA (standardized to 1,000 ng) from control and FA-treated mouse livers were reverse transcribed into complementary DNA (cDNA) using the Maxima H Minus First Strand cDNA Synthesis Kit (Thermo-Fisher Scientific, K1652). qRT-PCR was performed using the PowerUp™ SYBR™ Green Master Mix (Thermo-Fisher Scientific, A25742) and the Applied Biosystems ViiA7 Real-Time PCR System (Thermo-Fisher Scientific). Thermocycler conditions were as follows: initial denaturation (95°C, 8 min), followed by 40 cycles of denaturation (95°C, 15 s), annealing (Table 1, 40 s), and extension (72°C, 30 s). GAPDH served as the endogenous control to normalize mRNA expression. The relative change in mRNA expression was calculated using the comparative threshold cycle (2^−ΔΔCt) method [37]. Primer sequences are in Table 1.

Table 1. Primer sequences used for qRT-PCR

Gene	Accession Number	Primer Sequence	Annealing Temperature (°C)
METTL3	NM_019721	F: CTGGGCACTTGGATTTAAGGAA R: TGAGAGGTGGTGTAGCAACTT	58
METTL14	NM_201638	F: GACTGGCATCACTGCGAATGA R: AGGTCCAATCCTTCCCCAGAA	60
WTAP	NM_175394	F: ATGGCACGGGATGAGTTAATTC R: TTCCCTTAAACCAGTCACATCG	60
FTO	NM_011936	F: CCGTCCTGCGATGATGAAGT R: CCCATGCCGAAATAGGGCTC	60
ALKBH5	NM_172943	F: GCATACGGCCTCAGGACATTA R: TTCCAATCGCGGTGCATCTAA	60
YTHDF1	NM_173761	F: ACAGTTACCCCTCGATGAGTG R: GGTAGTGAGATACGGGATGGGA	58
YTHDF2	NM_145393	F: ACAGGCAAGGCCGAATAATG R: GGCTGTGTCACCTCCAGTAG	58
YTHDF3	NM_172677	F: TACATGGGGAACAAGTGGATCT R: TAGGTGGATAGCCGT	58
YTHDC1	NM_177680	F: TTCAGGAGTTCGCCGAGATG R: CGGTGGTCCCATGTTATGAAAT	60
YTHDC2	NM_001163013	F: GAAGATCGCCGTCAACATCG R: GCTCTTTCCGTACTGGTCAAA	60
GAPDH	NM_001289726	F: ATGTGTCCGTCGTGGATCTGAC R: AGACAACCTGGTCCTCAGTGTAG	Same as the gene of interest

Data analysis

All data were presented as the mean ± standard error of the mean (SEM) (n = 6/group) and analysed using GraphPad Prism version 5.0. Statistical significance was considered at p< 0.05 using the unpaired t-test with Welch’s correction.

Results

FA increased global m6A RNA methylation in C57BL/6 mouse livers

We first determined the effect of FA on global m6A RNA methylation in C57BL/6 mouse livers. Levels of m6A were quantified from total RNA using a commercialized kit (Abcam, ab185912). The percentage of m6A was significantly increased in the FA-treated mouse livers compared to the control (1.20-fold, p < 0.0001; Figure 1). Although this difference in m6A levels may seem low, it is similar to that observed with aflatoxin B₁ in C57BL/6J mouse livers [20].

Figure 1. The effect of FA on global m6A RNA methylation in C57BL/6 mouse livers. Levels of m6A were measured in total RNA isolated from control and FA-treated mouse livers using an m6A RNA Methylation Quantification Kit. FA increased global m6A levels in mouse livers compared to the control. Results are presented as the mean ± SEM (n = 6/group). The unpaired t-test with Welch’s correction was used to determine statistical significance, ****p < 0.0001. CON: Control; FA: Fusaric acid; SEM: Standard error of the mean.

FA induced differential expression of m6A regulatory genes in C57BL/6 mouse livers

RNA m6A levels are dynamically and reversibly regulated by m6A writers and erasers. Additionally, m6A modified RNAs recruit m6A-dependent readers, which determine the fate of the RNA transcripts. To gain insight into how FA regulates global m6A RNA methylation in C57BL/6 mouse livers, we evaluated the mRNA expression levels of m6A writers, erasers, and readers. FA significantly increased the expression of m6A writers, METTL3 (3.18-fold, p = 0.0143; Figure 2(a)) and METTL14 (2.16-fold, p = 0.0281; Figure 2(a)), and decreased the expression of WTAP (0.44-fold, p = 0.0103; Figure 2(a)) in mouse livers compared to the control. The expression of m6A erasers, FTO (0.64-fold, p = 0.0036; Figure 2(b)) and ALKBH5 (0.71-fold, p = 0.0035; Figure 2(b)) were decreased by FA in the mouse livers compared to the control. FA also significantly increased the expression of m6A readers, YTHDF2 (4.97-fold, p = 0.0007; Figure 2(c)), YTHDF3 (2.14-fold, p = 0.0061; Figure 2(c)), and YTHDC2 (3.35-fold, p = 0.0258; Figure 2(c)) in mouse livers compared to the control. There were no significant changes in YTHDF1 (1.08-fold, p = 0.2306; Figure 2(c)) and YTHDC1 (1.00-fold, p = 0.7931; Figure 2(c)) expression in the mouse livers following FA treatment.

Figure 2. The effect of FA on the mRNA expression of m6A regulatory genes in C57BL/6 mouse livers. RNA isolated from control and FA-treated mouse livers were reverse transcribed into cDNA and analysed by qPCR. FA altered the expression of m6A writers (a), erasers (b), and readers (c) in mouse livers compared to the control. Results are presented as the mean ± SEM (n = 6/group). The unpaired t-test with Welch’s correction was used to determine statistical significance, *p < 0.05, **p < 0.005, ***p < 0.001. ALKBH5: Alkylation repair homolog 5; CON: Control; FA: Fusaric acid; FTO: Fat mass and obesity-associated protein; METTL3: Methyltransferase-like 3; METTL14: Methyltransferase-like 14; SEM: Standard error of the mean; WTAP: Wilm’s tumour 1-associated protein; YTHDC1: YT521-B homology domain containing protein 1; YTHDC2: YT521-B homology domain containing protein 2; YTHDF1: YT521-B homology domain family protein 1; YTHDF2: YT521-B homology domain family protein 2; YTHDF3: YT521-B homology domain family protein 3.

Discussion

Dysregulation in the m6A methylation pattern is associated with various diseases, including diabetes and cancer [38–40]. More recently, mycotoxins have also been shown to alter the m6A RNA methylation pattern in cell lines and animal models [18–23]. Aflatoxin B₁ increased global m6A RNA methylation and altered hepatic function leading to cytotoxicity in C57BL/6J mice [20]. In bovine mammary epithelial (BME) cells, aflatoxin B₁ and aflatoxin M₁ decreased global m6A methylation levels [23]. The corneal tissue of Balb/c mice treated with the fungus Fusarium solani, a common producer of T2-toxin, displayed increased global m6A levels and differential m6A methylation of 1,137 mRNAs that may be involved in fungal keratitis [19]. Deoxynivalenol treatment of porcine intestinal epithelial (IPEC-J2) cells induced differential m6A methylation and expression of 733 mRNAs associated with immunological signalling pathways [21]. Fumonisin B₁ increased global m6A RNA methylation in response to oxidative stress [22], and FA decreased p53 RNA m6A levels in the HepG2 cell line [18].

In the present study, we found that FA significantly increased the global level of m6A RNA methylation in mouse livers (Figure 1). This study examined bulk RNA, of which mRNA is only 1–3%, so effects on mRNA adenine methylation may have been far more pronounced. The increase in global m6A RNA levels may have occurred due to the increase in m6A writers (METTL3 and METTL14; Figure 2(a)) and decrease in m6A erasers (FTO and ALKBH5; Figure 2(b)). These results are in agreement with previous studies in which knockdown of METTL3 and/or METTL14 were associated with a substantial decrease in global m6A RNA methylation levels, whereas overexpression of METTL3 and/or METTL14 were associated with an increase in global m6A levels in vivo [41]. In contrast, the knockdown of FTO and ALKBH5 were associated with increased global m6A levels [41,42]. Studies on other mycotoxins showed similar results with fumonisin B₁ inducing increased global m6A RNA methylation, increased expression of METTL3 and METTL14, and decreased expression of FTO and ALKBH5 in HepG2 cells [22]; and aflatoxin B₁ increased global m6A methylation levels and METTL3 expression, and decreased FTO expression in C57BL/6J mouse livers [20].

Although FA increased the expression of METTL3 and METTL14, it decreased the expression of WTAP in the mouse livers (Figure 2(a)). WTAP, a component of the writer complex, lacks catalytic enzymatic activity and mediates nuclear localization of the METTL3-METTL14 complex [4,5]. Apart from its role in m6A RNA methylation, WTAP has been shown to regulate cell proliferation and apoptosis [43,44]. WTAP promotes cell proliferation by stabilizing cyclin A2 and mediating G2/M cell cycle progression [43]. The suppression of WTAP leads to cell cycle arrest in the G2 phase and inhibits cell proliferation [43,44]. The suppression of WTAP also induces apoptosis through increased expression of cleaved caspase-3 [44]. We speculate that the decrease in WTAP is an underlying mechanism for the increased expression of cleaved caspase-3 previously observed to be elicited by FA in C57BL/6 mouse livers [45]. Only one study on mycotoxins evaluated the expression of WTAP; deoxynivalenol decreased WTAP expression, reduced cell viability, and increased apoptosis in IPEC-J2 cells [21].

The overexpression of METTL3 and knockdown of FTO was also shown to increase global m6A levels and activate apoptosis in cisplatin-treated kidney cells [46]. These results were further supported by another study in which upregulation of METTL3 promoted apoptosis in hypoxia/reoxygenation-treated cardiomyocytes [41]. Exposure to aflatoxin B₁ also increased METTL3, decreased FTO, and increased cleaved caspase-3 expression in C57BL/6J mouse livers [20]. In contrast, deoxynivalenol treatment in IPEC-J2 cells decreased METTL3, increased ALKBH5, and showed significantly higher rates of apoptotic cell death [21].

The cellular effects of m6A are mediated by m6A-dependent readers [1,8–10]. We showed that FA increased the expression of m6A readers YTHDF2, YTHDF3, and YTHDC2, while the expression of YTHDF1 and YTHDC1 were unchanged (Figure 2(c)). These findings implied that FA may alter the stability and translation of m6A modified RNA transcripts in mouse livers. This is in keeping with our previous in vitro work, which showed that FA decreased p53 m6A levels and YTHDF1, YTHDF2, YTHDF3, and YTHDC2 expression as well as reduced p53 mRNA and protein expression in HepG2 cells [18]. In another study, the overexpression of YTHDF2 has been shown to suppress cell proliferation and induce apoptosis by destabilizing EGFR mRNA [47]. The overexpression of YTHDF3 and YTHDC2 promoted translation of FOXO3 [48] and HIF-1 alpha [49] mRNA, respectively. The altered expression of YTHDF1, YTHDF2, and YTHDC1 were also shown to be differentially expressed in liver cancer, modulating the expression of numerous cancer promoting genes, and significantly correlating with patient survival time [50]. Similar to our findings in this study, other mycotoxins were also shown to alter the expression of m6A readers with fumonisin B₁ increasing the expression of YTHDF1, YTHDF2, YTHDF3, and YTHDC2 in HepG2 cells [22]; and aflatoxin B₁ increasing the expression of YTHDF1 in BME cells [23]. On the contrary, aflatoxin B₁ exposure decreased YTHDF2 expression in C57BL/6J mouse livers [20]. Further studies are required to determine the effect of altered m6A readers on the fate of specific m6A modified mRNAs following exposure to mycotoxins.

In conclusion, this study demonstrated, for the first time, that FA exposure can modify the liver m6A epitranscriptome and mRNA expression levels of key m6A regulatory genes in C57BL/6 mice. These findings are summarized in Figure 3 and lay the foundation for future studies to determine the molecular mechanisms by which m6A RNA methylation can influence the cellular functions that lead to the adverse health effects following FA exposure.

Figure 3. Summary of results. FA increased global m6A RNA methylation by increasing the expression of m6A writers (METTL3 and METTL14) and decreasing the expression of m6A erasers (FTO and ALKBH5). FA also increased the expression of m6A readers (YTHDF2, YTHDF3, and YTHDC2). The alteration in global m6A RNA methylation and expression of m6A regulatory genes may be an underlying mechanism for FA-induced adverse health effects. ALKBH5: Alkylation repair homolog 5; FTO: Fat mass and obesity-associated protein; METTL3: Methyltransferase-like 3; METTL14: Methyltransferase-like 14; WTAP: Wilm’s tumour 1-associated protein; YTHDC2: YT521-B homology domain containing protein 2; YTHDF2: YT521-B homology domain family protein 2; YTHDF3: YT521-B homology domain family protein 3.

Author contributions

TG and AC conceptualized and designed the study. TG conducted all laboratory experiments, analysed the data, and wrote the manuscript. SN and AC revised the manuscript. All authors have read the manuscript prior to submission.

Ethics approval

Ethics approval was obtained from the Institutional Animal Research Ethics Committee at the University of KwaZulu-Natal and were in compliance with the ARRIVE guidelines.

Acknowledgments

We thank the Africa Health Research Institute for allowing us to use their animal housing facilities. We thank Dr. S.D. Singh and Dr. S. Baijnath for maintaining, treating, and sacrificing the mice. We also thank Dr. S. Dhani for her assistance in transporting the mice tissue.

Disclosure statement

The authors report no conflict of interest.

Data availability

All datasets used in this study are available from the corresponding author on reasonable request.

Additional information

Funding

This work was supported by the National Research Foundation under grant 120716.