The enzymatic machinery for RNA modifications includes 'writers', 'erasers', and 'readers'. Writers like METTL3/METTL14 for m6A add modifications, while erasers such as FTO and ALKBH5 remove them. Readers including YTH domain proteins recognize and bind these modifications to mediate downstream effects. In disease contexts like cancer, altered expression of writers and erasers can reshape oncogenic programs and influence stemness, immune evasion, or therapy resistance, with METTL3/METTL14 acting variably as oncogenes or tumor suppressors based on tissue context . In neurological disorders, dysregulation of modifications such as m6A impacts neuronal differentiation and synaptic plasticity, linking transcriptomic changes to disorders and neurodegeneration .

Measurement technologies for RNA modifications are evolving with advancements like nanopore direct RNA sequencing offering single-molecule, multi-modification detection at native RNA level, moving towards dynamic and spatial resolution at single-cell levels . These technological advancements promise to provide more detailed and high-resolution insights into RNA modification landscapes, capturing rapid dynamic changes during cellular signaling or environmental adaptation. Future research could see enhanced understanding of modification turnover rates and their roles in real-time physiological processes, opening large-scale explorations of epitranscriptomic roles across various biological fields .

Epitranscriptome mapping technologies include antibody-based enrichment methods like MeRIP-seq or miCLIP for m6A detection, which provide transcriptome-wide maps but have limitations in peak width and antibody specificity. Chemical derivatization techniques like bisulfite RNA-seq for m5C provide single-nucleotide resolution but face challenges like incomplete conversion . Nanopore direct RNA sequencing allows for multi-modified detection in native RNA at the single-molecule level but requires context-dependent caller training . Each technology offers unique strengths, such as transcriptomic robustness or base-level resolution, and is selected based on specific research needs .

In cardiovascular and metabolic diseases, modification enzymes like those involved in m6A or m5C respond to metabolic states, such as α-ketoglutarate dependencies, linking nutrient status directly to gene expression in tissues like the liver and heart. Aberrant control of RNA modifications can influence lipid metabolism, inflammation, and cellular stress responses . Therapies aim at targeting these modifications using small-molecule modulators or programmable RNA editors to correct dysregulated pathways potentially causing these diseases. The focus is on developing site-specific, precise RNA editing tools that minimize off-target effects to ensure safety and efficacy .

Cancer cells exploit RNA modifications to promote oncogenic pathways, influencing gene expression profiles that favor proliferation, therapy resistance, or immune evasion. METTL3 and METTL14, as parts of the m6A methyltransferase complex, can act as oncogenes or tumor suppressors, with their roles differing based on the cellular context and tissue type . Therapeutic strategies include developing small molecules to inhibit RNA-modifying enzymes or design site-specific RNA modifiers to correct detrimental modifications. Additionally, RNA-based therapeutics aim to bypass modification-mediated immune evasion or drug resistance . By targeting these cellular mechanisms, therapeutic interventions can offer new means to combat tumor progression and resistance .

'Readers' are proteins that recognize and bind specific RNA modifications to orchestrate numerous cellular processes. YTH domain proteins, for example, bind m6A and influence translation initiation, mRNA decay, and nuclear processing . IGF2BP proteins recognize m6A in different motifs, stabilizing selected transcripts . Additionally, HNRNPs are involved in splicing, while ALYREF mediates mRNA export, especially recognizing m5C modifications . The interaction of readers with modifications is critical for contextual regulation of gene expression, reflecting cellular adaptation and response mechanisms to external and internal stimuli .

RNA modifications such as m6A, pseudouridine, and m5C play critical roles in regulating various stages of the mRNA life cycle. m6A marks are involved in co-transcriptional splicing by influencing splice-site choice and alternative polyadenylation through the recruitment of co-transcriptional writers marking nascent transcripts . In translation control, m6A enables cap-independent translation under stress by binding to eIF3 at m6A sites in the 5' UTR, enhancing initiation via mRNA looping . For stability, m6A facilitates mRNA decay by recruiting the decay machinery via YTHDF2, while IGF2BP recognition can stabilize transcripts, creating context-dependent outcomes . Pseudouridine contributes by altering endonuclease sensitivity, and m5C is linked to nuclear export and mRNA stability .

mRNA modifications such as m6A and m5C play roles in immune response regulation by shaping innate sensing mechanisms like RIG-I and TLRs and influencing cytokine expression . Viral replication often exploits these modifications to evade immune detection. Therapeutically, modified nucleosides such as N1-methylpseudouridine in mRNA vaccines are used to reduce innate immune activation while enhancing translation, providing a basis for improved vaccine design . Developing small-molecule modulators targeting writers, readers, and erasers can offer potential interventions in immunity-related conditions, offering new avenues for therapies needing fine-tuned immune modulation .

Reproducibility in epitranscriptomic research is challenged by variability in antibody specificity, efficiency of chemical conversions, and sequence-context biases. Orthogonal validation using multiple distinct methods is necessary to confirm findings . Moreover, standardizing spike-ins across labs, documenting antibody lot variations, and releasing raw data are essential practices to enhance reproducibility . In computational analyses, preregistration of analytic plans and full code releases are recommended to add transparency and consistency, which are critical in validating modified site predictions and ensuring the robustness of reported findings .

Computational analysis for RNA modifications involves processing raw sequencing reads, calling peaks or modification sites, and controlling false discovery rates using models that are aware of replicate conditions . Machine learning models integrate features like sequence k-mers, structural elements, and evolutionary conservation to prioritize modification sites, evaluated using metrics like AUROC and AUPRC . Challenges include ensuring cross-condition generalization and managing biases such as GC content or fragment-specific issues . Despite these challenges, computational tools are crucial in identifying novel modification sites and validating their potential functional roles within cellular transcripts .