Tissue-specific mRNA m6A reprogramming unveils vitamin-driven post-transcriptional regulation in mice.

Under fluctuating conditions, plants have evolved complex systems for sensing environmental changes to maximize reproductive success. By utilizing seasonal cues, such as day length and temperatures, plants optimize the timing of flowering to adapt to local environments (Andrés & Coupland, 2012; He et al., 2020). This regulatory network involves many pathways with numerous genes (Srikanth & Schmid, 2011; Bouché et al., 2016). In the network, FIONA1 (FIO1), an RNA methyltransferase, has been identified as a regulator of the circadian rhythm and photoperiod pathways in Arabidopsis thaliana (Kim et al., 2008; Yeom et al., 2009). The natural genetic variation of FIO1 is associated with flowering time, highlighting the significant roles of FIO1 in plant adaptability (Sasaki et al., 2015). FIO1 is a homolog of vertebrate METHYLTRANSFERASE-LIKE 16 (METTL16) and Caenorhabditis elegans mett-10 and is broadly conserved from plants to animals (Pendleton et al., 2017; Mendel et al., 2021). In plants, while the precise mode of action of FIO1 remained unclear more than 10 years after the first reports (Kim et al., 2008; Yeom et al., 2009), several independent research articles on its molecular function were published in succession in 2022 (Cai et al., 2022; Parker et al., 2022; Sun et al., 2022; Wang et al., 2022; Xu et al., 2022). These studies proposed the primary molecular function of FIO1 can involve the regulation of alternative splicing through N6-methyladenosine (m6A) modification on U6 snRNA and the posttranscription regulation by methylation of m6A motifs on mRNA (Cai et al., 2022; Parker et al., 2022; Sun et al., 2022; Wang et al., 2022; Xu et al., 2022). Interestingly, their conclusions are not always consistent, even though they employed the same approach, that is RNA-Seq analyses of young seedlings in FIO1 loss-of-function mutants (fio1). Parker et al. (2022) reported that FIO1 affects flowering time through altering splice site selections by m6A modification in U6 snRNA. By contrast, Wang et al. (2022) and Xu et al. (2022) disagreed with the global effect of FIO1 on alternative splicing. They proposed that FIO1 methylates adenosine residues in specific mRNA motifs to form m6A, which affects mRNA stability and controls flowering time through changes in mRNA abundance. Cai et al. (2022) and Sun et al. (2022) supported the view that FIO1 has RNA m6A-methyltransferase activity on specific genes, including a major flowering repressor, FLOWERING LOCUS C (FLC), in the vernalization pathway, while they have not deeply investigated genome-wide alternative splicing. These inconsistencies in their conclusions have left the molecular basis of FIO1 functions unclear. What factors cause these different arguments? What are the targets of FIO1 in controlling flowering time? To address these questions, we revisited this issue by re-analyzing the datasets they published. Specifically, we concentrated on the role of FIO1 in alternative splicing, where there is a disagreement of point of view between Parker et al. (2022) and Wang et al. (2022), which analyze deep sequencing data for short-read RNA-Seq. First, we considered that the different conclusions for FIO1 function might be derived from variations in experimental conditions and data analysis pipelines rather than biological factors. To test this, we assessed the four previously published short-read RNA-Seq datasets (Cai et al., 2022; Parker et al., 2022; Sun et al., 2022; Wang et al., 2022), consisting of FIO1 loss-of-function mutants and the wild-type (WT) Col-0, using identical data analysis pipelines for read filtering, mapping, and counting transcripts (Table 1; Supporting Information Methods S1). Despite differences in transgenic line genotypes, growth temperatures, and sampling stages across studies, multidimensional scaling (MDS) clustering analysis revealed a distinct cluster of fio1-specific genome-wide expression patterns separated from WT, once batch effects attributed to each experimental condition were removed (Fig. S1a,b). This result confirmed that all four studies reflected the same phenotypes resulting from the lack of FIO1. Next, we assessed the performance of data analysis pipelines for differential splicing (DS) detection. Different tools often exhibit low concordances of DS genes, and the performances of the tools tend to vary depending on organisms and experimental conditions (Mehmood et al., 2020; Jiang et al., 2023). Differential splicing detection tools can be categorized into two major methods: exon-based and event-based. Exon-based tools identify differentially expressed exons between samples by counting reads for each exon, indirectly inferring DS events. Event-based tools directly quantify the frequency of each splicing event from reads covering splice sites as their Percentage Spliced In (PSI) values (Venables et al., 2008). We assessed two widely-used exon-based tools, DEXSeq (Anders et al., 2012) and edgeR (Robinson et al., 2010), and two event-based tools, rMATS (Shen et al., 2014) and SUPPA2 (Trincado et al., 2018). We compared the performance considering recall rates (proportion of true positive DS in all true DS) and precision rates (proportion of true positive DS in all detected DS) using RNA-Seq reads simulated by ASimulatoR (Manz et al., 2021) from the A. thaliana TAIR 10 genome (Methods S1). The exon-based and event-based tools showed distinct patterns. Exon-based tools, DEXSeq and edgeR, showed higher recall rates than event-based tools, rMATS and SUPPA2, indicating higher sensitivity, while event-based tools showed average 57.2% pt higher precision rates than the exon-based tools, suggesting more accurate results but with higher false-negative rates (Fig. 1a). Even within the same exon-based or event-based tools, there were differences of 0.13 to 0.20 in recall rates. We comprehensively assessed the four published datasets – Cai et al. (2022) as study A, Parker et al. (2022) as study B, Sun et al. (2022) as study C, and Wang et al. (2022) as study D –using these four tools (Tables 1, S1). Overall, the number of detected DS events for each tool aligns with the recall rates in our simulation. Exon-based tools typically identified more than double DS events (study A, DEXSeq: 2818; edgeR: 1331) compared to the event-based tools (study A, rMATS: 173; SUPPA2: 590) although with greater difference than in the simulation (Fig. 1b). Notably, SUPPA2, an event-based tool, identified over double the DS events compared to rMATS. The genome-wide screening of DS events in previous studies has employed different event-based tools: SUPPA2 in study B and rMATS in study D (Table S1). These studies reported 3502 and 38 DS genes in their original results, respectively, and there were only 28 DS genes overlapping. This difference in data analysis pipelines used for the original studies may explain the different conclusions between two studies B and D. In addition, when we compared the number of DS events across studies using any tools, study B consistently provided the most DS events, while study D had the fewest in three out of four tools (Fig. 1b; Tables S2–S5). Therefore, we assessed the results based on the overlap rates of DS events identified in each study and the Jaccard index (Jaccard, 1901), which means similarity between DS event lists. As indicated by the simulation (Fig. 1a), there exists a trade-off between recall rate (sensitivity) and precision rates (accuracy). Between the three datasets from studies A, B, and C, both event-based tools showed relatively higher overlap rates (rMATS: 0.42–0.62; SUPPA2: 0.48–0.62) and the Jaccard index (rMATS: 0.26–0.31; SUPPA2: 0.29–0.36) than an exon-based tool, DEXSeq (overlap rate: 0.29–0.52; Jaccard index: 0.08–0.10), while detected DS events were <50% of the DS events identified by DEXSeq (Figs 1b,c, S2). However, regardless of the data analysis pipelines, study D consistently showed the lowest scores compared to other datasets (Figs 1b,c, S2, S3). These suggest that while SUPPA2 and rMATS, originally used in studies B and D, provided robust results in DS event detection in A. thaliana, the common DS events detected in study D show a significantly different pattern from the other three studies (P-value = 2.7E-06; Methods S1). The factor making study D an outlier is likely a combination of several factors. We assessed the impact of data quality on DS event detections by separately examining three parameters: (1) experimental replicates; (2) sequencing depth; and (3) read length through downsampling RNA-Seq data. The downsampled data were generated from RNA-Seq reads in study B (six replicates, average 71.7 M reads per replicate) to two-thirds or one-third of the original data. According to the reduction of all three parameter values, the number of detected DS events and recall rates consistently decreased, with the exception of experimental replicates in SUPPA2 (Fig. S4). However, in SUPPA2, fewer replicates reduced the accuracy of DS detection, leading to the decreased overlap of detected DS events with the original data. Regarding study D, the dataset had only two replicates (Table 1), partially accounting for its few DS event detection and low overlap rates with other studies (Figs 1b,c, S2, S3). Additionally, we noticed that dataset D had a higher percentage of adapter content in the raw reads near the 3′ ends (Fig. S5a). Adapter trimming shortened the read length, and the peak in read length appeared c. 100 bp (Fig. S5b) though the original length was 150 bp (Table 1). Based on estimation from conditions of dataset D (two replicates, c. 45 million reads as 2/3 reads number of dataset D, 100 bp read length after adaptor trimmed, analyzed by rMATS) using the downsampled dataset B, the detected DS events and recalled DS events compared with the performance of the original dataset B decreased dramatically to 50% and 39%, respectively. Such a limited data size can lead to the underestimation of DS events. Moreover, we investigated DS events that consistently changed in more than one study in fio1 by employing robust rank aggregation (Kolde et al., 2012) and identified a total of 1145 DS genes across the four tools (Table S6). The base frequencies of the 5′ splice site (5′SS) in these consistent DS genes showed enrichment of specific motifs, specifically//GURAG (R = A or G, alternative 5′SS: P-value = 0.031, alternative 3′SS: P-value = 0.018, retention intron: P-value = 0.018; Fig. S6). Since the ACAGA box of U6 snRNA recognizes the //GURAG motifs at 5′SS (Sawa & Abelson, 1992; Ishigami et al., 2021), the loss of FIO1 homologs, METTL16, and mett-10, which leads to hypomethylation of A at the third base in the ACAGA box of U6 snRNA, disrupts splicing with the A at the +4 position in the //GURAG motif at 5′SSs (Ishigami et al., 2021; Shen et al., 2023). These potential targets of FIO1, with A at the +4 position of the //GURAG motif at 5′SSs, were found in 60.9% and 39.5% of the DS events in fio1 detected by rMATS and SUPPA2, respectively (Table S6). Thus, our meta-analysis supports the role of FIO1 in U6 snRNA-mediated 5′SS selection, as proposed by Parker et al. (2022). In summary, the inconsistent viewpoint between Wang et al. (2022), study D, and Parker et al. (2022) study B, can be attributed to multiple factors: (1) differences in the performance of DS detection tools (Fig. 1a,b); (2) data quality including the number of replicates (Fig. S4); differential splicing (DS) and (3) the adapter content (Fig. S5). In addition, as a part of the DS detection workflow, the difference in splice site annotation used for the analysis may also result in the underestimation of de novo DS events in study D (Parker et al., 2022). We concluded that FIO1 globally affects alternative splicing events. Finally, we explored DS genes influencing flowering time phenotypes through FIO1 across studies A–D. Parker et al. (2022) observed fio1 disrupted splicing patterns of the MADS-box genes, FLOWERING LOCUS M (FLM), MADS AFFECTING FLOWERING 2 (MAF2), and MADS AFFECTING FLOWERING 3 (MAF3), which are temperature-sensitive flowering repressors. Functional products of FLM and MAF2 inhibit the expression of the flowering integrator FLOWERING LOCUS T (FT) under low temperatures, while an increase in nonfunctional transcripts results in early flowering under high temperatures (Lee et al., 2013; Posé et al., 2013; Rosloski et al., 2013). MAF3 also represses floral activators at low temperatures under the control of circadian rhythms (Gu et al., 2013). Among these three MADS-box genes, only FLM was consistently detected as a DS gene in multiple studies. This again highlights the sensitivity issue between the tools (Fig. 1a). Among flowering genes, including those involved in the circadian rhythm and floral transition pathways related to the reported fio1 phenotypes (Kim et al., 2008; Yeom et al., 2009), we identified 10 DS genes commonly detected in multiple studies (Figs 2, S7; FLOR-ID, Bouché et al., 2016). Circadian rhythm-responsible genes include NIGHT LIGHT-INDUCIBLE AND CLOCK-REGULATED 2 (LNK2) and LUX ARRHYTHMO (LUX). While disruption of circadian rhythm is one of the major phenotypes in fio1 (Kim et al., 2008), these genes do not have the //GURAG motif, suggesting a lack of preference for FIO1 through U6 snRNA. On the other hand, four genes in eight floral transition-responsible DS genes, FLOWERING LOCUS K (FLK), RRP6L1, TARGET OF EAT1 (TOE1), and FLM, were potential targets of FIO1 via m6A modification of U6 snRNA with //GURAG motif (Fig. 2; Table S6). Two out of the four DS genes are known FLC regulators, consistent with the well-known fio1 phenotype, low FLC expression. Briefly, RRP6L1 suppresses FLC expression through positive control of noncoding antisense transcripts (Shin & Chekanova, 2014; Zhang et al., 2014). FLK suppresses the splicing of a particular FLC isoform by binding to the m6A motif of the transcripts (Amara et al., 2023). Notably, this FLK-targeted FLC isoform was reported as a target of FIO1-mediated m6A modification by Cai et al. (2022) in study A. Thus, FLK-dependent splicing under FIO1 control may suppress FLC expression. While their effects on FLC expression were not consistently identified across the studies, this may be because Col-0 has low FLC expression due to an inactive FRIGIDA allele (Johanson et al., 2000). The other two out of the DS genes were FT repressors. TOE1 binds to the promoter regions of FT and prevents the activation (Zhang et al., 2015). As mentioned above, FLM is a temperature-sensitive flowering repressor. These results suggest that FIO1 may also affect several regulators, including those downstream of FLC and FT in the flowering pathway, both directly and indirectly. In conclusion, in this study, we compared DS events across four studies and the four DS detection tools to assess the function of FIO1 on alternative splicing events, particularly highlighted by the discrepancies in conclusions between Wang et al. (2022) and Parker et al. (2022). Detection of DS events is sensitive to both data analysis pipelines and RNA-Seq data quality, potentially leading to an underestimation of DS in fio1. Our results provide more conclusive evidence supporting the hypothesis that FIO1 regulates alternative splicing with high reproducibility. In addition, we demonstrated the possibility that FIO1 can directly influence flowering time by controlling the splicing of floral transition-responsible genes. However, it remains unclear the impact of the FIO1 function on phenotypes through m6A modification in mRNA. In our commonly detected DS lists, there was no overlap between m6A modification and splice sites. This suggests that the FIO1 effect of m6A modification in mRNA on alternative splicing might be limited. Nevertheless, we still cannot exclude potential experimental biases related to fio1 phenotypes, such as temperature (Parker et al., 2022), sampling time, and light conditions, which might influence DS detection. Interestingly, such biases might link to the environmental responses of plants by FIO1. Posttranscriptional regulations, such as those mediated by FIO1, including alternative splicing and mRNA stability, are significant sources of pleiotropy, which optimize phenotypic evolution (Hanemian et al., 2020; Govindan et al., 2022). The wide range of pleiotropy observed in fio1 likely results from this optimization. Further research is necessary to deepen our understanding of the precise mode of action of FIO1 and its role in environmental adaptation. This project was supported by Japan Society for the Promotion of Science (20 K22671 and 21H02538, to ES). We would like to thank Gordon G. Simpson and Matthew Parker for their critical reading and valuable comments on this manuscript. None declared. MY and ES planned and designed the research. MY analyzed data. MY and ES wrote the manuscript. Public RNA-Seq datasets using the analyses are available under the following accession numbers: GSE180768 (Cai et al., 2022) and GSE171926 (Sun et al., 2022) from the NCBI Gene Expression Omnibus, PRJEB51363 (Parker et al., 2022) from the European Nucleotide Archive, and CRA004052 (Wang et al., 2022) from the Genome Sequence Archive. Differential splicing analysis data in this article are available in GitHub at https://github.com/mio-ha/fio1_data. Fig. S1 Multidimensional scaling (MDS) plots for the gene expression profiles of the four analyzed datasets. Fig. S2 Overlaps of DS events between the datasets using DEXSeq and edgeR. Fig. S3 Overlaps of DS events between the tools for each dataset. Fig. S4 The impact of data quality on the detection of DS events. Fig. S5 Quality assessment of RNA-Seq data in dataset D. Fig. S6 Base frequencies at splice sites of DS genes. Fig. S7 Meta-analysis of DS patterns detected in fio1. Methods S1 Comprehensive RNA-seq data analysis workflow. Table S1 Overview of the compared DS analysis tools. Table S2 Detected DS events using DEXSeq. Table S3 Detected DS events using edgeR. Table S4 Detected DS events using rMATS. Table S5 Detected DS events using SUPPA2. Table S6 DS events consistently changed in more than one study in fio1. Please note: Wiley is not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Introduction: Middle ear cholesteatoma is characterized by the hyperproliferation of keratinocytes. In recent decades, N6-methyladenosine (m6A) modification has been shown to play an essential role in the pathogenesis of many proliferative diseases. However, neither the m6A modification profile nor its potential role in the pathogenesis of middle ear cholesteatoma has currently been investigated. Therefore, this study aimed to explore m6A modification patterns in middle ear cholesteatoma. Materials and methods: An m6A mRNA epitranscriptomic microarray analysis was performed to analyze m6A modification patterns in middle ear cholesteatoma tissue (n = 5) and normal post-auricular skin samples (n = 5). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed to predict the potential biological functions and signaling pathways underlying the pathogenesis of middle ear cholesteatoma. Subsequently, m6A modification levels were verified by methylated RNA immunoprecipitation-qPCR (MeRIP-qPCR) in middle ear cholesteatoma tissue and normal skin samples, respectively. Results: A total of 6,865 distinctive m6A-modified mRNAs were identified, including 4,620 hypermethylated and 2,245 hypomethylated mRNAs, as well as 9,162 differentially expressed mRNAs, including 4,891 upregulated and 4,271 downregulated mRNAs, in the middle ear cholesteatoma group relative to the normal skin group. An association analysis between methylation and gene expression demonstrated that expression of 1,926 hypermethylated mRNAs was upregulated, while expression of 2,187 hypomethylated mRNAs and 38 hypermethylated mRNAs was downregulated. Moreover, GO analysis suggested that differentially methylated mRNAs might influence cellular processes and biological behaviors, such as cell differentiation, biosynthetic processes, regulation of molecular functions, and keratinization. KEGG pathway analysis demonstrated that the hypermethylated transcripts were involved in 26 pathways, including the Hippo signaling pathway, the p53 signaling pathway, and the inflammatory mediator regulation of transient receptor potential (TRP) channels, while the hypomethylated transcripts were involved in 13 pathways, including bacterial invasion of epithelial cells, steroid biosynthesis, and the Hippo signaling pathway. Conclusion: Our study presents m6A modification patterns in middle ear cholesteatoma, which may exert regulatory roles in middle ear cholesteatoma. The present study provides directions for mRNA m6A modification-based research on the epigenetic etiology and pathogenesis of middle ear cholesteatoma.

Comprehensive analysis of m6A regulator-based methylation modification patterns characterized by distinct immune profiles in colon adenocarcinomas

PRRSV alters m6A methylation and alternative splicing to regulate immune, extracellular matrix-associated function

Paraquat-induced oxidative stress regulates N6-methyladenosine (m6A) modification of long noncoding RNAs in Neuro-2a cells

&lt;p&gt;RNA N6-methyladenosine (m&lt;sup&gt;6&lt;/sup&gt;A) modification in tumorigenesis and progression has been highlighted and discovered in recent years. However, the molecular and clinical implications of m&lt;sup&gt;6&lt;/sup&gt;A modification in melanoma tumor microenvironment (TME) and immune infiltration remain largely unknown. Here, we utilized consensus molecular clustering with nonnegative matrix factorization based on the melanoma transcriptomic profiles of 23 m&lt;sup&gt;6&lt;/sup&gt;A regulators to determine the m&lt;sup&gt;6&lt;/sup&gt;A modification clusters and m&lt;sup&gt;6&lt;/sup&gt;A-related gene signature. Three distinct m&lt;sup&gt;6&lt;/sup&gt;A modification patterns (m&lt;sup&gt;6&lt;/sup&gt;A-C1, C2, and C3), which are characterized by specific m&lt;sup&gt;6&lt;/sup&gt;A regulator expression, survival outcomes, and biological pathways, were identified in more than 1,000 melanoma samples. The immune profile analyses showed that these three m&lt;sup&gt;6&lt;/sup&gt;A modification subtypes were highly consistent with the three known immune phenotypes: immune-desert (C1), immune-excluded (C2), and immune-inflamed (C3). Tumor digital cytometry (CIBERSORT, ssGSEA) algorithm revealed an upregulated infiltration of CD8&lt;sup&gt;+&lt;/sup&gt; T cell and NK cell in m&lt;sup&gt;6&lt;/sup&gt;A-C3 subtype. An m&lt;sup&gt;6&lt;/sup&gt;A scoring scheme calculated by principal component of m&lt;sup&gt;6&lt;/sup&gt;A signatures stratified melanoma patients into high- and low-m&lt;sup&gt;6&lt;/sup&gt;sig score subgroups; a high score was significantly associated with prolonged survival and enhanced immune infiltration. Furthermore, fewer somatic copy number alternations (SCNA) and PD-L1 expression were found in patients with high m&lt;sup&gt;6&lt;/sup&gt;Sig score. In addition, patients with high m&lt;sup&gt;6&lt;/sup&gt;Sig score demonstrated marked immune responses and durable clinical benefits in two independent immunotherapy cohorts. Overall, this study indicated that m&lt;sup&gt;6&lt;/sup&gt;A modification is involved in melanoma tumor microenvironment immune regulation and contributes to formation of tumor immunogenicity. Comprehensive evaluation of the m&lt;sup&gt;6&lt;/sup&gt;A modification pattern of individual tumors will provide more insights into molecular mechanisms of TME characterization and promote more effective personalized biotherapy strategies.&lt;/p&gt;

RNA N6-methyladenosine (m6A) modification in tumorigenesis and progression has been highlighted and discovered in recent years. However, the molecular and clinical implications of m6A modification in melanoma tumor microenvironment (TME) and immune infiltration remain largely unknown. Here, we utilized consensus molecular clustering with nonnegative matrix factorization based on the melanoma transcriptomic profiles of 23 m6A regulators to determine the m6A modification clusters and m6A-related gene signature. Three distinct m6A modification patterns (m6A-C1, C2, and C3), which are characterized by specific m6A regulator expression, survival outcomes, and biological pathways, were identified in more than 1,000 melanoma samples. The immune profile analyses showed that these three m6A modification subtypes were highly consistent with the three known immune phenotypes: immune-desert (C1), immune-excluded (C2), and immune-inflamed (C3). Tumor digital cytometry (CIBERSORT, ssGSEA) algorithm revealed an upregulated infiltration of CD8+ T cell and NK cell in m6A-C3 subtype. An m6A scoring scheme calculated by principal component of m6A signatures stratified melanoma patients into high- and low-m6sig score subgroups; a high score was significantly associated with prolonged survival and enhanced immune infiltration. Furthermore, fewer somatic copy number alternations (SCNA) and PD-L1 expression were found in patients with high m6Sig score. In addition, patients with high m6Sig score demonstrated marked immune responses and durable clinical benefits in two independent immunotherapy cohorts. Overall, this study indicated that m6A modification is involved in melanoma tumor microenvironment immune regulation and contributes to formation of tumor immunogenicity. Comprehensive evaluation of the m6A modification pattern of individual tumors will provide more insights into molecular mechanisms of TME characterization and promote more effective personalized biotherapy strategies.

Analysis of N6-methyladenosine reveals a new important mechanism regulating the salt tolerance of sweet sorghum

IntroductionPseudorabies virus (PRV) is the pathogenic virus of porcine pseudorabies (PR), belonging to the Herpesviridae family. PRV has a wide range of hosts and in recent years has also been reported to infect humans. N6-methyladenosine (m6A) modification is the major pathway of RNA post-transcriptional modification. Whether m6A modification participates in the regulation of PRV replication is unknown.MethodsHere, we investigated that the m6A modification was abundant in the PRV transcripts and PRV infection affected the epitranscriptome of host cells. Knockdown of cellular m6A methyltransferases METTL3 and METTL14 and the specific binding proteins YTHDF2 and YTHDF3 inhibited PRV replication, while silencing of demethylase ALKBH5 promoted PRV output. The overexpression of METTL14 induced more efficient virus proliferation in PRV-infected PK15 cells. Inhibition of m6A modification by 3-deazaadenosine (3-DAA), a m6A modification inhibitor, could significantly reduce viral replication.Results and DiscussionTaken together, m6A modification played a positive role in the regulation of PRV replication and gene expression. Our research revealed m6A modification sites in PRV transcripts and determined that m6A modification dynamically mediated the interaction between PRV and host.

Chinese sacbrood virus (CSBV) is highly lethal to honeybee larvae (especially the larva of Apis cerana) and causes considerable losses to beekeeping industry. N6-methyladenine (m6A) modification of mRNA is a predominant post-transcriptional modification in eukaryotes and plays a role in viral infection. However, the role of m6A modification in CSBV infection remains unclear. Herein, we performed high-throughput sequencing for m6A-seq in CSBV-infected and non-infected larvae to investigate host transcriptome-wide m6A modifications and identify m6A-modified genes. A total of 671 variant peaks were identified. Combined analysis of m6A modification and mRNA expression revealed that a significant correlation between mRNA methylation modifications and expression levels observed for 668 Genes. It was proved that CSBV infection can cause important m6A modification changes in host. We examined the effects of CSBV infection on expression of two methylation regulatory genes by qPCR. At the same time, we verified the effect of two methylation regulatory genes on CSBV replication using RNAi technology. This study demonstrated for the first time that CSBV infection can cause m6A modification changes in A. cerana larvae, and comprehensively analyzed the m6A modification pattern of its mRNA, and CSBV infection significantly promoted the expression of AcMETTL3 (Ac represents A. cerana, p = 0.007), but had no effect on the expression of AcMETTL14. It was further confirmed that AcMETTL3 had a significant negative regulatory effect on CSBV replication (p = 0.0432). These results lay a foundation for further exploration of the role of m6A modification in CSBV infection.

N6-methyladenosine (m6A) modification has been implicated in ischemia-reperfusion injury in various systems and in several neurodegenerative diseases. Glaucoma is characterized by degeneration of retinal ganglion cells (RGCs) and shares similar pathologic injury characteristics with retinal ischemia-reperfusion (RIR) injury. However, the specific role of m6A modification in RIR injury is unclear, and the involvement of autophagy in RIR injury also remains controversial. Therefore, our study explored the role of m6A modification and autophagy in RIR injury. Male wild-type C57BL/6J mice (6-8 weeks old) were used to induce RIR injury. Retinal flat-mount immunofluorescence was performed to assess RGC survival rate. Electroretinogram and optomotor response were conducted to evaluate the retinal electrophysiologic function and visual acuity. Autophagy level was reflected by Western blot and transmission electron microscope images. M6A modification levels were determined via m6A dot blot. Methyltransferase-like protein 3 (Mettl3) and forkhead box O1 (FoxO1) protein expressions were tested by Western blot. Methylated RNA immunoprecipitation-quantitative PCR was conducted to examine m6A modification level on FoxO1 mRNA. We also employed 3-methyladenine and rapamycin to regulate autophagy level in RIR injury. Inhibiting autophagy ameliorated RGC loss and preserved retinal electrophysiologic function in RIR injury. Additionally, a decrease in Mettl3-mediated m6A modification was observed in RIR injury mice. By overexpressing Mettl3 via intravitreal injection of type 2 recombinant adeno-associated virus before RIR injury, we established that Mettl3 overexpression can also ameliorate RGC loss and retinal electrophysiologic dysfunction induced by RIR injury. Furthermore, Mettl3 overexpression inhibited autophagy and reduced FoxO1 expression by upregulating m6A modifications on FoxO1 mRNA. Mettl3-mediated m6A modification mitigates RGC loss and retinal electrophysiologic dysfunction by inhibiting FoxO1-mediated autophagy in RIR injury.

Purpose This study aimed to provide direct evidence of the potential role of N6-methyladenosine (m6A) modification in the progression of myopia. We focused on identifying genes that may be involved in scleral remodeling through m6A regulation in myopia. Methods We utilized m6A methylation immunoprecipitation sequencing (MeRIP-seq) alongside RNA sequencing (RNA-seq) to investigate the levels of m6A modification and mRNA expression in the scleras of form-deprived myopic (FDM) guinea pigs. Subsequent bioinformatics analysis was performed to identify the enriched pathways and genes associated with m6A modification. Results Bioinformatic analyses indicated that hypermethylated mRNAs were predominantly associated with the calcium signaling pathway and may participate in extracellular matrix (ECM) remodeling. Through integrated analysis of MeRIP-seq and RNA-seq data, it was found that more than half of the differentially expressed modified genes (DEGs) exhibiting increased mRNA levels also showed an upregulation of m6A modification levels. These genes may play significant roles in the process of myopic scleral remodeling in response to elevated levels of methyltransferase METTL14. Conclusion This study highlights the role of m6A methylation, mediated by METTL14, in the regulating of key genes involved in calcium signaling and ECM remodeling during myopia progression. These findings suggest that targeting m6A modifications may could offer new therapeutic strategies for the treatment of myopia.

Alternative Splicing: New Insights from Global Analyses

Recent studies have highlighted the biological significance of RNA N6-methyladenosine (m6A) modification in tumorigenicity and progression. However, it remains unclear whether m6A modifications also have potential roles in immune regulation and tumor microenvironment (TME) formation.Methods: In this study, we curated 23 m6A regulators and performed consensus molecular subtyping with NMF algorithm to determine m6A modification patterns and the m6A-related gene signature in colon cancer (CC). The ssGSEA and CIBERSORT algorithms were employed to quantify the relative infiltration levels of various immune cell subsets. An PCA algorithm based m6Sig scoring scheme was used to evaluate the m6A modification patterns of individual tumors with an immune response.Results: Three distinct m6A modification patterns were identified among 1307 CC samples, which were also associated with different clinical outcomes and biological pathways. The TME characterization revealed that the identified m6A patterns were highly consistent with three known immune profiles: immune-inflamed, immune-excluded, and immune-desert, respectively. Based on the m6Sig score, which was extracted from the m6A-related signature genes, CC patients can be divided into high and low score subgroups. Patients with lower m6Sig score was characterized by prolonged survival time and enhanced immune infiltration. Further analysis indicated that lower m6Sig score also correlated with greater tumor mutation loads, PD-L1 expression, and higher mutation rates in SMGs (e.g., PIK3CA and SMAD4). In addition, patients with lower m6Sig scores showed a better immune responses and durable clinical benefits in three independent immunotherapy cohorts.Conclusions: This study highlights that m6A modification is significantly associated with TME diversity and complexity. Quantitatively evaluating the m6A modification patterns of individual tumors will strengthen our understanding of TME characteristics and promote more effective immunotherapy strategies.

Acute myeloid leukemia (AML) is a type of blood cancer that is identified by the unrestricted growth of immature myeloid cells within the bone marrow. Despite therapeutic advances, AML prognosis remains highly variable, and there is a lack of biomarkers for customizing treatment. RNA N6-methyladenosine (m6A) modification is a reversible and dynamic process that plays a critical role in cancer progression and drug resistance. To investigate the m6A modification patterns in AML and their potential clinical significance, we used the AUCell method to describe the m6A modification activity of cells in AML patients based on 23 m6A modification enzymes and further integrated with bulk RNA-seq data. We found that m6A modification was more effective in leukemic cells than in immune cells and induced significant changes in gene expression in leukemic cells rather than immune cells. Furthermore, network analysis revealed a correlation between transcription factor activation and the m6A modification status in leukemia cells, while active m6A-modified immune cells exhibited a higher interaction density in their gene regulatory networks. Hierarchical clustering based on m6A-related genes identified three distinct AML subtypes. The immune dysregulation subtype, characterized by RUNX1 mutation and KMT2A copy number variation, was associated with a worse prognosis and exhibited a specific gene expression pattern with high expression level of IGF2BP3 and FMR1, and low expression level of ELAVL1 and YTHDF2. Notably, patients with the immune dysregulation subtype were sensitive to immunotherapy and chemotherapy. Collectively, our findings suggest that m6A modification could be a potential therapeutic target for AML, and the identified subtypes could guide personalized therapy.