Oligonucleotide mass spectrometry (MS-Seq) is emerging as a powerful approach for sequence-resolved RNA modification analysis, yet the field lacks standards for experimental workflows, data analysis and reporting. To assess current capabilities, the Human RNome Project Consortium conducted a cross-platform benchmarking study using a common RNA sample. A partial RNase T1 digest of human 28S rRNA was distributed to participating laboratories and analysed using existing LC-MS/MS workflows spanning different chromatographic strategies and mass spectrometers. To enable direct comparison, datasets were analysed using a harmonized NucleicAcidSearchEngine (NASE) workflow. Despite substantial methodological differences, laboratories recovered highly overlapping oligonucleotide sets and generated similar sequence coverage maps with a global coverage of 54.16%, demonstrating reproducible sequence information across platforms under standardized sample and analysis conditions. The benchmark further revealed incomplete sequence coverage, platform-specific differences in data architecture and increased assignment ambiguity during dynamic modification searches. Together with the community consensus developed during the HRPC workshop, these findings define priorities for the field, including improved sensitivity, standardized data analysis and reporting, community repositories, and robust bioinformatic workflows for confident de novo RNA modification discovery. This study provides an experimental benchmark and roadmap toward routine MS-based mapping of the human RNome.

More than 180 chemically distinct RNA modifications – the epitranscriptome – have now been reported across all kingdoms of life and all forms of RNA. Given the emerging evidence for critical roles played by RNA modifications in many diseases, it is imperative to identify the function of all RNA modifications in the nearly 250,000 types of RNA in human cells – the human RNome. Here we take a step toward this goal by reviewing the literature to date and assembling a catalog of all established or putative human RNA modifications. While the identities and locations of many human RNA modifications are well established, emerging technologies are revealing new ones as well as old ones in unexpected locations, with low abundance and poorly characterized structures hampering their validation. To this end, we introduce an analytical validation framework that scores each reported human RNA modification for chemical identity and transcriptomic localization and then assigns it to one of three confidence tiers: Authenticated, Provisional, or Putative. Applying this framework, we critically re-examine the primary literature underlying 24 recently reported additions to the human epitranscriptome, evaluating the mass spectrometric, spectroscopic, and sequencing-based evidence supporting each as well as discussing their emerging biological roles and disease associations. We further extend this scoring system to modifications previously summarized in reviews and databases but not yet subjected to this level of scrutiny, compiling a comprehensive, tier-ranked inventory of the human epitranscriptome as of July 2026. Together, this catalog and framework provide a benchmark for evaluating future epitranscriptome discoveries, mapping the human RNome, and for derisking the translation of validated modifications as diagnostic and therapeutic targets.

N6-threonylcarbamoyladenosine (t6A) is a universal transfer RNA (tRNA) modification essential for translational fidelity. The modification is installed at position 37 of ANN-decoding tRNAs (N is A, U, G, or C) by transfer of a threonylcarbamoyl moiety from a pathway intermediate, catalyzed in bacteria by the TsaBD complex. Despite the strict requirement for the 36-UAA-38 sequence in substrate tRNAs, the structural basis for this specificity has remained unclear. We determined cryo-EM structures of the Thermotoga maritima t6A synthase bound to unmodified and to natively modified tRNA carrying the t6A37 modification, at a nominal resolution of 3.2 Å. In both structures, A37 is positioned in the active site, and the anticodon loop is remodeled into a zig-zag conformation not previously observed in tRNA, stabilized by conserved interactions with the RNA backbone at the TsaD/TsaB interface. The wobble and middle anticodon bases occupy shallow surface pockets without base-specific contacts, explaining tolerance to base identity at these positions. In contrast, U36 and A38 are recognized indirectly through formation of a base triple with U32. Additional D-stem contacts provide a second mode of indirect readout. Together with mutagenesis data, the structures reveal the molecular basis for restriction of t6A37 to ANN-decoding tRNAs and the requirement for A38.

BackgroundThe degeneracy of the genetic code is increasingly recognized for roles in regulating translation rate, protein folding, and cell response. However, the functional genomics of codon usage patterns remains poorly defined. We previously showed that prokaryotic and eukaryotic cells respond to individual stresses by uniquely reprogramming the tRNA pool and the dozens of tRNA modifications comprising the tRNA epitranscriptome to cause selective translation of mRNAs from codon-biased stress response genes. Here, we tested the hypothesis that functional gene families have distinct values of codon bias in the Saccharomyces cerevisiae genome by modeling isoacceptor codon distributions using a new approach-analysis of synonymous codon signatures (ASCS).ResultsApplication of ASCS to the S. cerevisiae genome revealed linear relationships between patterns of codon bias and gene function using canonical correlation analysis. By mapping codon-biased open reading frames (ORFs) onto a functional network of gene ontology (GO) categories, we identified 91 gene families distinguished by unique codon usage signatures. The codon usage patterns were found to strongly predict functional clusters of genes, such as translational machinery, transcription, and metabolic processes.ConclusionsThe ASCS-derived model of codon usage patterns in S. cerevisiae reveals functional codon bias signatures and captures more biologically meaningful information when compared to other codon analytical approaches.

Abstract Escherichia coli uses wobble uridine (U34) modifications to tune codon decoding, but how individual tRNA writer enzymes shape gene expression remains unclear. Here, we identify MnmA, the U34 thiolation enzyme for tRNALys, tRNAGln, and tRNAGlu, as a central regulator linking codon-directed translation to regulon control and stress response. Loss of MnmA depleted s 2 U-dependent wobble modifications, preventing geranyl-(ges 2 U) and seleno-(se 2 U)-based modifications, causing growth defects, reduced catalase activity, and multi-level gene expression dysregulation. The Δ mnmA cells showed broad adaptive transcriptional reprogramming associated with RpoS- and OxyR-regulated pathways, which was accompanied by compromised protein output. Endogenous and tagged-protein analyses revealed specific impairment of transcriptional regulators, adaptive and detoxification proteins, including RpoS, OxyR, FliA, KatE, and KatG. Polysome profiling and polysome-associated RNA sequencing showed that MnmA deficiency globally reduces translational capacity and uncouples mRNA abundance from translational efficiency, which is exacerbated during oxidative stress. We developed genome-wide codon-usage mapping analytics to identify five codon-defined gene clusters, with specific clusters enriched for Lys, Gln, and Glu codons disproportionately affected by MnmA loss. Together, these findings support that wobble uridine thiolation and downstream modifications pair with corresponding codon architecture to coordinate the translation of regulon controllers and stress-response networks linked to bacterial fitness. Graphical Abstract

RNA modification analysis by LC-MS/MS is central to epitranscriptomics, yet quantitative comparison across laboratories and instrument platforms remains poorly standardized. Here, we performed a community-driven benchmarking study during the first Human RNome Project workshop to systematically evaluate cross-platform reproducibility of ribonucleoside mass spectrometry workflows. Using the same analytical column and gradient, standardized RNA samples, and shared reagents, we compared nucleoside quantification across quadrupole, time-of-flight, and orbitrap-based LC-MS platforms employing distinct acquisition strategies. While chromatographic separation was highly reproducible across systems, nucleoside-specific MS response behavior differed substantially between platforms and limited direct comparability of relative signal intensities. These response differences varied across analytes and concentration ranges, demonstrating that harmonized chromatography alone is insufficient for transferable quantitative analysis. Stable isotope-labeled internal standard (SILIS) normalization substantially reduced platform- and method-dependent response and improved agreement for most evaluated modifications. External calibration improved agreement between qTOF and Orbitrap workflows for a subset of modifications but did not fully resolve residual intersystem differences. Based on these findings, we establish benchmark-derived recommendations for harmonized relative and absolute RNA modification quantification, including guidance for calibration design, quality control, and data reporting. Together, this work provides a methodological framework for reproducible nucleoside LC-MS/MS workflows and establishes a foundation for large-scale comparative epitranscriptomic studies.

Synthetic genetic circuits, engineered networks of genetic components that perform user-defined logical functions in living cells, enable precise control over cellular behavior. As circuits become more complex to support sophisticated functions, implementing them within a single cell population becomes difficult. Tasks can instead be distributed among different populations in engineered consortia. However, selecting appropriate consortia and harnessing the natural traits of each species for such a division of labor remain unsolved challenges. Here, we report the creation of SINERGY, an engineered synthetic sourdough starter in which functional modules are distributed between the two constituent species. These modules, consisting of sensing, cross-kingdom communication, and response, enable in vitro biosensing and in vivo drug delivery and modulate the gut ecosystem in an animal disease model. SINERGY not only preserves the long-standing safety and health-promoting properties of the natural components but also endows symbiotic microbes with programmable functionalities, facilitating broad applications in biomedicine.

Stress promotes phenotypic changes in bacteria that allow them to survive antibiotic treatment. This phenomenon, known as antibiotic tolerance, can cause antibiotic treatment failure, making it important to define the bacterial pathways promote survival. Previously, we found Yersinia pseudotuberculosis downregulates tusB , a gene involved in modifying glutamate, glutamine, and lysine tRNAs with s2U, in response to doxycycline. Here, we find deletion of tusB induces tolerance to ribosome and RNA polymerase- targeting antibiotics. We hypothesize that loss of s2U induces slowed growth and antibiotic tolerance by slowing down translation of proteins enriched in glutamate, glutamine, and lysine codons. In line with this, we find that ribosomes pause more frequently at these codons in Δ tusB . Ribosomal proteins themselves are highly enriched in these codons and significantly reduced in protein abundance. Our results indicate that reduction in ribosomal proteins may be globally reducing translation and could signify a previously unknown mechanism of antibiotic tolerance. Highlights ### Competing Interest Statement The authors have declared no competing interest. National Institute of Allergy and Infectious Diseases, https://ror.org/043z4tv69, AI154116, AI175307, AI007417-26

By integrating a literature review with transcriptomic, proteomic, and phenotypic data from two model bacteria, Escherichia coli and Vibrio cholerae, we put forward the hypothesis that defects in tRNA modification broadly impact processes that are evolutionarily tuned to be sensitive to translation speed. These include the translation of regulatory proteins associated with motility, iron homeostasis, and leader peptide-driven attenuation mechanisms. Some of these translation speed-dependent processes are influenced by the absence of a single modification, while others are affected by the absence of multiple modifications. Although further experiments are needed to clarify the mechanisms involved in each case, this work provides a foundational framework to guide future research.

The human RNome comprises all forms of RNA and the 50 + chemical structures-the epitranscriptome-that modify them. Understanding the diverse functions of RNA modifications in regulating gene expression and cell phenotype requires technologies such as RNA sequencing-based modification mapping and mass spectrometry-based quantification of modified ribonucleosides. Liquid chromatography-coupled tandem quadrupole mass spectrometry (LC-MS/MS) is the gold standard for detecting and quantifying modified ribonucleosides with accuracy and precision. However, variations in RNA isolation, processing, and LC-MS/MS analysis have hindered reproducibility across laboratories, which is essential for accurate quantification of RNA modifications. As guidance toward harmonization, we report a multi-laboratory comparison of workflows for LC-MS/MS RNA modification analysis. We compared protocols for sample shipment, RNA hydrolysis, LC-MS/MS analysis, and data processing among three laboratories working with the same total RNA samples. We detected and quantified 17 modifications consistently across protocols and operators, with another 7 that were sensitive to experimental conditions, reagent contamination, and ribonucleoside instability, leading to poor precision among laboratories. Agreement among the three labs was strong, with coefficients of variation of 20% and 10% for relative and absolute quantification, respectively. These findings establish a robust and readily adoptable epitranscriptome analytical platform that enables reliable comparisons across laboratories.

Bacteria encode diverse anti-phage systems, such as CRISPR-Cas and restriction modification (RM), which limit infection by targeting phage DNA. We identified a DNA modification in phages, i.e., 5-arabinosyl-hydroxy-cytosine (5ara-hC), which adds arabinose to cytosines via a hydroxy linkage and protects phage from DNA targeting. The hydroxy linkage was common among arabinoslyated phages, with some arabinosylated phages encoding arabinose-5ara-hC transferases (Aat) that add a second or third arabinose to DNA. DNA arabinosylation enables evasion from DNA-targeting type I CRISPR-Cas and type II RM systems. However, arabinosylated phages remain sensitive to RNA-targeting CRISPR-Cas (type III and VI) and promiscuous type IV restriction endonucleases. 5ara-hC enables evasion of glycosylase defenses that target phages with glucosylated hydroxymethyl cytosines, and 5ara-ara-hC protects against some defenses capable of targeting 5ara-hC-modified phages. Collectively, this work identifies DNA modifications that enable phages to evade multiple defenses yet remain vulnerable to some systems that target RNA or modified nucleobases.

Stress promotes phenotypic changes in bacteria that allow them to survive antibiotic treatment. This phenomenon, termed antibiotic tolerance, can cause treatment failure, highlighting a need to define bacterial pathways that promote survival. Previously, we found Yersinia pseudotuberculosis downregulates tusB , a gene involved in modifying tRNAs with s 2 U, in response to doxycycline. Here we find that deletion of tusB results in loss of s 2 U and induces antibiotic tolerance. Using a combination of sequencing-based approaches and analysis of gene codon usage, our data show that loss of s 2 U decreases translation of ribosomal proteins. Ribosomal proteins are highly enriched in codons that require s 2 U-modified tRNAs for efficient translation, and loss of s 2 U results in ribosome pausing at these codons. Our results highlight a previously unknown mechanism of antibiotic tolerance where reduction in ribosomal protein abundance can globally reduce translation, and describes a novel strategy bacteria use to slow growth by modulating s 2 U levels.

Epigenetic regulation of gene expression and host defense is well established in microbial communities, with dozens of DNA modifications comprising the epigenomes of prokaryotes and bacteriophage. Phosphorothioation (PT) of DNA, in which a chemically reactive sulfur atom replaces a non-bridging oxygen in the sugar-phosphate backbone, is catalyzed by dnd and ssp gene families widespread in bacteria and archaea. However, little is known about the role of PTs or other microbial epigenetic modifications in the human microbiome. Here we optimized and applied fecal DNA extraction, mass spectrometric, and metagenomics technologies to characterize the landscape and temporal dynamics of gut microbes possessing PT modifications. Exploiting the nuclease-resistance of PTs, mass spectrometric analysis of limit digests of PT-containing DNA reveals PT dinucleotides as part of genomic consensus sequences, with 16 possible dinucleotide combinations. Analysis of mouse fecal DNA revealed a highly uniform spectrum of 11 PT dinucleotides in all littermates, with PTs estimated to occur in 5–10 The results of our studies provide a benchmark for understanding the behavior of an abundant and chemically reactive epigenetic mark in the human gut microbiome, with implications for inflammatory conditions of the gut.

Artemisinin has long been a first-line antimalarial. Yet, its mode of action is still poorly understood. Emergence of artemisinin-resistant strains highlight the importance of addressing this question so as to develop better drugs and overcome resistance. In this study, we performed RNA-sequencing and proteomics studies on artemisinin treated parasites indicated a striking difference in the codon-usage pattern of differentially translated genes. Using a liquid chromatography-coupled mass spectrometry (LC-MS)-based platform, we have quantified the full spectrum of modified ribonucleosides on tRNA in P. falciparum in response to the drug. We found that N6-threonyl-carbomyladenosine (t6A), a universal tRNA modification found at position 37 is hypomodified in response to artemisinin induced stress. Additionally, we also found that artemisinin treatment resulted in a downregulation of PfSua5, an enzyme involved in the t6A biosynthesis machinery. These findings provide new insights into how artemisinin works. More broadly, the findings exposes the tRNA epitranscriptome as a vulnerability in the parasite that can be exploited for new drugs.

Building on decades of work in characterizing the dozens of RNA modifications in the microbial epitranscriptome, recent advances in analytical technology and genetics have revealed systems-level functions for many tRNA modifications. The tRNA (uracil-5-)-methyltransferase TrmA and its product, 5-methyl uridine (m5U) at position 54 in the T-loop, however, has not been linked to a specific phenotype. Here, we defined the functional and biological roles of TrmA in Pseudomonas aeruginosa (PA14), a major multidrug-resistant pathogen. Surprisingly, though TrmA was found to site-specifically catalyze m5U54 on all PA14 tRNAs, loss of TrmA had no effect on the levels of any of 36 tRNA modifications except m5U and had minimal effects on multiple phenotypic parameters, including growth rate, morphology, motility, and biofilm formation. However, loss of TrmA conferred a striking polymyxin antibiotic resistance. mRNA and tRNA profiling and proteomics analyses revealed that TrmA regulates the expression of codon-biased gene families at the level of translation, including components of a type III secretion system (T3SS). Loss of TrmA upregulated T3SS, leading to increased macrophage IL-1β in bacterial challenge tests. Altogether, these results revealed novel biological functions of TrmA and its roles in modulating gene expression at multiple levels in P. aeruginosa.

DNA single strand breaks (SSBs) are abundant lesions due to cell metabolism and reactive oxygen species (ROS) and can lead to replication folk collapse, double strand break formation, genome rearrangements, cell death and disease. Among numerous chemical forms of SSBs, 5′-aldehyde terminus are the most abundant generated by hydroxyl radicals and pose significant challenge for cellular repair machinery due to the lack of specific end processing process, potentially leading to more severe biological consequences than other readily repairable SSBs. Herein we developed a new strategy to locate 5′-aldehyde terminus in genomic DNA at single-nucleotide resolution. The principle involves labelling the 5′-aldehyde terminus with an aminooxy-functionalized oligonucleotide, giving rise to a biocompatible altered DNA linkage and allowing labelled sites to be amplified by polymerase chain reaction. We sequenced the 5′-aldehyde terminus distribution in genomic DNA, nuclei, and cells following activation of the Fenton reaction. The results revealed a significant preference for adenine bases in DNA lesions and provided insights into the genome-wide distribution of such DNA damage, correlating with genomic features and chromatin accessibility. This method provide a new strategy for studies aiming to understand the biological and toxicological impacts of 5′-aldehyde termini in DNA as the form species of single strand break induced by reactive oxygen species from a human genome. ### Competing Interest Statement The authors have declared no competing interest.

Sudden environmental changes are a recurring challenge for unicellular organisms, but a necessity for many to progress through their lifecycle. To transmit from its human host to mosquito vector, malaria parasites differentiate into male and female, semi-quiescent stages that can re-initiate development within seconds after transmission. Here, we identify the RNA modification N6-methyladenosine (m6A) as the mediator of a rapid, sex-specific, and temperature-sensitive mechanism to restructure protein synthesis during transmission. We find that male parasites maintain high levels of translation during their semi-quiescence that are rapidly repressed following mosquito uptake. This translational shutdown is essential for the continuation of male parasite development and depends on the m6A-binding protein YTH.2. We further show that m6A and YTH.2 are already present prior to transmission, but that their repressive interaction requires a temperature drop accompanying the exit from the human host. Hence, m6A appears to prime the parasite transcriptome and subsequently converts an environmental shift into a rapid translational response. ### Competing Interest Statement The authors have declared no competing interest.

Arginine, glutamic acid and selenocysteine based codon bias has been shown to regulate the translation of specific mRNAs for proteins that participate in stress responses, cell cycle and transcriptional regulation. Defining codon-bias in gene networks has the potential to identify other pathways under translational control. Here we have used computational methods to analyze the ORFeome of all unique human (19,711) and mouse (22,138) open-reading frames (ORFs) to characterize codon-usage and codon-bias in genes and biological processes. We show that ORFeome-wide clustering of gene-specific codon frequency data can be used to identify ontology-enriched biological processes and gene networks, with developmental and immunological programs well represented for both humans and mice. We developed codon over-use ontology mapping and hierarchical clustering to identify multi-codon bias signatures in human and mouse genes linked to signaling, development, mitochondria and metabolism, among others. The most distinct multi-codon bias signatures were identified in human genes linked to skin development and RNA metabolism, and in mouse genes linked to olfactory transduction and ribosome, highlighting species-specific pathways potentially regulated by translation. Extreme codon bias was identified in genes that included transcription factors and histone variants. We show that re-engineering extreme usage of C- or U-ending codons for aspartic acid, asparagine, histidine and tyrosine in the transcription factors CEBPB and MIER1, respectively, significantly regulates protein levels. Our study highlights that multi-codon bias signatures can be linked to specific biological pathways and that extreme codon bias with regulatory potential exists in transcription factors for immune response and development.

Mitochondrial small open reading frame (ORF)-encoded microproteins (SEPs) are key regulators and components of the electron transport chain (ETC). Although ETC complex I assembly is tightly coupled to nutrient availability, including serine, the coordinating mechanism remains unknown. A genome-wide CRISPR screen targeting SEPs revealed that deletion of the LINC00493-encoded microprotein SMIM26 sensitizes cells to one-carbon restriction. SMIM26 interacts with mitochondrial serine transporters SFXN1/2 and the mitoribosome, forming a functional triad that facilitates translation of the complex I subunit mt-ND5. SMIM26 loss impairs serine import, reduces folate intermediates, and disrupts key mitochondrial tRNA modifications (τm5U and τm5s²U), resulting in ND5 translation failure and complex I deficiency. SMIM26 deletion is embryonic lethal in mice and impedes tumor growth in a xenograft model of folate-dependent acute myeloid leukemia. These findings define SMIM26 as a critical integrator of one-carbon flux and complex I biogenesis and establish a paradigm for localized mitochondrial translation through transporter-ribosome interactions.