Effect of Humantenine on mRNA m6A Modification and Expression in Human Colon Cancer Cell Line HCT116

Humantenine, an alkaloid isolated from the medicinal herb Gelsemium elegans (Gardner & Chapm.) Benth., has been reported to induce intestinal irritation, but the underlying toxicological mechanisms remain unclear. The object of the present study was to investigate the RNA N6-methyladenosine (m6A) modification and distinct mRNA transcriptome profiles in humantenine-treated HCT116 human colon cancer cells. High-throughput MeRIP-seq and mRNA-seq were performed, and bioinformatic analysis was performed to reveal the role of abnormal RNA m6A modification and mRNA expression in humantenine-induced intestinal cell toxicity. After humantenine treatment of HCT116 cells, 1401 genes were in the overlap of differentially m6A-modified mRNA and differentially expressed mRNA. The Kyoto Encyclopedia of Genes and Genomes and Gene Ontology annotation terms for actin cytoskeleton, tight junctions, and adherens junctions were enriched. A total of 11 kinds of RNA m6A methylation regulators were differentially expressed. The m6A methylation levels of target genes were disordered in the humantenine group. In conclusion, this study suggested that the HCT116 cell injury induced by humantenine was associated with the abnormal mRNA expression of m6A regulators, as well as disordered m6A methylation levels of target genes.


Introduction
Gelsemium elegans (Gardner & Chapm.) Benth. (G. elegans), an evergreen woody vine belonging to the family Loganiaceae, is a traditional medicinal herb with a drug use history of over 1800 years [1]. It is widely distributed in Southeast Asia, especially in Hunan, Guangdong, Fujian, and Yunnan provinces of China. In traditional Chinese medicine, G. elegans is used for the treatment of neuropathic pain, rheumatoid pain, swelling and skin ulcers, and also exerts anti-inflammatory, anti-tumor, anti-anxiety, and immunomodulatory effects [2]. However, G. elegans has been reported to be associated with severe gastrointestinal toxicity, such as vomiting, abdominal pain, diarrhea, or constipation, and even to induce symptoms of the nervous, circulatory, and respiratory systems, such as dizziness, drops in blood pressure, and respiratory paralysis [3]. To date, more than 2000 poisoning deaths have been reported in China [4]. According to the statistics from the Chinese Center for

Cell Culture and Treatment
Human colon cancer cell line HCT116 was originally purchased from American Type Culture Collection. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM, GE Healthcare Hyclone, Logan, UT, USA) supplemented with 10% fetal bovine serum (Gibco, New York, NY, USA), 100 μg/mL streptomycin sulfate and 100 U/mL penicillin G sodium salt, and then maintained in an incubator (Thermo Fisher, Langenselbold, Germany) with an atmosphere of 5% CO2 at 37 °C. Humantenine was dissolved in dimethyl sulfoxide (DMSO) to a concentration of 400 μM. The final working concentration of DMSO in cell culture experiments was 1%.

Cell Culture and Treatment
Human colon cancer cell line HCT116 was originally purchased from American Type Culture Collection. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM, GE Healthcare Hyclone, Logan, UT, USA) supplemented with 10% fetal bovine serum (Gibco, New York, NY, USA), 100 µg/mL streptomycin sulfate and 100 U/mL penicillin G sodium salt, and then maintained in an incubator (Thermo Fisher, Langenselbold, Germany) with an atmosphere of 5% CO 2 at 37 • C. Humantenine was dissolved in dimethyl sulfoxide (DMSO) to a concentration of 400 µM. The final working concentration of DMSO in cell culture experiments was 1%.

RNA Preparation
The cells were cultured in 9 cm dishes. After humanteinine treatment for 48 h, the cells were washed twice in PBS, and Trizol reagent (Invitrogen, Carlsbad, CA, USA) was added to lyse cells. Total RNA was extracted, and the concentration and integrity were measured via Qubit RNA HS Assay and Agilent 2100 Bioanalyzer (Santa Clara, CA, USA), respectively.

High-Throughput m6A MeRIP-seq and mRNA-seq
The m6A-modified RNA enrichment and sequencing was performed by E-GENE Tech Co., Ltd. (Shenzhen, China). Firstly, we used the Dynabeads™ mRNA Purification Kit (Invitrogen) to separate mRNA from the 44 µg total RNA. Subsequently, 1× RNA Fragmentation Buffer (Thermo Fisher, Langenselbold, Germany) was used at 70 • C for 5 min to turn purified mRNA into fragments. After purification, 5 µg of m6A antibody (Synaptic Systems, Goettingen, Germany) that had been incubated with 25 µL of protein A/G magnetic beads mixture (Thermo Fisher, Langenselbold, Germany) was taken out and used to enrich RNA. The incubation was then rotated at 4 • C for 4 h. After immunoprecipitation, the beads were washed three times with IP buffer (150 mM NaCl, 10 mM Tris-HCl, 0.1% IGEPAL CA-630 in nuclease-free H 2 O) for 10 min each time, rotating at 4 • C. Finally, m6Amodified mRNA fragments were extracted with Trizol reagent and purified using an RNA Clean & Concentrator Kit (Zymo Research, Orange, CA, USA). The purified m6A antibodyenriched RNA and some unenriched mRNA fragments were used as input to construct libraries with the VAHTS mRNA-seq V3 Library Prep Kit for Illumina (Vazyme, Nanjing, China) according to the manufacturer's protocol. The established libraries were analyzed using an Agilent 2100 Bioanalyzer, quantified by real-time PCR, and then sequenced on the NovaSeq 6000 platform (Illumina, San Diego, CA, USA). Three independent samples were used in each group of m6A MeRIP-seq and mRNA-seq.

Sequencing Data Analysis
The raw data were processed by Trim Galore (Cambridge, UK). The sequencing results were aligned with the human genome reference hg19 using Hisat2 [25], the RNA expression level was analyzed using StringTie (Baltimore, MD, USA) [26], and differential expression was calculated using DEseq [27]. The exomepeak2 (Suzhou, China) [28,29] was applied for m6A peak calling and the detection of differential methylation. In exomepeak2, the Poisson Generalized Linear Model was used to estimate the methylation level and detect differential methylation regions. The exomePeak2 s estimation of sequencing depth size factors was on non-methylation background regions. The consensus of m6A motif sequences was identified by STREME [30]. The STREME algorithm integrated the position weight matrix Markov model to report a useful estimate of the statistical significance of each motif it discovered. The Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) annotations were done by KOBAS (Beijing, China) [31] and DAVID (Frederick, MD, USA) [32], respectively. The distribution of epitranscriptome profiles was visualized by MetaTX (Suzhou, China) [33]. The m6A conservation and disease association were obtained from ConsRM (Suzhou, China) [34] and RMDisease (Suzhou, China) [35]; the substrates of m6A regulators identified by CLIP techniques were downloaded from starBase v2.0 (Guangzhou, China) [36]. The m6A patterns of gene methylated sites were visualized using Integrative Genomics Viewer (IGV) software. Protein-protein interaction (PPI) analysis was performed using STRING and visualized using Cytoscape.

Molecular Docking
To explore whether humantenine could interact with differentially expressed regulators of m6A modification, the SYBYL-X 2.0 software (Tripos, St Louis, MO, USA) was applied to perform molecular docking. The 3D structures of proteins were retrieved from the Protein Data Bank (PDB). The protein structures were prepared using SYBYL-X 2.0 to remove water molecules and heteroatoms, add hydrogen atoms, and repair side chains [37]. The 2D structure of humantenine was downloaded from the PubChem compound database and was imported into Chem3D (Waltham, MA, USA) to generate 3D structures according to the energy minimization principle. After being stored in mol2 format, the 3D structure of humantenine was set to the lowest energy conformation to mimic the stable molecular conformation in the natural system [38]. Proteins were docked in Surflex-Dock GeomX mode through the semi-flexible docking method. The total score, which is a comprehensive evaluation of hydrophobic complementarity, polar complementarity, solvation terms, and entropic terms, was deemed a stable interaction when the value was higher than 5 [39].

Transcriptome-Wide Detection of m6A Modification after Humantenine Treatment of HCT116 Cells
To explore the mechanism of action of humantenine in HCT116 human colon cancer cells, MeRIP-seq analysis and RNA-seq were performed. In the humantenine group, 94,863 m6A peaks containing transcripts of 13,577 genes were identified by the R package exomePeak. Similarly, 105,690 m6A peaks were identified in the control group, representing transcripts of 13,293 genes. In addition, 82,592 peaks corresponding to 12,895 genes were found at the intersection of the humantenine and control groups (Figure 2A,B). We then used STREME to determine whether the m6A consensus sequence of RRACH (where R represents purine, A is m6A, and H is a non-guanine base) was reported in the detection of m6A ( Figure 2C). The result showed the existence of some motifs, including the classical consensus sequence in the control and humantenine groups. Notably, the majority of genes had 1-3 m6A methylation peaks, whereas a relatively small number of genes contained four or more m6A methylation peaks in both humantenine and control groups ( Figure 2D). database and was imported into Chem3D (Waltham, MA, USA) to generate 3D structures according to the energy minimization principle. After being stored in mol2 format, the 3D structure of humantenine was set to the lowest energy conformation to mimic the stable molecular conformation in the natural system [38]. Proteins were docked in Surflex-Dock GeomX mode through the semi-flexible docking method. The total score, which is a comprehensive evaluation of hydrophobic complementarity, polar complementarity, solvation terms, and entropic terms, was deemed a stable interaction when the value was higher than 5 [39].

Transcriptome-Wide Detection of m6A Modification after Humantenine Treatment of HCT116 Cells
To explore the mechanism of action of humantenine in HCT116 human colon cancer cells, MeRIP-seq analysis and RNA-seq were performed. In the humantenine group, 94,863 m6A peaks containing transcripts of 13,577 genes were identified by the R package exomePeak. Similarly, 105,690 m6A peaks were identified in the control group, representing transcripts of 13,293 genes. In addition, 82,592 peaks corresponding to 12,895 genes were found at the intersection of the humantenine and control groups (Figure 2A,B). We then used STREME to determine whether the m6A consensus sequence of RRACH (where R represents purine, A is m6A, and H is a non-guanine base) was reported in the detection of m6A ( Figure 2C). The result showed the existence of some motifs, including the classical consensus sequence in the control and humantenine groups. Notably, the majority of genes had 1-3 m6A methylation peaks, whereas a relatively small number of genes contained four or more m6A methylation peaks in both humantenine and control groups ( Figure 2D).

Distribution of m6A Modification in the Transcriptome
The distributions of m6A methylation in the whole transcriptomes of the humantenine and control groups were analyzed. The result showed that m6A modification tended to enrich in five transcript segments: 5 untranslated region (5 UTR), the start codon segment, coding sequence (CDS), the stop codon segment, and 3 UTR. The m6A peak density increased rapidly between the 5 UTR and the start codon and was relatively gentle in the CDS region. The highest density region was near the stop codon. In the 3 UTR region, the density of the m6A peaks decreased rapidly ( Figure 3). nine and control groups were analyzed. The result showed that m6A modification tende to enrich in five transcript segments: 5′untranslated region (5′UTR), the start codon seg ment, coding sequence (CDS), the stop codon segment, and 3′UTR. The m6A peak densit increased rapidly between the 5′UTR and the start codon and was relatively gentle in th CDS region. The highest density region was near the stop codon. In the 3′UTR region, th density of the m6A peaks decreased rapidly ( Figure 3).

Differentially Methylated Genes and Differentially Expressed Genes
Setting the statistical standard as p ≤ 0.05, a total of 4516 differential m6A-modifie mRNAs were found through m6A-seq data, and 3430 mRNAs were determined to be dif ferentially expressed between humantenine and control groups after the analysis of RNA seq data. In addition, 1401 genes were consistently observed at the overlap of differentiall m6A-modified mRNA and differentially expressed mRNA ( Figure 4A). When the statis tical standard was set as fold change ≥ 1 and p ≤ 0.05, 681 differentially expressed mRNA were screened out, of which 433 were up-regulated and 248 were down-regulated (Figur 4B).

KEGG and GO Annotation of the Overlap of Differentially m6A-Modified mRNA and Differentially Expressed mRNA
Using KEGG pathway and GO enrichment analysis on the DAVID web server, th 1401 overlapping genes of differentially m6A-modified mRNA and differentially ex pressed mRNAs were associated with significant pathways and biological functions. Th

Differentially Methylated Genes and Differentially Expressed Genes
Setting the statistical standard as p ≤ 0.05, a total of 4516 differential m6A-modified mRNAs were found through m6A-seq data, and 3430 mRNAs were determined to be differentially expressed between humantenine and control groups after the analysis of RNA-seq data. In addition, 1401 genes were consistently observed at the overlap of differentially m6A-modified mRNA and differentially expressed mRNA ( Figure 4A). When the statistical standard was set as fold change ≥ 1 and p ≤ 0.05, 681 differentially expressed mRNAs were screened out, of which 248 were up-regulated and 433 were down-regulated ( Figure 4B).

Distribution of m6A Modification in the Transcriptome
The distributions of m6A methylation in the whole transcriptomes of the humantenine and control groups were analyzed. The result showed that m6A modification tended to enrich in five transcript segments: 5′untranslated region (5′UTR), the start codon segment, coding sequence (CDS), the stop codon segment, and 3′UTR. The m6A peak density increased rapidly between the 5′UTR and the start codon and was relatively gentle in the CDS region. The highest density region was near the stop codon. In the 3′UTR region, the density of the m6A peaks decreased rapidly ( Figure 3).

Differentially Methylated Genes and Differentially Expressed Genes
Setting the statistical standard as p ≤ 0.05, a total of 4516 differential m6A-modified mRNAs were found through m6A-seq data, and 3430 mRNAs were determined to be differentially expressed between humantenine and control groups after the analysis of RNAseq data. In addition, 1401 genes were consistently observed at the overlap of differentially m6A-modified mRNA and differentially expressed mRNA ( Figure 4A). When the statistical standard was set as fold change ≥ 1 and p ≤ 0.05, 681 differentially expressed mRNAs were screened out, of which 433 were up-regulated and 248 were down-regulated ( Figure  4B).

KEGG and GO Annotation of the Overlap of Differentially m6A-Modified mRNA and Differentially Expressed mRNA
Using KEGG pathway and GO enrichment analysis on the DAVID web server, the 1401 overlapping genes of differentially m6A-modified mRNA and differentially expressed mRNAs were associated with significant pathways and biological functions. The

KEGG and GO Annotation of the Overlap of Differentially m6A-Modified mRNA and Differentially Expressed mRNA
Using KEGG pathway and GO enrichment analysis on the DAVID web server, the 1401 overlapping genes of differentially m6A-modified mRNA and differentially expressed mRNAs were associated with significant pathways and biological functions. The KEGG pathway analysis revealed that these genes were mainly involved in tight junction, regulation of actin cytoskeleton, metabolic pathways, focal adhesion, and adherens junction pathways, and so on ( Figure 5). The GO enrichment analysis was classified into three functional types: molecular function (MF), cellular component (CC), and biological process (BP). The MF terms included actin filament binding and actin binding. The CC terms included zonula adherens, cytoskeleton, adherens junction, and actin cytoskeleton. The BP terms included actin cytoskeleton organization ( Figure 6).
KEGG pathway analysis revealed that these genes were mainly involved in tight junction, regulation of actin cytoskeleton, metabolic pathways, focal adhesion, and adherens junction pathways, and so on ( Figure 5). The GO enrichment analysis was classified into three functional types: molecular function (MF), cellular component (CC), and biological process (BP). The MF terms included actin filament binding and actin binding. The CC terms included zonula adherens, cytoskeleton, adherens junction, and actin cytoskeleton. The BP terms included actin cytoskeleton organization ( Figure 6).  tion pathways, and so on ( Figure 5). The GO enrichment analysis was classified into thre functional types: molecular function (MF), cellular component (CC), and biological pro cess (BP). The MF terms included actin filament binding and actin binding. The CC term included zonula adherens, cytoskeleton, adherens junction, and actin cytoskeleton. Th BP terms included actin cytoskeleton organization ( Figure 6).

The m6A Conservation and Disease Association
As shown in Figure 7A, based on the analysis of ConsRM, 86.7% of m6A-modified sites were nonconservative in the 1401 overlapping genes of differentially m6A-modified mRNAs and differentially expressed mRNAs. As shown in Figure 7B, the results of the RMDisease analysis showed that 13.6% of m6A genes and 10.5% of m6A peaks were associated with diseases.

The m6A Conservation and Disease Association
As shown in Figure 7A, based on the analysis of ConsRM, 86.7% of m6A-modified sites were nonconservative in the 1401 overlapping genes of differentially m6A-modified mRNAs and differentially expressed mRNAs. As shown in Figure 7B, the results of the RMDisease analysis showed that 13.6% of m6A genes and 10.5% of m6A peaks were as sociated with diseases.

Gene Expression of Tight Junctions, Adherens Junctions, and Regulation of Actin Cytoskeleton, and Their Potential Regulators
According to the RNA-seq data in HCT116 cells, after treatment with humantenine the mRNA expression levels of genes associated with tight junctions (CLDN4, CDK4, TJP3 MAGI1), adherens junctions (ACTN1, NECTIN2, WASF1), and regulation of actin cyto skeleton (MYH9, IQGAP3, DIAPH3, WASF1, ARPC5L, DOCK1) were significantly differ entially expressed ( Table 1). The potential regulators, including parts of writers, readers and erasers, were identified using CLIP techniques. The results showed that the reade IFG2BP3 was involved in the regulation of all these genes.

Gene Expression of Tight Junctions, Adherens Junctions, and Regulation of Actin Cytoskeleton, and Their Potential Regulators
According to the RNA-seq data in HCT116 cells, after treatment with humantenine, the mRNA expression levels of genes associated with tight junctions (CLDN4, CDK4, TJP3, MAGI1), adherens junctions (ACTN1, NECTIN2, WASF1), and regulation of actin cytoskeleton (MYH9, IQGAP3, DIAPH3, WASF1, ARPC5L, DOCK1) were significantly differentially expressed ( Table 1). The potential regulators, including parts of writers, readers, and erasers, were identified using CLIP techniques. The results showed that the reader IGF2BP3 was involved in the regulation of all these genes.
Y means the mRNA was the substrate of the RNA m6A regulators, and N means not.

Potential RNA m6A Regulators of Differentially Methylated Genes
In order to identify the potential regulators of RNA m6A methylation, we analyzed the mRNA expression levels of 22 RNA m6A methylation writers, readers, and erasers. As shown in Table 2, four writers, six readers, and one eraser were significantly differentially expressed (padj ≤ 0.05). Among them, writers RBM15 and METTL3, reader YTHDF3, and eraser ALKBH5 were up-regulated; writers RBM15B and ZC3H13, and readers HNRNPA2B1, YTHDF2, YTHDC2, IGF2BP3, and YTHDC1 were down-regulated. Since the GO and KEGG terms included tight junctions, adherens junctions, and actin cytoskeleton, as shown in Figures 5 and 6, the genes enriched in these terms were collected. We then analyzed the regulatory relationship between these genes and 11 differentially expressed RNA m6A methylation regulators in a PPI network ( Figure 8A). The degrees of the 11 regulators ALKBH5, HNRNPA2B1, YTHDF2, YTHDC2, IGF2BP3, RBM15B, YTHDF3, YTHDC1, RBM15, METTL3, and ZC3H13 were 9, 11, 10, 9, 3, 9, 9, 9, 9, 11, and 9, respectively. To explore the substrates of RNA m6A methylation regulators, the writer METTL3, readers HNRNPA2B1, IGF2BP3, YTHDC1, YTHDC2, and YTHDF2, and eraser ALKBH5 were selected, as well as the actin cytoskeleton genes IQGAP3, DIAPH3, DOCK1, WASF1, and ARPC5L. The result showed that insulin-like growth factor 2 mRNA-binding protein 3 (IGF2BP3) was involved in the regulation of all five of these actin cytoskeleton genes ( Figure 8B). As indicated by the IGV display of IGF2BP3, the m6A methylation level on mRNA transcripts was decreased after humantenine treatment ( Figure 8C).

Molecular Interactions of Humantenine with Differently Expressed Regulators of RNA m6A Modification
We evaluated the molecular interactions of humantenine with differentially expressed regulators of m6A modification, including ALKBH5, HNRNPA2B1, IGF2BP3, METTL3, YTHDC1, YTHDC2, and YTHDF2. The results are shown in Figure 9 and Table 3. All the total scores were over 5, which suggested stabilized interactions between humantenine and these regulators.

Discussion
The traditional medicine G. elegans exerts excellent pharmacological therapeutic effects in the treatment of a variety of clinical symptoms, especially neuralgia and inflammation. On the other hand, the severe side effects in multiple systems limit its wider application. Therefore, it is in urgent need to reveal the molecular mechanism of G. elegans-induced intestinal irritation, for the clinic drug safety. In the present study, the indole alkaloid humantenine was used in human intestinal cells to analyze the potential epigenetic alteration of RNA m6A modification, and the mRNA transcriptional profile, based on high-throughput m6A MeRIP-seq and mRNA-seq.
Through analysis of m6A MeRIP-seq data, the motif sequence was successfully identified with statistical significance in the region around the m6A peak of the control group and humantenine group. The consensus motif sequence RRACH was observed in the m6A motif region, as proven previously [40,41]. As to the distribution of m6A modification peaks, most genes had 1~3 peaks, and parts of genes contained more peaks, which was consistent with the study by Chen et al. [42]. When the genes were divided into 5 UTR, start codon, CDS, stop codon, and 3 UTR, most m6A modification peaks were in the CDS region, and the highest value was around the stop codon, as frequently reported in the past [40,43]. A higher number of m6A modification sites at the stop codon may be associated with mRNA stability and translation [44].
In the present study, humantenine treatment induced 86.7% RNA m6A modification in un-conserved sites, and 13.6% of genes were associated with diseases, suggesting that this xenobiotic compound might cause intestinal diseases. To explore the biological function change induced by RNA m6A modification, GO and KEGG pathway enrichment analysis was performed. After humantenine treatment in HCT116, a total of 1041 genes was reported in the overlap of the differentially m6A-modified mRNA and differentially expressed mRNA. The top 20 KEGG terms included tight junction, regulation of actin cytoskeleton, and adherens junction. In the GO terms, cadherin binding, actin filament binding, and actin binding were enriched in MF; zonula adherens, focal adhesion, cytoskeleton, adherens junction, and actin cytoskeleton were enriched in CC; actin cytoskeleton organization was enriched in BP. In clinic, diarrhea is the main intestinal side effect of G. elegans, manifested by increased intestinal permeability, which is regulated by intestinal intercellular junction and actin cytoskeleton. The intercellular junction mainly comprises tight junctions and adherens junctions, and acts as the mechanical barrier in defense against exogenous toxic compounds [45]. It also regulates the paracellular transport pathway to maintain intestinal cell integrity [46,47]. Kumar et al. found that Cryptosporidium parvum disrupted intestinal barrier function by downregulating the expression of tight junctions and adherens junctions, resulting in diarrhea [48]. In this study, expression levels of multiple mRNAs involved in tight junctions, adherens junctions, and regulation of actin cytoskeleton were significantly differentially expressed in the humantenine-treated group. These genes encoded proteins of claudin 4, zonula occludens protein 3, actinin α 1, nectin cell adhesion molecule 2, Wiskott-Aldrich syndrome protein family member 1, myosin heavy chain 9, IQ motif containing GTPase activating protein 3 (IQGAP3), diaphanous related formin 3, wasp family member 1, actin related protein 2/3 complex subunit 5, dedicator of cytokinesis 1, and so on. If the junction proteins and actin cytoskeleton were destroyed by humantenine, the paracellular pathway transport would be in disorder, characterized by abnormal osmotic pressure and permeability in intestinal cells [49,50].
RNA m6A modifications were mediated by three specific categories of proteins, namely m6A methyltransferases, m6A demethylases, and m6A-biding proteins. The biological effects of m6A methylation modifications depend on the recognition and binding of m6A-binding proteins [51]. Using the computational prediction method, humantenine was combined with the m6A regulators with total scores higher than 5. Therefore, the present study further tried to identify the m6A regulators of the differentially expressed genes related to intercellular junctions and actin cytoskeleton. A total of 11 m6A regulators were differentially expressed, of which IGF2BP3, an m6A reader exerting the main function of enhancing the stability of mRNA [52], was regarded as the core factor in the regulation of mRNA translation, decay, and storage. For example, IQGAP3 is a Ras GTPase-activatinglike protein that belongs to the family of scaffolding proteins and regulates cell adhesion by mediating the actin cytoskeleton [53]. The m6A methylated mRNA of IQGAP3 can be recognized by IGF2BP3, thus maintaining a stable state in nuclear and cytoplasmic conditions. As downstream biological effects, the translation and storage of IQGAP3 mRNA were promoted and the decay of IQGAP3 mRNA was inhibited. In this study, humantenine inhibited the mRNA m6A methylation level of IQGAP3, making it difficult for IGF2BP3 to recognize and bind IQGAP3. As a result, the mRNA stability and expression level of IQGAP3 decreased, which would cause the actin cytoskeleton to collapse and thus enhance intestinal permeability.

Conclusions
In summary, the present study investigated the mRNA m6A modification and expression profiles of HCT116 human colon cancer cells treated with humantenine. The abnormal mRNA expression of m6A regulators, together with disordered m6A methylation levels of target genes, led to the disruption of mRNA stability, translation, storage, and decay ( Figure 10). This study has therefore revealed a possible mechanism for the increased intestinal permeability induced by humantenine.
Genes 2022, 13, x FOR PEER REVIEW 14 of 17 cytoplasmic conditions. As downstream biological effects, the translation and storage of IQGAP3 mRNA were promoted and the decay of IQGAP3 mRNA was inhibited. In this study, humantenine inhibited the mRNA m6A methylation level of IQGAP3, making it difficult for IGF2BP3 to recognize and bind IQGAP3. As a result, the mRNA stability and expression level of IQGAP3 decreased, which would cause the actin cytoskeleton to collapse and thus enhance intestinal permeability.

Conclusions
In summary, the present study investigated the mRNA m6A modification and expression profiles of HCT116 human colon cancer cells treated with humantenine. The abnormal mRNA expression of m6A regulators, together with disordered m6A methylation levels of target genes, led to the disruption of mRNA stability, translation, storage, and decay ( Figure 10). This study has therefore revealed a possible mechanism for the increased intestinal permeability induced by humantenine.